Physics aspects of safety assurance in high dose rate brachytherapy: quality control testing and implementation of dosimetry audit

Physics aspects of safety assurance

in high dose rate brachytherapy:

quality control testing and

implementation of dosimetry audit

by

Antony Lee Palmer

Submitted for the degree of Doctor of Philosophy in Physics

Department of Physics University of Surrey January 2015 ©Antony Lee Palmer 2015

ii

i

Contents

Abstract vii

Acknowledgments ix

List of Figures xi

List of Tables xv

List of Abbreviations xvii

CHAPTER 1 Introduction and Overview

1.1 Introduction 2

1.1.1 Brachytherapy 2

1.1.2 Requirements for Accuracy of Dose Delivery 2

1.1.3 Quality Control, Dosimetry Audits and Treatment Errors 3

1.2 Research Questions and Objectives 4

1.3 Summary of Thesis 5

1.4 List of Publications and Presentations Arising from this Work 7

1.4.1 Quality Control and Accuracy of Brachytherapy 7

1.4.2 Dosimeters for Brachytherapy Audit 8

1.4.3 Radiochromic Film Methodology 8

1.4.4 Brachytherapy Audit 9

1.4.5 Monte Carlo Calculations 10

1.4.6 Miscellaneous Publications Relating to PhD Work 10

CHAPTER 2 Theory

2.1 High Dose Rate (HDR) Brachytherapy Equipment 12

2.2 HDR Brachytherapy Dose Distribution Measurement 12

2.2.1 Review of Contemporary Dosimetry Systems for Brachytherapy 13

2.2.2 Optical Fibre Dosimetry 15

2.2.3 Radiochromic Film Dosimetry 15

2.2.4 Radiochromic Plastic Dosimetry 17

ii

2.3 Brachytherapy Audit 17

2.3.1 Definition of Audit and the Need for Dosimetric Audit in Radiotherapy 17

2.3.2 The Need for Dosimetric Audit in Brachytherapy 19

2.4 Monte Carlo MCNP5 Simulation 20

2.4.1 The Monte Carlo Method 20

2.4.2 MCNP5 Applied to Brachytherapy 20

2.4.3 MCNP5 Input and Output Files 21

CHAPTER 3 Quality Control of High Dose Rate (HDR) Brachytherapy

Treatment Equipment

3.1 The Need to Review Quality Control and Commissioning Procedures

and Establish Performance Requirements 24

3.1.1 Quality Control Testing of HDR Brachytherapy Systems 24

3.1.2 Commissioning of HDR Brachytherapy Systems 25

3.1.3 Performance Requirements: Effect of Simulated Source Position Errors 26

3.2 Methodology 26

3.2.1 Survey of Quality Control Practices in the United Kingdom (UK) 26

3.2.2 HDR Treatment Unit Commissioning Tests, Dwell Position and Transit Dose 26

3.2.3 Treatment Planning Study to Determine the Effect of Simulated Source Position Errors 28

3.3 Results 28

3.3.1 UK Survey: Brachytherapy Equipment Profile and Physics Processes 28

3.3.2 UK Survey: Quality Control Testing 32

3.3.3 Commissioning and QC of HDR Treatment Units 33

3.3.3.1 Source Movement Profile 44

3.3.3.2 Transit Dosimetry 45

3.3.3.3 Dwell Position Accuracy with Transfer Tube Curvature 48

3.3.4 Treatment Planning Study of Simulated Source Position Errors 49

3.4 Discussion and Conclusions 54

3.4.1 UK Survey 54

3.4.2 Treatment Unit Commissioning and QC: Dwell Position Accuracy and Transit Dose 55

3.4.3 Equipment Performance Requirements 55

3.4.4 Future Directions for Quality Control Testing of HDR Brachytherapy 56

iii

CHAPTER 4 Candidate Dosimeters for Brachytherapy Applicator

Dosimetry Audit

4.1 Requirements of Dosimetry Systems for Brachytherapy Audit 60

4.2 Methodology 61

4.2.1 Dosimetry Systems 61

4.2.1.1 Doped Silica Glass Optical Fibres 61

4.2.1.2 Radiochromic Film 62

4.2.1.3 Solid Radiochromic Polymer 63

4.2.2 Test Objects and Irradiation Conditions 63

4.3 Results 65

4.3.1 Initial Processing and Calibration of Dosimeters 65

4.3.2 Isolated Source Radial Dose Measurements 66

4.3.3 Dose Distribution Measurements for Multi-Dwell Treatment Plans 68

4.3.4 Uncertainty Analysis 71


4.4.1 Doped Silica Glass Optical Fibres 73

4.4.2 Radiochromic Film 74

4.4.3 Solid Radiochromic Polymer 74

4.4.4 Dosimetric and Practical Considerations for Dosimetry System Choice 75

4.4.5 Selection of Dosimeter 76

CHAPTER 5 Development of Radiochromic Film Dosimetry for

Brachytherapy Audit

5.1 Objectives for the Development and Evaluation of Film Dosimetry for

Brachytherapy 78

5.2 Methodology 80

5.2.1 Film Dosimetry Equipment, Calibration and Scanning 80

5.2.2 Investigation of Film Dosimetry Performance Parameters

and Evaluation of Triple-Channel Dosimetry 83

5.2.3 Validation of Film Dosimetry Response to HDR Brachytherapy Sources 87

5.3 Results 87

5.3.1 Film Calibration, Scanning and Processing 87

iv

5.3.2 Calibration Function Linear-Scaling 90

5.3.3 Post-Irradiation Film Darkening 91

5.3.4 Lateral Position of Film on Scanner 92

5.3.5 Film Surface Perturbation 95

5.3.6 Film Active Layer Thickness 96

5.3.7 The Effect of Film Curvature at Scanning 96

5.3.8 Film Measurement and Validation of Radial Dose from an Ir-192 HDR

Brachytherapy Source 100

5.3.9 Film Measurement and Validation of Radial Dose from a Co-60 HDR

Brachytherapy Source 100


5.4.1 Dosimeter Performance Characteristics 103

5.4.2 Use of Radiochromic Film Dosimetry for Brachytherapy Dosimetric Audit 105

CHAPTER 6 Development of an ‘End to End’ Brachytherapy Dosimetric

Audit using Radiochromic Film

6.1 Review and Evaluation of Previous Brachytherapy Dosimetric Audits 108

6.2 Objectives for a Dosimetric Audit in Brachytherapy 109

6.3 Methodology 112

6.3.1 Design of an Audit Phantom 112

6.3.2 ‘End to End’ Audit Procedure Development 113

6.3.3 Sensitivity to Simulated Errors 113

6.4 Results 113

6.4.1 BRachytherapy Applicator Dosimetry (BRAD) Phantom Design 113

6.4.2 ‘End to End’ Audit Procedure 116

6.4.3 Film Dosimetry and Data Analysis Procedure 119

6.4.4 Sensitivity to Simulated Errors 121

6.4.5 BRAD Audit Uncertainty Budget 121


v

CHAPTER 7 Monte Carlo Simulations for the Brachytherapy Film

Dosimetry Audit

7.1 Objectives for MCNP5 Simulations 128

7.2 Methodology 128

7.2.1 MCNP5 Input Files 128

7.2.2 MCNP5 Validation of Implementation 130

7.2.3 Phantom Scatter: Spherical Scatter Volume with Point Source 131

7.2.4 Effect of the Presence of EBT3 Film on Measured Dose 133

7.2.5 Applicator Attenuation 134

7.3 Results 136

7.3.1 Validation 136

7.3.2 Phantom Scatter 137

7.3.3 Film Dosimetry Dose Perturbation 137

7.3.4 Applicator Attenuation 141

7.4 Conclusions and Discussion 142

CHAPTER 8 ‘End to End’ Audit of Clinical Brachytherapy Dosimetry

in the United Kingdom

8.1 Objectives for the UK Brachytherapy Audit 144

8.2 Methodology 144

8.2.1 Establishment of a Working Party of the Institute of Physics and Engineering in

Medicine (IPEM) 144

8.2.2 Brachytherapy Applicator Dosimetry ‘End to End’ Audit Methodology 144

8.2.3 Data Analysis 145

8.2.4 Schedule of Audits 146

8.3 Results 147

8.3.1 Point A Dose Measurement for Cervix Treatment Applicators 147

8.3.2 Dose Distribution Measurement for Cervix Treatment Applicators 151

8.3.3 Review of Physics Procedures for HDR Brachytherapy Treatment 154

8.3.4 Results Requiring Further Investigation 156

8.3.4.1 Library Applicator Misalignment 156

8.3.4.2 Prescription Dose Error 158

8.3.5 Feedback from Audited Centres 159


vi

CHAPTER 9 Summary Conclusions and Future Work

9.1 Summary Conclusions 166

9.1.1 Quality Control of HDR Brachytherapy Equipment 166

9.1.1.1 UK Survey and Analysis 166

9.1.1.2 Development of QC and Commissioning Tests and Performance Requirements 166

9.1.2 HDR Brachytherapy Audit Dosimetry 167

9.1.2.1 Evaluation of Dosimeters 167

9.1.2.2 Optimisation of Radiochromic Film Dosimetry 168

9.1.3 ‘End to End’ Audit of Brachytherapy Dosimetry 169

9.1.3.1 Design of an Audit Phantom and Audit Methodology 169

9.1.3.2 Monte Carlo Simulations 169

9.1.3.3 UK Audit 170

9.2 Future Work 171

9.2.1 Quality Control and Commissioning Testing for HDR Brachytherapy 171

9.2.2 Brachytherapy Film Dosimetry Development 172

9.2.3 Brachytherapy Dosimetric Audit 172

9.2.3.1 Development of the BRAD System for ‘End to End’ Dosimetry Audit 172

9.2.3.2 Future Directions for Audit 173

References

Reference list 176

Appendices

A. Abstracts of Peer Reviewed Journal Papers Resulting from this Research 192

B. Abstracts of Conference Presentations and Posters Resulting from this Research 197

C. Risk Assessment for Brachytherapy Dosimetric Audit with the BRAD Phantom 203

D. Protocol for Brachytherapy Dosimetric Audit with the BRAD Phantom 204

E. BRAD Phantom Availability in IPEM Virtual Equipment Library (Internet site) 211

F. MCNP5 Input File for Validation Test 212

G. MCNP5 Input File for Phantom Scatter Evaluation 216

H. MCNP5 Input File for the Effect of the Presence of EBT3 Film on Measured Dose 217

I. MCNP5 Input File for the Evaluation of Treatment Applicator Attenuation 219

J. Questionnaire Used for the Survey of HDR/PDR Quality Control Practice in the UK 221

vii

Abstract

This work is concerned with physics-aspects of safety, quality control (QC) and

dosimetry audit in high dose rate (HDR) gynaecological brachytherapy. A survey of

brachytherapy QC practice across the UK was conducted. Areas of least consistency were

addressed, including test method development and establishment of clinical performance

requirements. ‘End to end’ dosimetry auditing was not being utilised and its implementation

was the main focus of this work. Three candidate dosimeters were evaluated for use in audit:

Fibre optic thermoluminescence detector, Gafchromic EBT3® radiochromic film, and

Presage® radiochromic plastic. Film dosimetry was selected, fully characterised, triple-

channel dosimetry evaluated, and uncertainty reduction methods implemented.

A novel ‘end to end’ audit methodology was developed, the BRachytherapy

Applicator Dosimetry (BRAD) system, to measure dose distributions around clinical

brachytherapy applicators and compare to treatment planning system calculations. MCNP5

Monte Carlo code was used to support the design of the BRAD system and validate the use

of film dosimetry. 46 radiotherapy centres in the UK were audited. Delivery of the intended

prescription dose was confirmed to be within clinically acceptable levels at all centres, mean

difference 0.6% for plastic and 3.0% for metal applicators (±3.0% k=1). The intended dose

distribution was faithfully delivered to the film-measured dose planes with a mean gamma

passing rate of 97.8% at 3% (local) 2 mm criteria. Two audits had results that required follow-

up and both were resolved. Each audit included a review of local brachytherapy physics

practice and opportunities for improvement were reported, including imaging, applicator

reconstruction, planning procedures, QC tests, and staff training.

The brachytherapy audit provided the first comprehensive validation of ‘end to

end’ clinical brachytherapy dosimetry, from applicator imaging to treatment delivery,

combined with a review of clinical physics practice. The BRAD system is retained in the

Institute of Physics and Engineering in Medicine (IPEM) phantom library.

viii

Statement of originality

This thesis and the work to which it refers are the results of my own efforts. Any ideas, data, images or text resulting from the work of others

(whether published or unpublished) are fully identified as such within the work and attributed to their originator in the text, bibliography or in

footnotes. This thesis has not been submitted in whole or in part for any other academic degree or professional qualification. I agree that the

University has the right to submit my work to the plagiarism detection service TurnitinUK for originality checks. Whether or not drafts have been

so-assessed, the University reserves the right to require an electronic version of the final document (as submitted) for assessment as above.

© Antony Lee Palmer 2015

ix

Acknowledgments

I would like to thank the following…

My PhD supervisors, Andy Nisbet and David Bradley, for their guidance, encouragement,

scientific knowledge, wealth of experience, and allowing me the freedom to pursue my

interests.

Chris Lee and Ailsa Ratcliffe for discussions of brachytherapy physics and audit phantoms.

Patty Diez for helping me organise audits at 46 radiotherapy centres.

Andre Micke for assistance with film dosimetry from the manufacturer’s perspective.

Laura Gandon, Andrea Wynn-Jones and Peter Bownes, for expertly performing several audits

for me.

Neda Shiravand for many reference film dosimetry exposures while I was auditing.

The staff at all of the audited radiotherapy centres, for their hospitality and interest in my

work.

Edwin Aird, Margaret Bidmead, Peter Bownes, and Gerry Lowe, for informal discussions of

brachytherapy and audit.

Simon Doran for assistance with radiochromic plastic dosimetry.

My two MSc students at the University of Surrey, Poppy Di Pietro and Sheaka Alobaidli, for

their work on radiochromic plastic dosimetry and optical fibre dosimetry, respectively.

Sarah Muscat and John Kearton for covering my ‘day job’ while I was out auditing, and

everyone at the Portsmouth Medical Physics Department for support.

Funding from: NHS Health Education Wessex for PhD fees and backfill; from IPEM, UK, for

costs of audit phantom construction and travel expenses to conduct the audits; from Ashland

ISP Inc., USA, for supply of radiochromic film for research and audits; from Eckert & Ziegler

Bebig GmbH, Germany, Ashland ISP Inc., USA, and University of Surrey, for sponsorship of

conference attendances; and from Portsmouth Hospitals NHS Trust for various expenses.

I would particularly like to thank my wife, Sam, for her encouragement and love, my

wonderful daughter, Erin, for welcome distractions from the thesis, my mother, Fay, father,

Terry, and brother, Andy, for their love and support. Finally, my late grandad, Fred, for views

through his old telescope, cherished earliest memories that sparked my wonder of science.

x

xi

List of Figures

Figure

Page

Title (abbreviated)

2.1 13 HDR brachytherapy physics processes with dosimetric tools that have been applied to measure dose, accuracy and uncertainty, as reported in recent literature.

3.1 27 Transit dose calculation points, D10 and D20, at 10 mm and 20 mm respectively, perpendicular distance from the centre of the intended dwell position.

3.2 46 Position and speed of source during transit from first to second dwell positions, from a series of three dwells at 10.0, 15.0 and 20.0 mm, with new control software.

3.3 46 Position and speed of source during transit from the EZ BEBIG Multisource® on approach to the first dwell position at 10.0 mm, with (a) the new control software and (b) the old control software.

3.4 49 Autoradiographs of actual source dwell positions compared to planned positions (vertical lines) as a function of curvature of the transfer tube.

3.5 51 3D projection illustration of the treatment applicator guide tubes (red), the clinical target volume (orange), the 100% isodose surface (purple) the bladder (green), rectum (yellow) and sigmoid (blue).

3.6 52 The effect on a typical cervix plan isodose distribution of simulated error dwell position shifts in (a) sagittal and (b) coronal projections.

3.7 53 Effect on DVH curves for a typical cervix patient plan for HR-CTV and OARs of simulated dwell position errors of 1, 2, 5 and 10 mm proximal shift.

3.8 53 Effect of systematic proximal dwell position shifts, 0.2 to 6.0 mm, on clinical DVH treatment plan quality parameters.

4.1 64 Schematic diagram of Solid Water test object, shown with (a) lower slab containing one catheter and upper slab with optical fibre cavities, and (b) lower slab with three catheters and upper slab holding radiochromic film.

4.2 64 Measurement of radial dose from a single HDR source, within a plastic catheter, using EBT3 film, secured in a Perspex frame within a full scatter water tank

4.3 65 Presage sample machined with three cavities each containing an HDR brachytherapy catheter for a typical cervix treatment dose distribution irradiation.

4.4 66 TL yield from Ge-doped optical fibres as a function of radial distance from the centre of a Co-60 HDR source, compared to Monte Carlo data.

4.5 67 Dose as function of radial distance from the centre of a Co-60 HDR source measured with EBT3 Gafchromic film, with comparison to Monte Carlo source data.

4.6 68 Presage signal as a function of radial distance from the centre of a Co-60 HDR source, with comparison to Monte Carlo source data.

4.7 69 TL yield from Ge-doped optical fibres as a function of distance across an HDR dose distribution, compared to treatment planning system (TPS) calculation.

4.8 70 Dose measured with EBT3 Gafchromic film as a function of distance across an HDR dose distribution, compared to treatment planning system (TPS) calculation.

4.9 71 Dose measured with Presage as a function of distance across and HDR dose distribution, compared to treatment planning system (TPS) calculation.

xii

Figure

Page

Title (abbreviated)

5.1 82 Arrangement of calibration and test films on flatbed scanner for simultaneous scanning.

5.2 85 Process of delamination and restacking to produce a double-thickness active layer.

5.3 85 Photograph showing natural curvature of a 10 x 10 cm Gafchromic EBT3 film placed on a flatbed scanner glass plate.

5.4 89 Eight Gafchromic EBT3 film calibration strips, irradiated to doses in the range 0 to 16 Gy and scanned 48 hour post exposure.

5.5 89 Gafchromic EBT3 calibration curves, in red, green and blue colour channels (16 bit), for two film batches, scanned 48 hour post exposure, with rational (linear) dose fit equations.

5.6 90 Isodose overlay of two scans, taken 10 minutes apart (thick and thin lines), of the same EBT3 film piece that had previously been exposed by a typical brachytherapy dose distribution, with and without a non-reflective matt at scanning.

5.7 92 EBT3 film net optical density (as a ratio to a simultaneously scanned reference density sample) as a function of time post-irradiation, up to three months (2277 hrs), over a dose range 0 to 14 Gy.

5.8 93 Change in film dose as a function of lateral distance on scanner, using single-channel (red lines) and triple-channel (green lines) film dosimetry, over range 1 to 14 Gy.

5.9 94 Comparison of film-measured (thick lines) and treatment planning system-calculated (thin lines) dose distributions from a typical brachytherapy cervix applicator, with (a) single (red)-channel dosimetry and (b) triple-channel dosimetry.

5.10 95 Effect of film surface perturbation (grease and scratches) on calculated film dose using single-channel and triple-channel dosimetry.

5.11 97 EBT3 film measured and treatment planning system calculated isodose distributions from a typical cervix brachytherapy treatment delivery, with (a) film naturally curved on scanner glass plate and (b) with the film flat on the scanner plate, compressed under glass.

5.12 98 Normalised film dose profiles using single (red)-channel dosimetry across a single EBT3 film irradiated by a 6 MV linac beam, with the film in various flat or curved positions.

5.13 98 Normalised film dose profiles using triple-channel dosimetry across a single EBT3 film irradiated by a 6 MV linac beam, with the film in various flat or curved positions.

5.14 99 Film dose profiles using red and green single-channel dosimetry across an EBT3 film exposed to three dose level regions, with the film flat in contact with the scanner glass plate and with the film raised uniformly by 5 mm.

5.15 101 Normalised dose as a function of radial distance from a Nucletron mHDR-v2 Ir-192 brachytherapy source; EBT3 measured compared to HEBD consensus data.

5.16 101 EBT3-film measured dose with radial distance from an Ir-192 source, at three different dose levels (irradiation times) compared to treatment planning system (TPS) calculated dose.

5.17 102 Normalised dose as a function of radial distance from an Eckert & Ziegler Bebig Co0-A86 Co-60 HDR brachytherapy source; EBT3 measured compared to HEBD consensus data.

5.18 102 EBT3-film measured dose with radial distance from a Co-60 source, at three different dose levels (irradiation times) compared to treatment planning system (TPS) calculated dose.

xiii

Figure

Page

Title (abbreviated)

6.1 115 Photographs of BRAD phantom with (a) Nucletron plastic ring and IU applicator, and (b) with Eckert & Ziegler BEBIG plastic split-ring and IU applicator.

6.2 115 Wireframe CAD design drawing of BRAD national brachytherapy audit test object.

6.3 116 3D rendered CAD design drawings of BRAD phantom.

6.4 118 Process flow diagram of brachytherapy ‘end to end’ dosimetric audit using the BRAD phantom.

6.5 122 BRAD isodose comparison of normal delivery and simulated errors: TPS-planned and film-measured for (a) normal delivery, (b) all dwells shifted 5 mm proximal, and (c) 8 Gy delivered instead of planned 7 Gy.

6.6 126 Dose gradients around cervix treatment applicator for typical clinical treatment plan.

7.1 132 MCNP5 simplified geometry source models, for (1) Nucletron mHDR-v2 Ir-192 source and (2) Eckert & Ziegler Bebig Multisource Co0.A86 Co-60 source.

7.2 132 MCNP5 validation geometry; several cylindrical tally cells concentric with the centre of the HDR source (Ir-192 or Co-60) in full scatter water.

7.3 133 MCNP5 geometry to investigate the effect of phantom size on measured dose from a point brachytherapy source, with various spherical shell tally cell radii and water sphere radii.

7.4 134 MCNP5 geometry used to investigate the dose perturbation caused by the presence of Gafchromic EBT3 film compared to a uniform water medium.

7.5 136 MCNP5 geometry used to investigate the dose attenuation caused by metallic brachytherapy treatment applicators compared to a homogeneous water phantom.

7.6 138 (a) MCNP5 calculation of dose rate with distance from an Ir-192 source compared to consensus data from the High Energy Brachytherapy Dosimetry Working Group, as a function of number of particles and tally type. (b) Percentage difference for tally F6:p. (c) Percentage difference for tally *F8:p,e.

7.7 139 (a) MCNP5 calculation of dose rate with distance from an Co-60 source compared to consensus data from the High Energy Brachytherapy Dosimetry Working Group, as a function of number of particles and tally type. (b) Percentage difference for tally F6:p. (c) Percentage difference for tally *F8:p,e.

7.8 140 MCNP5 calculated percentage difference in dose at 20 to 75 mm radial distance from (a) Ir-192 and (b) Co-60 point sources, for spherical water phantoms of radii 100 to 400 mm compared to unbounded water phantom (600 mm radius); Compared to data from Perez-Calatayud et al. (2004) for Ir-192.

7.9 141 Results of MCNP5 calculations of the percentage reduction in dose deposition due to metallic applicator attenuation, for an intrauterine applicator of 60 mm active source length (AL), at levels of mid-AL, 10 mm from mid-AL, and 30 mm beyond AL, for radial distances of 20, 40 and 60 mm.

8.1 146 Map of 46 radiotherapy centres that participated in the BRAD national audit of clinical brachytherapy dosimetry in the United Kingdom during 2013 and 2014.

8.2 148 Percentage dose difference, film-measured compared to TPS-planned, at Point A left and Point A right, for cervix treatment applicators, at 46 audited centres.

xiv

Figure

Page

Title (abbreviated)

8.3 148 Percentage dose difference, film-measured compared to TPS-planned, mean Point A for cervix treatment applicator, by applicator type, at 46 audited centres.

8.4 150 Distance to agreement, film-measured compared to TPS-planned, at Point A left and Point A right, for cervix treatment applicators, at 46 audited centres.

8.5 150 Distance to agreement, film-measured compared to TPS-planned, mean Point A for cervix treatment applicator, by applicator type, at 46 audited centres.

8.6 152 BRAD dose distribution comparison between TPS-calculated and film-measured, isodose lines in the range 0.7 Gy to 16 Gy.

8.7 153 Mean gamma passing rate, at three different criteria, comparing film-measured to TPS-planned 2D dose distributions in four planes around cervix treatment applicators, at 46 audited centres.

8.8 157 BRAD audit isodose comparison between TPS-calculated and film-measured for (a) audit with good agreement, and (b) audit with discrepancy at the distal end of the IU applicator.

8.9 158 BRAD audit with incorrect dose normalisation. Section of plan parameters hard copy from TPS, showing (a) prescription dose of 7 Gy and (b) calculated dose to Point A average 10.2 Gy.

E.1 211 IPEM webpage advertising BRAD phantom in Virtual Phantom Library.

xv

List of Tables

Table

Page

Title (abbreviated)

2.1 14 Dosimeter selection considerations for brachytherapy dose distribution measurement in the context of an external audit

3.1 30 Methods employed, and their popularity, for source strength measurement of HDR & PDR sources at centres in UK.

3.2 30 Methods employed, and their popularity, to independently verify treatment planning system (TPS) calculations at centres in UK.

3.3 31 Primary sources of guidance for establishing QC schedules, and their popularity at centres in UK.

3.4 34 HDR & PDR QC survey: Response to questionnaire on test popularity, measurement frequency and tolerance values.

3.5 40 UK QC survey: Additional tests identified by responding centres as being included in local QC procedures but not included in original survey questionnaire.

3.6 41 QC tests included in all UK centres’ schedules, with comparison to published recommendations.

3.7 42 Comparison of UK practice and published recommendations for two QC tests included in only half of UK centres’ schedules.

3.8 43 Suggested QC tests not included in any UK centres’ schedules (at the time of the UK survey).

3.9 44 Required areas of investigation for HDR brachytherapy system commissioning, or for consideration after re-commissioning following hardware or software update.

3.10 47 Comparison of actual calculated transit dose and Eckert & Ziegler BEBIG Multisource® compensated transit dose, for each dwell position in a series of three dwells, at 10.0, 15.0 and 20.0 mm, evaluated at D10 and D20.

3.11 51 Effect of dwell position shifts, 0.2 to 2.0 mm, on DVH treatment plan metrics, modes (a) and (b) see text.

4.1 72 Uncertainty budget for optical fibre, Gafchromic film, and Presage, for the experimental methodology used in this work, for isolated source radial dose rate measurement.

4.2 75 Dosimetric and practical considerations of the dosimetry system for brachytherapy audit, comparing optical fibre, Gafchromic film, and Presage.

5.1 88 Rational (linear) calibration equation parameters, see text, for two Gafchromic EBT3 film batches (lot a=#01171401, lot b=#12171303), scanned 48 hour post exposure.

5.2 91 Calculated mean film dose from ten test films, at four irradiation dose levels, using different linear-scaling calibration dose references.

5.3 96 Calculated film dose for original and ‘double active layer’ film using single- and triple-channel analysis with Gafchromic EBT3 film in the range 1 to 14 Gy.

xvi

Table

Page

Title (abbreviated)

6.1 110 Summary of published HDR/PDR brachytherapy dosimetric audits.

6.2 124 Uncertainty budget for the experimental determination of the dose difference at Point A between treatment planning system calculated and film-measured dose using the BRAD ‘end to end’ audit system.

7.1 135 Specifications of metallic intrauterine treatment applicators in use during the UK brachytherapy audit, and availability of attenuation correction in manufacturers’ treatment planning systems.

7.2 137 Results of MCNP5 calculation comparing dose deposition to a water cell of dimensions equivalent to EBT3 film active layer, with and without a 15 mm width film between the source and tally cell, for Ir-192 and Co-60.

8.1 149 p-values and statistical significance of unpaired two-tailed t-test comparing dose difference at Point A (film-measured to TPS-planned) for various grouped data from the 46 audits.

8.2 153 Mean and standard deviation (sd) gamma passing rate, at three different criteria, for 46 audited centres.

C.1 203 Risk assessment for BRAD audit.

xvii

List of Abbreviations

Abbreviation

Meaning

2D Two dimensional

3D Three dimensional

AAPM American Association of Physicists in Medicine

ABS American Brachytherapy Society

ASTRO American Society for Radiation Oncology

BRAD Brachytherapy Applicator Dosimetry

BRAPHYQS Brachytherapy Physics Quality Assurance System Working Group of ESTRO

CAD Computer aided design

CT Computed tomography (imaging)

D90 Minimum dose delivered to 90% of the volume

D2cc Highest dose to 2 cm3 of the volume

DH Department of Health, UK

DICOM Digital imaging and communications in medicine (standard)

DTA Distance to agreement

DVH Dose volume histogram

ESTRO European Society for Radiotherapy and Oncology

EQUAL ESTRO quality assurance network

GEC-ESTRO Gynaecological ESTRO Working Group

HDR High dose rate

HEBD High Energy Brachytherapy Source Dosimetry Working Group of AAPM and ESTRO

HR-CTV High risk clinical target volume

HSE Health and Safety Executive, UK

IAEA International Atomic Energy Agency

ICRU International Commission on Radiation Units

IMRT Intensity modulated radiotherapy

IPEM Institute of Physics and Engineering in Medicine

IU Intrauterine

kV Kilovoltage

LDR Low dose rate

MCNP Monte Carlo N-Particle

MCNP5 Monte Carlo N-Particle code version 5

xviii

Abbreviation

Meaning

MOSFET Metal oxide semiconductor field effect transistor

MRI Magnetic resonance imaging

MV Megavoltage

NHS National Health Service, UK

OAR Organ at risk

PDR Pulsed dose rate

QA Quality assurance

QC Quality control

rgb Red green blue

RTDose Radiotherapy dose expansion of DICOM standard (3D dose grid)

sd Standard deviation

sem Standard error of the mean

TL Thermoluminescent

TLD Thermoluminescent dosimetry

TPS Treatment planning system

TRAK Total reference air kerma

UK United Kingdom

VMAT Volumetric modulated radiotherapy

1

CHAPTER 1

Introduction and Overview

2

1.1. Introduction

1.1.1. Brachytherapy

Radiotherapy is the medical use of ionising radiation primarily in cancer

treatment, with curative or palliative intent, and may be combined with other treatment

modalities, depending on a number of factors including the tumour site and extent.

Brachytherapy is a form of radiotherapy concerning the treatment of cancer using small

radioactive sources inserted within or very close to the tumour. There is a rapid fall-off of

dose around the radioactive sources enabling a high dose to be delivered to the tumour with

sparing of surrounding normal structures. However, this requires high accuracy and precision

of the source delivery mechanisms. The technique is sub-divided into low dose rate

treatments, usually permanent implantation of radioactive material with dose rates in the

range 0.4 to 2 Gy per hour, high dose rate (HDR), which utilise the temporary introduction of

a radioactive source within the body with dose rates in excess of 12 Gy per hour, and pulsed

dose rate (PDR), which uses similar sources and equipment to HDR but with repeating low

exposures pulses to simulate a low dose rate approach (Gerbaulet et al. 2002, Venselaar et

al. 2013). This work considers HDR brachytherapy treatment, for gynaecological applications.

The most common radionuclide used for HDR brachytherapy is Ir-192. Recently, HDR

treatments using Co-60 sources have also been introduced in a minority of centres, including

one centre in the United Kingdom (UK) (Andrassy et al. 2012, Palmer et al. 2012d). Co-60 has

a half-life of 5.26 years compared to 73.8 days for Ir-192, and has emissions at 1.17 and 1.33

MeV, compared to a complex emission spectrum for Ir-192 with mean energy 0.38 MeV. The

HDR brachytherapy treatment unit consists of a HDR source, of a few mm length, attached

to a wire. The source is stepped through treatment applicators and catheters using computer

software control such that a planned dose distribution is delivered.

1.1.2. Requirements for Accuracy of Dose Delivery

The need for dosimetric accuracy in external beam radiotherapy is well

established (ICRU 1976, Brahme 1984). A standard requirement of the combined

uncertainties in absorbed dose delivery is often stated as 3.5% at one standard deviation,

first quoted by Mijnheer et al. (1987), or more recently expressed as ±3% (1 sd) at the

prescription point, ±3-5% (1 sd) across the entire target volume, and geometric accuracy 2-4

mm (1 sd), for the majority of external beam treatments (Thwaites 2013). A standard

accuracy requirement for advanced external beam radiotherapy is often quoted in terms of

a 3%, 3 mm criteria. Dosimetric accuracy of HDR brachytherapy may also be expected to

contribute to the achievement of clinical treatment aims, tumour control and minimised

normal tissue toxicity, due to the steepness of the clinical dose effect curves. Van Dyk et al.

(1993) defined a requirement for brachytherapy treatment delivery of 3% accuracy in dose

at distances of 0.5 cm or more at any point for any radiation source. IAEA (2004) have

recommended that 5% dose agreement and 2 mm spatial agreement is required for

brachytherapy. The dosimetric challenge is particularly difficult in brachytherapy due to small

3

treatment distances, very high dose gradients and orders of magnitude variation in dose

deposition across volumes of interest.

Brachytherapy is currently undergoing a period of significant innovation and

rapid modernisation (Potter and Kirisits 2012), including a shift from 2D to 3D basis (Hepel

and Wazer 2012), the enhanced use of imaging (Haie-Meder et al. 2011), patient-specific

treatment plan optimisation, fully volume-based prescribing (Wanderas et al. 2012), inverse-

planning (Siauw et al. 2011), advanced planning algorithms (Rivard et al. 2009), use of

advanced treatment applicators (Bernstein et al. 2012, Song et al. 2012), and in-vivo

dosimetry verification systems (Suchowerska et al. 2011). It is essential that fundamental

assurances of quality and dose delivery accuracy are not lost amongst these developments.

1.1.3. Quality Control, Dosimetry Audits and Treatment Errors

Quality control (QC) of HDR brachytherapy techniques has received only

moderate research attention in recent years compared with external beam radiotherapy,

however with current and imminent changes to the complexity and patient-specific

optimisation opportunities in HDR brachytherapy (Thomadsen et al. 2008, Ibbott et al. 2008),

as described above, the appropriateness of current QC techniques requires review. It is

essential that physics-aspects of quality verification keep pace with the changing technology

and clinical practice. This includes re-assessing the use of historic QC tests that are no longer

fit for purpose and replacing with more relevant QC, or where system performance is verified

by other means, avoiding unnecessary redundancy. Guidance on HDR brachytherapy QC

from the professional body of Medical Physicists in the UK, IPEM, does exist but may be ‘out

of date’ having been published in 1999 (Mayles et al. 1999) and is currently being revised.

Similar guidance from outside the UK also requires review and modernisation, dating from

1997 in USA (Nath et al. 1997) and 2004 in Europe (Vanselaar and Perez-Calatayud 2004).

Control of dose delivery is particularly difficult to achieve in brachytherapy due

to small treatment distances, very high dose gradients and a multitude of aspects that affect

accuracy (Palmer et al. 2012b). A quality assurance system in radiotherapy is essential to

ensure treatment delivery is consistent and as intended. This will include a multitude of

quality control (QC) tests designed to evaluate actual operating performance in comparison

to goal values, and to enable rectification/reconciliation of any differences. A comprehensive,

robust QC system should include complementary, independent external audit. With the

exception of a basic ESTRO mailed dosimetry service, there has been little brachytherapy

dosimetric audit activity reported in the literature, when compared with external beam

radiotherapy verification which has been active in the UK for over 20 years (Thwaites et al.

1991, Nisbet and Thwaites 1997). External quality audits in brachytherapy are therefore not

common practice but it is clear that future advancements in brachytherapy should be

underpinned by the reassurance of comprehensive dosimetric audit. Assuring confidence in

the clinical utility of brachytherapy requires many aspects of clinical audit, of which

dosimetric audit is an essential component.

4

The need for dosimetric audit in brachytherapy exactly mirrors the need in

external beam radiotherapy: to detect any errors, to provide reassurance, to enable

improvements, and to demonstrate compliance. There are of course numerous sources of

uncertainty in brachytherapy (Kirisits et al. 2014), and there have been previous errors in

brachytherapy delivery (ICRP 2000, ICRP 2005a, ICRP 2005b, Daigh 2010) including: well

chamber calibration error (Dempsey 2011); confusion over units (USNRC 2009); media

reports of at least two incidents involving incorrect dwell positions in treatment applicators

(ASAHI SHIMBUN 2013, Sydney Morning Herald 2003a and 2003b); and manufacturer safety

notifications, including potentially incorrect HDR source dwell times under special

circumstances (Varian 2010), potential incorrect dose/number of treatment fractions for

standard plans (Nucletron 2014), potential source positioning errors in applicators (Eckert &

Ziegler BEBIG 2014). It is not sufficient to just investigate the cause of problems that have

occurred in the past, but one must investigate systems for problems that may occur in the

future. The great benefit of ‘on-site’ external audit is that it also allows discussion of

processes with the staff that undertake them, and hence it is one of only a few methods to

provide quality assurance of the ‘qualitative’ issues in safety which are normally ‘resistant’

to discovery by conventional, technical device-centric QC.

1.2. Research Questions and Objectives

To ensure safety and optimisation in medical radiation exposures, there is a

fundamental requirement for “adequate tests of equipment performance at appropriate

levels” (HSE 2006). For radiotherapy, the tests, methods, frequencies and tolerances are the

responsibility of Medical Physicists and their judgment. The purpose of this research work is

to contribute to this challenging endeavour, specifically to contribute to the development of

physics aspects that improve patient safety and best practice in HDR brachytherapy. The

specific research questions are summarised below:

What is the consistency of practice for QC of HDR treatment equipment in the UK?

Guidance from IPEM is over 20 years old and there have been significant technological

and clinical practice changes in this time. A new consensus benchmark is required.

Are there any specific QC tests that should be championed or techniques developed? Is

there an efficient method for dwell position, source movement and transit dose

assessment?

What is the required accuracy for HDR treatment equipment performance, particularly

dwell position tolerance? How does this impact treatment plan quality parameters?

What is the status of interdepartmental/external audit in the UK and does this need to

be developed for brachytherapy? What is the value of external audit alongside routine

QC, and is this worthwhile in brachytherapy?

5

Can a phantom and dosimetry system be developed to provide a convenient external

audit measurement for brachytherapy? Can this be implemented as an ‘end to end’

system check for brachytherapy physics dosimetry?

What form of dosimetric measurement is necessary for ‘end to end’ audit in

brachytherapy? Which detectors could be used and how would results be analysed and

interpreted?

How might radiochromic film dosimetry be applied for an ‘end to end’ brachytherapy

dosimetry audit? Can film dosimetry be sufficiently well characterised, accuracy

improved and uncertainties minimised for a brachytherapy audit?

Is it feasible to undertake an ‘end to end’ dosimetry audit at every brachytherapy centre

in the UK and how might this be organised?

Are there any variations in physics practice for HDR gynaecology brachytherapy and any

opportunities for improvement?

The initial objective of the research was to review and improve QC practices for

HDR brachytherapy. It became clear there was a particular need to improve brachytherapy

audit, and a subsequent focus of the research was then to design and implement an ‘end to

end’ dosimetry audit for brachytherapy in the UK. Dosimetric audit is recommended in many

authoritative documents, is a mandatory requirement in many countries, and advocated by

the majority of physicists working in radiotherapy. Whether for improvements in patient

care, reassurance of accuracy, fulfilling a legal requirement, credentialing for clinical trials,

simple best-practice approach, minimising the risk of error, avoiding litigation, or adding

security in a high pressure environment, audit is a valuable tool and its application in

brachytherapy is worthwhile. ‘End to end’ testing includes the entire preparation and

delivery process, with a quantitative measurement that can be evaluated at the end of the

test to confirm accurate delivery of the planned treatment (ASTRO 2012). Brachytherapy is

undergoing significant clinical and technical developments, the quality of which should be

underpinned by the reassurance of comprehensive dosimetric audit.

1.3. Summary of Thesis

This thesis documents research conducted into physics-aspects of safety

assurance in HDR brachytherapy. There are four linked themes to this work; brachytherapy

QC; dosimeters for brachytherapy audit; development and implementation of a

brachytherapy audit in the UK; and supporting Monte Carlo simulations. The thesis is

necessarily split into self-contained chapters, with joint sections for introduction, theory,

summary conclusions, and future work.

6

Chapter 2 provides a succinct review of relevant theory, including brachytherapy

physics, dosimetric accuracy, definition and need for brachytherapy audit, brachytherapy

dosimetry measurement, optical fibre dosimeters, radiochromic film dosimeters,

radiochromic plastic dosimeters, and Monte Carlo methods.

Chapter 3 opens with a review of QC testing and commissioning of

brachytherapy equipment and highlights a number of brachytherapy treatment errors that

have occurred in recent years. There are three elements to the research presented in this

chapter. First, a UK survey of brachytherapy QC was undertaken to establish benchmark

consensus and inform on areas requiring development. This led to the second section which

provides a methodology for the measurement of dwell position accuracy and transit dose,

and led to consultation with a manufacturer to improve their HDR equipment performance.

The final section is a treatment planning study to establish requirements for source dwell

positioning accuracy, the impact of systematic errors on treatment planning quality

parameters, and suitability of dose measuring points to detect errors.

Chapter 4 is concerned with the evaluation and selection of a dosimetry system

for brachytherapy applicator dose measurement, appropriate for a national audit. Three

systems are evaluated: point-dose measurement with doped silica glass optical fibres; 2D

measurement with Gafchromic EBT3 radiochromic film; and 3D measurement with Presage

solid radiochromic polymer. A Solid Water (RMI457) test object was designed for the fibre

and film measurements, while cavities were machined directly into the Presage sample for

irradiation. Each dosimeter is evaluated as follows: initial processing and calibration; radial

dose measurements from an HDR source; measurement of typical brachytherapy dose

distribution; practical considerations; and an uncertainty budget estimate.

Chapter 5 presents research and development into the characterisation and

optimal use of radiochromic film for brachytherapy audit applications, building on the results

presented in Chapter 4. The objectives and need for film dosimetry development are first

outlined. Film-related and scanner-related performance characteristics are then evaluated,

as well as the use of triple-channel compared to single-channel dosimetry, and

recommendations made for the reduction of uncertainties.

Chapter 6 presents research on the development of a phantom and

methodology for brachytherapy audit using radiochromic film, enabling an ‘end to end’

system test approach, as advocated in the joint American societies document ‘Safety is No

Accident’ (ASTRO 2012). A review and evaluation of previous brachytherapy audits is

provided to give context and inform development objectives for the UK brachytherapy audit.

The phantom design, audit methodology, and its sensitivity to simulated errors are discussed.

A detailed protocol for the optimal use of film dosimetry for brachytherapy audit is provided,

building on research work presented in Chapter 5. An uncertainty budget for the audit

approach is evaluated.

Chapter 7 describes the Monte Carlo work undertaken to support the

brachytherapy dosimetry audit development. MCNP5 is used to determine the required audit

phantom size, to confirm water equivalence of EBT3 film in the context of the brachytherapy

7

audit measurement, and to estimate the dose attenuation due to metallic treatment

applicators to interpret results of the national brachytherapy audit.

Chapter 8 discusses the scope and results of a national brachytherapy audit

undertaken as a key element of the thesis work. The audit phantom and methodology

described in Chapter 6 is utilised. Film is used to assess the agreement between delivered

dose and planned dose as an ‘end to end’ audit of the clinical brachytherapy treatment

process, in terms of dose to the prescription point and dose distribution in clinically relevant

regions. Differences in physics brachytherapy practice between centres are discussed,

highlighting potential opportunities for improvement, and audit results requiring further

investigation are described.

Chapter 9 brings the body of work together with a summary review of the key

results from the thesis, as well as highlighting potential areas for future work.

1.4. List of Publications and Presentations Arising From this Work

1.4.1. Quality Control and Accuracy of Brachytherapy

Journal Publications:

J 1. Palmer A. Impact of software changes: Transit dose and source position

accuracy of the Eckert & Ziegler BEBIG GmbH Multisource high dose rate (HDR)

brachytherapy treatment unit. J. Radiother. Pract. 2013:12;80-87.

J 2. Palmer A.L., Bidmead M., Nisbet A. A survey of quality control practices for high

dose rate (HDR) and pulsed dose rate (PDR) brachytherapy in the United

Kingdom. J. Contemp. Brachyther. 2012;4:232-240.

J 3. Palmer A., Bradley D., Nisbet A. Physics-aspects of dose accuracy in high dose

rate (HDR) brachytherapy: source dosimetry, treatment planning, equipment

performance and in-vivo verification techniques. J. Contemp. Brachyther.

2012;4(2):81-91.

Conference Publications, Posters and Presentation:

C 1. Palmer A.L., Bradley D.A., Nisbet A. Improving quality assurance of HDR

brachytherapy: Verifying agreement between planned and delivered dose

distributions using DICOM RTDose and advanced film dosimetry. Poster at

AAPM 2014 annual meeting, Austin, Texas, July 2014.

C 2. Patel I., Palmer A. Revision of IPEM guidance on quality control of radiotherapy

equipment. Med. Phys. Int. 2013;1(2):179. Invited oral presentation at

ICMP2013, Brighton, UK.

8

C 3. Palmer A.L., Nisbet A., Bradley D.A. A new standard for HDR brachytherapy

quality control: practical and advanced film dosimetry for treatment applicators

and sources. Proceedings of the 2nd Estro Forum, Geneva, Switzerland.

Radiother. Oncol. 2013;106:S368-9.

C 4. Nisbet A., Palmer A.L.*, Bradley D.A. Available guidance, current UK practice,

and future directions for HDR brachytherapy quality control. Proceedings of the

2nd Estro Forum, Geneva, Switzerland. Radiother. Oncol. 2013;106:S370. (* main

and presenting author)

C 5. Palmer A., Ioannou L., Hayman O., Nagar Y.S. Is HDR equipment performance

suitable for modern brachytherapy? Positional errors, dosimetric impact & case

study. Proceedings of the World Congress of Brachytherapy, Barcelona, Spain.

Radiother. Oncol. 2012;103:S132-3.

1.4.2. Dosimeters for Brachytherapy Audit

Journal Publication:

J 4. Palmer A.L., Di Pietro P., Alobaidli S., Issa F., Doran S., Bradley D., Nisbet A.

Comparison of methods for the measurement of radiation dose distributions in

high dose rate (HDR) brachytherapy: Ge-doped optical fibre, EBT3 Gafchromic

film, and PRESAGE® radiochromic plastic. Med. Phys. 2013;40(6):061707-1-11

1.4.3. Radiochromic Film Methodology


J 5. Palmer A.L., Bradley D., Nisbet A. Evaluation and mitigation of potential errors

in radiochromic film dosimetry due to film curvature at scanning. J. Appl. Clin.

Med. Phys. 2015 (in press)

J 6. Palmer A.L., Bradley D., Nisbet A. Evaluation and implementation of triple-

channel radiochromic film dosimetry in brachytherapy. J. Appl. Clin. Med. Phys.

2014;15(4):280-296.

J 7. Palmer A.L., Nisbet A., Bradley D. Verification of high dose rate brachytherapy

dose distributions with EBT3 Gafchromic film quality control techniques. Phys.

Med. Biol. 2013;58:497-511.


C 6. Palmer A.L. Advanced Film Dosimetry for Brachytherapy. Invited oral

presentation as Guest Speaker at Gafchromc Symposium, Ashland Inc, held

during AAPM 2014 annual conference, Austin, Texas, July 2014. Available at:

http://www.filmqapro.com/documents/Palmer_Brachy%20Audit_Gafchromic_

AAPM_2014.pdf

9

C 7. Palmer A.L., Nisbet A., Bradley D.A. Semi-3D dosimetry of high dose rate

brachytherapy using a novel Gafchromic EBT3 film-array water phantom. J.

Phys.: Conf. Ser. 2013;444:012101. Published abstract, poster and oral

presentation at the conference: Proceedings of the International Conference on

3D Dosimetry (IC3DDose), Sydney, Australia, November 2012.

1.4.4. Brachytherapy Audit


J 8. Palmer A.L., Diez P., Gandon L., Wynn-Jones A., Bownes P., Lee C., Aird E.,

Bidmead M., Lowe G., Bradley D., Nisbet A. A multicentre ‘end to end’ dosimetry

audit for cervix HDR brachytherapy treatment. Radiother. Oncol. 2015 (in press,

http://dx.doi.org/10.1016/j.radonc.2014.12.006).

J 9. Palmer A.L., Bradley D., Nisbet A. Dosimetric audit in brachytherapy. Br. J.

Radiol. 2014;87:20140105. (http://dx.doi.org/10.1259/bjr.20140105)

J 10. Palmer A.L., Lee C., Ratcliffe A.J., Bradley D., Nisbet A. Design and

implementation of a film dosimetry audit tool for comparison of planned and

delivered dose distributions in high dose rate (HDR) brachytherapy. Phys. Med.

Biol. 2013;58:6623-6640. This article was listed as a “Featured Article” (“recent

articles of high-interest across IOP content”) on the PMB website.


C 8. Palmer A.L. Auditing HDR/PDR Brachytherapy Physics in UK. Invited oral

presentation at the IPEM Biennial Radiotherapy Meeting, Sept 2014, Glasgow.

C 9. Palmer A.L., Bradley D.A., Nisbet A. Comprehensive audit of brachytherapy dose

distributions: A methodology and UK audit results. Poster at ESTRO33

conference, Vienna, Austria, April 2014. (Radiother. Oncol. supplement in press)

C 10. Palmer A.L. Group hugs in brachytherapy physics. Invited oral presentation at

the Medical Physics Research Away Day, Portsmouth Hospitals NHS Trust,

February 2014.

C 11. Palmer A.L. UK Brachytherapy Audit: origin, funding and collaboration. Invited

oral presentation at the IPEM, NPL, CTRad meeting on “Reaching a Consensus

on Verification of Radiotherapy Deliver”, held at NPL, December 2013.

C 12. Palmer A.L., Diez P., Aird E., Lee C., Radcliffe A., Gouldstone C., Sander T., Bradley

D., Nisbet A. Development of a UK dosimetry audit for HDR/PDR brachytherapy.

Med. Phys. Int. 2013;1(2):233. Oral presentation at ICMP2013, Brighton, UK.

10

1.4.5. Monte Carlo Calculations


C13. Palmer A.L., Bradley D., Nisbet A. Monte Carlo derived correction factors for

brachytherapy film dosimetry for audit and QC. Accepted for poster

presentation at the ESTRO 3rd Forum conference, Barcelona, April 2015.

1.4.6. Miscellaneous Publications Relating to PhD Work


J 11. Palmer A.L. BJR brachytherapy dosimetry special feature. Br. J. Radiol.

2014;87:20140506.

11

CHAPTER 2

Theory

12

2.1. High Dose Rate (HDR) Brachytherapy Equipment

High dose rate (HDR) brachytherapy is delivered by a cylindrical Ir-192 or Co-60

radioactive source of a few mm length and 1 mm diameter, which is attached to a drive cable

of 1 to 1.5 m length, whose position is computer controlled. The cable is used to position the

source inside catheters or applicators that have previously been positioned and imaged

within a patient. The source dwells at a predefined position for a predefined time before

stepping along the catheter to the next planned dwell position. This process is repeated to

create the required dose distribution with multiple dwell positions. By varying the position

and dwell time of the radiation source, the cumulative dose distribution may be neatly

sculpted to conform to the shape of the target and avoid organs at risk. The HDR equipment

is termed an afterloader because the catheters or applicators are positioned and imaged

prior to treatment delivery. Treatment duration is typically 5 to 20 minutes, with treatment

usually fractionated over one or more weeks. It is critical for the success of treatment that

the applicators are correctly positioned, that the applicators do not move between imaging,

treatment planning and treatment, that the treatment is carefully planned and that the

source dwell positions and dwell times are accurately controlled. For cervix cancer

treatments, the dose is historically prescribed at a position termed ‘Point A’, which is defined

as being 20 mm along the intrauterine (IU) source tube, from the level of the ring or ovoid

applicator surface, and 20 mm lateral to the IU (Viswanathan and Thomadsen 2012). While

dose distributions may now be prescribed and planned using 3D DVH parameters, with

patient-specific plan optimisation, the dose to Point A is still an important parameter

recorded for clinical dose reporting consistency.

2.2. HDR Brachytherapy Dose Distribution Measurement

Due to the small source size in HDR brachytherapy, and its proximity to the

treatment volume, there is rapid fall-off of dose enabling a high dose to be delivered to the

tumour with sparing of surrounding normal structures. However, this requires high accuracy

and precision of the source delivery mechanisms and also the dosimetry systems. Accurate

confirmation of dose distributions actually delivered by treatment equipment is critical to

confirm the intended prescribed radiation treatment is achieved. However, the

measurement of radiation dose distributions around clinical brachytherapy sources, both

isolated and within treatment applicators, is a challenging endeavour. The volume of interest

contains orders of magnitude variation in dose levels, extremely steep dose gradients near

the source, around 6% per mm at the edges of the target volume in typical cervix cancer

treatments, and low dose rates further away (Viswanathan and Thomadsen 2012).

Conventional radiotherapy dosimetry for external beam treatments often involve relatively

large detectors, but when these are used close to a brachytherapy source inaccuracies are

introduced due to dose averaging over the active volume: a 0.6cc ionisation chamber will

exhibit a non-uniformity correction of around 10% at 20 mm and 30% at 10 mm from a

brachytherapy source (Tolli and Johansson 1993, Majumdar et al. 2006).

13

2.2.1. Review of Contemporary Dosimetry Systems for Brachytherapy

Figure 2.1 summarises dosimetry tools that have been used recently in HDR

brachytherapy physics, for the measurement of dose parameters or to assess accuracy and

uncertainty. At least ten dosimetry systems have been discussed in publications over the last

five years for basic source and equipment dosimetry (Palmer et al. 2012b), including

thermoluminescent dosimeters (Karaiskos et al. 1998, Kirov et al. 1995, Zhang et al. 2010),

ionisation chambers (Mishra et al. 1997), semiconductor diodes (Kirov et al. 1995), MOSFETS

(Zilio et al. 2006, Toye et al. 2007, Mason et al. 2013), radiochromic film (Sharma et al. 2004,

Sureka et al. 2007, Aldelaijan et al. 2011), and Polymer or Fricke gel dosimetry (Baras et al.

2002). There is no clear consensus on the optimum dosimetry systems to use for the various

measurement situations.

Figure 2.1. HDR brachytherapy physics processes with dosimetric tools that have been applied to

measure dose, accuracy and uncertainty, as reported in recent literature.

LiF thermoluminescent dosimetry (TLD) has historically been the method of

choice for the experimental determination of source dosimetry parameters defined by the

American Association of Physicists in Medicine (AAPM) Task Group report TG-43 (Rivard et

al. 2004) and for the measurement of brachytherapy dose. Indeed Rivard et al. (2004)

comments the validity for using other detectors has not yet been convincingly demonstrated,

and that “multiple publications of results in peer-reviewed journals by independent

14

investigators are desirable to demonstrate independence and consistency. Therefore, use of

other experimental dosimeters is an area for future research of significant scientific value”.

The selection of a dosimetry system to sample the dose distribution around

clinical brachytherapy treatment applicators in the context of an interdepartmental

dosimetry audit (Chapter 4 and Chapter 8) is a trade-off between desirable characteristics,

with no one detector performing optimally in all criteria. Table 2.1 provides a summary of

considerations for potential dosimeters for use in brachytherapy dosimetry audit.

Table 2.1 Dosimeter selection considerations for brachytherapy dose distribution measurement

in the context of an external audit

Criteria Considerations

Detector response to HDR brachytherapy sources

Accuracy, precision, dose rate response, energy response, angular response, dynamic range, detection limit

Spatial resolution Detector size, readout resolution

Cost Individual detector cost, re-usable?, other equipment

Availability Commercially available or ease of local production

Handling Time to set-up and ease of operation during audit

Geometric accuracy Accuracy and speed of positioning with respect to brachytherapy applicator

Readout method Time required for read-out process, and ability to delay read-out may be required

Robustness Appropriateness of detector for UK-wide dosimetry audit visits, susceptibility to transport damage

Maturity of technology Level of prior evidence of application in brachytherapy dose measurement or audit

Taking account of the required characteristics in Table 2.1 and the available

detectors in Figure 2.1, one dosimetry system was chosen for each of three potential

measurement modes: multiple point detectors, 2D detector, or 3D detector. Optical fibre

dosimeters are a potential very low cost, high spatial resolution multiple point detector

system. Radiochromic film is a potential very high spatial resolution, commercially available

2D system with ease of geometric positioning. Radiochromic plastic is a robust detector with

potential for good spatial resolution in 3D. Key aspects of the theoretical considerations of

operation of these chosen detectors is provided in Sections 2.2.2 to 2.2.4.

15

2.2.2. Optical Fibre Dosimetry

A comprehensive review of the thermoluminescent (TL) properties of doped

silica glass optical fibres and their application as dosimeters in radiation therapy has been

conducted by Bradley et al. (2012). Primary advantages include significantly improved spatial

resolution in one dimension (typically 9 µm active core diameter within 100 µm diameter

fibre) compared to conventional LiF TL dosimeters (several mm), hygroscopic nature

(impermeability to water) facilitating possible use as in vivo detector, a linear response of TL

yield as a function of dose, and a high dynamic range. Disadvantages include a current

requirement for manual preparation from commercial optical fibre cables, their fragility, and

lack of water-equivalent atomic number (Zeff = 11.4 for SiO2 with Ge-doping 0.15-0.19 mol%

(Bradley et al. 2012) compared to Zeff = 7.3 for water). To date, the application of doped

optical fibres in brachytherapy dosimetry has been limited to measurements at less than 50

mm from low dose rate (LDR) I-125 seeds (Bradley et al. 2012), with similar work by Issa et

al. (2012) who conducted measurements at less than 20 mm from Ba-133 and Co-60

laboratory sources.

2.2.3. Radiochromic Film Dosimetry

Radiochromic film has a long history in radiation dosimetry (Butson et al. 2003)

and provides a number of advantages over other dosimetry methods, including high spatial

resolution, low energy dependence, and near water equivalence. The film is self-developing

after exposure to ionising radiation, by a topochemical (solid-state) photopolymerisation

process with diacetylene monomers. The rod shaped (2 µm x 15 µm) polymers are highly

anisotropic, being preferentially aligned parallel to the coating direction during manufacture,

and hence act as visible light polarizers. Gafchromic EBT31 film, available from late 2011, is

composed of an active radiochromic layer 26-28 µm thickness, laminated between two 100

µm polyester layers, creating a symmetric structure, different to the asymmetric structure of

its predecessor EBT2 (Reinhardt et al. 2012). The total film thickness is 0.23 mm. The matte

polyester contains microscopic silica spheres at the surface eliminating Newton’s Rings

artefacts at scanning. A marker dye within the active layer is included for film uniformity

correction, via the blue scanner channel. The film has a near water-equivalent effective

atomic number, Zeff (EBT3) = 6.84, compared to Zeff (water) = 7.3 (Arjomandy et al. 2010). An

absorbed-dose energy dependence is only significant below 100 kV (Sutherland and Rogers

2010, Massillon et al. 2012, Bekerat 2012). Brown et al. (2012) have demonstrated weaker

energy dependence with EBT3 Gafchromic film at low keV, compared to the older Gafchromic

films (EBT, EBT2).

Film readout can be performed with specialised densitometers, however it is

now accepted practice that conventional ‘high-end’ office flatbed film or document scanners

may be used for film dosimetry (Butson et al. 2003, Devic et al. 2005, Lewis et al. 2012).

Radiochromic reactions by definition are a direct colouration of a media by the absorption of

1 Ashland ISP Inc, Wayne, New Jersey, USA

16

radiation. The attenuation of visible light traversing the exposed film is a measure of its

optical density. It is essential that no automatic software colour correction is made to the

scanned image. Red, green and blue image channels are available, and various approaches

have been used to generate the calibrated dose image, either red channel alone (Bouchard

et al. 2009), different channels for different dose ranges (Andres et al. 2010), ratios between

channels (Mayer et al. 2012), or triple-channel approaches (Devic et al. 2009, Tamponi et al.

2014, Mendez et al. 2014). It is also possible to use a correction/subtraction method with a

pre-irradiation scan of the film, for the single channel approach (Devic et al. 2005, Paelinck

et al. 2007, Menegotti et al. 2008). Van Hoof et al. (2012) found the triple-channel

radiochromic film read-out method performed at least as well as the single-channel method

with inclusion of a pre-irradiation film scan, and that it reduced film non-uniformity and

saved time with the elimination of the pre-irradiation scan. Mendez et al. (2013) studied

various aspects of radiochromic dosimetry, and stated “three-channel dosimetry was found

to be substantially superior to red-channel dosimetry”. In this thesis, triple-channel

dosimetry is the favoured method, with comparisons to single-channel dosimetry in some

experimental work. The triple-channel algorithm developed by Micke et al. (2011) is used ‘as

is’ as an established and accepted film dosimetry method (van Hoof et al. 2012, Lewis et al.

2012, Hayashi et al. 2012, Mendez et al. 2013, Mendez et al. 2014). Although its application

for brachytherapy is evaluated and characterised, the algorithm itself is not further

developed.

Until recently, the majority of film dosimetry applications have been in the

verification of relative dose distribution for external beam radiotherapy (Molineu et al. 2013).

However, recent improvements in the latest radiochromic film technology (Gafchromic EBT3)

and advances in scanning and analysis methods using triple-channel dosimetry (as above)

may enable calibrated absolute dose measurement of brachytherapy dose distributions. The

majority of prior published work using film in brachytherapy has studied single source Ir-192

dose rate distributions (Sellakumar et al. 2009, Aldelaijan et al. 2011) or low dose rate seeds

(Acar et al. 2013), usually single-channel film analysis (Bouchard et al. 2009) and

measurements with older EBT and EBT2 Gafchromic film versions (Carrasco et al. 2013).

Reinhardt et al. (2012) have demonstrated the validity of EBT3 in photon and proton beams.

DeWerd et al. (2011) states there have been conflicting results for radiochromic

film in the literature, which requires further research. Perez-Calatayud et al. (2012a) reported

that radiochromic film must be considered “under development at this time because of

numerous artefacts which require rigorous correction”. For these reasons, radiochromic film

requires careful characterisation and validation when used, particularly for novel

applications.

17

2.2.4. Radiochromic Plastic Dosimetry

The use of polymer gels and plastics for radiation dosimetry has evolved over

the last two decades, reviewed by Doran (2009) and Baldock et al. (2010), and has a long

history (McJury et al. 1999, De Deene et al. 2001). However, only recently has the technology

developed to give the potential for routine clinical dosimetry applications (Adamson et al.

2012, Gorjiara et al. 2012). Probably the most promising material for brachytherapy

applications is Presage2 which has a number of advantages over gel radiochromic dosimetry

materials including; an insensitivity to oxygen, which frustrated early users of gel dosimeters

(De Deene et al. 2002) and is key for the insertion of catheters into the dosimeter; light

absorption rather than scattering as a contrast mechanism, which is favourable for optical CT

readout using a pixelated detector rather than a scanning laser beam; a machineable

dosimeter material that does not require an external container (Guo et al. 2006, Jordan

2010), and is water equivalent (Zeff = 7.4). Of the publications that have used 3D radiochromic

material for radiotherapy dosimetry, a clear majority have considered only external beam

radiotherapy dosimetry, with a few investigations considering brachytherapy sources

(Austerlitz et al. 2007, Wai et al. 2009, Pierquet et al. 2010, Massillon et al. 2012).

2.3. Brachytherapy Audit

2.3.1. Definition of Audit and the Need for Dosimetric Audit in Radiotherapy

Audit is required in a multitude of scenarios in medicine, and the term has

acquired different meanings over time in relation to health care quality. Clinical audit, for

example, may involve systematically looking at the procedures for diagnosis, care and

treatment, examining how resources are used, investigating the effect care has on the

outcome for the patient, and importantly recognising audit as a quality improvement process

not just a monitoring system. Audit may consider any aspect of infrastructure, procedure or

outcome to ensure safe, effective and best-practice processes and enable improvements.

NHS England has defined clinical audit as “a way to find out if healthcare is being provided in

line with standards”, and importantly “the aim is to allow quality improvement to take place

where it will be most helpful and will improve outcomes for patients” (NHS England 2014).

Audit therefore needs to be undertaken, and it needs to be directed appropriately. In this

work, the scope is limited to specific consideration of dosimetric audit of brachytherapy.

In radiotherapy physics, a key component of auditing is to review the most

fundamental of requirements; that prescribed radiation doses are being accurately delivered.

This may involve testing the dissemination of dosimetry calibration from national standards

laboratories, verifying dose or dose distribution for particular treatment techniques, or

assuring dose delivery for compliance with clinical trial protocols (EORTC 2014). With regard

to the latter it has been demonstrated that the number of patients required in a randomised

clinical trial may be reduced by introducing appropriate dosimetry quality assurance as the

risk of under-powering the study is minimized (Pattersen et al. 2008). The largest dosimetric

2 Heuris Inc, Skillman, New Jersey, USA

18

audit networks at present are operated by the International Atomic Energy Agency (IAEA

2007), the American Radiological Physics Centre (RPC 2014), and in Europe the European

Society for Radiotherapy and Oncology ESTRO Quality Assurance network (EQUAL) (Ferreira

et al. 2000). The reader is directed to the proceedings of a 2010 IAEA meeting on Standards,

Applications and Quality Assurance in Medical Radiation Dosimetry (IDOS) for a number of

papers on external beam audit (IAEA 2011) and a review of audits for advanced treatment

dosimetry by Ibbott and Thwaites (2015). There are also many national audit groups, for

example in the UK standard auditing is coordinated by the Institute of Physics and

Engineering in Medicine (IPEM) via a number of regional groups (Bolton 2009, Palmer et al.

2011). This network arose following an IPEM coordinated national megavoltage photon

beam dosimetry audit (Thwaites et al. 1991) and a later national electron beam dosimetry

audit (Nisbet and Thwaites 1997). Audits for clinical trials in the UK are conducted by the

Radiotherapy Trials Quality Assurance Group (RTTQA), Mount Vernon Cancer Centre, UK. In

2012, the IAEA surveyed the worldwide coverage of dosimetry audit programmes for

radiotherapy (Izewska et al. 2012), finding audit activity in 45 countries, of which 16 had a

mandatory requirement for participation, but with around 1/3 of world radiotherapy centres

having no independent assessment. In the UK a consortium of professional bodies published

‘Towards Safer Radiotherapy’ (RCR 2008) which recommends that “all centres should

participate in dosimetric audit networks” and that “comparative audits between

departments can provide valuable opportunities to ensure safe delivery of radiotherapy and

consistency of patient outcomes”. The National Health Service (NHS) National Cancer Peer

Review Programme Manual for Cancer Services: Radiotherapy Measures (NHS DH 2013),

requires centres to take part in local audit networks. This is typical of publications from

several bodies in recent years suggesting how radiotherapy could be made safer. Dunscombe

(2012), has analysed seven authoritative documents, including ‘Towards Safer Radiotherapy’,

to find commonalities between the recommendations. “Dosimetric audit” was one of twelve

topics identified in three or more documents as being pertinent to the improvement of

patient safety in radiotherapy. Dunscombe (2012) states that “organisations like the RPC and

EQUAL-ESTRO have had, and continue to have, a huge positive influence on the safety and

quality of radiotherapy”. However, Dunscombe also discusses that dosimetric audits are not

always carried out appropriately, stating audits should “take place prior to the first clinical

use”, and enable “testing the device under conditions other than those used to calibrate it”,

citing a treatment error from Ontario as an example that might be avoided if audits were

optimally used (Dunscombe et al. 2008).

The increasing complexity of radiotherapy planning and delivery makes

dosimetric audits challenging, and it is no longer sufficient to verify only the absolute dose

delivery at a reference point, which has been one of the standard approach in the last two

decades. Kron et al. (2013) states the focus of current research is to adapt dosimetry audit

for ever more diversified radiotherapy procedures including image guided/adaptive

radiotherapy, motion management and brachytherapy.

19

2.3.2. The Need for Dosimetric Audit in Brachytherapy

In comparison to external beam radiotherapy, the physical processes by which

the majority of brachytherapy equipment calculates and delivers treatment is relatively

simple. However, this does not mean that dosimetric audit is without complexity. Indeed, the

high dose gradients, orders of magnitude variation in dose deposition across clinical regions

of interest, and small spatial scales, mean measurements to verify absolute dose and dose

distribution are challenging. Haworth et al. (2013) states “to date, dosimetric audits of HDR

facilities have not been conducted in Australia despite the high risks associated with these

treatments due to the challenges presented by measuring doses in steep dose gradients”. In

the UK, the National Cancer Peer Review Programme Manual for Cancer Services:

Radiotherapy Measures (NHS DH 2013), states there is a requirement that “the department

should have taken part in the External Quality Control programme”, but this is only

specifically listed within the external beam radiotherapy measures, not within the

brachytherapy measures. This may be due to the then lack of availability of brachytherapy

dosimetric audit, difficulties in implementation, or prioritisation of need. However, in the

more recent NHS England Service Specification document for 2013/14 for brachytherapy

(NHS England 2013) it is stated that “to ensure that the services being delivered offer high

quality brachytherapy to patients”, one of the specific requirements is that “the provider

department must participate in the national inter-departmental dosimetry audit programme

(National audit of high dose-rate (HDR) brachytherapy)”. In addition to the audit work

presented in this thesis, two other brachytherapy audits are ongoing in the UK (currently

unpublished), one a well-chamber intercomparison, and the other an alanine and Farmer-

type ionisation chamber measurement of dose to a point from a straight catheter.

In the last few years, numerous commercial detectors and phantoms have been

specifically developed to verify dose distributions in external beam radiotherapy, partly

driven by the adoption of intensity modulated radiotherapy (IMRT) and volumetric

modulated radiotherapy (VMAT) techniques. These active detectors may conveniently be

adopted for dosimetric audit (Hussein et al. 2013). However, there have been no similar

commercial developments for brachytherapy. Well-type ionisation chambers have been

adopted for source strength determination, and commonly TLDs have been used for point-

dose measurements, but there is no clear consensus on techniques for verification of dose

distribution measurement in brachytherapy. In the last decade, many dosimetry systems

have been investigated for brachytherapy measurement, including gel dosimetry,

thermoluminescent detectors, semiconductor diodes, ionisation chambers, MOSFET

detectors, alanine, radiochromic film, radiochromic plastic, calorimetry, and optically

stimulated or radioluminescent detectors (Palmer et al. 2012b).

The need for dosimetric audit in brachytherapy exactly mirrors the need in

external beam radiotherapy: to detect any errors, to provide reassurance, to enable

improvements, and to demonstrate compliance. In brachytherapy, a dosimetry system audit

may be useful to test many aspects of dosimetric accuracy, for example assessing the extent

by which simplifications in the TG-43 dosimetry formulism (Zehtabian et al. 2012) give rise to

differences between calculation and delivered doses in HDR treatment delivery, including

20

scatter, tissue homogeneity, and applicator attenuation. There are of course numerous

sources of uncertainty in brachytherapy (Kirisits et al. 2013) that could be assessed by

dosimetry audit and there have been previous errors in brachytherapy delivery that may be

discovered through audit (ICRP 2000, ICRP 2005a, ICRP 2005b, USNRC 2009, Daigh 2010,

Dempsey 2011). The pace of change in brachytherapy equipment, physics and clinical

processes has also been rapid in recent years with the integration of multi-modality 3D

imaging and voluming, improved dosimetry (Sander 2014), and new equipment and

applicators (Thomadsen et al. 2008, Rivard et al. 2010, Beaulieu et al. 2012). Treatment

planning algorithms in brachytherapy may also be on the verge of a step-change in

complexity (Papagiannis et al. 2014) and a move from standard planning to fully flexible

optimisation may have widespread uptake (Thomadsen et al. 2008, Rivard et al. 2009, Rivard

et al. 2010, Beaulieu et al. 2012, Lee et al. 2014). To contribute to overall quality, safety and

reassurance, brachytherapy should be subject to the same rigor of local quality checks

(Palmer et al. 2012a) and audit mechanisms as external beam radiotherapy.

2.4. Monte Carlo MCNP5 Simulation

2.4.1. The Monte Carlo Method

Monte Carlo software codes can be used to duplicate theoretically a statistical

process, such as the interaction of nuclear particles with materials. They are particularly

useful for complex problems that cannot be modelled by computer codes that use

deterministic methods. The individual probabilistic events that comprise a process are

simulated sequentially, with probability distributions governing the events statistically

sampled to describe the total phenomenon. The sampling process is based on the selection

of random numbers (X-5 Monte Carlo Team, 2008a). Monte Carlo methods are very

convenient to estimate the result of measurement situations that would be problematic to

undertake practically, such as dose measurements in the steep dose gradients and small

spatial scales of brachytherapy sources. There are many Monte Carlo codes available for

radiation transport simulation (COG, EGSnrc, GEANT4, MCNP, MCSHAPE, PENELOPE, SCALE,

SRIM, TRIPOLI). The Monte Carlo N-Particle (MCNP)3 code has been chosen because it is a

widely used general-purpose code with capabilities including photon and electron transport

(Dunn and Shultis 2012). MCNP5 version 1.60 is used in this thesis.

2.4.2. MCNP5 Applied to Brachytherapy

MCNP has been widely used and validated in numerous medical physics

applications which are relevant to the present work. Examples of MCNP5 application for

brachytherapy dosimetry are provided by the following authors; MCNP5 Development Team

(2005), Yang and Rivard (2009), Gerardy et al. (2010), Adamson et al. (2012), Mosleh-Shirazi

et al. (2012), Reda (2013), Fronseca et al. (2013), Rivard et al. (2014). MCNP5 code may

3 Los Alamos National Laboratory, Los Alamos, New Mexico, USA

21

therefore be used ‘as is’ in this thesis with no further requirement to demonstrate

applicability of the code. However, it is essential to validate local implementation of the code

and input files to ensure correct operation. Advice on the application of Monte Carlo

methods in brachytherapy is provided in a report of the AAPM and ESTRO, published by

Perez-Calatayud et al. (2012a).

2.4.3. MCNP5 Input and Output Files

MCNP input files are used to define the problem; source, structures, tallies, and

other physics or simulation control data. There is a strict format required for the input file (X-

5 Monte Carlo Team 2008b), consisting of message block, problem title card with cell cards,

surface cards, and data cards, each delimited with single blank lines. Input lines have a

maximum of 80 columns, command mnemonics beginning in the first 5 columns or if blank

assumed a continuation of the previous line. Comment lines, ignored by the code, commence

with ‘c ‘.

Tallies are used in MCNP to specify and collect the required output data, in terms

of particle current, particle flux or energy deposition for specified geometries. Tallies are

normalized to be per source particle. There are two tallies in MCNP5 that may be used to

estimate energy deposition for brachytherapy problems. The F6 tally calculates the deposited

energy in a cell, via the transport of photons only, with local deposition of electron energy. It

consists of a track length tally multiplied by a reaction rate convolved with an energy-

dependent heating function. The *F8 pulse height tally provides the energy distribution of

pulses (or total energy if no bins are specified) created in a detector by radiation. *F8 is most

rigorous as both photons and electrons may be transported. However, F6 calculation is more

efficient, faster, and with reduced statistical uncertainty for the same number of particle

histories. *F8:p,e is the theoretically preferred method for photon and electron transport

problems as all details of the transport are followed (Goorley 2007). F6 tally estimates energy

deposition by integrating the track-length photon flux weighted by photon heating numbers.

These numbers represent the average kinetic energy given to electrons along the photon

path. Therefore, this tally is approximately valid only when most of the electrons are trapped

in the tallied cells. If the cells are small enough that a significant amount of electron energy

can escape, then the F6 tally will overestimate the energy deposition. In contrast, the *F8

tally performs a detailed accounting of the energy entering a cell minus the energy leaving a

cell for each history in a mode p,e problem.

A wealth of information is available in the MCNP output file, the content of

which is controlled by the inclusion of optional commands in the input file. The output file

always contains the input file listing, summary of particle loss/creation, tallies and tally

fluctuation charts. Data in the output file provides information for the user to assess the

precision (not the accuracy) of the tally results, particularly the ‘ten statistical indices’

calculated by MCNP. The relative error, R, is the fractional 1-sigma estimated uncertainty in

the tally mean, i.e. the ratio of the standard deviation of the tally mean to the mean. If R <0.1

the tally is reliable except for point/ring detectors, if R <0.05 the tally is reliable even for

22

point/ring detectors (Shultis and Faw 2011). The relative error must not be consulted in

isolation, and all 10 statistical tests should pass (X-5 Monte Carlo Team 2008b). If any of the

tests are failed, MCNP automatically produces additional output to aid the user in

interpreting the seriousness of the failure.

23

CHAPTER 3

Quality Control of High Dose Rate

(HDR) Brachytherapy Treatment

Equipment

24

3.1. The Need to Review Quality Control and Commissioning

Procedures and Establish Performance Requirements

3.1.1. Quality Control Testing of HDR Brachytherapy Systems

In comparison to external beam radiotherapy, the physical process by which

high dose rate (HDR) brachytherapy treatment units deliver treatment is simple. However,

this does not mean the quality control task is without complexity, and this must be

undertaken comprehensively to mitigate risk of treatment error. Quality assurance of all

brachytherapy techniques has recently received increased attention following a low dose

rate brachytherapy incident in 2009 at the Philadelphia VA Medical Centre in which a number

of patients received poor quality prostate seed brachytherapy treatment (Daigh 2010). There

have been many reported errors in brachytherapy delivery (ICRP 2000, ICRP 2005a, ICRP

2005b), including well chamber calibration error (Dempsey 2011), confusion over units

(USNRC 2009), media report of at least two incidents involving incorrect dwell positions

(ASAHI SHIMBUN 2013, Sydney Morning Herald 2003a and 2003b), and ‘field safety notices’

from equipment manufacturers, see Section 1.1.3. It is important that QC testing is robust

and comprehensive and meets the needs of modern equipment and treatment techniques.

It is over twenty years since a comprehensive assessment was undertaken of

brachytherapy QC practice in the United Kingdom (UK) (IPEM Radiotherapy Physics Topic

Group 1992), reproduced in the Institute of Physics and Engineering in Medicine (IPEM)

Report 81 (Mayles et al. 1999). The IPEM guidance on QC for radiotherapy is currently being

revised, and it is therefore timely to undertake a repeat benchmark exercise of current QC

practice for brachytherapy. The present survey has been endorsed by the IPEM Radiotherapy

Special Interest Group. A similar survey was conducted in 2002 in the Netherlands and

Belgium (Elfrink et al. 2001a), which reported large variations in test frequencies and

methods, and differences in QC-philosophy and available equipment. The author is not aware

of any other recent surveys of brachytherapy QC practice between centres.

The presentation here of a comprehensive assessment of current QC practice

has several potential benefits for individual radiotherapy departments: centres may be

reassured that their QC systems are in-line with accepted practice; alternatively, centres may

identify discrepancies against standards of practice. Following investigation, this may lead to

either reduction of tests or frequencies and hence efficiency savings, or resolution of

deficiencies and potential improvements in safety and quality. However, the details of QC

tests presented here should not be interpreted as guidelines or recommendations, but as a

‘snapshot’ of current UK practice against which individual centres may benchmark

themselves. It is important to be aware that specific QC testing is a local decision, based on

many local factors, and should ideally be based on risk-assessment approaches.

25

3.1.2. Commissioning of HDR Brachytherapy Systems

The performance of an HDR unit and its ability to accurately implement a series

of planned source dwells is critical to the quality of the treatment. The accuracy of delivered

doses is particularly dependent on source positioning due to the short distances between

target and source, steep dose gradients and large inverse-square law corrections for any

geographic errors. Prior to clinical use, new HDR treatment units require a complete and

thorough assessment to verify that the manufacturer’s and clinical users’ performance and

accuracy requirements are met, and that recommended baseline QC and dosimetric criteria

are satisfied. Previous performance evaluations (McDermott et al. 1996, Wallace 1997) have

shown that there can be appreciable source control and dosimetric differences between

different models of HDR systems. There are several publications on the general quality

control requirements for HDR units (Nath et al. 1997, Mayles et al. 1999, Venselaar et al.

2004, Wilkinson 2006), but no known contemporary review of proposals for an optimum

commissioning schedule.

A previous publication by the author presented baseline comparative

commissioning data for the first Eckert & Ziegler Bebig4 HDR brachytherapy system in the UK

(Palmer and Mzenda 2009). The work concluded with three suggestions for technical

improvements to (a) the compensation calculation for transit dose, (b) the source transit

movement profile to the first dwell point, and (c) the effect of curvature of source transfer

tubes on source position accuracy. Positioning errors of the source, of up to 1.0 mm for slight

bends, 2.0 mm for moderate bends and 5.0 mm for extreme curvature (depending on

applicators and transfer tube used) were reported in the study. The current work has

involved communication with the manufacturer to optimise the performance of the control

system software, addressing the technical issues above. Following issue of new software,

detailed measurements were performed on the source dwell positioning accuracy, using

both video analysis and autoradiography techniques, and on the transit dose and associated

corrections made by the HDR unit.

It is accepted practice that HDR treatment planning systems do not make any

correction for radiation dose resulting from transits of the source between dwells, presuming

these can safely be ignored due to the low actual transit time and dose, and that the HDR

unit will itself make correction to the planned dwell time to account for the transit dose. The

magnitude of the transit dose and the actual corrected dwell time are not normally presented

to the clinical user. The practice and validity of not considering transit doses in treatment

calculations must be verified for each manufacturer’s HDR unit, relating to its particular

source velocity, resulting transit dose, and the accuracy of transit dose corrections that are

made by the system. Similarly, the ability of a particular HDR unit to accurately and

reproducibly position the source at a series of dwell positions must be tested in a range of

clinical situations, in order to provide reassurance of the ability to deliver high quality

treatments. An action level of 2 mm source positioning error has been proposed as an upper

limit in clinical conditions (Venselaar et al. 2004). It is also important to consider the

4 Eckert & Ziegler Bebig GmbH, Berlin, Germany

26

possibility of any effect on source positioning and dwell timing of variations in the clinical use

of the HDR system.

3.1.3. Performance Requirements: Effect of Simulated Source Position Errors

There is also a need to determine performance requirement of brachytherapy

equipment in terms of basic source positioning accuracy and their effect on clinical treatment

planning parameters. An objective of the work below was to undertake a retrospective

treatment planning study to investigate the effect of simulated HDR source dwell position

errors on the dose to the (high risk) clinical target volume (HR-CTV) and organs at risk (OAR)

in cervix cancer. The aim was to determine the clinically relevant positional accuracy

requirements for source dwells in modern brachytherapy techniques, and hence decide

appropriate acceptable limits for routine QC tests for modern brachytherapy. An indication

of treatment dose uncertainty resulting from dwell position uncertainty was also derived.

3.2. Methodology

3.2.1. Survey of Quality Control Practices in the United Kingdom (UK)

All 64 radiotherapy centres in the UK were contacted by email in June 2012 to

request their contribution to the study, with collation of responses taking place during June

to August 2012. Centres were asked to complete a detailed questionnaire on their routine

QC practices and other aspects of high and pulsed dose rate (HDR and PDR) service provision.

The questionnaire is provided in Appendix J. To enable a contextual review of QC practice,

initial questions asked included equipment type, average patient workload, sites treated,

source strength calibration methods, level of image guidance, and prescribing practice. A

spreadsheet containing a comprehensive list of possible quality control tests for treatment

and planning equipment was also provided. Centres were asked to document whether they

routinely perform each test, at what frequency, by which staff group, and the acceptable

tolerance values used. They were also asked to comment whether the target test frequencies

were actually achieved in practice.

3.2.2. HDR Treatment Unit Commissioning Tests, Dwell Position and Transit Dose

Published guidance for the initial commissioning, or re-commissioning after

hardware or software change, for HDR treatment units was compiled and a list of activities

produced. This was then applied to the re-commissioning of an Eckert & Ziegler Bebig

brachytherapy system following a software change. The system consists of the treatment

unit, MultiSource, and a separate treatment planning system, HDRplus, which uses TG-43

formalism based dosimetry (Rivard et al. 2004). Importantly, the BEBIG HDR unit moves the

source through the planned dwells in a distal to proximal direction. An Eckert & Ziegler Bebig

MultiSource HDR treatment unit with software version 7.4.1, firmware version 4.14.1, was

used throughout this study. The results of all analyses were compared to data obtained with

27

the previous HDR control software, version 7.4.0, firmware version 4. All results are based

on use of the HDR unit with an Ir-192 source, model Ir2.A85-2.

To assess source positioning and movements over large distances a treatment

consisting of three dwells at 5.0, 35.0 and 105.0 mm was evaluated using x-ray film

autoradiography (Kodak EDR2 Ready Pack film), using dwells of 1 s for optimum film density.

A scanning optical densitometer (Vidar VXR-16) was used to locate the centre of the source.

The actual dwell position may be affected by curvature of the transfer tube between the HDR

unit and the applicator. Therefore measurements were made with a curvature induced in the

transfer tube by displacing the distal end of the tube by distances of 100, 300 and 900 mm,

while maintaining a stationary proximal end at the HDR unit. The 900 mm displacement was

included as an extreme ‘physically limiting’ case and not an expected realistic clinical

situation.

To assess source positioning and movements over smaller distances a high

definition video camera (Panasonic HDC-SD10) was used to image a catheter during a

treatment prescription consisting of three dwells at 10.0, 15.0 and 20.0 mm. The centre of

the source was identified visually in each video frame, at a resolution of 25 frames per

second, and the position recorded against a ruler also located within the video image. This

data set was used to evaluate the location of the source as a function of time between dwell

positions and to calculate the speed of source movement by the simple division of

displacement and elapsed time.

Using the above data set, the source position was recorded at 1/25 s intervals in

the approach to the first dwell position, between dwells, and from the last dwell back to the

HDR unit. An approximation of the transit dose was calculated by the summation of the dose

delivered by the source considered to be stationary for 1/25 s at each of the imaged

positions, using published dose-rate distribution data (Granero and Perez-Calatayud 2007).

The transit dose was calculated separately for the movement of the source to and from the

dwell position, and evaluated at two interest points, D10 and D20, located at 10 mm and 20

mm respectively from the centre of the dwell position, perpendicular to the source

movement axis, shown in Figure 3.1.

Figure 3.1. Transit dose calculation points, D10 and D20, at 10 mm and 20 mm respectively,

perpendicular distance from the centre of the intended dwell position.

28

The Eckert & Ziegler Bebig Multisource makes corrections for transit doses by

reducing the actual dwell time for each dwell position using the following algorithm (Spiller

2012):

100 cd T- T DT - TpDT rfrom dwellrto dwellr

where pDT is the performed dwell time (ms), DT is the planned dwell time (ms),

Tr is the time reduction applied to the dwell position for each transit (ms), d is the distance

between dwell positions (mm), and c is a constant which equals 3 for dwell separations of

less than or equal to 100 mm, and equals 2 for dwell separations of greater than 100 mm. Tr

for the first and last dwell positions is fixed at 450 ms. The fixed time reduction applied to

the first and last dwells was introduced with the new software, all other algorithm

parameters being consistent with the previous software version.

The transit dose correction implemented by the Eckert & Ziegler Bebig

Multisource was evaluated for the specific treatment situation detailed above, and the

‘equivalent dose reduction’ resulting from this dwell time reduction was compared to the

calculated transit dose, using the video analysis data.

3.2.3. Treatment Planning Study to Determine the Effect of Simulated Source

Position Errors

Simulated dwell position errors, in the range 0.2 to 10.0 mm, were introduced

in the treatment plans of a consecutive series of eight HDR cervix patients previously treated

at Portsmouth Hospitals NHS Trust. The dosimetric effect on ICRU reference point doses

(Chassagne et al. 1985) and GEC-ESTRO 3D dose volume histogram (DVH) parameters (Potter

et al. 2006) were calculated for the HR-CTV and OARs. Two error modes were simulated: (a)

incorrect calibration of the source position, modelled by the systematic proximal shift

(towards the HDR unit) of all dwell positions, separately evaluated with error magnitudes of

0.2, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0 and 10.0 mm; and (b) a potential error mode of source cable

take-up lag, in which the first dwell position is unaffected and all others are shifted distally

(Palmer and Mzenda 2009). The CT-based treatment planning simulations were performed

on the Eckert & Ziegler Bebig HDRplus treatment planning system (version 2.5) employing a

TG-43 dose calculation, with an Ir-192 HDR source (model Ir.A85-2) and split-ring and IU

clinical treatment applicator.

3.3. Results

3.3.1. UK Survey: Brachytherapy Equipment Profile and Physics Processes

47 (73%) of the 64 UK radiotherapy centres that were invited to take part in the

survey of HDR and PDR QC responded. 31 centres had appropriate brachytherapy equipment

and provided fully completed questionnaires on their QC practice. 13 centres reported no

HDR or PDR brachytherapy facilities, and a further 3 intended to commence HDR

brachytherapy in the near future (at June 2012).

29

The majority (29) of radiotherapy centres had HDR units, with only three having

access to PDR treatments, two having exclusively PDR. The equipment profile of the

responding centres included 20 Nucletron/Elekta microSelectron, 7 Varian GammaMed, 4

Nucletron/Elekta/Isodose Control Flexitron, 1 Eckert & Ziegler Bebig HDR Multisource, and 1

Varian Varisource. One centre had 3 treatment units, 1 HDR and 2 PDR, all others had 1 unit.

One centre had a Co-60 source (HDR), the others were all using Ir-192. The latter radionuclide

being exchanged at 3-monthly intervals in all but two centres: one at 4-monthly intervals and

another PDR centre at between 3 and 6 months. One centre stated they were considering

moving to 4-monthly intervals to reduce cost. The planned frequency of exchange for the Co-

60 source was 4 years. The mean number of HDR fractions delivered per year at each centre

was 281, with interquartile range 173 to 359 (minimum 80 and maximum 730).

There was a lack of agreement as to whether the locally measured value or the

manufacturer’s supplied source certificate should be used for the source strength value in

treatment planning calculations; 18 centres (58%) preferring to use their own measurement.

The current UK Code of Practice for HDR brachytherapy dosimetry (Bidmead et al. 2010)

recommends a well chamber for the primary source strength measurement, but allows some

flexibility in the method used to obtain the second independent verification value. Table 3.1

lists methods used for source calibration and their relative popularity within UK centres (at

June 2012).

The quoted origin of the TG-43 (Rivard et al. 2004) source model data used in

the treatment planning systems also varied between centres. 14 (45%) used the supplied

manufacturer data, 8 (26%) used journal published data, and 8 (26%) used manufacturer data

and verified this against publications, (with 1 (3%) not answering the question). There was

also a variety of methods quoted as an independent check of the output of the treatment

planning system. The methods and their popularity are given in Table 3.2.

30

Table 3.1 Methods employed, and their popularity, for source strength measurement of HDR and

PDR sources at centres in the UK (at June 2012).

Technique for determination of source strength Centres citing method,

Number Percentage

Initial method Well chamber 29 94%

Manufacturer supplied source certificate 1 3%

NE2571 chamber with in-air jig 1 3%

Verification method

Well chamber (second unit) 14 45%

NE2571 chamber with in-air jig 12 39%

NE2571 chamber in solid phantom 2 6%

Manufacturer supplied source certificate 1 3%

I-125 seed device with adaptor 1 3%

Gafchromic film calibrated via 260kV x-rays 1 3%

Table 3.2. Methods employed, and their popularity, to independently verify treatment planning

system (TPS) calculations at centres in the UK (at June 2012).

Technique for verification of TPS calculation Centres citing method,

Number Percentage

Locally developed check software (including systems based on Matlab, Excel, Java, visual basic; usually employing either TG43 (Rivard et al. 2004) or BIR/IPSM 1993 (Aird et al. 1993)

16 52%

Commercial check software or additional TPS (including IMSure QA, Radcalcbrachy, Lifeline)

7 23%

Manual calculation or use of data tables 3 10%

Nomogram (prostate treatment) or TRAK relationship to target volume 2 6%

Use of standard plans only with initial independent calculation, no per-patient plan verification

2 6%

Consistency check performed with standard plan on same day 2 3%

31

Table 3.3. Primary sources of guidance for establishing QC schedules, and their popularity at

centres in the UK (at June 2012).

Documents providing guidance on HDR or PDR QC Centres citing document,

Number Percentage

Physics aspects of quality control in radiotherapy, IPEM Report 81 (Mayles et al. 1999)

19 61%

A practical guide to quality control of brachytherapy equipment, ESTRO Booklet No. 8 (Venselaar and Perez-Calatayud 2004)

15 48%

The IPEM code of practice for the determination of the reference air kerma rate for HDR (192)Ir brachytherapy sources based on the NPL air kerma standard, (Bidmead et al. 2010)

10 32%

Code of practice for brachytherapy physics, AAPM TG-56, 1997 (Nath et al. 1997)

8 26%

Discussion with colleagues and other centres’ documents 6 19%

High dose-rate brachytherapy treatment delivery, AAPM TG-59, (Kubo et al. 1998)

4 13%

Recommendations for Brachytherapy Dosimetry, BIR/IPSM Report 1993 [Aird et al. 1993]

3 10%

Calibration of photon and beta ray sources used in brachytherapy, IAEA TecDoc 1274, (IAEA 2002)

3 10%

Quality assurance for clinical radiotherapy treatment planning, AAPM TG-53, (Fraass et al. 1998)

2 6%

Quality assurance tests for prostate brachytherapy ultrasound systems, AAPM TG-128, (Pfeiffeera et al. 2008)

2 6%

Manufacturer’s guidance or manual (unspecified) 2 6%

A revised AAPM protocol for brachytherapy dose calculations, AAPM TG-43U1, (Rivard et al. 2004)

2 6%

Thomadsen BR ‘Achieving quality in brachytherapy’, (Thomadsen 1999) 2 6%

Other radiotherapy or brachytherapy text books, each n=1 2 6%

Remote afterloading technology, AAPM TG-41, (Glasgow et al. 1993) 1 3%

Towards Safer Radiotherapy, joint report of RCR, SoR, CoR, IPEM, NPSA, BIR, (RCR 2008)

1 3%

In-house experience with treatment unit 1 3%

Quoted “Relevant regulations” 1 3%

Quoted “Unsure of origin” 1 3%

32

Treatment plan optimisation in some form was used in 23 centres (73%),

including for cervix (majority), prostate (next most common), skin/limb moulds, interstitial

anus, vaginal vault, lung, head & neck, multilumen mammosite breast, intraluminal, and

keloid scars. 27 centres (87%) optimised treatment plans for individual patients; 19 (61%)

employing manual methods and the others inverse planning optimisation, often with final

manual adjustment. 6 centres (19%) stated they used pre-optimised standard plan libraries.

The level of image-guidance varied significantly between centres. In cervix

treatments, 16 (52%) used CT alone for treatment planning, 12 (39%) MRI with CT, 2 (6%)

MRI alone, and 1 (3%) c-arm 2D imaging alone. When MRI was available this was often used

for the first fraction, with CT used in subsequent treatments. For vaginal vault treatments,

13 (42%) did not image, 10 (32%) used orthogonal 2D x-ray, and 8 (26%) used CT. Some

centres responded they would only image vault treatments for complex cases or if

individualised plans were required. There were an insufficient number of responses on

imaging used for other treatment sites for statistical significance.

Gynaecology cancers were the most commonly treated. In cervix, 22 centres

(71%) still prescribe treatment doses to Manchester Point A. For those prescribing instead to

high-risk clinical target volume, HR-CTV (Potter et al. 2006), all centres additionally record

the Point A dose. Only 2 centres (6%) exclusively recorded ICRU organ at risk (OAR) point

doses, likely when only orthogonal imaging is used, the others recorded either just GEC-

ESTRO dose-volume histogram (DVH) data (48%), or both ICRU point dose and DVH data

(46%).

Centres were asked to list the primary sources of guidance used in establishing

their HDR or PDR quality control schedules. Table 3.3 provides a list of the documents that

were indicated and their popularity, quoted as the percentage of centres citing the

document. All centres stated they had reviewed the content of their HDR/PDR QC schedule

within the last two years, except two which did not answer the question.

3.3.2. UK Survey: Quality Control Testing

Table 3.4 shows the percentage of centres that include each of the specific tests

in their planned QC schedules, from the UK survey. A centre is deemed to have included the

test in their regular QC if it is performed within the department whether it is in the specific

‘physics QC documentation’ or other ‘standard operating procedures’, for plan checking for

example. A test is deemed not to be in regular QC if it was only intended to be performed

once at initial equipment commissioning. The mean and range of frequencies of testing and

acceptable tolerance levels are provided in the table. There is a significant variation in

consistency between centres across the range of tests. Inclusion of tests in QC schedules

varies from 31 centres (100%) for source strength measurements to just 2 centres (6%) for

MRI tests, the latter of course being due to limitations of access to MRI for brachytherapy-

specific clinical use and QC testing. There is also a lack of consistency between centres in the

frequency of testing and the acceptable tolerance levels used for individual QC tests. Of the

45 tests included in the survey, 21 were performed at greater than 75% of centres, and 4

33

were performed at less than 25% of centres. Only 2 centres (6%) reported that achieved

measurement frequencies were below planned measurement frequencies, and then only for

up to two tests each.

Table 3.5 lists additional QC tests suggested by responding centres, which were

not included in the original set of tests in the distributed survey. These are generally

proposals made by single centres and there is no information of the popularity of these tests

across UK, however they are included for completeness. Quality control testing of dosimetry

equipment associated with HDR and PDR use has not been included in the results tables;

secondary standard calibrations and consistency testing of well chambers and Farmer-type

ionisation chambers are covered elsewhere and in published codes of practice.

It is valuable to consider in further depth those tests that are: (a) included in QC

schedules at all UK centres, Table 3.6; (b) those at around half of centres, Table 3.7; and (c)

those not currently in any centres QC schedule, Table 3.8. The tests of primary importance,

source strength, timer accuracy, and basic source position (in straight catheter), were

included in all UK centres’ QC schedules, and the published guidance documents are

generally in good agreement, see Table 3.6. This was a valuable validation of basic quality

checking. However, there was disagreement among UK centres, and in published guidance

documents, as to the value of including other tests in routine QC. Two such tests are given in

Table 3.7, accuracy of dwell positions in clinical applicators and measurement of source

transit time/dose. A method for the evaluation of transit dose is provided in Sections 3.3.3.2

along with discussion indicating the value of this test. The need to test the position of source

dwells in clinical treatment applicators is indicated in Sections 3.3.3.3, 3.3.4, and also due to

prior errors of source positioning having been reported, Section 1.1.3.

Table 3.8 lists three performance tests that were not included in any UK centres’

QC schedules at the time of the survey, and are not included in published guidelines: ‘end to

end’ testing, measurement of dose distributions around clinical treatment applicators, and

checking of advanced planning system features available for modern brachytherapy, such as

inverse optimisation and inhomogeneity correction

3.3.3. Commissioning and QC of HDR Treatment Units

A list of activities that comprise the required commissioning test for HDR

brachytherapy systems is provided in Table 3.9. This provides a contemporary collation of

areas of investigation that are required, derived from several publications (Bastin et al. 1993,

McDermott et al. 1996, Wallace 1997, Mayles et al. 1999, Sahoo et al. 2001, Venselaar and

Perez-Calatayud 2004, Wilkinson 2006, Evans et al. 2007), suggestions from other UK

radiotherapy centres via personal communication with the author, and the author’s own

opinion. This table is useful to put into context the detailed results in the following sections.

34

Table 3.4. HDR and PDR QC survey: Response to questionnaire on test popularity, measurement

frequency and tolerance values (at June 2012).

QC test % of centres including in routine QC

Measurement intervals, % using mean value (and range of responses)

Tolerance, % using mean value (and range of responses)

Comments

Sou

rce

str

en

gth

Initial measurement after source installation

100% 100% at source change

(all centres)

52% use 3%

(2% to 5%) Achievable

tolerance depends on test

method and whether result is compared to

certificate or 1st measurement

Independent measurement after source installation


(all centres)

35% use 3%

(0.5% to 5%)

Repeat measurements during life of source

75% 83% at 1m

(1 d to 1m)

56% use 3%

(0.5% to 5%)

Leak testing of source


(1m to source change)

71% not > background

(zero to 200 Bq)

Tre

atm

en

t u

nit

fu

nct

ion

Confirm accuracy of source data at treatment unit

97%

41% at each patient

(each patient to commissioning only)

55% use exact match

(exact to 4%)

Confirm accuracy of decay correction at treatment unit for plans

91%

86% at each patient

(weekly to commissioning only

22% use 1%

(exact match to 3%)

Tolerance depends on

how frequently unit makes

decay correction

Plan data transfer from TPS

83%

96% at each patient

(weekly to commissioning only)

63% use exact match

(0.1 s to 2%)

Some standard template plans

only, no transfer

Simulated treatment functionality test

68% 68% at 1d

(1d to 12m)

50% use 1mm

(1 mm to 2 mm)

May be independent

test or combined

System display and print-out accurate and in agreement

96%

83% at each patient

(each patient to commissioning only)

88% use exact match

(exact to 2%)

Test of function with mains power loss

58% 50% at 3m

(1m to 12m)

Test of UPS power supply

44% 31% at 3m

(1d to 12m)

d = day, w = week, m = month, y = year, TPS = treatment planning system

Table 3.4 continued on next page…

35

…Table 3.4 continued from previous page




Comments

Tre

atm

en

t p

lan

nin

g sy

ste

m

Accuracy of source model data used by TPS (e.g. check TG-43 data against reference values)

36% 65% at commissioning

(1m to commissioning)

No consistency

(Interpolation to 5%)

Normally undertaken at

software updates

Accuracy of individual source data used by TPS (e.g. source strength, calibration date)

89% 64% at each patient

(each patient to 3m)

No consistency, responses were:

exact match, rounding error, 1

day correction, 0.5%, 1%, 2%, 3%.

Tolerance may depend on how frequently unit

makes decay correction

Calculation of standard plans compared to reference data

63% 28% at 3m

(1d to commissioning)


exact match, 1%, 2%, 3%, 5%, 1 mm

idsodose lines

Independent check calculation of TPS patient plans or standard plans.

62%

74% at each patient

(each patient to commissioning)

50% use 3%

(1% to 5%)

Depends whether

patient-specific optimised or

standard plans are in use

Repeat of tests performed at TPS commissioning (e.g. DVH accuracy, geometric tests)

32% 39% at 3m


33% use 2 mm

(variety of definitions including

mm, % of DVH or dose points)

Often performed at

software updates only



36





Comments

Ap

plic

ato

rs, s

ou

rce

po

siti

on

an

d s

ou

rce

mo

vem

en

t

Visual inspection of applicators and transfer tubes for damage

97% 70% at 1d

(1d to 12m)

Measurement of dimensions and angles of applicators and transfer tubes


(each patient to 12m)

75% use 1 mm

(0.5 mm to 1 mm)

Often rely on image match to

TPS library

X-ray imaging of applicators


(3m to 12m) 100% use 1 mm

Most commonly only when suspected

damage

Verification of source dwell timer accuracy

97% 38% at 1d

(1d to 12m)

38% use 1s

(0.1 s to 2 s)

Large variation in definition of

test methodology

Measurement of source dwell positions in straight catheter (not clinical applicator)

100% 42% at 1d

(1d to 4m)

78% use 1 mm

(0.5 mm to 2 mm)

Multiple techniques

often in use at each centre

Measurement of source dwell positions in clinical applicators


(2w to 12m)

79% use 1 mm

(1 mm to 2 mm)

Measurement of actual source dwell positions compared to TPS stated position in complex geometry e.g. ring applicator

41% 35% at 3m

(2w to commissioning)

67% use 1 mm

(1 mm to 3 mm)

Absence of test often due to

ring applicator not being used

Source position relative to dummy source or marker wire


(1d to 12m)

72% use 1 mm

(0.5 mm to 2 mm)

Absence of test normally due to marker wire not

being used



37





Comments

Ap

plic

ato

rs, s

ou

rce

po

siti

on

an

d s

ou

rce

mo

vem

en

t (c

on

t.)

Applicator/transfer tube connection interlock and simulated error

73% 68% at 1d

(1d to 6m)

Verification of expected position of internal applicator shielding

15%

50% at each patient


Not commonly

in use in UK

Measurement of source transit times

47% 41% at 3m

(1w to commissioning)


0.1 s, < 0.5 s, <1 s, not > baseline, not

> 0.05 s dwell equivalent

Large variety of techniques (well

chamber to stop watch) and tolerance values

Confirm error code ‘meanings and actions’ are available at treatment unit

36%

25% each at 1d and 12m


Historic test, mostly replaced

with improved software interface

information

Radiation monitor of applicators after use

26% 86% at each patient

(each patient to 1w)

100% not above background

Majority rely on in-room

radiation monitor

Availability of in-vivo dosimetry system for brachytherapy?

19%

Not in clinical use in the

majority of centres.

Included TLD, diode, MOSFET.

Calibration of in-vivo dosimetry system

60% (of 19%) 50% at 1m

(1m to ‘as required’) 100% use 5%

Test of in-vivo measurement against expected /planned dose measured in phantom

60% (of 19%) 50% each at 1w and

12m 100% use 5%



38





Comments

Faci

litie

s

Visual (CCTV) and audible (intercom) patient monitoring

100% 93% at 1d

(1d to 3m) functional

Radiation warning lights

100% 90% at 1d


Independent radiation monitor (room monitor)

100% 83% at 1d


Interlocks (e.g. door, timer delay)

100% 90% at 1d


Emergency stop control

100% 81% at 1d


Co

nti

nge

nci

es

Practice of simulated emergency (e.g. source stuck)

97% 46% at 12m

(1w to 12m)

Presence of emergency equipment (e.g. source container, forceps, shield, monitor)

97% 91% at 1d

(1d to 12m)

Review of responsibilities (e.g. who removes applicator if source stuck)

84% 50% at 1d

(1d to 12m)



39





Comments

Imag

ing 2D kV imaging

tests, including applicator reconstruction

24% 29% at 1d


50% use 2%

(1% to 2%, or 2 mm)

Absence of test often due to 2D

imaging not being used

CT imaging tests, including applicator reconstruction

38%

50% at commissioning



1%, 2%, 3%, 5%, 1 mm, 2 mm

Absence of specific tests often due to

reliance on TPS applicator

library

MR imaging tests, including applicator reconstruction and distortion



50% use 2 mm

(1 mm to 2 mm)


brachytherapy-specific MR imaging not being used.

Access to MR for QC often a

problem.

Ultrasound imaging tests, including applicator reconstruction and grid alignment

26% 43% at 3m



1 mm, 2 mm, 1 cc, 5%


ultrasound imaging not being used

Accuracy of image data transfer to TPS

60%

27% at each patient


43% use exact match

(exact to 2 mm, or 2%)

Image-based data only


40

Table 3.5. UK QC survey: Additional tests identified by responding centres as being included in local

QC procedures but not included in original survey questionnaire.

Additional possible QC tests not included in the survey

Equ

ipm

en

t p

erf

orm

ance

Dwell time linearity

Check behaviour if transfer tube loop/curvature too tight

Treatment interruption behaviour

Source drive motor operational (check audible indication of movement)

Satisfactory performance of system self-test

PD

R

PDR pulse timing

Check of nurses’ station and remote control panels

Catheter integrity and connectivity to PDR unit

Partial treatment completion

TPS

Consistency of plans between software version

Data security including backup (patient information, source data, system settings)

Imag

ing

Image fusion CT/MR

MR image scaling

Rad

iati

on

pro

tect

ion

Radiation monitoring of treatment unit (e.g. dose rate at 5cm or 1m)

Radiation monitor empty treatment unit during source change

Receipt and return of source paperwork

Confirm controls in place for source security

Mis

c.

External audit of system quality control/performance

41

Table 3.6. QC tests included in all surveyed UK centres’ schedules (at June 2012), with comparison

to published recommendations.

QC test UK survey:

Measurement

frequency,

mean with

range in

brackets.

UK survey:

Acceptable

tolerance,

mean with

range in

brackets

Comparison to published recommendations

Initial and

independent

measurement

of source

strength

At source

change

(all)

3%

(2% to 5%)

AAPM: 3-monthly, 3%

ACPSEM: 3-month/source change, 3% tolerance, 5%

action

CAPCA: 3-month/source change, 3% tolerance, 5%

action

ESTRO: source change, action level >5%

IPEM: source change, investigate 3%, do not use >5%

Source dwell

timer accuracy

Daily

(daily to

annually)

1s

(0.1s to 2s)

AAPM: no value given

ACPSEM: daily and source change, 1% tolerance, 2%

action

CAPCA: 3-monthly, 1% tolerance, 2% action

ESTRO: annually, action >1%

IPEM: no value given

Source dwell

positions in

straight

catheter

Daily

(daily to at

source change)

1 mm

(1 mm to 2

mm)

AAPM, daily, 1 mm

ACPSEM: daily and source change, 1 mm tolerance, 2

mm action

CAPCA: daily, 1 mm tolerance, 2 mm action

ESTRO: daily, action >2 mm

IPEM: no value given

AAPM (Nath et al. 1997), IPEM (Mayles et al. 1999), ESTRO (Venselaar and Perez-Calatayud 2004), CAPCA

(Arsenault et al. 2006), ACPSEM (Dempsey et al. 2013).

42

Table 3.7. Comparison of UK practice and published recommendations for two QC tests included

in only half of surveyed UK centres’ schedules. (Source dwells in clinical applicators in

55% of schedules and transit dose in 47%)

QC test UK survey:

Measurement

frequency, mean

with range in

brackets.

UK survey:

Acceptable

tolerance,

mean with

range in

brackets


Source dwell

positions in

clinical

applicators

At commissioning

(2-weekly to

annually)

1 mm

(1 mm to 2

mm)

AAPM: 3-monthly, ±2 mm position relative to

applicator, 0.5 mm relative to dummy markers

ACPSEM: X-ray marker positional accuracy,

annually, 1 mm tolerance, 2 mm action

CAPCA: annually, 1 mm tolerance, 2 mm action

ESTRO: not explicitly required, suggestion of jig

IPEM: “preferably less than 1 mm”, no frequency

given

Measurement

of source transit

time or dose

3-monthly

(weekly to at

commissioning)

Large

variations in

definition

AAPM: annually, “influence of transit dose must be

evaluated and corrected for if necessary”

ACPSEM: annually, “transit dose reproducibility”,

1% action, 2% tolerance

CAPCA: annually, 1% action, 2% tolerance

ESTRO: annually, no tolerance given

IPEM: not stated



43

Table 3.8. Suggested QC tests not included in any surveyed UK centres’ schedules (at the time of

the UK survey).

QC test UK survey:

Measurement

frequency

UK survey:

Tolerance

value


‘End-to-end’

check,

including

imaging,

planning and

delivery

Not included in

any routine QC

Not included

in any routine

QC

AAPM: No specific recommendation

ACPSEM: No specific recommendation

CAPCA: No specific recommendation

ESTRO: No specific recommendation

IPEM: No specific recommendation

Verify dose

distribution

around

clinical

treatment

applicators

Not included in

any routine QC

Not included

in any routine

QC

AAPM: No specific recommendation





Check of plan

optimisation

software

performance

(and other

advanced TPS

features)

Not included in

any routine QC

Not included

in any routine

QC

AAPM : initially and new software versions, “series of

test cases based on idealised implant geometries,

develop sense of what optimisation does to an

implant compared to uniform loading”







44

Table 3.9. Required areas of investigation for HDR brachytherapy system commissioning, or for

consideration after re-commissioning following hardware or software update.

Process

Critical Examination, radiation shielding survey, local rules

Manufacturer’s acceptance testing schedule

Mechanical and electrical safety tests

Interlocks, safety features and basic operational checks, including timer accuracy checks

Source positioning: in straight catheter and in all clinical treatment applicators, with comparison to

manufacturer specification and to actual TPS positioning.

Source dwell time: accuracy over full clinical range.

Source transit time

Source specification data

Manufacturer’s source certificate

Source strength measurement with a well chamber

Source strength measurement with an independent method, e.g. thimble ionisation chamber in-phantom

Data transfer from the treatment planning system (TPS)

TPS algorithm, functionality tests, and source data

Applicator commissioning including labelling, rigidity, consistency with TPS library of applicators and

transmission measurements

Definition of, and baseline results for, routine quality checks

Training materials and documentation

3.3.3.1. Source Movement Profile

Figure 3.2 shows the movement profile, position of the source as a function of

time at 1/25 s resolution imaged with a video camera, between 10.0 and 15.0 mm dwell

positions, which is similar to that between 15.0 and 20.0 mm dwells (not shown). Movement

between all dwells appeared smooth with equal phases of acceleration and deceleration, no

‘overshoot’ in dwell positioning, nor any fine corrections required to achieve the final dwell

position.

Figure 3.3 shows the movement profile on approach to the first dwell, when the

Eckert & Ziegler BEBIG MultiSource® system was operated with (a) new control system

(software version 7.4.1, firmware 4.14.1), and (b) previous version control system (software

version 7.4.0, firmware 4.13.0). The newer software moves the source directly to the first

dwell position, rather than intentionally driving beyond the first position and pausing as

controlled by the old software. The intentional pause was implemented in the previous

software to enable any slack or ‘snaking’ of the cable to be released prior to fine final

45

positioning. The new software implements an empirical correction (applicator and transfer

tube specific) to account for this and achieves the same final positional accuracy. The

maximum recorded source transit speed was 400 (±20) mm s-1 in both data sets. The

maximum speed achieved between dwells of 5.0 mm separation was 63 (±4) mm s-1 with the

new control software, which is consistent with measurements made with the old software at

62 (±4) mm s-1 (Palmer and Mzenda 2009).

The largest positional uncertainty from the video analysis technique was ±2 mm

at maximum source speed, which reduced as the source slowed; this was estimated from the

extent of blurring in the video frames. The final intended source dwell position was always

achieved within ±0.5 mm. The uncertainty in speed is quoted as the quadrature sum of

positional and time uncertainty.

3.3.3.2. Transit Dosimetry

Table 3.10 presents transit doses calculated from video analysis of source

movement compared to the equivalent dose reduction implemented from the Eckert &

Ziegler BEBIG MultiSource® system correction algorithm, at D10 and D20 (10.0 and 20.0 mm

from the centre of the dwell position perpendicular to the source movement axis), for the

previous and updated control software. The magnitude of the actual transit dose at the first

dwell is reduced with the new software due to the absence of the preceding ‘pause’ in

approaching the first dwell, see Figure 3.2. All other actual transit doses are consistent within

experimental uncertainty between the two software versions.

At both D10 and D20 the applied correction for a dwell within a series (15.0 mm

dwell point in this example) is consistent between the two software versions, and is

equivalent to the actual transit dose calculated from the video analysis. The majority of

dwells in a clinical treatment will be mid-series dwells, hence this result is most significant.

The newer control software applies a larger transit dose correction for the first

dwell compared to the old software, 0.45 s compared to 0.115 s, and this is in closer

agreement to the actual transit dose in the cases considered; At D10 the actual transit dose is

2.94 (±0.05) cGy for the new software and 7.45 (±0.05) cGy for the old software, resulting in

transit dose correction errors of -2.6% and 6.0% respectively. At D20 the actual transit dose is

0.91 (±0.05) cGy and 3.00 (±0.05) cGy for the new and old software, resulting in errors of -

0.5% and 2.6% respectively.

The magnitude of the actual transit dose will reduce as the source decays since

the transit time is unchanged but the dose rate from the source will decrease. At D20, for a

dwell point within a series of 5.0 mm separation dwells, as above, there will be a software

time correction (both new and old versions) resulting in an equivalent dose reduction of 0.72

(±0.02) cGy. For a new Ir-192 source the actual transit dose in this case is 0.73 (±0.05) cGy.

However, when the source has decayed to the end of its normal clinical use period, the actual

transit dose is reduced to 0.31 (±0.05) cGy at 3 months or 0.23 (±0.05) cGy at 4 months,

leading to a small overcorrection by the software.

46

Figure 3.2. Position and speed of source during transit from first to second dwell positions, from a

series of three dwells at 10.0, 15.0 and 20.0 mm, with new control software.

Figure 3.3. Position and speed of source during transit from the Eckert & Ziegler BEBIG Multisource®

on approach to the first dwell position at 10.0 mm, with (a) the new control software

and (b) the old control software.

47

Table 3.10. Comparison of actual calculated transit dose and Eckert & Ziegler BEBIG Multisource®

compensated transit dose, for each dwell position in a series of three dwells, at 10.0,

15.0 and 20.0 mm, evaluated at D10 and D20 (10.0 and 20.0 mm from the centre of the

dwell position perpendicular to the source movement axis), for the apparent activity of

a new Ir-192 source. The data is presented for both the newer and older control system

software (see text).

D10, New control software:

Dwell point Treatment unit calculated dwell time reduction (s)

Equivalent dose reduction implemented (cGy) (±0.02)

Actual transit dose (cGy) (±0.05)

Transit dose error ‘actual – corrected’ dose (cGy) (±0.1)

10.0 mm (first) 0.45 5.49 2.94 -2.6

15.0 mm 0.23 2.81 2.66 -0.2

20.0 mm (last) 0.45 5.49 2.65 -2.8

D10, Old control software:





10.0 mm (first) 0.115 1.41 7.45 6.0

15.0 mm 0.23 2.81 2.62 -0.2

20.0 mm (last) 0.115 1.41 2.72 1.3


48


D20, New control software:





10.0 mm (first) 0.45 1.41 0.91 -0.5

15.0 mm 0.23 0.72 0.73 0.0

20.0 mm (last) 0.45 1.41 0.90 -0.5

D20, Old control software:





10.0 mm (first) 0.115 0.36 3.00 2.6

15.0 mm 0.23 0.72 0.73 0.0

20.0 mm (last) 0.115 0.36 0.85 0.5

3.3.3.3. Dwell Position Accuracy with Transfer Tube Curvature

Figure 3.4 compares the effect of source transfer tube curvature on actual dwell

positions between the old and new Eckert & Ziegler BEBIG MultiSource® control system

software. There is a clear and significant improvement in accuracy of the actual compared to

intended dwell positions with the new software. In all but the extreme curvature case (90 cm

displacement), the dwells are accurate within 1.0 (±0.5) mm with the new software,

compared to typical errors of 3.0 (±0.5) mm at 100 mm displacement and up to 6.5 (±0.5)

mm at 300 mm displacement, with the old software.

Previous work by the author (Palmer and Mzenda 2009) proposed that dwell

position errors due to transfer tube curvature are the result of the drive cable taking the

outer radius of the source transfer tube during the outward drive motion, as well as potential

‘snaking’, and then taking up this ‘slack’ during distal-to-proximal stepping of the source

during the initial few dwells; the drive cable taking the internal radius on return to the Eckert

& Ziegler BEBIG MultiSource® unit. The new control software has mitigated this error by using

empiric applicator factors to account for resistance and moving properties of the source

cable, for each applicator and transfer tube combination (Spiller 2012). An unintended dose

enhancement at the first dwell position can be seen on the autoradiograph with the old

software, which is no longer present with the new software, Figure 3.4. This is due to the

elimination of the transit pause prior to the first dwell.

49

Figure 3.4. Autoradiographs of actual source dwell positions compared to planned positions

(vertical lines) as a function of curvature of the transfer tube, for the 1400 universal

applicator (140 cm length), both for the previous EZ BEBIG Multisource® control system

software, “old”, and the latest software, “new”. The autoradiographs are for a straight

transfer tube and for bends induced by displacements of the distal end by 10, 30 and 90

cm towards the treatment unit.

3.3.4. Treatment Planning Study of Simulated Source Positioning Errors

Figure 3.5 shows a 3D projection of the applicator, HR-CTV and OARs in order to

provide context of a typical HDR brachytherapy cervix treatment to the presented results.

The close proximity of the target and critical radiosensitive normal structures confirms the

need for accurate source dwell positioning.

An illustration of the isodose shifts resulting from the simulated source

positioning errors for a typical cervix treatment plan is shown in Figure 3.6. In the sagittal

projection, Figure 3.6(a), noticeable isodose shifts can be seen across the 2, 5 and 10 mm

magnitudes at the tip of the IU applicator, and at the location of the bladder balloon and

50

rectum. In the coronal projection, Figure 3.6(b), no effect on isodose position is observed

even for the 10 mm shift in the region of the Manchester Point A location, which is

traditionally the dose prescription point for cervix treatment.

Figure 3.7 presents the DVH for the bladder, rectum and HR-CTV for the case in

Figures 3.5 and 3.6, with intended dose delivery and with simulated dwell position errors. It

is problematic to infer radiobiological implications of small changes in DVH curves, but it is

apparent that changes occur at 2 mm shift and are significant at 5 and 10 mm shift in terms

of the changed dose to structures.

Figure 3.8 represents combined data from eight cervix patient plans with

simulated errors, showing the mean and interquartile range, for the dose received by 90% of

the HR-CTV, D90, and the maximum dose received by a 2 cm3 volume, D2cc, of the OARs

bladder, rectum and sigmoid. There is an approximate linear relationship; shifts of 1.0 mm

result in a 2.0% change, and 4.0 mm in 10.0% change. Point A exhibits less change than DVH

metrics, 0.4% and 0.7% respectively. Bladder doses increase while rectum, sigmoid and HR-

CTV decrease due to the proximal shift direction of dwells. This would be reversed for distal

shifts.

Table 3.11 provides detailed DVH data for simulated dwell position errors in the

range 0.2 to 2.0 mm, for two error modes (a) systematic proximal shift of all dwell positions,

and (b) source cable take-up lag, in which the first dwell position is unaffected and all others

are shifted distally (Palmer and Mzenda 2009). In addition to D90, the volume receiving 100,

150 and 200% of the prescribed dose is provided for the HR-CTV. The maximum dose received

by 1 and 2 cm3 of OARs bladder, rectum and sigmoid are provided. Only at ±0.2 mm accuracy

between simulated and intended dwell positions were the effects on all considered

parameters within 1.0%. An accuracy of at least ±1.0 mm was required to limit the dose

uncertainty on the most important clinically relevant parameters, D90, D2cc, to within an

acceptable level of 3.0%.

51

Table 3.11. Effect of dwell position shifts, 0.2 to 2.0 mm, on DVH treatment plan metrics, modes (a)

and (b) see text.

Dwell position

shift

Mean percentage change in GEC-ESTRO DVH parameters for eight cervix plans

HR-CTV Bladder Rectum Sigmoid

Mode Distance

(mm)

D90 V100 V150 V200 D2cc D1cc D2cc D1cc D2cc D1cc

(a) 0.2 0.0 -0.1 0.0 0.1 0.6 0.9 -0.5 -0.4 -0.5 -0.3

0.5 -0.3 -0.3 -0.1 -0.4 1.5 2.1 -1.1 -0.8 -0.9 -1.1

1.0 -2.0 -1.0 -0.7 -1.1 2.6 2.8 -2.7 -2.9 -1.8 -2.6

2.0 -4.2 -2.3 -2.0 -1.9 5.3 5.6 -5.0 -5.7 -3.4 -3.6

(b) -1.0 0.9 0.6 0.7 0.3 -2.4 -1.9 1.4 4.0 1.4 1.7

-2.0 2.0 1.3 0.7 0.5 -4.6 -3.9 2.8 6.0 3.0 3.4

Figure 3.5. 3D projection illustration of the treatment applicator guide tubes (red), the clinical

target volume (orange), the 100% isodose surface (purple) the bladder (green), rectum

(yellow) and sigmoid (blue). The Manchester Point A prescription location is indicated

by a yellow marker. View is from patient right to patient left.

52

Figure 3.6. The effect on a typical cervix plan isodose distribution (40% to 200% normalised

prescription isodose lines shown) of simulated error dwell position shifts in (a) sagittal

and (b) coronal projections. The larger images shows the intended isodose distribution,

the inset detail shows the effect of dwell position shifts. The yellow crosses indicate the

Manchester Point A traditional dose prescription points.

53

Figure 3.7. Effect on DVH curves for a typical cervix patient plan for HR-CTV and OARs of simulated

dwell position errors of 1, 2, 5 and 10 mm proximal shift.

Figure 3.8. Effect of systematic proximal dwell position shifts, 0.2 to 6.0 mm, on clinical DVH

treatment plan quality parameters. Data is the mean and interquartile range from eight

cervix plans, for HR-CTV target and maximum dose to OAR D2cc.

54

3.4. Discussion and Conclusions

3.4.1. UK Survey

The survey data presented in this report represents the current practice in the

majority of brachytherapy centres within the UK (at the survey date of June 2012). Whilst

there is a high level of consistency in inclusion, frequency and tolerance values for some tests,

such as source strength measurement, there are varied responses to other tests. This is likely

due to differences in local planning and treatment procedures in clinical use, availability of

equipment, and differing functionality or performance of equipment. Local assessment of

QC needs is essential in determining schedules, rather than simple reliance on the ‘majority

view’. However, benchmarking against accepted practice is a good starting point for local

review. A risk assessment approach including local known factors is advocated for final

decisions on QC testing. All schedules must include measurement of source strength, source

position and dwell time, but the specific details require knowledge of local clinical practice

and equipment in use. There are difference in the published recommendations for QC testing

frequencies and tolerances, shown in Tables 3.6 to 3.8. This may in part be attributed to the

range in publication dates of the guidance documents, as equipment, clinical techniques and

knowledge evolved over time (AAPM published in 1997, IPEM in 1999, ESTRO in 2004, CAPCA

in 2006, and ACPSEM in 2013).

While there was some variation in staff groups involved in QC testing between

centres, physics staff most commonly performed all of the QC tests except daily facilities

testing, which was almost exclusively performed by radiographers. Within each centre, a

specific QC test may be performed in multiple ways, including staff group involved,

equipment used, frequency of measurement, and tolerance value. Each different

measurement method has been included in the results table. For example, ‘decay correction

accuracy at treatment unit’ may be performed prior to each patient treatment by

radiographers, and separately by radiotherapy physics after each source change but to a

tighter investigation tolerance level. The achievable tolerance value for this test is also

dependent on the equipment design, whether the software makes hourly corrections for

source decay or 12-hourly for example. The standard operating procedures of individual

departments also have a significant effect on the QC testing that is performed. This includes

all aspects such as whether optimised or standard/tabulated planning is used, whether 2D or

3D imaging is utilised, and whether electronic transfer of data is available. An independent

method for the verification of the accuracy of treatment plans is required for individually-

optimised treatments, but may not be required for each patient if a standard plan is used

that has previously been verified and is checked for consistency.

Some tests are adopted by all centres such as ‘source strength measurement’

and ‘source position in a straight catheter’. However others such as ‘x-ray imaging of

applicators’ is undertaken by only 32% of centres. The difference may be attributed to

whether the process is already being assessed by alternative means, and there is some

evidence from the survey to support this. For applicator dimensions and angles, a specific

measurement may not be necessary if the consistency of shape is evaluated through

55

agreement to planning system library applicators used for each individually-planned

treatment (provided both physical applicator and library applicator have already been tested

at commissioning). Checking of the 1st dwell position should however be verified in this case.

The number of centres including a routine measurement of actual source dwell

positions in clinical applicators was surprisingly low, at 17 centres (55%). Such testing should

be performed at commissioning and at regular intervals, comparing the actual and TPS

planned dwell positions.

IPEM Report 81 (Mayles et al. 1999) was the most frequently cited document

used for guidance on required HDR or PDR QC tests, but it is surprising that only 61% of UK

centres cited this document, being the UK professional body’s recommendations for QC. This

may be because the document is now quite dated. More recent, but again surprisingly cited

by only 48% of centres is the ESTRO Booklet No. 8 (Venselaar and Perez-Calatayud 2004).

There is a large range in the documents identified by individual centres as their primary

sources of guidance for QC testing, supporting the need for an update of IPEM Report 81 in

the UK.

A benchmark data set of brachytherapy HDR and PDR QC testing has been

presented which is representative of practice across the UK. This updates a previous survey

conducted over twenty years ago.

3.4.2. Treatment Unit Commissioning and QC: Dwell Position Accuracy and Transit

Dose

The results of this work demonstrate how a change in equipment software can

significantly affect the performance of a medical device. The need for robust quality control

checks after software changes is therefore reinforced, either to confirm improvements or to

detect any unexpected changes in performance.

In the HDR brachytherapy treatment unit case presented, the accuracy of dwell

positions with transfer tube curvature, transit dose at the first dwell, and transit dose

corrections have been markedly improved with the latest internal operating software. This

was following discussions between the author and the manufacturer.

3.4.3. Equipment Performance Requirements

The study on simulated dwell position errors has shown even small deviations

in source position can have a demonstrable impact on 3D dose delivery parameters. Only at

±0.2 mm accuracy between actual and intended dwell positions were the effects on all

considered parameters within 1.0%. An accuracy of at least ±1.0 mm was required to limit

the dose uncertainty on clinically relevant parameters (D90, D2cc) to within an acceptable

level of 3.0%. While the adoption of a QC action level of ±0.5 mm is perhaps indicated from

the results, taking account of all other uncertainties in brachytherapy and practical

limitations, a ±1.0 mm maximum tolerance is adequate to ensure equipment performance

56

does not adversely impact on dose delivery accuracy to target and OARs. A 2.0 mm maximum

tolerance (as prescribed in ACPSEM, CAPCA, ESTRO guidance, Table 3.6) is too large according

to the results of the presented work. Instead, it is proposed that ±0.5 mm action level and

±1.0 mm maximum tolerance level (as prescribed by AAPM, Table 3.6) for source position

accuracy in straight or simple curvature applicators (IU tube or ovoids) is appropriate for

modern brachytherapy. It is accepted that the mechanical performance in ring applicators

may be insufficient to achieve this suggested tolerance, however custom dwell-paths should

be used to account for any deviations, where possible.

The potential impact of dwell position errors on dose delivery accuracy has been

demonstrated in this work for simulated inaccuracies (position calibration errors). The

magnitude of dose uncertainties observed is generally applicable but actual values in each

clinic will be dependent on the specific HDR applicator, HDR equipment, and control system

software in use.

Importantly, the results demonstrate that measurements of dose at the

traditional prescription Point A, are insensitive to dwell position errors. This is because the

isodose distribution is relatively uniform at this point along the axis of the treatment

applicator. QC measurements of dose at Point A are not sufficient to infer accurate intended

dose delivery at other locations in the 3D dose distribution, particularly at the distal end of

the applicator, and at the bladder and rectum OAR positions. An improved system for QC of

HDR dose delivery is required rather than simple measurements of dose in the vicinity of

Point A. The same conclusion is drawn for dosimetric audit measurements of point dose or

dose distribution. An independent dose assessment in the region of Point A is likely to be

insensitive to all but gross errors in source positioning accuracy. A more comprehensive

assessment of dose distribution around treatment applicators would be required to be

sensitive to dwell position errors.

3.4.4. Future Directions for Quality Control Testing of HDR Brachytherapy

Based on the developments that have taken place in brachytherapy treatment

techniques in recent years, comparisons to external beam QC and judgement of value-added

testing, it is the author’s opinion the tests in Tables 3.7 and 3.8 should be considered by

brachytherapy physicists for inclusion in routine QC schedules for brachytherapy equipment.

Transit dose has therefore been studied in Section 3.3.3.2, and a focus of the remainder of

this thesis is concerned with improved verification that any inaccuracy in source positions in

clinical applicators has an acceptable impact on measured dose distributions, via an ‘end to

end’ test. This is applied for the purposes of dosimetric audit in this thesis, but could also be

adopted for local routine QC.

There is a general need to review fundamental approaches to QC in

radiotherapy, improving its effectiveness and reducing its resource-intensive burden, which

is likely to include moving from an exclusively device-centric approach, to a more system-

based evaluation of quality confirmation. A recent ‘point/counterpoint’ debate published in

the journal Medical Physics hypothesised that “QA procedures in radiation therapy are

57

outdated and negatively impact the reduction of errors” (Amols and Klein 2011). The planned

‘end to end’ audit for brachytherapy is consistent with this proposal. In the joint American

societies document ‘Safety is No Accident’ (ASTRO 2012), it is recommended for external

beam radiotherapy to undertake “end-to-end testing for representative treatments,

performing the entire process, with dosimetric or other quantitative tests that can be

evaluated at the end of the test to confirm accurate delivery of the planed treatment”.

Although not stated in the document, it is a reasonable assumption that an ‘end to end’ test

would be as valuable in brachytherapy as for external beam radiotherapy.

58

59

CHAPTER 4

Candidate Dosimeters for

Brachytherapy Applicator

Dosimetry Audit

60

4.1. Requirements of Dosimetry Systems for Brachytherapy Audit

In selecting an appropriate detector for brachytherapy dosimetry, the nature of

the dose distributions to be measured must be considered. Brachytherapy uses radioactive

sources of several mm length inserted within or very close to the tumour. Because of the

small source size and its proximity to the treatment volume there is rapid fall-off of dose

around the radioactive source. The measurement of radiation dose distributions around

clinical brachytherapy sources, both isolated and within treatment applicators, is therefore

a challenging endeavour. The volume of potential interest for an audit contains orders of

magnitude variation in dose levels, extremely steep dose gradients near the source, around

6% mm-1 at the edges of the target volume in typical cervix cancer treatments, and low dose

rates further away (Viswanathan and Thomadsen 2012). Conventional measurement

techniques in radiotherapy often involve relatively large detectors, but when these are used

close to brachytherapy sources they introduce inaccuracies due to dose averaging over the

active volume: A 0.6cc ionisation chamber will exhibit a non-uniformity correction of around

10% at 20 mm and 30% at 10 mm from a brachytherapy source (Tolli and Johansson 1993,

Majumdar et al. 2006).

Perez-Calatayud et al. (2012a) defined four characteristic requirements of

detectors for brachytherapy: wide dynamic range; small active volume, flat or well-

characterized energy response; sufficient precision and reproducibility to permit dose rate

measurements, quoting k=1 Type A (statistical) uncertainties ≤ 3% and k=1 Type B (systematic

or calculable) uncertainties ≤ 6%. These requirements are more demanding than those

previously quoted by Rivard et al. (2004) of ≤ 5% and ≤ 7% respectively. For routine

brachytherapy quality control measurements in clinics or interdepartmental audit, one

should also consider practicalities, cost and availability of the dosimetry system. Perez-

Calatayud et al. (2012a) also make recommendations of phantom material for brachytherapy

dosimetry. Consequently, dosimeters should be used in test objects manufactured from

appropriate materials such as Solid Water (RMI457, Gammex, Middleton, WI, USA), or liquid-

water tanks, of sufficient volume to provide full scatter conditions, or appropriate corrections

made.

For the purposes of brachytherapy dosimetry audit, the measurement of dose

around clinical treatment applicators could be accomplished by dosimetry system operating

in one of three different sampling modes: (1) multiple point detectors, (2) one or several 2D

detectors, or (3) full 3D dose measurement. The considerations for selection of an optimum

measurement system are not related solely to the performance of the detector in the

particular brachytherapy dose field, as may be of primary concern in a laboratory setting, but

practicalities of the audit measurement process, at many sites across the UK, is relevant; such

as cost, robustness, transport convenience, and potential read-out delay. This study

compares the three potential measurement sampling modes (multiple point, 2D and 3D), by

selecting and evaluating one dosimetry system in each category. The chosen detectors are

(1) Ge-doped optical fibres as thermoluminescent dosimeters; (2) the latest Gafchromic EBT3

radiochromic film with advanced triple-channel optical scanning and analysis; and (3) Presage

radiochromic plastic with optical CT scanning. These detectors were chosen as potential

61

candidates for high spatial resolution measurement of dose distributions in brachytherapy

audit around treatment applicators, in the setting of a dosimetry audit visit. Optical fibres

and Presage have very limited prior application in brachytherapy. While radiochromic film

has been widely used in brachytherapy, the new EBT3 film with triple-channel analysis has

not. In order to validate each dosimetry system, an investigation of the ability of each

dosimeter to measure the dose (or signal proportional to dose) as a function of radial

distance from an HDR source was undertaken, with comparison to Monte Carlo source model

calculation as a gold-standard. Each system was then used to sample the dose distribution

along one or more lines perpendicular to three brachytherapy treatment catheters, with

source dwell positions and a dose plan typical of cervix cancer treatments. This tested the

dosimetry systems’ response to typical clinical dose distribution measurement situations and

was compared to a brachytherapy treatment planning system (TPS) calculation as a gold-

standard.

4.2. Methodology

All HDR brachytherapy exposures were performed with the Eckert & Ziegler

Bebig GmbH MultiSource® HDR treatment unit with a Co-60 source. Comparative Monte

Carlo derived TG-43 data (Rivard et al. 2004) of the Co-60 source was obtained from data

tables published by Granero et al. (2007). Treatment planning system (TPS) dose calculations

were made with the Eckert & Ziegler Bebig GmbH HDR Plus® software, utilizing a TG-43 based

dose calculation. The TPS plans used for dose distribution measurement comprised three

parallel catheters to deliver typical ‘pear-shaped’ isodose curves used for cervix treatments

(Viswanathan and Thomadsen 2012). The line of dosimetric measurement was orthogonal to

the catheters at the proximal end of the high dose region. The plan used for Presage

measurement was modified to include additional modulation within the measurement

volume by the removal of several dwells in the central catheter.

4.2.1. Dosimetry Systems

4.2.1.1. Doped Silica Glass Optical Fibres

The optical fibres used in this work were described by Bradley et al. (2012) and

consist of Ge-doped SiO2 telecommunication fibre (CorActive High Tech., Canada). Dosimeter

rods were cut to 5.0 ± 0.5 mm lengths from a continuous reel of fibre optic cable, and their

protective coating removed with an optical fibre stripper. A conventional heating cycle and

annealing process was utilised with a standard TL reader system; preheat 160 °C for 10 s,

readout 300 °C for 25 s, heating rate 25 °C s-1. The sensitivity of each dosimeter rod is

described as the TL yield per unit dose per unit mass, and is expected to vary along the length

of the fibre optic cable. Fibre dosimeters of a 3.5% sensitivity range were selected via a

screening process in which large sample groups were irradiated with a dose of 3 Gy from a

linear accelerator of nominal 6 MV photon energy, at 5 cm depth in a water phantom, and

the TL yield from each individual dosimeter separately read out.

62

Because of their small physical size, it was possible to use a group of seven

dosimeter rods at each measurement position (a central rod surrounded by a circle of six

other rods) to reduce uncertainty, with the resulting data value being the arithmetic mean

TL yield of all fibres at a particular location. It was not considered necessary to establish an

absolute calibration of the fibres for the scope of this work, and instead they have been used

to give normalised relative dose results. There are also practical difficulties in absolute dose

measurement with fibres including TL fading.

4.2.1.2. Radiochromic Film

All radiochromic film measurements were performed with Gafchromic EBT3 film

(Ashland ISP Advanced Materials, NJ, USA) from the same batch (Lot #A12141101). The

procedure summary recommendations for handling radiochromic film as defined in AAPM

TG-55 (Niroomand-Rad et al. 1998) were adopted. All EBT3 films were scanned in red-green-

blue (rgb) format using a 48-bit (16-bit per channel) scanner (Epson Perfection V750 Pro) at

72 dpi, in transmission mode, with no colour or sharpness corrections and consistent

orientation and time-since-exposure protocol. A nominal 6 MV linear accelerator, with

calibration traceable to a primary standard at the National Physical Laboratory, Teddington,

UK, was used for the film calibration. Ten separate film strips of 10 cm x 5 cm were exposed

to ten doses in the range 0 Gy to 50.8 Gy, positioned on the central axis in a 10 cm x 10 cm

field at 5 cm depth in Solid Water (RMI457, Gammex, Wisconsin, USA). The average film pixel

values in a 4 cm x 4 cm region centred on the axis of the beam were used to derive the

average film response at each dose level, establishing a calibration of the film as a function

of the three colour channels. Test images were converted to dose maps using a rgb triple-

channel algorithm via FilmQAPro® software (Ashland ISP Advanced Materials, NJ, USA,

version 2.0.4631.19736) using the method proposed by Micke et al. (2011). This multi-

channel method enables the separation of the scanned signal into a dose-dependent part

and a dose-independent part that enables corrections for a variety of disturbances including

non-uniformities in the film active layer and artefacts from the film scanner. All HDR test films

were exposed contemporaneously with a suitable dose reference patch at the linac, to a

known dose level approximately 75% of the expected test film maximum dose, from the same

film batch. This allowed linear scaling of film calibration functions to correct for any residual

time-since-exposure darkening kinetics of the film calibration and any inconsistencies in the

film scanner compared to the calibration scans (Lewis et al. 2012). 75% was chosen to force

consistency at an appropriate dose level (close to the prescription dose) within the large dose

range of the brachytherapy dose distributions.

63

4.2.1.3. Solid Radiochromic Polymer

Cylindrical Presage® samples (Heuris Inc. LLC, Skillman, NJ, USA) of 60 mm

diameter and 100 mm height were used in this study. A parallel beam optical CT readout

system, described by Krstajic and Doran (2006) was utilised, with samples held within an

optical matching fluid tank during rotation and image acquisition. 400 projections were

acquired for each sample and reconstructed by filtered back-projection into images of matrix

size 256 x 256 using software written in IDL by S.J. Doran at University of Surrey

(unpublished). A region of the Presage sample that approximates to being un-irradiated,

distant from the HDR brachytherapy sources, was used for background subtraction. The

resulting Presage signal was then expected to be directly proportional to the dose received.

The optical density response was known to evolve with time and in a fashion dependent on

storage temperature. Since it was not the primary purpose of this comparative study to

provide an absolute dose measurement, no dose calibration was performed. Instead relative

dose-proportional distributions were measured, the suitability of this method having been

demonstrated by Wai et al. (2009). 10 Presage cuvettes were irradiated in the range 0 Gy to

20 Gy using a nominal 6 MV linear accelerator, in order to verify a linear dose-response

relationship and determine an optimum dose response level for the particular Presage

formulation.

4.2.2. Test Objects and Irradiation Conditions

A Solid Water (RMI457, Gammex, Middleton, WI, USA) test object was purpose-

designed for this work, as shown in Figure 4.1. Interchangeable slabs were constructed, with

a lower block containing either a single or three catheter channels and an upper block either

flat to hold film or with cavities for optical fibres. Fibre cavities were offset from the plane

containing the catheter to ensure no perturbation effect from each measuring position on

the others. This test object was used for treatment plan dose distribution measurements

using optical fibres and film, and for the single source measurement with fibres. 13 cavities

were located each side of the single catheter at radial distances of between 5 and 140 mm,

to match standard TG-43 tabulated data resolution. Measurements of radial dose from a

single source using film were conducted by carefully attaching a catheter along a film edge,

held in a water tank, shown in Figure 4.2. For Presage measurements, no external test object

was required and either one or three cavities were machined directly into the Presage

material samples. Figure 4.3 is a photograph of a Presage sample containing three cavities

with inserted HDR brachytherapy treatment catheters. The Presage samples, Solid Water test

objects, and ‘single catheter with film’ were all placed at the centre of 120 litre cubic liquid

water tank during irradiation to ensure full scatter conditions.

64

Figure 4.1. Schematic diagram of Solid Water test object, shown with (a) lower slab containing one

catheter and upper slab with optical fibre cavities, and (b) lower slab with three

catheters and upper slab holding radiochromic film. In all cases, the measurement plane

is 15 mm from the source catheter plane.

Figure 4.2. Measurement of radial dose from a single HDR source, within a plastic catheter, using

EBT3 film, secured in a Perspex frame within a full scatter water tank.

65

Figure 4.3. Presage sample machined with three cavities each containing an HDR brachytherapy

catheter for a typical cervix treatment dose distribution irradiation. The grey marker

indicates the approximate location of the plane of considered dose profiles.

4.3. Results

4.3.1. Initial Processing and Calibration of Dosimeters

Screening of the TL yield from optical fibres showed considerable variation

between individual dosimeters, up to 60%, attributed to variations in doping concentration

and other non-uniformities in the fibres. To achieve a batch variation of 3.5% deviation from

a nominal mean, a rejection rate of 75% of fibres was required. It was impractical to achieve

an improved consistency of sensitivity due to the number of fibres required: 7 fibres were

used at each of 13 measuring positions, with three experiment runs. A total of 273 fibres

were used, selected for consistency from an initial set of 1500 fibres. For EBT3 Gafchromic

film calibration, the multi-channel dosimetry method was successfully utilised to establish a

dose-response relationship in three colour channels over a range 0 to 50 Gy with a rational

fit function. For Presage dosimetry, optical spectral analysis of ten cuvettes was used to verify

the absence of any saturation in the range 0 Gy to 20 Gy. Initial results indicated a linear

dose-response relationship.

66

4.3.2. Isolated Source Radial Dose Measurements

The response as a function of radial distance from the centre of the Co-60 HDR

source measured with optical fibres, EBT3 Gafchromic film and Presage is given in Figures 4.4

to 4.6 and compared to source model data from Monte Carlo calculation. The ordinate data

have been multiplied by the radial distance squared to remove the inverse square law

dependence to improve visualisation. Uncertainty analysis is given in Section 4.3.4.

Figure 4.4 presents three repeated measurements of dose as a function of radial

distance from a Co-60 HDR source using the TL yield from Ge-doped optical fibre dosimeters,

compared to Monte Carlo source model data, all normalised at 20 mm from the source.

Different sets of fibre dosimeters were used in each measurement, and each data point is

the arithmetic mean of 7 fibres. There is good agreement between optical fibre and Monte

Carlo data in the range 5 mm to 40 mm, the mean relative TL yield being within 8% of Monte

Carlo data. There is a 4 to 9% repeatability variation in the fibre response (magnitude of

absolute difference (globally normalised response), rather than relative difference (local

percentage difference). There is a consistent over-response of the fibres at lower dose levels,

apparent at greater than 60 mm from the source, equivalent to less than 14 cGy in this data,

with the magnitude of discrepancy increasing with increasing distance from the source, i.e.

reduced dose-rate and dose level (20% over-response at 80 mm, 30% at 140 mm).

Figure 4.4. TL yield from Ge-doped optical fibres as a function of radial distance from the centre of

a Co-60 HDR source, compared to Monte Carlo data (Granero et al. 2007), all normalised

at 20 mm, showing repeatability from three separate measurements. (Vertical blue lines

indicate error bars, expanded uncertainty estimate at k=2, shown for trial 1 only for

clarity).

67

Figure 4.5 presents three repeated measurements of dose as a function of radial

distance from a Co-60 HDR source using Gafchromic EBT3 film, compared to Monte Carlo

source model data, normalised at 20 mm from the source. There is good agreement between

EBT3 film and Monte Carlo data over the range 3 mm to 60 mm, the mean film dose being

within 5% of Monte Carlo data. There is a 1 to 3% repeatability variation in the film response

(absolute difference). There is no consistent over- or under-response of the film within 3 mm

to 50 mm from the source, with some over-response at greater distances (equivalent to less

than 20 cGy in these results), in comparison to the Monte Carlo data.

Figure 4.5. Dose as function of radial distance from the centre of a Co-60 HDR source measured

with EBT3 Gafchromic film, with comparison to Monte Carlo source data (Granero et al.

2007), MC normalised to the average film dose at 20 mm, showing repeatability from

three separate measurements. (Vertical blue lines indicate error bars, expanded

uncertainty estimate at k=2, shown for trial 1 only for clarity).

Figure 4.6 presents a measurement of dose as a function of radial distance from

a Co-60 HDR source using Presage, compared to Monte Carlo source model data, all

normalised at 20 mm from the source. The data points are (circular) radial averages through

the reconstructed 2D plane centred on the source. The distance range is reduced compared

to Figures 4.5 and 4.6 due to physical size constraints of the Presage sample and optical-CT

equipment scanning field of view. There is good agreement between Presage and Monte

Carlo data in the range 4 mm to 22 mm, being within 8% of Monte Carlo data (percentage

68

difference). There is an expected under-response of the Presage signal at distances beyond

22 mm from the source, which relates to a combination of a low signal-to-noise ratio (random

error) and the well-known “wall effect” of imperfect refractive index matching between the

sample and the surrounding bath of index-matching fluid.

Figure 4.6. Presage signal as a function of radial distance from the centre of a Co-60 HDR source,

with comparison to Monte Carlo source data (Granero et al. 2007), all normalised at 20

mm. A single data set is presented, incorporating a radial average centred on the source.

(Vertical blue lines indicate error bars, expanded uncertainty estimate at k=2).

4.3.3. Dose Distribution Measurements for Multi-Dwell Treatment Plans

Figures 4.7 to 4.9 present measurements using optical fibres, EBT3 film, and

Presage, respectively, along lines through typical clinical dose distributions involving multiple

HDR source dwell positions in three catheters, compared to treatment planning system (TPS)

calculated dose points.

Figure 4.7 shows optical fibre TL yield measurements compared to TPS

calculated data, normalised at 30 mm. The TL yield reproduces the general shape of the TPS

dose profile but with an apparent under-response at the central high dose region. There is

also an over-response within the low dose region which is consistent with the data presented

in Figure 4.4 for the single source case. The discrepancy within the high dose region may be

explained as a consequence of the physical size of the optical fibres and a volume-measuring

69

effect within a high dose-gradient region of the treatment dose distribution (average 3% mm-

1). In the high dose region, the dose gradient led to a change from 8 Gy to 9 Gy along the 5.0

mm length of the optical fibre and a calculation of the average volume dose across the optical

fibre indicates agreement with the measured values. The dose gradient in the low dose

regions was negligible along the length of the fibres. Figure 4.7 has been normalised at 30

mm as a reasonable balance between acceptable dose gradient and sufficient signal strength.

Figure 4.7. TL yield from Ge-doped optical fibres as a function of distance across an HDR dose

distribution, compared to treatment planning system (TPS) calculation, in a plane 15 mm

from three source catheters with multiple source dwell positions. TL yield has been re-

normalised to the TPS calculated dose level at 30 mm. (Discussion of uncertainty in data

values given in the text; error bars indicate expanded uncertainty estimate at k=2).

Figure 4.8 presents an EBT3 film measurement of dose distribution, showing

excellent agreement with TPS calculated data, within 1.5% measurement uncertainty across

the plateau high dose region. In the steep gradient and lower dose regions, the EBT3 derived

dose is indistinguishable from the planning system calculation. The EBT3 dose has not been

scaled to match the TPS data and represents the un-normalised absolute dose as determined

by the multi-channel film analysis technique.

70

Figure 4.8. Dose measured with EBT3 Gafchromic film as a function of distance across an HDR dose

distribution, compared to treatment planning system (TPS) calculation, in a plane 15 mm

from three source catheters with multiple source dwell positions. No re-normalisation

of either EBT3 or TPS calculated dose has been performed. (Vertical blue lines indicate

error bars, expanded uncertainty estimate at k=2).

Figure 4.9 presents results of a Presage measurement of the dose distribution

from multiple HDR source dwells in three catheters. A reduction in the abscissa range

compared to Figures 4.7 and 4.8 is required due to limitations of the physical size of the

Presage sample. A different treatment plan is used compared to film and optical fibre

measurements, to ensure sufficient modulation of the dose distribution within the measured

region. Taking advantage of the full 3D data set available, line profiles at several distances, 5

mm to 20 mm orthogonal distance from the source catheters are analysed in comparison to

TPS data. The Presage-derived normalised dose distribution data at all considered distances

from the catheters is in good agreement with the general shape of the TPS calculation,

although there is significant noise or artefacts present in the measurement data, leading to

errors of between 3% and 14% across the curves. The principal systematic error in the

Presage measurement occurs for the profile acquired 5 mm from the source plane, where

there is a significant deviation in the region between the source catheters. Inspection of the

original optical CT image data reveals a significant reconstruction artefact along, and

immediately adjacent to, the line between the catheters. This was caused by perturbation of

the light transmission through the catheter walls.

71

Figure 4.9. Dose measured with Presage as a function of distance across and HDR dose distribution,

compared to treatment planning system (TPS) calculation, in planes at distances

between 5 mm and 20 mm from three source channels with multiple source dwell

positions. Presage value has been re-normalisation to the TPS calculated dose level at 0

mm. (Error bars indicate expanded uncertainty estimate at k=2, shown for one data set

only, for clarity. Note, the treatment plan used in Figure 4.8 is different from that used

in Figures 4.6 and 4.7, see text).

4.3.4. Uncertainty Analysis

The primary sources of uncertainty using the experimental methodologies in

Section 4.3.2 are presented in Table 4.1, quoted at a coverage factor of 2 (k=2), for the case

of isolated source radial dose rate measurements. This analysis has been conducted following

the recommendations by DeWerd et al. (2011). The calculated dose uncertainty due to

detector-to-source positional uncertainty is modified by the inverse square law resulting in a

variable uncertainty with distance from the source. An estimate of the combined uncertainty

has been calculated using a simple quadrature sum of individual components for each

dosimetry system: The combined expanded uncertainty (k=2) for the optical fibre dosimeter

is estimated as 14% at a distance from the source of 10 mm (high dose gradient) and 8.3% at

30 mm from the source (low dose gradient). For EBT3 film dosimeter the values are 7.1% and

2.3%, and for the Presage dosimeter are 8.2% and 6%, at 10 mm and 30 mm respectively (all

k=2). For measurements in typical dose distributions, Section 4.3.3, the uncertainty of source

to detector distance has less impact due to reduced dose gradients, compared to the single

isolated source case.

72

Table 4.1. Uncertainty budget for optical fibre, Gafchromic film, and Presage, for the experimental

methodology used in this work, for isolated source radial dose rate measurement.

Uncertainty of activity distribution within the source, as discussed by DeWerd et al. (2011), and uncertainties

in the determination of absolute dose (such as energy dependence) are omitted from this budget analysis.

a. Effect of positional uncertainty on relative dose rate. Includes estimate of all distance-related uncertainties

including source position within catheter (0.2 mm), position of dosimeter from catheter, and readout position

within the dosimeter, where applicable, (the latter two combined as 0.6 mm for optical fibre, 0.3 mm for film,

and 0.2 mm for Presage) (all k=2).

b. Statistical evaluation of multiple individual fibres through batching process.

c. Local experience of TL reader.

d. Film scanning and analysis method proposed by Micke et al. (2011) has been adopted to reduce uncertainties

resulting from film non-uniformities, darkening kinetics and scanning process variations, validated in

brachytherapy applications by Palmer et al. (2013b).

e. Local experience of Presage and optical CT equipment.

f. While not included for the single source radial dose rate measurement, it is worth noting that for multiple

catheters within the Presage, increased uncertainty will be present up to 50% along the line directly between

source catheters, reducing to 4% at 8 mm perpendicular distance. This type of reconstruction artefact is well

known in x-ray CT, manifesting as a streaking artefact in the presence of x-ray opaque inclusions.

Dosimetry system

Source of uncertainty Type Expanded uncertainty (coverage k=2)

Notes

Optical fibre Source to detector distance B 12% at 10 mm from source, 4% at 30 mm from source

a

Variation in TL yield dose sensitivity between fibres

A 7% b

Variation in annealing regime and readout

B 2% c

Gafchromic film

Source to detector distance B 7% at 10 mm from source, 2% at 30 mm from source

a

Film calibration fit function B 1% d

Film scanning process A 0.4% d

Presage Source to detector distance B 6% at 10 mm from source, 2% at 30 mm from source

a

Sample consistency B 4% e

Optical CT readout (single catheter)

B 4% e, f

73


4.4.1. Doped Silica Glass Optical Fibres

The variation in TL yield for individual Ge-doped optical fibres was significant

and required a large percentage rejection rate (75%) to achieve a reasonable batch size of

similarly responding dosimeters (±3.5%) from a nominal mean value. However, the

‘discarded’ fibres can themselves be grouped around other means and hence multiple

batches of similarly responding dosimeters can be obtained from the preparation process.

Fibres can of course also be re-used multiple times once the batching process has been

completed. However, the inherent variation between fibres leads to an unacceptable

uncertainty and reproducibility of measurement for practical batch sizes from an initial

sample set. To mitigate, several fibres were used at each measuring position and an

arithmetic mean response used. However, this multiple dosimeter approach is not an ideal

experimental proposal, compromises the attractive spatial resolution, and leads to additional

uncertainties in position. The fibres are fragile and difficult to manipulate and position

accurately due to their small size.

For measurements of dose as a function of radial distance from an isolated HDR

source, optical fibres provide a reasonable representation of the Monte Carlo source model

data, Figure 4.4, especially at higher dose levels. In this experimental measurement, the

fibres were placed orthogonal to the radial source vector, and hence the smallest physical

dimension, and therefore highest resolution, was aligned with the dimension of interest.

When used in this mode, optical fibres demonstrate a significant spatial resolution advantage

over other detectors. However, when used to measure the dose distribution in a typical

clinical brachytherapy treatment, Figure 4.7, there was a significant dose gradient in three

dimensions close to the source, and hence the longer dimension of the fibres led to an

unacceptable dose-averaging along the length of the fibre, and a resulting significant

disagreement with TPS calculation (under-response in Figure 4.7). The fibre optic dosimeter

cannot be considered a ‘point-detector’ in the context of brachytherapy dosimetry. Whilst

current accepted practice is for optical fibres of several mm length, it would be possible to

reduce this dimension to perhaps 1 mm, but with reduced sensitivity and challenging

practical handling, especially at readout. In both the isolated source and treatment plan

measurements, fibres exhibited an over-response at low dose levels. This may be due to

exposure to ambient light during the preparation and cutting, phantom-loading, and readout

of the fibres. While this is not a large effect, it could add appreciably to the net signal at low

dose exposure levels and is difficult to avoid in the manual preparation and read-out process.

Fibre optic material is relatively inexpensive, and standard TL readers are widely available.

However, due to inefficiencies in batch production, inherent variation in sensitivity, excessive

size in longest dimension, the need for external geometry fixation test objects, and the time-

consuming nature of the experiments, optical fibres as utilised in this study are not currently

recommended for clinical brachytherapy dosimetry and further research and development

is required to fully realise their other dosimetric advantages.

74

4.4.2. Radiochromic Film

The use of Gafchromic EBT3 film with triple-channel analysis has shown

excellent agreement with both Monte Carlo source models in measurements of radial dose

from an isolated source, and with TPS calculations in measurements of complex

brachytherapy treatment dose distributions. A dose-response calibration for the film was

implemented over a range 0 Gy to 50 Gy and applied in the clinical dose distribution

measurement, with resulting agreement of absolute dose measurement using film to within

1.5% of TPS calculation. However, to obtain high quality dosimetric results with film, careful

attention is required to scanning protocols and triple-channel analysis is advocated (Micke et

al. 2011). Of the three dosimeters investigated for use in brachytherapy dosimetry in this

work, only EBT3 film with multi-channel analysis satisfies the investigated characteristics of

detectors as specified by Perez-Calatayud et al. (2012a). The energy dependence of film

(Brown et al. 2012, Masillon et al. 2012) has not been explicitly investigated in the work

presented in Section 4, however the overall response of the dosimeters, which includes all

sources of uncertainty, has been successfully evaluated against Monte Carlo data.

4.4.3. Solid Radiochromic Polymer

The use of Presage radiochromic material with optical CT readout has shown

good agreement to both Monte Carlo data of radial dose function from a single source and

with TPS calculations of a complex multi-dwell dose distribution. A significant advantage of

the method is the ability to validate complete 3-D dose distributions with a single

measurement and without the need to construct a specially designed phantom for film

positioning. Data can be analysed in any orientation, as might be most convenient for a

particular problem, whereas film measurements are restricted to pre-defined planes and it

is extremely difficult to generate isotropic 3-D information via film measurements.

Nevertheless, despite these attractive features, the preliminary experimental results

presented here indicate a significant level of noise or artefact in the measurement of dose

distributions using Presage. However, there are at least three available mitigation strategies

that may be explored in future research development: with existing equipment, combine

repeated readouts at different levels of illumination (Krstajic and Doran 2007a), and at

different dose levels, to extend the dynamic range; modify existing equipment to sample the

optical absorption spectrum at two different wavelengths and hence two different dose

sensitivities; use a laser scanner rather than a CCD-based device to improve signal-to-noise

and reduce artefact (Krstajic and Doran 2007b). Whilst the Presage samples used in this study

were limited in physical size, thus reducing the clinical utility of the results presented, much

larger samples may be obtained from the manufacturer, albeit at significant cost. The optical

CT equipment used in this study was a bespoke lab-based solution, but commercial options

are now available (example Vista Optical CT Scanner, Modus Medical, USA) that could be

used in a hospital medical physics department.

75

4.4.4. Dosimetric and Practical Considerations for Dosimetry System Choice

The characterisation of the response of a detector to brachytherapy source

irradiation, as evaluated in preceding sections, is the primary consideration in determining

suitability. However, other practical considerations are also relevant for application of the

dosimetry system in brachytherapy dosimetric audit. A summary of key considerations of

both a dosimetric and a practical nature are summarised in Table 4.2, for the experimental

methods and materials used in this work.

Table 4.2. Dosimetric and practical considerations of the dosimetry system for brachytherapy

audit, comparing optical fibre, Gafchromic EBT3 film, and Presage.

Consideration Optical fibre

thermoluminescence

(point detector)

Gafchromic EBT3

radiochromic film

(2D detector)

Presage radiochromic

plastic

(3D detector)

HDR source radial dose vs. Monte Carlo, in high dose region

Disadvantage

Normalised dose within 8%, repeatability 9%, high noise

Advantage

Absolute dose within 5%, repeatability 3%, low noise

Neutral

Normalised dose within 8%, repeatability n/a, medium noise

Spatial resolution

Neutral

9 µm diameter x 5 mm length (in this study)

Advantage

26-28 µm active layer thickness x scanner resolution (typically 72dpi, 0.35 mm)

Advantage

0.25 mm voxels (in this study) dependent on optical CT system

Effective atomic number

Disadvantage

11.4

Advantage

6.84

Advantage

7.4

Potential geometric uncertainty

Disadvantage

Accuracy of placement problematic

Advantage

Fixed 2D array of detection

Advantage

Fixed 3D matrix of detection

Cost Advantage

Low individual detector cost per detector (pence), but labour intensive production

Advantage

Moderate cost per sheet (approx. £25) sufficient for several brachytherapy-sized detectors

Disadvantage

Expensive purchase cost per sample (approx. £100 for 6cm diameter cylinder)

Availability Disadvantage

Need manual preparation of dosimeter, high initial rejection rate

Advantage

Purchase direct from supplier, readily available

Disadvantage

Limited supply, not ‘off the shelf’ item

Ease of use & readout device

Disadvantage

Requires TL readout. Need many individual detectors to assess dose distribution.

Neutral

Requires careful and advanced methodology to reduce uncertainties, but only requires (high-end) flatbed scanner

Disadvantage

Requires specialised optical CT, not widely available, and careful handling

76

4.4.5. Selection of Dosimeter

The suitability of three emerging dosimetry systems has been investigated for

the measurement of dose distributions in clinical HDR brachytherapy applications for (a)

measurement of radial dose distribution from an isolated HDR source and (b) measurement

of dose distribution along a profile in a typical clinical treatment plan. Thermoluminescent

properties of Ge-doped optical fibres, the dose response of 2D EBT3 Gafchromic film with

multi-channel dosimetry, and radiochromic properties of 3D Presage with optical-CT readout

have been investigated (Palmer et al. 2013d)5. EBT3 film is a commercial product, subject to

robust quality assurance by the manufacturer, this is not the case for the other two detectors

studied, which are more exploratory media at this time. Based on the dosimetric results from

this investigation and practicalities for use, EBT3 Gafchromic film is advocated as a suitable

dosimetry system for commissioning activities, routine QC measurements, or

interdepartmental audit dose distribution measurements in HDR brachytherapy applications.

5 This publication is co-authored by Poppy Di Pietro, Sheaka Alobaidli, Fatma Issa and Simon Doran, who contributed to initial work on optical fibre and Presage dosimetry. The first two are MSc students who were supervised by the author in projects related to the PhD thesis work. The last two are physicists at the University of Surrey who had prior experience in these techniques and contributed theoretical and practical background knowledge. This contribution relates to initial experimental work on these two techniques, but not the concept design, data analysis nor conclusions presented in this chapter, which are the author’s. The film dosimetry work was wholly instigated and conducted by the author.

77

CHAPTER 5

Development of Radiochromic Film

Dosimetry for Brachytherapy Audit

78

5.1. Objectives for the Development and Evaluation of Film

Dosimetry for Brachytherapy

In Chapter 4, the use of Gafchromic EBT3 film was advocated in general as a

suitable dosimetry system for the measurement of brachytherapy dose and dose

distributions. In this chapter, further research is presented that enabled its use for the

national brachytherapy dosimetry audit. This includes testing the film’s useable dose range,

checking for any dose-rate variations, and validating dose-response with radial distance from

Ir-192 and Co-60 HDR brachytherapy sources, in comparison to Monte Carlo and TPS data.

Simple tests are performed to confirm orientation dependence, scanner warm-up

characteristics, and repeatability. The use of film dosimetry for brachytherapy audit

applications requires further technique development and research, including optimisation of

calibration and scanning methodology, evaluation of advanced triple-channel dosimetry

techniques, and a detailed evaluation and understanding of uncertainties and strategies for

their reduction.

The resulting pixel value in a scanner-produced image of a radiochromic film is

a complex convolution of scanning lamp emission, absorption of the film, sensitivity of CCD

array and, importantly, optical properties of the scanner along the light path influenced by

polarisation caused by the film. All of these may change as a function of position on the

flatbed scanner, and each have further dependencies including scanner warm-up

characteristics and fluctuations, film orientation, film temperature, film humidity level, post-

irradiation film darkening kinetics, dose-dependent effect on polarisation, and scan and

analysis protocols. There are likely to be large dose variations across films used in

brachytherapy dosimetry, which may exacerbate the above issues. For brachytherapy

dosimetry, Perez-Calatayud et al. (2012a) reports that radiochromic film must be considered

“under development at this time, because of numerous artefacts which require rigorous

correction”. Advanced film dosimetry techniques, including triple-channel scan processing

(Micke et al. 2011, Mayer et al. 2012) has the potential to improve the accuracy and reliability

of film dosimetry and mitigate sources of uncertainty. The use of all three colour channels of

a flatbed scanner has been proposed to correct for deviations from the calibrated average

film-scanner response. Hence the non-dose–dependent signal component can be separated

from the signal and compensated for. The triple-channel film dosimetry implemented by

Ashland Inc. (Micke et al. 2011), is stated to have the following features and advantages: a)

separates dose-dependent and dose-independent parts of the signal and signal disturbances,

such as film thickness variation, scanner distortion, and background correction, which can

then be accounted for; b) enables use of the entire dynamic dose range of the film; c)

improves dose map accuracy; and d) indicates any inconsistencies between film and

calibration and estimates the dose uncertainty of the measurement. Triple-channel film

techniques have been studied in external beam radiotherapy applications, typically up to 2

to 3 Gy (Van Hoof et al. 2012, Sim et al. 2013), but their application in brachytherapy, with

prescription doses of 7 to 8 Gy and peak doses that are significantly higher, as well as

different dose distribution and different energy spectrum, needs further investigation.

79

In brachytherapy, film dosimetry applications are usually for routine quality

control, commissioning or audit, where there is generally some freedom in the time between

irradiation and scanning. However, post-irradiation film darkening kinetics is often

considered a potentially significant uncertainty in film dosimetry (Devic et al. 2010).

Polymerization of the active component in radiochromic film continues after irradiation, but

the rate of polymer growth decreases with time. Casanova Borca et al. (2013) studied post-

irradiation darkening appropriate for external beam radiotherapy, up to 4 Gy over three days.

The effect is studied in the present work at dose levels and scan times appropriate for

brachytherapy applications, up to 14 Gy, and darkening kinetics are evaluated during a three-

month post-irradiation period. Irrespective of whether multiple or single-channel dosimetry

is used, it is important to characterize post-irradiation darkening to minimise uncertainties

in film dosimetry.

Several characteristics of film dosimetry that must be appreciated for accurate

use are well-understood and have been documented in the literature, including film

orientation, batch consistency, and disabling scanner image correction (Niroomand-Rad et

al. 1998, Martisikova et al. 2008). These considerations are not extensively repeated in the

current work. There are, however, factors that have not been sufficiently researched, nor

their impact in brachytherapy applications assessed: post-irradiation darkening, lateral

scanner effect, film surface perturbation, film active layer thickness, measurement of clinical

brachytherapy dose distributions, film curling, and proposed film dosimetry methodologies,

and where appropriate, whether triple-channel dosimetry can improve dosimetric accuracy.

A potentially significant artefact is the response of the scanner to the film as a function of

lateral distance on the scan plane, the artefact increasing with dose level, described by

Menegotti et al. (2008) in the range 0 to 7 Gy. The response artefact is caused by the

polarisation of transmitted light by the near-linear array of polymer rods in the film, and the

varying transmission on reflection at mirrors in the scanner being a function of angle for

polarised light, which changes with lateral position on the scanner. The longer wavelength

red light is affected significantly more than green and blue light; hence, it is expected the

multichannel process will mitigate the artefact. In this work, we investigate the effect at up

to 14 Gy, appropriate for brachytherapy, and compare the calculated film dose using single-

and triple-channel dosimetry.

Inadvertent surface contamination, such as fingerprint grease and scratches

from contact with phantoms and jigs, particularly of concern in the context of a national

audit, can cause artefacts in film dosimetry, changing the optical density in the scanned

image. The measured density will also vary proportionally with the thickness of the active

layer, which for Gafchromic EBT3 has a design specification of 28 microns. Manufacturing

tolerances of the active layer thickness are typically up to 1.5% (personal communication

with manufacturer), which would result in an uncertainty in reported dose of 1.5% for single-

channel conventional dosimetry. Triple-channel film dosimetry is expected to mitigate such

dose-independent signal components. The film-dose response using single- and triple-

channel dosimetry in the presence of such film surface perturbations and extreme active

layer thickness variations in doses typical for brachytherapy applications is investigated.

80

A methodology for efficient dosimetry using radiochromic film has been

proposed by Lewis et al. (2012) in which test films are scanned together with a reference

dose film strip and an unexposed film strip in order to eliminate, by normalizing, any scan-

dependent uncertainty, such as scanner lamp output. This protocol is evaluated for external

beam radiotherapy (IMRT and VMAT) by Lewis et al. (2012). The evaluation is extended to

typical clinical brachytherapy situations below, with a suggested modification.

In this chapter, the perceived advantages of triple-channel compared to single-

channel analysis are evaluated in dosimetric test situations applicable for brachytherapy,

using the latest Gafchromic EBT3 film, which is structurally different to its predecessors, EBT

and EBT2; the latter having had greater coverage in the published literature to date

(Reinhardt et al. 2012). Film dose maps calculated using single- and triple-channel film

dosimetry for brachytherapy exposures are compared with brachytherapy treatment

planning system intended dose distributions, to evaluate any benefit of increased film dose

range and dose map accuracy of the triple-channel technique. Finally, taking account of the

above work, an optimum methodology for film dose distribution measurement in

brachytherapy is proposed.

5.2. Methodology

Radiation doses were delivered to an estimated precision of ±0.5% and

distances measured to a precision of ±0.05 cm throughout this work (irrespective of the

number of significant figures presented).

5.2.1. Film Dosimetry Equipment, Calibration and Scanning

All film measurements were performed with Gafchromic EBT3 (Ashland ISP

Advanced Materials, NJ) from two batches (Lot #01171401 and #12171303), with each test

using film from a single batch, unless otherwise stated. The procedure summary

recommendations for handling radiochromic film as defined by Niroomand-Rad et al. (1998)

in AAPM TG-55 were adopted. Initially a set of basic film dosimetry usage tests were

undertaken. Film samples exposed to 1 and 8 Gy were scanned at 0, 90, 180, and 270°

rotation on the flatbed scanner, and with each face of the film in turn towards the scanner

detectors. Multiple scans were performed to evaluate scanner warm-up characteristics and

repeatability in response. The effect of water submersion and cutting of the film was also

examined visually and in terms of scanner response. A film exposed to a typical

brachytherapy dose distribution was scanned with and without a non-reflective matt

surround (black card) to evaluate the necessity of minimising stray light in the scanner.

Film scanning was performed in red-green-blue (rgb) format using a 48-bit (16-

bit per channel) scanner (two identical scanners were initially compared for consistency;

Epson Perfection V750 Pro; US Epson, Long Beach, CA) at 72 dpi, in transmission mode, with

no colour or sharpness corrections, consistent orientation on the scanner, and 48 hours from

exposure to scanning, unless otherwise stated in the methodology. Dose-response

81

calibration of the film was undertaken within FilmQAPro® software (Ashland ISP Advanced

Materials, versions 2.0.4631 (2012), 3.0.4835 (2013) and 4.0.5323 (2014)). A nominal 6 MV

linear accelerator, traceably calibrated to a primary standard at the National Physical

Laboratory (Teddington, UK) and measured using an ionization chamber calibrated for

absorbed dose to water, was used for film calibrations, and all test film dose exposures. Film

strips of 10 × 5 cm were positioned on the central axis in a 10 × 10 cm field at 5 cm depth in

Solid Water® (RMI457, Gammex, Middleton, WI) and each exposed to a different dose level:

0, 1, 2, 4, 7, 10, 13, and 16 Gy, corresponding to the range expected for films in the BRAD

phantom (Section 6.4.1). The average film pixel values in a 4 × 5 cm region centred on the

axis of the beam were used to derive the average film response at each dose level. Scanned

images of irradiated films in TIFF format were converted to dose maps using both a single-

channel method (red channel) and a triple-channel method (red, green, and blue channels),

via FilmQAPro software. The single-channel dose conversion utilizes a simple calibration

function; the red channel was chosen, as this is the most commonly used in simple

radiochromic film dosimetry since this wavelength has the highest absorption spectra (Devic

2011). The multichannel analysis method uses an algorithm described by Micke et al. (2011)

to separate the scanned signal into a dose-dependent part and a dose-independent part. The

algorithm essentially determines and subtracts a disturbance function that is independent of

dose by fitting the measured colour signal to allowed colours in the dose-to-rgb signal

calibration. Calibration fits, based on a rational (linear) fit function of the form in Equation

5.1 were derived for each colour channel:

X(D) = (A+BD)/(C+D) Equation 5.1

where X(D) is the scanner response, ranging between 0 and 1, at dose D, and A,

B, and C are the fitted function constants. The derived constants were compared for two film

batches.

Reference dose films were also used for linear rescaling of the calibration

function and this methodology was evaluated. Lewis et al. (2012) have proposed an efficient

methodology for film dosimetry in external beam radiotherapy, in which one reference dose

film and an unexposed film are scanned simultaneously with the test film, to account for

scanner-related variations and time-since-exposure darkening compensation from the

calibration condition. In brachytherapy, the dose range across films may be somewhat

greater than that in external beam radiotherapy, and the dose level for the reference dose

film may need different considerations. Ten EBT3 test films were each exposed to accurately

known dose levels of 5, 7.5, 10, and 13 Gy, representative of dose ranges expected in

brachytherapy film dosimetry applications, at 5 cm depth in a Solid Water phantom. The film

dose in 4 × 4 cm regions of interest at each dose level for each film was calculated using

FilmQAPro software, triple-channel dosimetry, using 0 and 7.5 Gy, and then 0 and 13 Gy

reference doses for linear rescaling of the film calibration function. The average percentage

difference of the film-calculated dose at each dose level from the anticipated dose for the

two calibration systems was recorded and compared.

82

Figure 5.1 shows the arrangement of test films with reference films for

simultaneous scanning on the flatbed scanner. The regions of interest are shown as red

rectangles and were placed centrally left-to-right in the scan window to mitigate effects of

lateral scanner response, Section 5.3.4. For the brachytherapy audit test films, the region of

interest is the corner of the film which was closest to the clinical applicator and received the

highest dose (see Section 6.3.1). This region is aligned centrally with the regions used for

calibration and validation. Due to the large dose range across the film, it is proposed that two

reference calibration films, in addition to the unexposed film, are required to provide

assurance of accuracy across the range of doses: it is suggested that one film be accurately

exposed to the brachytherapy prescription dose, as this is generally of primary interest, and

another to the expected maximum dose to be recorded on the film, to confirm accuracy

across the dose range.

Figure 5.1. Arrangement of calibration and test films on flatbed scanner for simultaneous scanning.

The regions of interest, shown as red rectangles in the figure, were placed centrally left-

to-right in the scan window, to mitigate lateral scanner artefact.

83

5.2.2. Investigation of Film Dosimetry Performance Parameters and Evaluation of

Triple-Channel Dosimetry

Post-Irradiation Film Darkening

It is known that darkening of radiochromic film after irradiation continues for

some time post-exposure (Devic et al. 2010). It is important to take account of the time delay

to scanning in all dosimetry evaluations. The exposure of reference dose films along with test

films which are then subsequently scanned together can be used for calibration scaling as

described above. In external beam therapy dosimetry, test and reference films can be

irradiated consecutively from the same radiation source. However in brachytherapy, and

especially in the context of a national dosimetry audit, there may be significant delays

between the irradiation of test films on a brachytherapy unit and corresponding reference

films from an external beam linear accelerator (used to provide uniform film dose irradiation)

due to practical equipment access issues and logistics of the audit. To establish the post-

exposure darkening kinetics of EBT3 film, nine samples were exposed to doses up to 14 Gy

and then repeatedly scanned over a period of three months. Reference films were also

acquired and scanned only once at three months to confirm no darkening of the test films

due to exposure to multiple scans. The data was used to determine the required delay

between test exposure and scanning, as a function of the time difference between test film

and reference film, in order not to introduce error due to differential darkening between the

two films.

Lateral Scanner Effect

Four 4 × 4 cm EBT3 film pieces were exposed to 1, 8, 12, and 14 Gy, respectively,

at 5 cm depth in a Solid Water phantom. The films were cut in half to provide two identical

film samples at each dose level. One of each dose level film was positioned along the central

axis of the scanner, while the other piece was displaced laterally by 1, 2, 3, 4, 6, 7 and 9 cm,

in both positive and negative directions. The lateral direction is defined as being

perpendicular to the direction of travel of the scanning lamp, with zero lateral displacement

being the centre of the scan plane. Film dosimetry, using single-channel and triple-channel

analysis, was performed for each film piece in each scan. To account for any scan-dependent

variations, the calculated film dose of the laterally displaced piece was corrected for any

variations in the reported dose of the central piece. The resulting change in dose was a

function of the lateral scanner effect only, and this was compared for the two dosimetry

methods.

The potential mitigation of lateral scanner effect by the use of triple-channel

dosimetry was evaluated for a typical brachytherapy dose distribution, similar to films

subsequently acquired for the national brachytherapy audit (Chapter 8). With a prescription

dose of 7 Gy to Point A, the dose range across the films was 0.3 to 13 Gy, with the highest

doses at the corner of the film closest to the applicator. Films were converted to dose maps

with single-channel and triple-channel dosimetry and compared to the Nucletron Oncentra

Brachy (version v4.1.0.132) (Nucletron) TPS-calculated 3D dose grid (1.0mm resolution),

84

using FilmQAPro software. Isodose overlay was used to compare the film-measured dose

with the TPS dose.

Film Surface Perturbation

Three 4 × 4 cm GAFCHROMIC EBT3 film samples were uniformly exposed to 0,

4, and 10 Gy respectively, at 5 cm depth in a Solid Water phantom. Each film had surface

perturbations applied on one face: scratches were made with medium hand pressure using

a solid block corner at one end and a thin layer of grease applied at the other end, typical of

fingerprint marks. The change in colour signal caused by these disturbances was between 6%

and 13% (at 16 bits per channel, in the red channel; for the 0 Gy film a maximum reduction

from 40420 to 35280 (relative pixel values), and for the 10 Gy film, a maximum reduction

from 13900 to 11860). Variations in surface perturbations between the films are

inconsequential, as this test examined the qualitative performance of single-channel and

triple-channel dosimetry at each dose level. Film dose maps were created using single-

channel and triple-channel dosimetry. Dose profiles were taken through the two dose maps,

and the effect of surface perturbations was compared.

Film Active Layer Thickness

The measurement of the effect of variations in film active layer thickness on the

scanned image and calculated film dose is limited by difficulties in measuring the actual

thickness of the active layer, sandwiched between two polyester layers. However, to test the

concept of thickness compensation by triple-channel film dosimetry, the active layer of

irradiated film samples was doubled through a process of delamination of irradiated EBT3

film and restacking to produce an effective double active layer, as shown in Figure 5.2. While

there are significant uncertainties in the physical process of de-laminating and re-laminating,

the process is sufficient to test the relative performance of single-channel and triple-channel

dosimetry to significant variations in active layer thickness, accepting the additional error

sources introduced by this process. One can then infer relative performance of the two

systems to more modest variations that may be encountered in commercially available film.

Four film samples were exposed uniformly to 1, 8, 12, and 14 Gy, at 5 cm depth in a Solid

Water phantom. Each sample was cut into three pieces, two of which were carefully

delaminated. In de-laminating, the active layer remained adhered to one of the polyester

sheets, the other being essentially clear, see Figure 5.2. The polyester layers with the active

layer attached were stacked together to produce a sample with an active layer of effectively

double the original thickness. A scan was made of the two configurations: original film and

restacked delaminated with a double active layer. The film dose for each of these at each

dose level was determined using single-channel and multichannel dosimetry.

Film Curvature at Scanning

No prior publications have considered the potential uncertainty introduced in

film dosimetry by film-curling at the time of scanning, which is often undetected or deemed

insignificant by the operator, but could introduce a substantial dose error. Due to the

manufacture process, film naturally curls. Smaller film sample sizes, such as those used in the

brachytherapy national audit (Chapter 8), are more susceptible to curling than large films

85

typical in external beam verification (in which the weight of the film supresses the tendency

to curl). A 100 mm width film is shown in Figure 5.3, exhibiting a natural curl of 1.5 mm

maximum height from the flatbed scanner glass plate, along the film central axis.

Figure 5.2. Process of delamination and restacking to produce a double-thickness active layer: (a)

original films, (b) delaminated with active layer adhered to one polyester, (c) restacked

with effective two active layer thickness.

Figure 5.3. Photograph showing natural curvature of a 100 x 100 mm Gafchromic EBT3 film placed

on a flatbed scanner glass plate.

86

The methodology to investigate film curvature at scanning utilised three test

situations. The first to evaluate the effect on brachytherapy dose distributions, and the

others to establish the magnitude of the effect under controlled conditions of uniform

irradiation. (1) An EBT3 film of a clinical dose distribution was scanned flat and with natural

film curvature. A 100 mm x 100 mm EBT3 film previously irradiated with a typical clinical

cervix brachytherapy dose distribution as for the national brachytherapy audit (Chapter 8)

was used, resulting in a film dose range 2 to 13 Gy. The film was scanned flat under a glass

compression plate and also scanned allowing the natural curve in the film to be exhibited, as

shown in Figure 5.3. Dose maps were calculated for the flat and naturally curved cases and

compared to the treatment plan. Gamma evaluation index (Low et al. 1998) was calculated

using FilmQAPro® software comparing the treatment planning system intended dose

distribution and the film-measured dose distributions, for the flat and naturally curved film

scans.

(2) An EBT3 test film was irradiated with a nominal 6 MV linear accelerator, at 5

cm deep in Solid Water (RMI457, Gammex, WI, USA), with a 200 mm x 200 mm field to 7.5

Gy. The film was scanned with controlled curvature, held in contact with the scanner plate at

two opposite edges and raised at the centre by 1, 2, and 4 mm above the scanner plate. The

film was also scanned perfectly flat at 0 mm height, achieved by taping the edges of the film

to the scanner plate, and then achieved with a compression glass plate of 2 mm thickness

above the film. Finally, the film was scanned horizontally flat raised uniformly 1 mm from the

scanner plate. The film was held under tension to prevent any curvature at 1 mm height.

Dose profiles using single red-channel and triple-channel dosimetry were compared for the

flat and raised film scans.

(3) An EBT3 test film was irradiated with a nominal 6MV linear accelerator in

three 30 mm x 30 mm fields to doses of 1.8, 3.5, and 6 Gy respectively. The film was scanned

horizontally flat with 5 mm vertical displacement above the scanner plate, and flat with no

vertical displacement, the latter with and without a glass compression plate. The scans were

taken with the film horizontally flat, at height above the scanner, to evaluate whether the

change in pixel value is a result of film curvature only or height of film above the conventional

scan plane. 5 mm height represents an extreme case of film positioning difference from the

conventional position in contact with the scanner. 5 mm is the maximum gap between

scanner glass and scanner lid for the Epson V750 scanner model. The three dose level squares

were aligned along the scan direction axis of the scanner to prevent any effects of lateral

scanner uncertainty. Dose profiles were separately created using single-channel red and

green dosimetry (the colours most commonly used for film dosimetry) and compared, for the

flat and raised film scans.

87

5.2.3. Validation of Film Dosimetry Response to HDR Brachytherapy Sources

The response of EBT3 film to irradiation from a single HDR source was compared

to Monte Carlo derived consensus data sets (Perez-Calatayud et al. 2012b) as a simple

method to experimentally validate the use of the film dosimetry system for HDR

brachytherapy: combined effect of all variables including dose range, dose rate, energy

dependency, triple-channel dosimetry, dose scaling. This was done for both Ir-192 and Co-60

HDR sources using a custom made Perspex film holder in a near full-scatter water tank

(Palmer et al. 2013b) using plastic source transfer tubes. The repeatability of results was

assessed for both the Ir-192 and Co-60 sources using a second measurement for each.

Any dose-rate effect on film response was evaluated by exposing films to the

same dose level but at different distances from the source and hence different dose rates,

using appropriately calculated irradiation times. The expected dose was calculated from a

TPS using Monte Carlo derived TG-43 data, with knowledge of the source strength and

exposure time.

5.3. Results

5.3.1. Film Calibration, Scanning and Processing

The film orientation on the scanner had a significant effect on the scan signal.

For red, green and blue channels at 1 and 8 Gy, films scanned at 0° and 180° were identical,

and those scanned at 90° and 270° were identical, but the two sets differed by between 4.7%

and 8.4% depending on colour channel and dose level, red high dose exhibiting the largest

difference. Modest scanner warm-up characteristics were observed, with the first scan after

switch on being between 0.5 to 1.5% below steady-state values, which were achieved within

0.2% by the fourth scan and then consistently for the remaining scans in the series.

Repeatability of multiple scanning was found to generally be within 0.2%, occasionally up to

0.5%. Any change in signal depending on which side of the film was towards the light source

was within the general scan-to-scan variation and could not be statistically demonstrated.

Submersion into water affected the edges but not the faces of the EBT3 film, with an opaque

edge appearing of width 0.2 mm after 1 h, 0.4 mm after 2 h, up to a maximum 0.5 mm after

3 h. Following a drying period of around two hours at ambient room temperature, the

opaque edge faded, with no impact on scanner response. A more significant effect was

occasional de-lamination of the film at cut edges. Depending on the cutting technique used

a dysfunctional edge of 0.2 to 1.0 mm, occasionally up to 3.0 mm, was observed along cut

lines. Delamination of the film increased the standard deviation of scanned pixel values and

was unpredictable in terms of scanner response; hence films with apparent delamination

were discarded. An optimum cutting method which minimised de-lamination was achieved

with a rotary paper cutter (Staples A5 Precision Trimmer), which compressed the film from

above and below at the point of cutting.

88

Figure 5.4 shows eight EBT3 film strips, from the same batch, used for the dose-

response calibration that were exposed to doses in the range 0 to 16 Gy. The level of

colourisation darkening increases with dose, with no evidence of saturation. Figure 5.5 shows

the corresponding pixel values in three colour channels as a function of irradiated dose, for

two film batches, scanned 48 hours post-exposure. There is an apparent difference between

the colour response of the two batches, confirming the need for careful batch control and

re-calibration for each new batch. The systematic difference in the blue channel could be the

result of a difference in thickness of the active layer between the two batches.

Table 5.1 provides the calibration function constants of Equation 5.1 for the two

EBT3 film batch calibration curves in Figure 5.5. The calibration fit was derived via FilmQAPro

(Ashland, NJ, USA) software. It is clear that different film batches must have their own unique

calibration. The two film batches presented were used for the film dosimetry research in this

thesis and for the national brachytherapy audit (Chapter 8).

Table 5.1. Rational (linear) calibration equation parameters, see text, for two Gafchromic EBT3 film

batches (lot a=#01171401, lot b=#12171303), scanned 48 hour post exposure.

Colour channel Red Green Blue

EBT3 film batch Lot a Lot b Lot a Lot b Lot a Lot b

Calibration parameter A 2.031 2.334 3.655 4.754 3.570 4.675

Calibration parameter B 0.063 0.034 0.021 -0.047 0.070 0.020

Calibration parameter C 3.098 3.694 6.215 8.568 11.406 16.725

Figure 5.6 shows dose-map isodose lines from a scanned test irradiation of EBT3

film, exposed to a typical brachytherapy dose distribution with Ir-192 source (using the BRAD

phantom, Section 6.4.1) scanned 48 hours post exposure, and processed using FilmQAPro

software. The film was scanned twice, 10 minutes apart, separately represented by thick and

thin isodose lines, once with a non-reflective matt surrounding the film piece on the flatbed

scanner, and once without. The two sets of isodose lines are identical at isodose levels 150

to 2200 cGy, with deviations <1.0 mm seen at 70 and 100 cGy isodoses, demonstrating

excellent scanner repeatability and negating the need for a matt surround to the film for

scanning.

89

Figure 5.4. Eight Gafchromic EBT3 film calibration strips (film batch lot b=#12171303), irradiated

to doses in the range 0 to 16 Gy and scanned 48 hour post exposure.

Figure 5.5. Gafchromic EBT3 calibration curves, in red, green and blue colour channels (16 bit), for

two film batches (lot a=#01171401, lot b=#12171303), scanned 48 hour post exposure,

with rational (linear) dose fit equations.

90

Figure 5.6. Isodose overlay of two scans, taken 10 minutes apart (thick and thin lines), of the same

EBT3 film piece that had previously been exposed by a typical brachytherapy dose

distribution with Ir-192 source, with and without a non-reflective matt at scanning.

5.3.2. Calibration Function Linear-Scaling

Table 5.2 shows the average calculated film dose at four known dose levels

across a wide dose range from a series of ten test films. The primary calibration function was

derived over a dose range 0 to 16 Gy. Within FilmQAPro software, the calibration function

was rescaled to an unexposed film and a known dose level film, scanned within the same

image as the test film. This allows correction for any scanner-related response changes

compared to film calibration scans. The calculated film dose is forced into agreement with

the expected dose at the reference dose level, with the percentage difference increasing at

other dose levels. While maintaining the 0 to 16 Gy calibration function, two different sets of

reference dose films, scanned with the test film, were used. With reference doses of 0 and

7.5 Gy, a maximum discrepancy of 1.4% across the range 5 to 10 Gy, increasing to 2.3% at 13

Gy, was found. Using reference doses of 0 and 13 Gy gave agreement of 0.1% at 13 Gy, but a

maximum discrepancy of 2.1% across 5 to 10 Gy.

91

Table 5.2. Calculated mean film dose from ten test films, at four irradiation dose levels, using

different linear-scaling calibration dose references. The percentage difference of the

reported film dose to the actual irradiated dose level is shown in the table.

Dose references for calibration

function linear scaling

Percentage difference of calculated film-dose from true dose,

at each dose level

5 Gy 7.5 Gy 10 Gy 13 Gy

0 Gy and 7.5 Gy 0.1% 0.4% 1.4% 2.3%

0 Gy and 13 Gy -1.0% -2.1% -1.1% 0.1%

5.3.3. Post-Irradiation Film Darkening

Figure 5.7 shows the net optical density of EBT3 film as a function of time, post-

irradiation, over a range 0.7 to 2277 h (approximately three months), at nine dose levels

between 0 and 14 Gy. All curves fit a logarithmic function with time, of increasing gradient

with dose level. The data in Figure 5.7 has been used to determine the required delay before

scanning of simultaneous test and reference films, such that post-exposure darkening of both

films does not introduce significant uncertainty into the dosimetry results, if the test and

reference films were exposed at different times. The rate of change of optical density post-

exposure increases with dose level; hence, the longest delay before scanning is required for

the highest dose level. Considering the measured rate of change at 14 Gy exposure, the

required delay for typical brachytherapy dosimetry situations has been evaluated. A worst

case delay between a brachytherapy test exposure and a reference film exposure from a

different (external beam) treatment unit may be typically 24 hrs, for the national

brachytherapy dosimetry audit (Chapter 8). If the test film irradiation is performed at time t

= 0 h and the reference film irradiation at t = 24 h for the same 14 Gy dose level, the difference

in net optical density of the films due to post-exposure darkening would be 5% if scanned at

0.1 h following the reference film exposure. This reduces to 3% at 1 h, 1% at 12 h, 0.6% at 48

h, 0.4% at 72 h, and 0.2% at 96 h following reference film exposure. For a time difference of

6 hrs between test and reference films, the net optical density difference between the films

at a dose of 14 Gy is 3.8% at 0.1 h post-exposure, reducing to 0.3% at 24 hrs. For a time

difference of 2 hrs, the difference in net optical density at 14 Gy is 2.8% at 0.1 hr, reducing

to 0.3% at 8 hrs post-exposure.

92

Figure 5.7. EBT3 film net optical density (as a ratio to a simultaneously scanned reference density

sample) as a function of time post-irradiation, up to three months (2277 hrs), over a

dose range 0 to 14 Gy. (Error bars indicate one standard deviation of the sampled pixels).

5.3.4. Lateral Position of Film on Scanner

Figure 5.8 shows film dose obtained using triple-channel and single-channel

dosimetry as a function of the lateral position of the film on the scanner, at dose levels 1, 8,

12, and 14 Gy. For both triple-channel and single-channel dosimetry, there is a lateral-

scanner effect, which increases the reported film dose with increasing lateral position of the

film on the scanner. The effect is significantly larger for single-channel compared to triple-

channel. The percentage increase in film dose at 4 cm lateral distance compared to the film

dose on the central axis of the scanner, at 1 Gy, was 6% with single-channel dosimetry and

1% with triple-channel dosimetry; at 8 to 12 Gy was 7% with single-channel and 1% with

triple-channel dosimetry; and at 14 Gy was 11% with single-channel and 2% with triple-

channel dosimetry. At 9 cm lateral distance, at 1 Gy, the dose increase was 20% with single-

channel dosimetry and 5% with triple-channel dosimetry; and at 14 Gy, the dose increase

was 24% with single-channel dosimetry and 9% with triple-channel dosimetry. The results for

positive and negative displacements in the lateral direction were consistent within

experimental error.

93

Figure 5.8. Change in film dose as a function of lateral distance on scanner, using single-channel

(red lines) and triple-channel (green lines) film dosimetry, over range 1 to 14 Gy. (Error

bars indicate one standard deviation of the sampled pixels).

94

Figure 5.9 shows isodose comparisons between TPS-calculated and film-

measured dose planes, from a film exposed in the national brachytherapy audit phantom

(Section 6.4.1), using single-channel and triple-channel dosimetry. Both single- and triple-

channel dosimetry provide good agreement close to the applicator axis, up to 20 mm lateral

(abscissa), corresponding to 3 to 13 Gy, and good agreement up to 105 mm from the

applicator base (ordinate), corresponding to 5 to 13 Gy. While triple-channel film dosimetry

maintains good agreement with TPS calculation across the entire film, the single-channel

analysis exhibits an increasing difference between film dose and TPS dose with increasing

lateral distance (abscissa). This corresponds to increased lateral distance on the scanner.

Comparing the film and TPS doses in Figure 5.9 using gamma analysis, with criteria of 3%

(local normalisation) and 2 mm (with zero threshold for analysis across the full film), the

passing rate was 90.6% for triple-channel and 37.5% for single-channel. The gamma passing

rate is dramatically improved when triple-channel dosimetric analysis is used.

Figure 5.9. Comparison of film-measured (thick lines) and treatment planning system-calculated

(thin lines) dose distributions (isodose lines from 30 cGy to 1300 cGy shown), from a

typical brachytherapy cervix applicator, with (a) single (red)-channel dosimetry and (b)

triple-channel dosimetry. Abscissa aligned with lateral direction of the scanner, where

x=0 represents the middle of the scanner.

95

5.3.5. Film Surface Perturbation

Figure 5.10 shows film dose profiles through three EBT3 film samples exposed

to 0, 4, and 10 Gy, respectively, each with surface perturbations caused by the application of

grease and scratches. Film dose calculated using triple-channel and single-channel dosimetry

is compared in the Figure. The surface perturbations have significantly greater effect on the

reported dose using single-channel, compared to triple-channel dosimetry, with the latter

accurately reporting the expected dose values for all three films. Single-channel analysis

reported significantly higher dose values due to the presence of grease and scratches.

Figure 5.10. Effect of film surface perturbation (grease and scratches) on calculated film dose using

single-channel and triple-channel dosimetry. Profile through three film samples

irradiated to 0, 4, and 10 Gy each, with grease and scratches.

96

5.3.6. Film Active Layer Thickness

Table 5.3 gives the calculated dose values for the original film and ‘doubled

active layer’ film scanned and analysed using single-channel and triple-channel dosimetry

methods, at dose levels of 1, 8, 12, and 14 Gy. The process of de-laminating and restacking

increases the noise in the scanned image, shown as an increase in the standard deviation of

the sample region. However, it is clear that an effective doubling of the active layer is

reported as an approximate doubling of the dose with single-channel dosimetry, but has

much less effect on the dose reported by triple-channel dosimetry, with doses being

approximately consistent with the original single active layer film. Triple-channel dosimetry

is clearly less sensitive to active layer thickness variations than single-channel dosimetry.

Table 5.3. Calculated film dose for original and ‘double active layer’ film using single- and triple-

channel analysis with Gafchromic EBT3 film in the range 1 to 14 Gy. (One standard

deviation of the dose values in the regions of interest are shown in brackets).

Film-dose (cGy) from original film Film-dose (cGy) from de-laminated

film restacked with two active layers

Dose level (Gy) Single-channel

dosimetry

Triple-channel

dosimetry

Single-channel

dosimetry

Triple-channel

dosimetry

1 102 (2) 100 (2) 247 (105) 92 (35)

8 785 (11) 799 (16) 1681 (269) 866 (118)

12 1166 (26) 1214 (28) 2452 (204) 1188 (223)

14 1401 (22) 1400 (35) 2601 (390)

1427 (261)

5.3.7. The Effect of Film Curvature at Scanning

Figure 5.11 provides isodose overlay comparisons of film-measured and TPS-

calculated dose distributions for a dose distribution typical of the national audit (Chapter 8),

with the film permitted to exhibit its natural curvature, Figure 5.11(a), and with the film

scanned flat under a glass compression plate, Figure 5.11(b). When scanned flat there was

good agreement between film-measured and TPS-calculated isodoses. With natural film

curvature, there was good agreement where the film was in contact with the glass plate but

poorer agreement where the film was displaced above the scanner, leading to a maximum 2

mm shift in isodose line for a 1.5 mm vertical height offset, for the dose gradient considered

in this example. For these dose distributions, the gamma passing rates comparing film-

measured and TPS-calculated isodose distribution for (a) film naturally curved and (b) held

flat under glass plate, were 95.7% and 100.0% respectively at 3% (local) 2 mm criteria, and

28.1% and 94.2% respectively at 2% (local) 1 mm criteria.

97

Figure 5.11. EBT3 film measured (thin lines) and treatment planning system calculated (thick lines)

isodose distributions from a typical cervix brachytherapy treatment delivery, with (a)

film naturally curved on scanner glass plate (in contact with the scanner plate at 0 mm

abscissa, rising to a maximum 1.5 mm above the scanner plate at the maximum of the

abscissa scale), and (b) with the film flat on the scanner plate, compressed under glass.

Figures 5.12 and 5.13 show the effect on dose profiles of controlled curvature

of EBT3 film irradiated by a 6 MV linear accelerator to approximately 7.5 Gy. The profile

direction is physically aligned with scan direction to avoid any lateral scanner effect due to

polarisation. Profiles across half of the film are shown in the figures, corresponding to the

range from edge to centre of the linear accelerator field. The film is in contact with the

scanner plate to the left of the graph, with upward curvature and maximum height achieved

at the right of the graph, corresponding to the centre of the film. Dose profiles were

generated with single red-channel dosimetry in Figure 5.12 and with triple-channel dosimetry

in Figure 5.13. In Figure 5.12 there is a clear dose reduction (increase in scanned optical

density pixel value) with curvature of the film above the scanner plate, 1% reduction at 1 mm

height, 3.5% reduction at 2 mm and 8% reduction at 4 mm height. For the equivalent case

with triple-channel dosimetry dose calibration, Figure 5.13, there are no systematic changes

to the dose profile with increasing height above scanner, and the triple-channel algorithm

effectively corrects the effect. There was no difference in profiles between films held flat

with tape or under a glass compression plate in either single-channel or triple-channel

dosimetry. The above investigation considers the consistency between film scans in the two

situations. It is assumed that the flat-film case is closer to the true value and is in better

agreement with the treatment planning system calculation.

98

Figure 5.12. Normalised film dose profiles using single (red)-channel dosimetry across a single EBT3

film irradiated by a 6 MV linac beam, with the film in various flat or curved positions.

(Profiles aligned with scan direction to avoid lateral scanner effect).

Figure 5.13. Normalised film dose profiles using triple-channel dosimetry across a single EBT3 film

irradiated by a 6 MV linac beam, with the film in various flat or curved positions. (Profiles

aligned with scan direction to avoid lateral scanner effect).

99

Figure 5.14 shows single-channel (red and green) dose profiles across three 30

mm x 30 mm fields from a linear accelerator at dose levels of 1.8, 3.5 and 6 Gy, for EBT3 film

held flat and raised uniformly by 5 mm above the scanner glass plate. The profile direction is

physically aligned with scan direction to avoid any lateral scanner effect due to polarisation.

There is a clear reduction in reported dose with the film raised 5 mm compared to the case

in contact with the scanner plate. The average dose reduction in the red channel at 1.8 Gy

was 21%, at 3.5 Gy was 17%, and at 6 Gy was 14%. The average dose reduction in the green

channel at 1.8 Gy was 33%, at 3.5 Gy was 21%, and at 6 Gy was 16%.

Figure 5.14. Film dose profiles using red and green single-channel dosimetry across an EBT3 film

exposed to three dose level regions (180, 350 and 600 cGy), with the film flat in contact

with the scanner glass plate and with the film raised uniformly by 5 mm. (Profiles aligned

with scan direction to avoid lateral scanner effect).

100

5.3.8. Film Measurement and Validation of Radial Dose from an Ir-192 HDR

Brachytherapy Source

Figure 5.15 shows EBT3 film measured dose as a function of radial distance from

the centre of an isolated Ir-192 HDR source. The response was measured with a single film

over the range 5.0 to 60.0 mm from the source axis, corresponding to a range 8 to 0.06 Gy in

this exposure. There is good agreement with Monte Carlo data, within experimental

uncertainty (1.8% at k=2, see Section 6.4.5, not including film positional uncertainty) across

the full distance range considered.

Figure 5.16 shows the dose derived from three EBT3 film measurements each

with different exposure times and dose rates (different distances from the source), achieving

similar dose ranges across the full width of each film. The film doses are compared to TPS

calculated dose points in the Figure. Neither re-normalization nor axis shifts of the film or

TPS dose data has been applied, showing true measured dose agreement directly with that

calculated by the TPS in the three dose ranges. The Figure demonstrates there is a negligible

dose-rate dependence on EBT3 response in the range considered. The three films were

exposed to the same dose, but at different distances from the source and hence different

dose rates, using appropriately calculated irradiation times. For example, 8.0 Gy was

delivered to the low, medium and high irradiation time films at distances of 8, 18 and 34 mm,

corresponding to dose rates of 350.5, 69.2 and 19.4 cGy min−1, respectively, for the HDR

source used in this study. The distance to agreement between film measured and TPS

calculated (interpolated) dose at the 8.0 Gy dose level was 0.5, 0.2 and 0.1 mm, respectively.

5.3.9. Film Measurement and Validation of Radial Dose from a Co-60 HDR

Brachytherapy Source

Figures 5.17 and 5.18 show film measured dose as a function of radial distance

from an isolated Co-60 source, using the same methodology as for the Ir-192 source in

Section 5.3.8. Agreement with Monte Carlo data is within experimental uncertainty. 8.0 Gy

was delivered to the low, medium and high irradiation time films at distances of 9, 27 and 43

mm, corresponding to dose rates of 391.6, 41.4 and 15.8 cGy min−1, respectively, for the HDR

source used in this study. The distance to agreement between film measured and TPS

calculated dose at the 8.0 Gy dose level was 0.5, 0.2 and 0.1 mm, respectively.

101

Figure 5.15. Normalised dose as a function of radial distance from a Nucletron mHDR-v2 Ir-192

brachytherapy source; EBT3 measured compared to HEBD consensus data (Perez-

Calatayud et al. 2012b).

Figure 5.16. EBT3-film measured dose with radial distance from a Nucletron mHDR-v2 Ir-192

brachytherapy source, at three different dose rates and irradiation times, compared to

treatment planning system (TPS) calculated dose.

102

Figure 5.17. Normalised dose as a function of radial distance from an Eckert & Ziegler Bebig Co0-

A86 Co-60 HDR brachytherapy source; EBT3 measured compared to HEBD consensus

data (Perez-Calatayud et al. 2012b).

Figure 5.18. EBT3-film measured dose with radial distance from an Eckert & Ziegler Bebig Co0-A86

Co-60 HDR brachytherapy source, at three different dose rates and irradiation times,

compared to treatment planning system (TPS) calculated dose.

103


5.4.1. Dosimeter Performance Characteristics

The post-irradiation darkening behaviour of Gafchromic EBT3 film has been

shown to be a logarithmic function, continuing at least three months post-exposure, with the

effect more significant at higher dose levels, shown in Figure 5.7. These results differ from

Borca et al. (2013) who considered post-irradiation colouration over 72 hrs and concluded

net optical density stabilized after nearly 0.5 h at dose levels up to 2 Gy, and 2 h would be

sufficient to guarantee stability to perform analysis at studied doses up to 4 Gy. Theoretically,

the film active layer polymer formed by exposure will continue to grow, but the rate of

growth will rapidly decrease as the distance from the growing polymer chain to the next

available monomer increases. Growth will only occur on random occasions when enough

thermal energy is acquired to bridge the gap. The results presented here demonstrate post-

irradiation colouration is a continual, logarithmic function. It is important to appreciate that

continual darkening of the film occurs, and reference dose films irradiated for purposes of

calibration scaling, scanned simultaneously with test films, must be irradiated at the same

time, or the time difference recorded and sufficient time allowed between test and reference

irradiations prior to scanning. The anecdotal recommendation of delaying scanning by a time

four times the interval between test and reference films has been shown to be valid (in this

work, up to 14 Gy and 24 hrs between test and reference films), in order to reduce dose

errors to less than 0.3%.

The lateral scanner artefact described by Menegotti et al. (2008) for doses up to

7 Gy, has been confirmed in this work and documentation of its effect extended up to 14 Gy,

appropriate for brachytherapy film dosimetry. Menegotti and colleagues showed a change

in pixel value in the range 9% to 19%, depending on scanner model, for a 7 Gy exposure at

10 cm lateral to the central axis of the scanner. The results presented in the current work

indicate a reported dose increase of 23% at 14 Gy, 9 cm lateral position, for single-channel

film dosimetry. However, this effect is reduced to 9% with triple-channel dosimetry. The

array of monomer and polymer rods in the film means the amount of polarization of light is

different depending on whether the film is 0° or 90° rotated on the scanner, and this may

explain the difference in results between the current work and Menegotti et al. (2008).

Schoenfeld et al. (2014) showed theoretically that radiation-induced polymers in EBT3 film

cause polarisation and scattering of scanner light, causing lateral scanner artefacts (termed

the parabola effect by Schoenfeld).

The advantage of triple-channel dosimetry over single-channel dosimetry in

mitigating the effect of lateral scanner artefact has been demonstrated at dose levels and

distributions typical for brachytherapy dosimetry. However, the useable scanner width is

limited even with triple-channel dosimetry to less than ±4 cm width to reduce lateral scanner

effect to less than 1% with the Epson V750 Pro scanner (the useable width may increase with

physically larger scanners). The effects of film surface perturbations and of variations in

active layer thickness have been shown to be significantly reduced with triple-channel

compared to single-channel dosimetry.

104

A proposed methodology for brachytherapy film dosimetry has been discussed,

using three reference strips (including zero dose), rather than two (including zero dose), in

order to confirm accurate dosimetry over the large dose ranges encountered in

brachytherapy. It has also been demonstrated that choice of reference film dose level for

linear calibration rescaling can improve the dose uncertainty at dose levels of interest. This

is particularly important in brachytherapy dosimetry in which there are often very large dose

ranges across films. It is recommended to use a reference dose level for rescaling of the film

response function around the dose level of particular interest, such as the prescription dose,

rather than the maximum dose in the film, as recommended by Lewis et al. (2012), for

external beam radiotherapy dosimetry applications, in which there is often a smaller dose

range. It is also recommended that film is perfectly flat on the scanner to provide reliable

dosimetric results, avoiding changes in scanner response which may be due to variations in

illumination, optical disturbances, and effects such as that described by Callier (Chavel and

Lowenthal 1978) in which light changes from a collimated to a diffuse source. A compression

glass plate positioned on top of the film is suggested to ensure sufficient flatness.

The proposed methodology for film dosimetry utilising the triple-channel

algorithm and linear rescaling, implemented within FilmQAPro software, as discussed in this

work, has been applied to the measurement of dose distributions around clinical

brachytherapy treatment applicators. Using gamma analysis to compare film-measured and

TPS-calculated dose distributions, at criteria 3% local normalisation and 2 mm distance to

agreement, over a region of interest of 115 x 80 mm (equivalent to 0.3 to 11 Gy dose range),

gamma passing rates exceeding 90% for triple-channel dosimetry have been reported, but

the passing rate reduced to exceeding 37% for single-channel dosimetry. (Due to very steep

dose gradients, the position sensitivity of the gamma map is high, and in each case the

relative position of the film dose and plan dose was optimised for maximum gamma pass

rate). The significant reduction in performance of film dosimetry for brachytherapy with

single-channel dosimetry compared to triple-channel dosimetry is likely to be primarily the

result of lateral scanner artefact, which is mitigated with triple-channel dosimetry.

At the currently achieved uncertainty levels for film dosimetry, quoted between

1% and 10% depending on the situation (Martisikova et al. 2008, Bouchard et al. 2009,

DeWerd et al. 2011), film curling during scanning can be a substantial error source if not

addressed. Displacement of film above the scanner plate by 1 mm to 2 mm can increase pixel

values and reduce reported dose by 1% to 4%. From the data presented, the physical height

of the film above the scanner plate appeared to cause changes in scanned pixel value, and

hence reported dose, irrespective of any inherent curvature of the film. Increasing the height

of the film above the scanner plate increased the effect on scanned pixel value. No difference

was found with films scanned flat using adhesive tape on the scanner or scanned under a

compression glass plate, with the latter being more convenient. Moderate natural film

curling has been observed, as may be experienced with small film pieces typical of the

national brachytherapy audit (Chapter 8), which may be a significant source of uncertainty

and error in film dosimetry. A glass compression plate may be conveniently used to flatten

small film pieces, and triple-channel dosimetry is able to mitigate the detrimental effect on

105

film dosimetry accuracy when scanning curved films. The magnitude of the effect is likely to

be dependent on the film type and the scanner model used as well as scanning protocols,

and results presented here should be confirmed for local film dosimetry situations. Film

curvature must be considered and controlled alongside the other uncertainties present in

film dosimetry for the national brachytherapy audit.

The mechanism by which deviations from perfectly flat film at scanning result in

variations in scanner response is not clear, and the magnitude of the effect has not been

reported to date in the literature. The flatbed scanner manufacturer in a private

communication has reported that “any [film] curve will cause a change in the resultant scan

due to the media potentially being outside the focal plane of the scan head. Additionally it

can affect the uniformity of lighting. We would recommend the film is held flat in a frame

during scanning. A fluid mount kit may be an option” (Epson 2014). There is no information

in the literature on film curvature or height above scan plane at scanning for dosimetry

applications, but there has been discussion of film height above imaging plane in

photographic film scanning and optical photographic developers and enlargers. In

photographic applications, small changes in film position can have marked changes in the

resulting signal due to the Callier effect (Chavel and Lowenthal 1978) which describes

changes in optical density resulting from variations in illumination, such as from a collimated

to diffuse source within the optical system. Any curved film will also have different effective

film thicknesses for light paths at different scan positions and will affect surface scattering

from the film. A glass compression plate has been used to achieve flat film scan conditions in

the above methodology. EBT3 film may be used in contact with the glass scanner plate and

the glass compression plate film without the formation of light reflection interference

patterns, such as Newton’s rings, since the film is formed of matte polyester layers with silica

particles at the surface creating a 5 µm air gap to the glass plates which is significantly larger

than the light wavelength. Other films, such as the predecessor EBT2, may suffer from

interference artefact when in contact with either the upper or lower glass plates since they

are constructed with smooth polyester layers. To achieve reduced dose uncertainty with a

glass plate compression technique, EBT3 film must be used for the national brachytherapy

dosimetry audit.

5.4.2. Use of Radiochromic Film Dosimetry for Brachytherapy Dosimetric Audit

The use of film dosimetry has been evaluated in brachytherapy applications in

both test cases and clinical dose distribution measurements. It has been demonstrated that

radiochromic film dosimetry with Gafchromic EBT3 film and a flatbed scanner is a viable

method for brachytherapy dosimetry, and uncertainties may be reduced with triple-channel

dosimetry, linear dose re-scaling, and specific film analysis methodologies including a glass

compression plate and minimising lateral displacement on the scanner. The separation of the

scanner signal into dose-dependent and dose-independent parts via triple-channel

dosimetry enables the mitigation of signal disturbances, such as variations in film active layer

thickness, film surface perturbations, and also to some extent lateral scanner effect. The

lateral effect is particularly significant for accurate dosimetry and must be considered in

106

brachytherapy exposures to high dose levels and, even with triple-channel dosimetry, limits

the usable lateral region of the scan plane. The use of simultaneous scanning of calibration

reference films with test films is advantageous to scale the pixel value for any scanner

fluctuations, provided post-exposure darkening kinetics are accounted for between the two

films. Darkening of the film continues after irradiation as a logarithmic function with time, at

least up to three months. The importance of keeping film flat when scanning has been

demonstrated, an effect overlooked in previous recommendations on film dosimetry, but it

is not necessary to use a non-reflective matte around the films.

107

CHAPTER 6

Development of an ‘End to End’

Brachytherapy Dosimetric Audit

using Radiochromic Film

108

6.1. Review and Evaluation of Previous Brachytherapy Dosimetric

Audits

Prior to developing the national brachytherapy dosimetry audit, a systematic

review was conducted of journal articles published over the last three decades, that have

discussed concepts, research and practices of dosimetric audit in brachytherapy. Medline

and Embase databases were searched, internet search engines utilised, and several

brachytherapy experts across the world consulted. Eight peer-reviewed journal publications

were obtained that presented results of fully completed audits, or pilot audits at several

centres, involving dosimetric measurement (Haworth et al. 2013, Venselaar et al. 1994, de

Almeida et al. 1999, Elfrink et al. 2001b, Heeney et al. 2005, Roue et al. 2007, Carlsson

Tedgren and Grindborg 2008, Casey et al. 2013), and one concerned with a geometric

applicator reconstruction audit (Roue et al. 2006). Each of these above audits deals with

afterloading equipment. There was also one completed dosimetric audit of electronic

brachytherapy in the UK (Eaton et al. 2013). There have also been conference presentations

on brachytherapy audit, however the full content is not always widely available in searchable

proceedings or publications (Ratcliffe et al. 1997, Sudom and Ratcliffe 1998, Aukett et al.

1999, Thwaites and Nisbet 2003a, Thwaites et al. 2003b). An IAEA survey of dosimetric audit

networks in 2013 indicated only 13 of 53 organisations worldwide undertook audits in

brachytherapy of any modality (Grochowska and Izewska 2013). A summary of published

dosimetric audits conducted on HDR/PDR brachytherapy equipment is provided in Table 6.1

(reproduced in Palmer et al. 2014b).

In approximately one-third of the published audits, errors were detected at one

centre in each audit, which appear to be clinically significant. Causes of the errors were stated

as calibration factor, calibration coefficient, certificate source strength, and step-size

mismatch between TPS and treatment delivery system. In 2001, Elfrink et al. (2001b) studied

agreement between ion-chamber measured dose and dose calculated by a treatment

planning system at the chamber measurement point from straight catheters and found one

clinic where a 6.8% deviation was recorded and attributed to an error in the source strength

value on the source certificate. The source position accuracy at seven clinics had errors larger

than 1 mm, three of which exceeded 2 mm. Roue et al. (2007) reported on audits with a

remote mailed phantom and TLD, developed by the ESTRO BRAPHYQS Physics Network and

the EQUAL-ESTRO laboratory, consisting of three straight catheters around a central TLD

holder. It is reported that one audit result was initially outside the tolerance level, with the

error being due to the use of an “inadequate calibration coefficient”. Once rectified, the audit

was repeated and a result within tolerance was obtained. The acceptable limit used by the

EQUAL-ESTRO Laboratory for dose deviation is 7% (Veres 2013). A dosimetric audit of source

strength using a well-type chamber reported by Carlsson Tedgren and Grindborg (2008)

found agreement within 1%, except one case which was within 3%. Another audit-detected

error was reported by Haworth et al. (2013), discussed below.

109

The only publication to date of a brachytherapy audit incorporating imaging and

reported as a full system test, was described by Haworth et al. (2013), however in common

with the above audits, only straight source catheters were used. The audit was conducted at

seven clinics in Australia as a pilot study to demonstrate feasibility prior to a more complex

audit to be devised. The current phantom was based on the design described by Roue et al.

(2007), using two source dwell positions (±30 mm from the dosimeter) in each of three

straight catheters with a central TLD measurement in a homogeneous dose region. The

phantom was filled with water and CT imaged using normal clinical protocols, identifying

source catheters and radio-opaque markers in the position of the central TLD. A host

treatment plan was produced to deliver 1 Gy to the TLD, and delivered using normal clinical

processes. All results were within the stated ‘optimal’ tolerance of ±5%, except one centre in

which a 25% error was reported, outside the ‘emergency’ tolerance level. This was caused by

the use of a 5 mm step-size at the treatment console for delivery, rather than the 2.5 mm

step size that has been used at the TPS. A repeat audit once this had been rectified was within

the ‘optimal’ tolerance.

6.2. Objectives for a Dosimetric Audit in Brachytherapy

There is no doubt that clinical and physics aspects of brachytherapy will continue

to develop at a significant rate, including the realisation of advanced and functional imaging

in brachytherapy, a new era of dose calculation algorithms for brachytherapy, and an

evolution away from traditional template planning to fully patient-specific optimisation

(Palmer 2014a). As a result, the scope of dosimetric audit in brachytherapy will continue to

be challenged to be fit for purpose. The dosimetric audits discussed in Section 6.1 had the

objective of accurate confirmation of source strength, using straight source catheters, with

dwell patterns designed to generate homogeneous dose deposition regions at the dosimeter.

Indeed, the majority of audits have been designed to minimise the effect of source

positioning errors on the measured dose: the design by Casey et al. (2013) stating “variations

in distal/proximal source positioning up to 10 mm had minimal effect on dose measurement

accuracy”. Verification of source strength for afterloading equipment was a key aim for

earlier dosimetric audits, since there was indeed a need to confirm the accuracy of this

parameter. There had been a lack of primary standards for Ir-192 and the uncertainty of

source strength quoted by manufactures some years ago was ±10% (Venselaar et al. 1994,

Baltas et al. 1999). The uncertainty quoted by manufacturers is now often ±5% at k=3, with

their source strength measurements traceable to national standards laboratories. The ability

for local radiotherapy clinics to determine source strength has also improved, with the

development of well-type ionisation chambers designed specifically for brachytherapy

(Goetsch et al. 1992), which are more robust, less sensitive to positional inaccuracies and

room-scatter effects, and produce higher ionisation currents compared to free in-air

measurements with smaller ionisation chambers (Carlsson Tedgren 2008). Early results from

a UK well-chamber intercomparison demonstrated excellent agreement in source strength

determination comparing local measurement, auditor measurement, and source certificate,

with 14 centres having agreement within 0.5% (Lee 2012).

110

Table 6.1 Summary of published HDR/PDR brachytherapy dosimetric audits.

Reference Dosimeter used Region Number

of audits

Results

Venselaar

et al. 1994

Thimble-type

ionisation chamber

with in-air jig

The

Netherlands

and Belgium

13 Auditor compared to local

measurement: Mean +1.3%,

range -0.4% to +3.0%

De Almeida

et al. 1999

Thimble ionisation

chamber with in-air jig

and/or well-type

ionisation chamber

Brazil 10 Visiting centres compared to

reference measurement: All

within ±3.0%, except one error

detected (-4.6%), calibration

factor

Elfrink

et al. 2001b

Thimble ionisation

chamber in solid

cylindrical PMMA

phantom with three

straight catheters

The

Netherlands

and Belgium

33 Measured compared to

prescribed: For HDR mean + 0.9%

±1.3% (1sd), for PDR mean +1.0%

±2.3% (1sd). One error detected

(6.8%) source strength certificate.

Heeney

et al. 2005

Thimble-ionisation

chambers in custom-

designed in air jig

Ireland,

Scotland and

the north of

England

11 (at 9

clinics)

Auditor compared to local

measurement: For HDR, mean

+0.8%, range -0.9% to +1.5%, For

PDR, mean +0.3%, range - 3.8% to

+2.4%.

Roue et al.

2007,

EQUAL-

ESTRO

Laboratory

Mailed TLD with

phantom comprising

three PMMA tubes

around central

dosimeter, in water

Europe 17 Measured compared to

treatment planning system:

Range -4.7% to +4.7%. One error

detected, calibration coefficient.

Carlsson

Tedgren

and

Grindborg

2008

Well ionisation

chamber

Sweden 14 (at 7

clinics)

Auditor compared to local

measurement: All within ±1.0%,

except one within +3%. Auditor

to certificate and local to

certificate: range -1.5% to +3%.

Haworth et

al. 2013

‘End-to-end’ audit.

TLD with phantom

comprising three

PMMA tubes around

central dosimeter, in

water

Australia 7 Measured compared to

treatment planning system. All

results within ±5%, except one

large error detected (25%),

incorrect step size used at

treatment unit compared to TPS.

Casey

et al.

2013

Optically stimulated

luminescence (OSL)

nanoDots in mailable

polystyrene phantom

USA 8 Measured compared to

treatment planning system.

Mean 0.0% ±1.2% (1 sd)

111

While previous concerns of source calibration may now essentially be resolved,

which should lead to reduced uncertainty in determination of source strength, it is important

not to abandon audit checks of this fundamental parameter, but it does enable a shift in

focus to more complex areas for audit. The alternative novel approach used in the present

work is to audit the dose delivered in typical clinical situations, assessing the combined effect

of all dosimetric uncertainties, including applicator reconstruction on planning scans,

treatment planning system calculation, data transfer and consistency, dwell position length

calibration, dwell paths in treatment applicators, attenuation by the applicators, as well as

source strength calibration. This has the benefit of a more ‘clinically realistic’ dosimetric

audit. Dosimetric measurements in clinical brachytherapy situations are problematic and

measurement uncertainties may increase as we move closer to the clinical treatment

situation. Therefore, results of such an audit may lack specificity of the cause of any error,

but the reassurance of auditing the full ‘patient path’ is particularly advantageous in a clinical

radiotherapy department. It is important that brachytherapy equipment is tested under

conditions other than those used for routine calibration, and assessment of dose

distributions rather than point doses is a particular requirement for robust assurance of

treatment dose delivery.

The advantage of a visiting auditor conducting audits rather than using mailed-

phantom/detector dosimetry approaches is the opportunity to discuss local physics clinical

procedures and to reduce operator error in set up. An aim of the national brachytherapy

dosimetric audit was to enable such discussion, hence site visits were required by an auditing

physicist experienced in brachytherapy physics. A simple review of brachytherapy

procedures and infrastructure during the course of an audit is designed to at least

qualitatively attempt assessment of the likelihood of errors, and offer any quality

improvement suggestions to mitigate risks.

Brachytherapy is an effective treatment modality for a number of sites, but it is

most frequently utilised for carcinoma of the cervix in the UK, followed by prostate. It is a

logical choice to undertake brachytherapy system audit on the most frequently used

treatments. Also, since cervix applicators generally have more challenging geometry for

accurate source positioning due to high levels of curvature (Stern and Liu 2010), compared

to other sites which utilise approximately straight catheters, cervix brachytherapy has been

chosen for the national audit. Brachytherapy treatment planning requires applicator

reconstruction, usually on patient CT scans, and this has the potential to be a significant

source of uncertainty (Hellebust et al. 2010, Kirisits et al. 2014). It is therefore a key objective

that the national brachytherapy audit incorporates this aspect within the ‘end to end’ audit,

the intention being to incorporate each stage of the patient imaging, planning and treatment

process.

The scope of the audit is to include only the primary aspects of the

brachytherapy process, and include only those that usually come under the responsibility of

Medical Physics Departments. Of course ‘physics-aspects’ are not the only contributors to

dose delivery uncertainty in brachytherapy and indeed the more ‘clinically’ related aspects

have the potential for more, or even overwhelming, influence on the treatment outcome

112

than ‘physics’ aspects, if physics aspects are delivered within expected quality levels and no

significant error. There are numerous clinical uncertainties in brachytherapy treatment

(Thomadsen et al. 2003, Kirisits et al. 2014), that are not tested by the current audit

approach, including target volume definition and contouring (Petric et al. 2013, Hellebust et

al. 2013), tissue heterogeneities, human factors in applicator positioning, internal organ

motion and inter- and intra-fraction applicator movement (Nesvacil et al. 2013, Anderson et

al. 2013). The magnitude and importance of various uncertainties were discussed in a special

volume of Radiotherapy & Oncology in April 2013. Tanderup et al. (2013) calculated the one

standard deviation combined uncertainty for a single fraction of intracavitary brachytherapy

as 12% for the HR CTV D90 target, and 21-26% for D2cc organs at risk, reducing with multiple

fractions.

6.3. Methodology

6.3.1. Design of an Audit Phantom

The purpose of this work is to approach HDR brachytherapy audit from a novel

perspective: to measure dose distributions around clinical treatment applicators and

compare to TPS isodose calculations, rather than traditional audits in brachytherapy of point

doses from straight catheters. The measurement is therefore more closely aligned with

external beam patient-specific QC for IMRT/VMAT delivery verification than conventional

brachytherapy QC tests. The intention of the phantom is to enable a comparison between

measured dose using film and TPS calculated dose distributions. To achieve this aim, a

phantom was required that securely held any cervix gynaecology HDR/PDR applicator and

held measurement film accurately at known positions around the applicator. Film positioning

was required that enabled dose to traditional prescription location Point A (Viswanathan and

Thomadsen 2012) to be determined, as well as measuring dose distribution in clinically

relevant regions. It was also important that the treatment applicator be imaged in

approximately the orientation used for patient imaging, such that assessment of applicator

reconstruction was realistic to the clinical pathway.

The reference medium recommended by the AAPM Task Group 43 for obtaining

the dose rate distribution around a brachytherapy source is liquid water (Rivard et al. 2004).

The requirement for a sufficiently large phantom for brachytherapy dosimetry, either for

practical measurement or theoretical calculation, is well known (Perez-Calatayud et al. 2004,

Melhus and Rivard 2006, Granero et al. 2008, Perez-Calatayud et al. 2012a). It is essential the

phantom for the national brachytherapy audit either provides full-scatter conditions or the

correction for lack of scatter is known. The scatter assessment is required for both Ir-192 and

Co-60 HDR sources, and to inform the practical decision on phantom size that can be

conveniently transported for a UK dosimetric survey. Monte Carlo calculations were

undertaken to inform the decision on audit phantom size, see Section 7.3.2.

113

6.3.2. ‘End to End’ Audit Procedure Development

The uncertainty in any measurement system should be significantly less than the

variations to be measured in the test signal. There are uncertainties in film dosimetry,

including positional, darkening kinetics and scanning characteristics, but these can be

mitigated by advanced film dosimetry methods, reported in Chapter 5, and any residual

uncertainty is expected to be acceptable for the intended purpose of this work. The audit

procedure is designed not to precisely establish TG-43 type dose-rate functions (Rivard et al.

2004) but to confirm the dose distributions around clinical treatment applicators is consistent

with TPS calculated intended distributions, within acceptable clinical tolerance. An increased

level of uncertainty is acceptable for this purpose.

6.3.3. Sensitivity to Simulated Errors

In order to confirm the BRAD audit process was sufficiently sensitive to detect

errors in treatment delivery, a typical brachytherapy treatment plan using ring and ovoid

applicator, was modified at the TPS to induce two different types of simulated errors: (1) All

dwells shifted 5 mm proximal from their planned position, (2) treatment times scaled for the

delivery of 8 Gy instead of planned 7 Gy. These error were chosen as conceivable failure

modes for a brachytherapy treatment, at a magnitude that would be of clinical significance.

The treatment plan was compared to the film-dose measurement using isodose overlay and

gamma analysis, for the normal delivery case and the two simulated error cases.

6.4. Results

6.4.1. BRachytherapy Applicator Dosimetry (BRAD) Phantom Design

A test object was designed to measure the dose distributions around any ring or

ovoids and intrauterine (IU) clinical treatment applicator for HDR or PDR brachytherapy for

cervix cancer. The use of additional interstitial needles was not accommodated in the design.

Several prototype phantoms were conceived and tested (constructed of polystyrene or

Perspex) prior to the final design being selected. An early concept design to test the use of

film for brachytherapy dosimetry is shown in Figure 4.2, which consisted of a three sided

Perspex frame with grooves to hold an array of several films. Alignment of the HDR treatment

applicator between film planes had unacceptable positional uncertainty, and films were

susceptible to curvature. Other designs included slots for film in water-equivalent plastic,

however accurate positioning of films was difficult and the design did not satisfy a primary

intention to use of liquid water around the films. The final design enabled accurate film

placement against thin water equivalent material sheets, within a water tank, with well-

defined applicator positioning with respect to the film, as described below.

114

The BRachytherapy Applicator Dosimetry (BRAD) was designed to maximize the

volume of liquid water around the applicator and film with all supporting and alignment

material constructed of Solid Water6 (Gammex RMI-457) which introduces acceptable

corrections from liquid water (Perez-Calatayud et al. 2012). Aldelaijan et al. (2011) has

quantified the ratio of dose delivered to radiochromic film in liquid water compared to the

dose delivered in Solid Water using Monte Carlo simulations to be 0.9941 ± 0.0007. Thin

sheets of Solid Water, 5 mm thickness, are used to accurately position the film.

There is a practical limitation on the size of a water tank that can be used for the

national dosimetric audit measurements due to transportation requirements. A full scatter

phantom may be required to reproduce TG-43 assumptions. A competing argument is that

the scattering volume should equate to approximate patient dimensions to compare

delivered and planned dosimetry, rather than unbound scattering conditions. It was decided

to not include limitations of TG-43 dosimetry in the audit and hence a full-scatter phantom

was designed. Monte Carlo modelling (Section 7.3.2), was used to establish the required

dimension of a phantom in order to reduce the magnitude of the scatter correction for dose

measurements at Point A. Balancing the theoretical requirement and practical consideration

of transportation, a scatter phantom comprising a cubic water tank with each side dimension

400 mm was selected.

The BRAD test object enabled four films to be placed in orthogonal planes

crossing the IU tube of the clinical treatment applicator. Figure 6.1 shows photographs of the

final BRAD phantom, Figure 6.2 a wireframe CAD design drawing, and Figure 6.3 3D rendered

CAD views. The applicator ring or ovoids were placed in contact with the upper surface of the

test object with the IU tube inserted into the central cavity. A range of collars and spacers

were available for the central cavity to ensure accurate positioning along the central axis of

the test object, for any HDR applicator in use in the UK. The appropriate inserts were selected

to match the external diameter and length of each IU tube. An applicator support arm and

clamp were used to rigidly hold the applicator in a consistent position for CT scanning and

radiation delivery. One Gafchromic EBT3 film of dimensions 100 × 120 mm was positioned

on each of four vanes in the test object, with the film moved up against end stops towards

the central axis and upper surface of the test object, giving a reproducible film-edge-to-IU-

axis distance of 10.0 mm and a film-edge-to-surface-of-applicator ring or ovoid distance of

5.0 mm. The evaluated region was film edge to 50 mm away from the IU and film edge to 70

mm along the IU, (the unused portion of film was for ID marking and handling). The film

positions were intended to sample the dose distribution through the majority of the HR-CTV

and into typical locations of OAR. Left lateral and right lateral films containing CTV and

conventionally defined Point A, the anterior film including CTV and the closest section of

typical bladder; and the posterior film including CTV and closest sections of typical rectum

and sigmoid, for cervical cancer brachytherapy.

6 Gammex Inc, Middleton, Wisconsin, USA

115

Figure 6.1 Photographs of BRAD phantom with (a) Nucletron plastic ring and IU applicator, and (b)

with Eckert & Ziegler Bebig plastic split-ring and IU applicator

Figure 6.2 Wireframe CAD design drawing of BRAD national brachytherapy audit test object. One

film shown in position, yellow shading, with one insert and spacer shown above prior to

insertion, applicator support arm shown without clamp detail. Position of applicator ring

and IU tube indicated in green with position of point A indicated by red dot. (All

dimensions in mm).

116

Figure 6.3 3D rendered CAD design drawings of BRAD phantom (left to right: oblique from above,

oblique from side, oblique from below).

117

6.4.2. ‘End to End’ Audit Procedure

To provide an ‘end to end’ system test of brachytherapy dosimetry, it was

essential that all normal operating procedures were utilised for the phantom as for patients,

including CT scan protocol, data transfer, planning process, export to treatment unit, and

treatment delivery. The steps in the audit process are shown in Figure 6.4 and described

below. The full work instructions used by the auditing physicist is provided in Appendix D.

During these process steps, local brachytherapy physics practice was discussed between

auditor and host physicists.

CT scan

For each audit, a clinical cervix treatment applicator was selected and

appropriate collars and spacers inserted into the central cavity of the BRAD phantom to

match the outer dimension of the IU tube and the distance from ring/ovoid surface to IU tip.

The applicator was inserted into the phantom and the clamp adjusted and secured to ensure

no movement, shown in Figure 6.3. The phantom was placed on its side such that the

applicator was in an approximate clinical position, shown in Figure 6.4, and then CT scanned

using the audit centre’s normal gynaecology brachytherapy protocol.

Treatment planning

Using normal clinical protocols, the applicators were reconstructed on the CT

images (with the assumption that reconstruction of an applicator scanned in-air would be

consistent with one scanned in-patient), and a typical cervix treatment plan prepared, either

prescribing to point A (usual) or noting its dose. Confirmation that the locally used definition

of point A was as defined by the 2012 ABS recommendations (Viswanathan and Thomadsen

2012) was established, or a correction applied. Local standard planning approach was used

to generate a suitable treatment plan, with typical local prescription dose. Following export

of the plan to the treatment delivery unit, the DICOM RTDose file (3D dose grid) was also

exported for later analysis.

Treatment delivery

An EBT3 film was positioned on each vane of the phantom, ensuring contact

with the central and upper end-stops, and secured in place using the film clamps. The

phantom was then placed vertically in the custom water tank and irradiated according to the

imported treatment plan.

118

Figure 6.4 Process flow diagram of brachytherapy ‘end to end’ dosimetric audit using the BRAD

phantom

119

6.4.3. Film Dosimetry and Data Analysis Procedure

The procedure for film dosimetry for the BRAD system has been developed

based on work reported in Chapter 5 on the optimisation of film measurements for

brachytherapy, taking account of the specific requirements for the national brachytherapy

audit, including anticipated dose rages. The procedure is outlined below:

Film calibration, with audit and reference dose films from the same film batch.

All films scanned with consistent orientation on central axis of scanner and under a glass

compression plate to prevent curvature, using rgb format tiff with 16-bits per channel.

Gafchromic EBT3 film dose-response calibration undertaken using triple-channel

dosimetry, at dose levels of 0, 1, 2, 4, 7, 10, 13, 16 Gy.

Gafchromic EBT3 film cut into four sheets of 100 x 120 mm for use in BRAD.

Audit films exposed in BRAD phantom to brachytherapy dose distribution at audited

radiotherapy centre, at approximately the same time (within 6 h) as reference dose

strips exposed to well-defined dose (7 Gy) from a nominal 6 MV linear accelerator at

Portsmouth Hospitals NHS Trust (the audit coordinating centre).

Audit and reference dose films scanned simultaneously (in same scan image) on a

flatbed scanner at a time following irradiation of at least a multiple of four of the time

difference between irradiation of test and reference films.

Triple-channel dosimetry and linear-scaling of calibration function using reference

doses, undertaken within FilmQAPro software (Ashland, USA).

Dose to Point A, and distance to agreement with TPS dose, obtained using geometric

distance from corner of film.

Isodose overlay (visual assessment) and gamma evaluation (including 3% local

normalisation, 2 mm, 2 Gy lower threshold) comparing film-measured dose to TPS

RTDose file.

The film dose to Point A was evaluated by first obtaining the scanned pixel

coordinate of the corner of the film closest to the treatment applicator. The BRAD system

accurately places the film corner at 10.0 mm from the applicator IU axis and 5.0 mm from

the surface of the applicator ring (or ovoid). Point A is conventionally defined as 20 mm from

the IU axis and 20 mm from the surface of the ring. Hence the dose at pixel coordinates 10

mm (20 – 10 mm) from the edge of the film away from the IU, and 15 mm (20 – 5 mm) from

the edge of the film away from the ring/ovoid, is evaluated to find Point A dose. Any deviation

of the locally defined position of Point A, or any special circumstances with applicator

alignment, may amend the above pixel positions and are taken into account. All pixels in the

vicinity of the geometric position of Point A are inspected in the film-dose image to find the

‘distance to agreement’ of the TPS planned dose.

120

Visual evaluation of isodose overlay between TPS-calculated and film-measured

dose distribution was used, which is a rapid qualitative method and provides information as

to the location of any particular discrepancies. A quantitative assessment was made using

gamma analysis (Low et al. 1998, Low and Dempsey 2003) in FilmQAPro (version 3.0.4835)

(Ashland, USA) software. The gamma calculation in FilmQAPro software uses the average film

dose calculated to the same resolution as the TPS dose grid, hence eliminating film noise

effects artificially improving the gamma score (Low 2010), and uses interpolation. The

determination of a ‘clinically relevant’ departure between intended and actually delivered

radiation dose distribution is difficult to define and depends on local factors for each

radiotherapy centre. It is likely to be in the region of a dose difference greater than 5.0%, or

a shift in isodose position of greater than 3.0 mm (Thwaites 2013). The clinically acceptable

passing rate for a gamma analysis in HDR brachytherapy is particularly problematic to

establish, with no prior published work having considered this concept. Gamma is therefore

used here as a benchmark, as an indicator of consistency between centres, and the results

for each audit may be compared to the normal range of gamma established for this particular

measurement, software, and analysis situation, to establish what is ‘acceptable.’

To constrain the gamma evaluation to regions of increased clinical significance,

a 2 Gy lower cut-off threshold and a region of interest of 50 mm (away) by 70 mm (along)

from the film corner, has been used for the analysis of BRAD data. A larger analysis region

including significantly lower doses can give a misleading result, particularly in brachytherapy

dose distributions. A significant difference between external beam and brachytherapy dose

distributions is the much larger range of dose levels and extent of low-dose regions present

in the latter, and it follows that a locally-normalised (rather than globally normalised) gamma

calculation is required for brachytherapy. If a single high-dose global normalisation point is

used for gamma calculations in brachytherapy, any deviations in the much lower dose regions

would be rendered insignificant. The gamma calculation passing rate could be artificially

increased simply by including a larger region of lower doses in the analysis. Local

normalisation allows the percentage dose difference to be applied at each dose level,

avoiding biasing the gamma calculation passing rate to only the high dose region.

It is not easy, nor rigorously defined, as to how to select criteria for the gamma

analysis for any given clinical situation. One approach is to use the detector uncertainty

estimate to establish the criteria, which may be derived from say positional uncertainty and

impact on dose (1 mm and 6% at the maximum dose gradient of Point A, see Figure 6.6). For

the national brachytherapy audit using BRAD, the emphasis is on assessment of dosimetric

accuracy across the ‘end to end’ treatment process, hence the decision on acceptability of

an audit result should be based on the largest errors that would be clinically acceptable.

Gamma criteria of 5% dose difference and 3 mm DTA is a reasonable estimate of the

maximum error acceptable in HDR/PDR brachytherapy. A tighter target level of 3% and 2 mm

is also reasonable. An additional tolerance has also been included, at 2% and 1.5 mm, to

discriminate results, with 1.5 mm being chosen as a maximum based on typical 1.0 mm

RTDose grid resolution in brachytherapy treatment planning systems. Consideration of

typical dose gradients is also informative in setting gamma criteria. In brachytherapy, the

121

dose gradient may change considerably across the region of interest; at Point A the gradient

is approximately 6% mm-1 radially away from the source and 1% mm-1 parallel with the source

axis, shown in Figure 6.6. The radial dose gradient decreases significantly with distance from

the source. A value of 1.5% mm-1 is a representative average across the regions of interest

for the BRAD film phantom. Gamma criteria set via this average dose gradient would give a

ratio of dose difference and DTA consistent with 3% and 2 mm criteria, above. Thwaites

(2013) suggests the DTA value should be numerically less than the dose value for gamma

criteria, which is also satisfied by the above criteria. As a general rule, the evaluated

distribution data spacing should be less than or equal to one-third of the DTA criteria in

gamma. Therefore, to use a DTA criteria of 1.5 mm, film scanning resolution of at least 72 dpi

is required (0.35 mm per pixel).

6.4.4. Sensitivity to Simulated Errors

Figure 6.5 shows isodose overlay comparisons of TPS-planned and film-

measured doses using the BRAD system for (a) normal delivery, (b) systematic dwell position

error, and (c) dwell time/prescription dose error, for an IU and ovoid applicator. The two

simulated errors are clearly evident in the visual isodose comparison. In Figure 6.5 (b) the

measured dose is shifted proximally at the applicator IU tip (top left of the figure) by

approximately 5 mm, which agrees with the induced error in the plan. Little effect is seen

near Point A, this position being fairly insensitive to dwell position errors (as modelled in

Section 3.3.4). A shift in isodose lines is also seen close to the ovoids (lower edge of the

figure), due to a shift in dwell positions left to right in the ring as seen in this image. Figure

6.5 (c) shows an apparent shift in all isodose lines, with the measured 7 Gy isodose overlaying

the planned 8 Gy isodose, exactly as expected from the magnitude of the induced dose error.

The gamma passing rate, with criteria 3% (local) 2 mm, with normal delivery case (a) was

95.1%, for dwell position error case (b) was 51.5%, and for dose delivery error case (c) was

38.9%.

6.4.5. BRAD Audit Uncertainty Budget

Brachytherapy dose distributions are characterised by steep dose gradients

close to sources. The positioning of detectors for brachytherapy measurement can be a

significant source of uncertainty and requires particular care. Figure 6.6 illustrates the dose

gradients at the prescription Point A for a typical cervix brachytherapy treatment. Parallel to

the intrauterine (IU) tube the dose gradient is 1% mm-1 at Point A, increasing to 6% mm-1

perpendicular to the IU tube.

The combined standard uncertainty (k=1) for the BRAD dosimetry audit has

been estimated as 3.0% using a simple quadrature sum of individual components of the

dosimetry system as detailed in Table 6.2. The expanded uncertainty, k=2, is estimated as

5.9%. This represents the uncertainty in the determination of point dose differences between

film-measured and TPS-calculated dose at Point A. Due to varying dose gradients across the

film, positional uncertainties have a variable effect on dose uncertainty, the prior values

122

having been evaluated in the vicinity of Point A. Delivered doses closer to the applicator than

Point A are of less clinical significance even though the dose gradient is higher and

uncertainties would be larger. Regions of lower dose would have associated lower

uncertainties due to lower dose gradients, however it is not appropriate to calculate a

separate uncertainty estimate at each point and hence the Point A uncertainty estimate is

taken as an upper limit for analysis.

The combined standard uncertainty of the reported dose value from film

dosimetry alone, not including uncertainties of the BRAD audit such as position and

brachytherapy treatment unit calibration and performance, is 0.9% (k=1), expanded

uncertainty estimated at 1.8% (k=2). This assumes robust film dosimetry, triple-channel

dosimetry and dose linear scaling and measurement close to the scanner central axis, i.e.

does not include lateral scanner effects (see Chapter 5).

Figure 6.5 BRAD isodose comparison of normal delivery and simulated errors: TPS-planned (thick

lines) and film-measured (thin lines) for (a) normal delivery, (b) all dwells shifted 5 mm

proximal, and (c) 8 Gy delivered instead of planned 7 Gy.

123


The BRachytherapy Applicator Dosimetry (BRAD) phantom has been designed,

built and evaluated for use in a national ‘end to end’ brachytherapy dosimetry audit (Palmer

et al. 2013c)7. The system consists of a Solid Water (RMI-457, Gammex USA) frame which

accurately holds any gynaecology cervix treatment applicator and four Gafchromic EBT3 films

at well-defined positions with respect to the applicator. The phantom is CT scanned on its

side with the treatment applicator in a near-clinical orientation; the phantom is then

positioned upright, with film dosimeters, in a full-scatter water tank for treatment delivery.

Advanced film dosimetry methods have been used and documented to obtain film-dose for

comparison to TPS-calculated dose via the export of DICOM RTDose file within FilmQAPro

software (Ashland, NJ, USA). Methods for analysis by visual isodose overlay and gamma

evaluation of the dose distribution, and percentage dose difference and distance to

agreement at Point A, have been described. The standard uncertainty of the evaluation of

the difference between film-dose and TPS-dose has been estimated at 3.0% (k=1).

7 Chris Lee and Ailsa Ratcliffe are co-authors of this work. As experts in brachytherapy physics in UK, my design ideas were discussed with these physicists to check appropriateness and validity of concept. Chris assisted in liaising with workshop engineers and Ailsa had previous, if historic, experience of small-scale auditing in brachytherapy.

124

Table 6.2. Uncertainty budget for the experimental determination of the dose difference at Point

A between TPS-calculated and film-measured dose using the BRAD ‘end to end’ audit

system. Values quoted are the ‘expected’ uncertainty under normal operating

conditions and good-practice processes.

Component Source of uncertainty Type Uncertainty

(k=1)

Note

Applicator

reconstruction

Reconstruction of applicator in CT data set A/B 2.25% a

TPS dose

calculation

Monte Carlo derived TG-43 source model

parameters

B 0.5% b

Source strength Prior determination of source strength at host

centre

A 0.4% c

Treatment

equipment

performance

Source dwell position and time performance

(normal operating conditions)

A/B 0.5% d

Film dosimetry Linear accelerator output calibration for

reference film exposures

A/B 0.5% e

Film calibration fit function B 0.5% f

Film position A 1.5% g

Energy dependence of film B 0.5% h

Phantom scatter compared to unbounded

case

B 0.0% i

Film scanner, effect of lateral position B 0.0% j

Film scanner, scan repeatability A 0.2% k

Combined standard uncertainty (k=1)

Expanded uncertainty (k=2)

3.0%

5.9%

Notes

a. Applicator reconstruction for the BRAD phantom CT images is expected to have less uncertainty than reported estimates of uncertainty in clinical practice (Tanderup et al. 2008, Hellebust et al. 2010). The uncertainty may depend on CT slice width, and the dose effect is dependent on proximity to the source; using the maximum typical dose gradient of 6% mm-1 at Point A, (Figure 6.6) a typical value of 4.5% is derived for a 0.75 mm uncertainty in applicator reconstruction at coverage k=2; 2.25% at k=1.

b. Estimate based on general Monte Carlo uncertainty, for relative dose distribution only not absolute cGy h -1

U-1 c. Uncertainty in calibration of well chamber at NPL, IPEM/NPL HDR code of practice (Bidmead et al. 2010).

Estimate of uncertainty in use of calibrated well chamber to determine source strength is not additionally included.

125

d. Uncertainty in HDR treatment equipment performance under normal operating conditions may include 1.0 mm systematic error in source position calibration, however the dose to Point A is insensitive to small dwell position errors, see Section 3.3.4 (Palmer et al. 2012c). An uncertainty of 1.0% at k=2 is included to account for all possible treatment unit performance issues, including dwell timer accuracy and transit dose; 0.5% at k=1.

e. Uncertainty taken from IPSM 1990 MV Code of Practice (Lillicrap et al. 1990). f. Consistency of triple-channel film calibration and estimated uncertainties from film analysis software, for

films scanned under glass compression plate g. Film position uncertainty estimated at 0.5 mm, at coverage k=2, in BRAD phantom. The dose effect is

dependent on proximity to the source, but a value of 3.0% is used, from maximum typical gradient of 6% mm-1 at Point A, at coverage k=2; 1.5% at k=1.

h. The energy dependence uncertainty of EBT3 film is taken as the correction from calibration in 6MV to measurement in HDR energies, estimated as 1% at coverage k=2 from Bekerat (2012) and Brown et al. (2012); 0.5% at k=1.

i. Scatter correction for bounded scatter in 400 mm side square BRAD phantom compared to unbounded case calculated using MCNP as 0.0%, see Section 7.3.2.

j. Lateral scanner effect at ±25 mm from scanner central axis, at 7 Gy, consistent with BRAD audit dosimetry requirements, see Section 5.3.4.

k. From author’s prior study of repeatability of film scanning, after lamp warm-up.

126

Figure 6.6. Dose gradients around cervix treatment applicator for typical clinical treatment plan. (a)

RTDose map from treatment planning computer, (b) 3D visualisation of dose gradients,

(c) Profiles through clinical prescription Point A.

127

CHAPTER 7

Monte Carlo Simulations for the

Brachytherapy Film Dosimetry

Audit

128

7.1. Objectives for MCNP5 Simulations

Monte Carlo calculations have been used to fulfil several objectives relating to

the UK brachytherapy dosimetry audit: (a) Determine the necessary dimensions for full-

scatter for an audit phantom, or a correction factor if a smaller phantom size were to be

used; (b) Evaluate uncertainties of using radiochromic film for brachytherapy dose

measurement in terms of any disturbance from the film itself when irradiated along the

length of the film; (c) Estimate the attenuation effect of metal treatment applicators on the

delivered dose. The findings have been employed to help interpret the results of the UK

brachytherapy audit. This chapter includes discussion of the choice of input file parameters

and appropriate tallies for Monte Carlo simulation with MCNP5 for the above cases, and

describes the process used for validation of the Monte Carlo implementation used in this

work.

7.2. Methodology

7.2.1. MCNP5 Input Files

The basic required format of input files for MCNP5 is described in Section 2.4.1.

The methodology discussed below is concerned with decisions on the specific details of the

input files applied for each brachytherapy problem, including how the calculation is

performed (mode, physics treatment, tally card) and how the problem is defined (origin of

material data, source energy spectra).

For HDR or PDR brachytherapy applications, it is often appropriate to use

photon-only transport and the F6 tally. However, there are cases where the collision kerma,

F6 tally, is not a good approximation for the absorbed dose, and in these cases the *F8 tally

with full photon and electron transport is more appropriate. In these cases, F6 and *F8 tallies

may give different results. Examples of such cases are (a) within the first few mm from an

HDR source where a beta radiation component may be present, (b) regions in which there is

a loss of charged particle equilibrium such as at a tissue inhomogeneity interface or the

surface of a patient, and (c) relevant to the present work, when calculating absorbed dose to

a very thin layer of medium, << 1 mm, such as within the active layer of Gafchromic EBT3

film, 25µm thick. Electrons generated from photon interactions in the polyester layers above

and below the active layer of the film may lead to energy deposition in the active layer, in

which case the collision kerma approximation in the active layer may not be appropriate. The

phantom scatter and applicator attenuation MCNP5 calculations, Sections 7.2.3 and 7.2.5,

use F6 tally with photon-only transport (mode p). The calculations including film in the

geometry, Section 7.2.4, use *F8 with photon and electron transport (mode p e). Both tallies

are used in the simple source geometry cases, Section 7.2.2, to evaluate differences in the

tallies and validate the MCNP5 implementation for this work. The F6 tally gives results in MeV

g-1 and are converted to J kg-1, equivalent to Gray (Gy), using a tally multiplier card, FM card,

multiplying by 1.602E-10. The *F8 tally gives results in MeV and are later converted to Gy by

129

dividing by the mass within the cell, and multiplying by 1.602E-10 to convert from MeV g-1

to J kg-1 (Gy).

The physics treatment of interactions is controlled by the PHYS card in MCNP:

Simple or full physics treatment of interactions is possible. The simple physics ignores

coherent (Thomson) scattering and fluorescent photons from photoelectric absorption. For

brachytherapy source modelling, full physics treatment is required, including coherent

scattering and Doppler energy broadening. Simple physics treatment is primarily intended

for photons of higher energy than modelled for brachytherapy. Full physics is the default

setting in MCNP5 for photon energies with upper energy limit of 100 MeV (X-5 Monte Carlo

Team, 2008), hence no PHYS card is required in the MCNP5 input files for the brachytherapy

cases considered in this work. The CUT card in MCNP5 can be used to set a lower energy level

below which particles will not be tracked. It is important not to cut-off electrons at low

energies since these contribute to the majority of dose deposition. However it is possible to

set an energy cut-off for photons if efficiencies need to be made in calculation process time.

The default value in MCNP5 is 1 keV cut-off for photons. A cut-off value of 10 keV was used

in this work only for calculations including electron transport in order to reduce the

calculation time (validated in Section 7.2.2). No energy cut-off was required for photon-only

calculations.

The number of source particle histories has been set to 1E9 (1 x 109) for the

brachytherapy simulations. Validation calculations have also been undertaken with fewer

particles to investigate the dependency on the number of source particles in the

brachytherapy calculations. In all cases, the relative error, R, given in the MCNP5 output file

has been inspected to ensure the number of particles is adequate for acceptable precision of

the tally results. Only MCNP5 calculation results in which the standard error, R < 0.05 (X-5

Monte Carlo Team, 2008b), and where the ten statistical tests provided in the output file are

passed, have the results been accepted as reliable, see Section 2.4.3.

Perez-Calatayud et al. (2012a) have published methodological

recommendations for Monte Carlo based dosimetry, which are adopted in the current work.

The recommendations advise that HDR/PDR sources may be specified with emitted photon

energies above 10 keV only, being adequate due to attenuation of lower energies in the

source encapsulation, and being the commonly used value in publications. It is reported that

a lower energy cut off does not produce more accurate results but prolongs calculation time

for the same uncertainty level. The emission energy spectra for Co-60 and Ir-192

brachytherapy sources used in the MCNP5 input files has been derived from the Chart of

Nuclides published by the National Nuclear Data Centre, Brookhaven National Laboratory,

NY, USA (NNDC cited 2014), omitting energies below 10 keV and those with very low

probability per disintegration, < 0.03%, since these contribute minimally to the overall dose

deposition. The resulting energy spectra, presented in the MCNP5 input files in Appendices

F to I, were compared to those listed by Rivard et al. (2010) and the Live Chart of Nuclides

provided by the Nuclear Data Section of the International Atomic Energy Agency, Vienna,

Austria (IAEA cited 2014), and were deemed acceptable for the purposes of the Monte Carlo

brachytherapy calculations of the current work. Rivard et al. (2010) proposed that the AAPM

130

consider recommending NNDC data for medical physics applications of all brachytherapy

radionuclides, and it is this data that is used for the MCNP input files used in the current

work.

The materials and dimensions of the HDR brachytherapy sources were taken

from the 2012 Report of the AAPM and ESTRO from the High Energy Brachytherapy Source

Dosimetry (HEBD) Working Group (Perez-Calatayud et al. 2012b). This includes definition of

the source encapsulation material as 316L steel for the Ir-192 and Co-60 sources considered

in the MCNP calculations in the current work. The brachytherapy sources were modelled

using simplified geometry, which accurately reproduces the radial thicknesses of materials

but omits any air gaps within the source. The sources were treated as cylinders with

simplified end caps of appropriate material thickness, with no attached cable. This geometry

was sufficient (validated in Section 7.2.2) for the purposes of the investigations presented

below, in which only relative dose rate with radial distance from the centre of the source in

a direction perpendicular to the source axis (or at a small angle from perpendicular) is

required. The source geometries are shown in Figure 7.1, created using the MCNP5 geometry

plotter. Material composition data for MCNP5 input files was obtained from the

compendium of material data published by the Pacific Northwest National Laboratory, WA,

USA, (McConn et al. 2011). Liquid water has been used in each of the MCNP calculations

presented below, and the material has been specified as recommended by AAPM TG-43U1

to be pure degassed liquid water (H2O) with a mass density of 0.998 g cm3 (Rivard et al. 2004).

MCNP5 code was installed and run on two personal computers, one with quad

Intel Core i3-3220 processor with 4 GB RAM within Windows XP operating system, and the

other with quad Intel Core i5-2450M with 6 GB RAM within Windows 7 operating system,

with calculation times up to 24 hours. The MCNP5 code was operated under a single-user

software licence issued to Antony Palmer by the Radiation Safety Information Computational

Centre (RSICC), Oak Ridge National Laboratory, United States Department of Energy.

7.2.2. MCNP5 Validation of Implementation

The implementation of the MCNP5 code, the choice of input card parameters,

and the accuracy of source geometries and material definitions, were validated by calculation

of the normalised dose rates as a function of radial distance from the centre of the source

and comparison to consensus data from the High Energy Brachytherapy Source Dosimetry

(HEBD) Working Group (Perez-Calatayud et al. 2012b). The validation exercise was also used

to investigate the required number of particle histories, differences between the F6 and *F8

tally results, and the associated relative errors, for brachytherapy source calculations.

The MCNP5 calculation geometry is shown in Figure 7.2. The source was placed

centrally in a water sphere of 600 mm radius, to provide full scatter conditions (Perez-

Calatayud et al. 2004). Nine cylindrical tally cells were defined of 0.5 mm radial width and 0.5

mm length, centred at radii of 7.5, 10, 15, 20, 30, 40, 50, 60 and 70 mm from the source,

aligned with the central axis of the source. The tally cell dimensions were chosen to minimise

the impact of voxel size effects while maintaining reasonable counting efficiency, consistent

131

with the recommendations of Perez-Calatayud et al. (2012b). Photon and electron transport

was considered, calculating F6:p and *F8:p,e for each tally cell, for 1E6, 1E7, 1E8 and 1E9

source particle histories. The MCNP5 input files for the Ir-192 and Co-60 source validation

are provided in Appendix F.

7.2.3. Phantom Scatter: Spherical Scatter Volume with Point Source

Figure 7.3 shows the MCNP5 geometry that was used to investigate the effect

of phantom size on delivered dose from a point source, compared to results for an essentially

unbound phantom of radius 600 mm, within a calculation sphere of radius 800 mm air.

Spherical shells of radial thickness 2 mm centred at radii of 20, 50, 60, and 75

mm were used as tally cells. 20 mm radius was of particular interest as this is the standard

prescription distance from applicators in clinical brachytherapy treatments (Point A). 50 and

60 mm were chosen as likely maximum radial distances from the source that would be of

interest in film audit measurements, being the maximum extent of clinically relevant dose

levels. 75 mm was chosen to enable a direct comparison of the results of this study to work

by Perez-Calatayud et al. (2004).

Spherical water phantom dimensions of radii 100, 150, 200, 230, 300, 400 and

600 mm were simulated, corresponding to the potential range of phantoms that could

practically be used for external audit of radiotherapy centres, and a full-scatter condition.

230 mm radius was also studied specifically as this corresponds to the equivalent scatter

conditions of a 400 mm side-length cube, for Ir-192 and Cs-137, according to Granero et al.

(2008):

Rsph = 1.1Rcub+10 mm

where Rsph is the radius of a sphere and 2Rcub is the side length of a cube of equivalent scatter,

for 100 mm < Rcub < 300 mm. A 400 mm cube was considered the maximum size for

convenient practical transportation. Photon transport only was considered, calculating F6

tally with 1E9 source particle histories. An example MCNP5 input file for the Ir-192 source is

provided in Appendix G.

132

Figure 7.1. MCNP5 simplified geometry source models, viewed from (a) side and (b) end, for (1)

Nucletron mHDR-v2 Ir-192 source and (2) Eckert & Ziegler Bebig Multisource Co0.A86

Co-60 source.

Figure 7.2. MCNP5 validation geometry; several cylindrical tally cells concentric with the centre of

the HDR source (Ir-192 or Co-60) in full scatter water, viewed from (a) side, with enlarged

central region, and (b) end of the source.

133

Figure 7.3. MCNP5 geometry to investigate the effect of phantom size on measured dose from a

point brachytherapy source, with various spherical shell tally cell radii and water sphere

radii.

7.2.4. Effect of the Presence of EBT3 Film on Measured Dose

Figure 7.4 shows the MCNP5 geometry that was used to investigate whether the

presence of Gafchromic EBT3 film perturbs the dose distribution that would be present in a

uniform water medium. A geometry simulating the use of film in the BRAD phantom for a

cervix treatment applicator (Section 6.4.1) was devised: An extended source of 10 mm

length, simulating several dwells in a ring or ovoid, was positioned 10 mm from the film edge,

the film itself having a width of 15 mm, being the dimensions of film between the source and

the prescription Point A in the BRAD phantom. A ring geometry was used to improve source

particle collection efficiency. A water tally cell outside and level with the film was used to

calculate dose per source particle. The tally cell was of the same thickness as the film active

layer. The dose was compared for the case with the film present and for a case with the film

replaced by water. Due to the thin tally cell dimensions, both photon and electron transport

were considered (mode p e), with calculation of *F8:p,e tally, using 1E9 source particle

histories. The MCNP5 input file is provided in Appendix H.

134

Figure 7.4. MCNP5 geometry used to investigate the dose perturbation caused by the presence of

Gafchromic EBT3 film compared to a uniform water medium (where the film is replaced

by water, not shown).

7.2.5. Applicator Attenuation

Applicator dimensions and materials were obtained by personal communication

with the brachytherapy equipment manufacturers (Varian, Nucletron and Eckert & Ziegler

Bebig), for applicators that were used in the UK brachytherapy audit, see Table 7.1. Figure

7.5 shows the MCNP5 geometry to investigate the radiation attenuation due to metallic

treatment applicators, to interpret results obtained during the UK brachytherapy dosimetry

audit, Section 8.3.1. An extended source was modelled, 60 mm length, simulating a series of

dwells in a typical cervix intrauterine treatment applicator. Ring tally cells were defined at

three radial distances from the source, 20, 40, and 60 mm, and at three positions along the

source train, mid-source, 10 and 60 mm from centre. The dose at the nine tally cells was

calculated with and without the treatment applicator present. Photon transport only was

considered, calculating F6 tally with 1E9 source particle histories. An example MCNP5 input

file with an Ir-192 source and Varian titanium applicator is provided in Appendix I.

135

Table 7.1. Specifications of metallic intrauterine treatment applicators in use during the UK

brachytherapy audit (data from personal communication with manufacturers, at stated

precision), and availability of attenuation correction in manufacturers’ treatment

planning systems.

Manufacturer Material Internal radius,

Wall thickness,

External radius,

(mm)

Brachytherapy

sources used

Attenuation correction

in treatment planning

system (TPS)?

Varian Titanium

(Stainless steel

applicator also

available but

not

encountered

during UK audit)

1.0,

0.5,

1.5

Ir-192 No

(Available in Acuros®

TPS, but not

encountered during UK

audit)

Nucletron Titanium 1.01,

0.48,

1.49

Ir-192 Optional: single

correction factor

(Used at only 1 centre

during UK audit)

Eckert &

Ziegler Bebig

Titanium

(Stainless steel

applicator also

available but

not

encountered

during UK audit)

1.75,

0.75,

2.5

Co-60

(Ir-192 also

available but

not

encountered

during UK audit)

Yes: single correction

factor applied

automatically

136

Figure 7.5. MCNP5 geometry used to investigate the dose attenuation caused by metallic

brachytherapy treatment applicators compared to a homogeneous water phantom

(where the applicator is replaced by water, not shown).

7.3. Results

7.3.1. Validation

Figure 7.6 and 7.7 show results of the calculations used to validate the local

implementation of MCNP5 and the simplified source models. The calculated dose rate with

radial distance from the source were compared to consensus data from Perez-Calatayud et

al. 2012b, normalised at 10 mm, for Ir-192 in Figure 7.6(a) and for Co-60 in Figure 7.7(a), over

a range 7.5 to 70 mm, (distance range required for the UK brachytherapy audit). The standard

error, R, for the MCNP5 tallies were all < 0.05 except for the case of 1E6 histories for *F8:p,e

tally. This curve showed the greatest departure from the consensus data and it was clear an

increased number of particles histories were required. Figures 7.6(b) and 7.7(b) show the

percentage difference of the MCNP5 F6:p tally from the consensus data, with 1E6 particle

histories within 2%, 1E7 within 1%, 1E8 within 0.5% and 1E9 within 0.2%. Figures 7.6(c) and

7.7(c) show greater uncertainty of *F8:p,e tally against the consensus data than for the F6:p

tally, with 1E6 particle histories having up to 25% disagreement, improving to 0.8% at 1E9

histories.

137

7.3.2. Phantom Scatter

Figure 7.8 shows MCNP5 results of the percentage change in dose at tally cell

radii in the range 20 to 75 mm, between an unbounded water phantom and spherical water

phantoms of radii 100 to 400 mm, with centrally located point sources of (a) Ir-192 and (b)

Co-60. The difference in dose between unbounded and finite sized water phantoms was

more significant at greater radial distances and for smaller phantoms, and of greater

magnitude for Ir-192 than for Co-60. At a radial measurement distance of 75 mm, a scatter

correction of 9.0% was required for Ir-192 within a 100 mm radius phantom, 1.1% for a 200

mm radius phantom and 0.1% for a 300 mm radius phantom. Perez-Calatayud et al. (2004)

modelled the same geometric conditions, finding 9%, 1%, 0%, respectively for the above

cases. MCNP5 calculations were also performed for a water sphere of radius 230 mm,

equivalent to a cube of side length 400 mm, this dimension having been selected based on

the above data, as a compromise between low scatter correction and practicality of

transportation for an audit phantom. For the BRAD audit phantom (Section 6.4.1), the dose

at 20 mm radial distance was of particular importance, corresponding to the clinical

prescription Point A, and also at the maximum extent of the measuring film, 60 mm radial

distance. The scatter corrections at radii 20 mm and 60 mm in a 230 mm radius sphere were

0.0% and 0.4% for Ir-192 and 0.0% and 0.2% for Co-60, respectively.

7.3.3. Film Dosimetry Dose Perturbation

Table 7.2 presents results of the MCNP5 calculations comparing the dose

delivered to a water tally cell, of dimensions equivalent to EBT3 film active layer, with and

without 15 mm width of EBT3 film between the source and tally cell. For both Ir-192 and Co-

60 sources, there is a negligible percentage change in dose due to the presence of the film.

Table 7.2. Results of MCNP5 calculation comparing dose deposition to a water cell of dimensions

equivalent to EBT3 film active layer, with and without a 15 mm width film between the

source and tally cell, for Ir-192 and Co-60.

Source type: Ir-192 Co-60

*F8:p,e tally

(MeV/source

particle)

[relative error, R]

With 15 mm width of EBT3

film between source and

tally cell

3.379E-5

[0.0027]

3.421E-5

[0.0016]

With water between source

and tally cell

3.380E-5

[0.0030]

3.425E-5

[0.0016]

Ratio of *F8:p,e tally with 15 mm film compared to

no film

1.000 0.999

% change in dose due to presence of film 0.0% -0.1%

138

Figure 7.6. (a) MCNP5 calculation of dose rate with distance from an Ir-192 source compared to

consensus data from the High Energy Brachytherapy Dosimetry Working Group, HEBD

WG (Perez-Calatayud et al. 2012b), as a function of number of particles and tally type.

(b) Percentage difference for tally F6:p. (c) Percentage difference for tally *F8:p,e.

139

Figure 7.7. (a) MCNP5 calculation of dose rate with distance from a Co-60 source compared to

consensus data from the High Energy Brachytherapy Dosimetry Working Group, HEBD

WG (Perez-Calatayud et al. 2012b), as a function of number of particles and tally type.

(b) Percentage difference for tally F6:p. (c) Percentage difference for tally *F8:p,e.

140

Figure 7.8. MCNP5 calculated percentage difference in dose at 20 to 75 mm radial distance from (a)

Ir-192 and (b) Co-60 point sources, for spherical water phantoms of radii 100 to 400 mm

compared to unbounded water phantom (600 mm radius); Compared to data from

Perez-Calatayud et al. (2004) for Ir-192 (note 25 mm rather than 20 mm in the latter

data set).

141

7.3.4. Applicator Attenuation

Figure 7.9 presents results of the MCNP5 calculations of applicator attenuation,

for Ir-192 with Nucletron titanium applicator and for Co-60 with Eckert & Ziegler Bebig

titanium applicator. The percentage reduction in dose at three radial distances is plotted for

three levels along a 60 mm active length of a typical intrauterine applicator, as involved in

the UK brachytherapy audit. The MCNP5 F6:p tally relative error was <0.003 for all

calculations. The effect of applicator attenuation is dependent on the level of obliquity of the

radiation path through the metallic applicator. This is increased for measurement points

radially closer to the applicator or for points closer to the end, or beyond, the active source

length (AL). At the clinical dose prescription Point A, shown as circled data points on Figure

7.9, there is a 1.5% reduction in dose for Co-60 irradiation, and 2.0% reduction for Ir-192. At

30 mm beyond the end of the active length, at the same radial distance, the attenuation

increases to 3.8% for Co-60 and 4.3% for Ir-192.

Figure 7.9. Results of MCNP5 calculations of the percentage reduction in dose deposition due to

metallic applicator attenuation, for an intrauterine applicator of 60 mm active source

length (AL), at levels of mid-AL, 10 mm from mid-AL, and 30 mm beyond AL, for radial

distances of 20, 40 and 60 mm. The data for clinical prescription Point A are circled.

142

7.4. Conclusions and Discussion

The local implementation of MCNP5 and the source geometries and materials

used were validated for use in the brachytherapy scenarios considered in this thesis, in which

radial dose alone is of significance. 1E9 histories were required for *F8:p,e tally MCNP5

calculations to provide agreement better than 1% of radial dose rate to published consensus

data. 1E8 was sufficient for F6:p tallies, however due to the significantly faster calculation

time of photon-transport only calculations, 1E9 was used for both scenarios.

Scatter corrections were calculated as a function of measurement position and

phantom size within spherical water phantoms compared to an unbounded case. The results

were consistent with data provided by Perez-Calatayud et al. (2004) for Ir-192, who also

reported lower scatter corrections for the higher energy Cs-137 compared to Ir-192 (mean

energies 0.6617 MeV and 0.3722 MeV, respectively). This is consistent with lower corrections

calculated for Co-60 (mean energy 1.2520 MeV) compared to Ir-192 in the present work. The

radial dose function for Co-60 is lower than Ir-192 up to 120 mm from the source but scatter

from the higher energy source is much more forward-biased, with back-scatter and hence

phantom-scatter to measuring points being more significant from Ir-192. For film dosimetry

measurements in the BRAD phantom (Section 6.4.1), at the prescription Point A, no scatter

correction was required and measurements could be assumed to be equivalent to an

unbound phantom. The scatter correction increases with radial distance and at the edge of

the measuring film, a correction of 0.4% for Ir-192 and 0.2% for Co-60 has been estimated.

This was included in the uncertainty budget but no correction made to measured film doses.

Films are held in the BRAD phantom such that a proportion of the radiation from

the HDR brachytherapy source travels along the length of the film. MCNP5 calculations have

shown that there is a negligible perturbation of the delivered dose due to the presence of

Gafchromic EBT3 film in these irradiation situations.

The attenuation of delivered dose due to metallic applicators was generally not

included in treatment planning system calculations encountered during the UK

brachytherapy audit. The results of MCNP5 calculations estimate the reduction in dose at

prescription Point A (20 mm radial distance) from an extended source active length, as 1.5%

for Co-60 and 2.0% for Ir-192. Ye et al. (2004) gave the applicator attenuation factor for a

single Ir-192 source dwell position as 1.3% at 20 mm radial distance. It is expected that a

series of dwells creating a longer active length would exhibit a larger correction, due to

source contributions with increasing obliquity through the applicator walls. The User Manual

for the Eckert & Ziegler Bebig treatment planning system, which unusually does include a

single global correction for metal applicator attenuation, gives a dose reduction factor for

titanium applicators of 1.3% for Co-60 and 2.0% for Ir-192, which is consistent with the above

MCNP5 calculations. The MCNP5 calculated attenuation factors were used to interpret

results of the UK brachytherapy audit for centres that used metallic applicators with no dose

correction at the treatment planning system.

143

CHAPTER 8

‘End to End’ Audit of Clinical

Brachytherapy Dosimetry in the

United Kingdom

144

8.1. Objectives for the UK Brachytherapy Audit

The objective of this work was to undertake the first comprehensive national

audit of physics-aspects of brachytherapy dosimetry in the UK. Concerned primarily with

dose measurement in comparison to planned, the audit also included discussion of clinical

brachytherapy processes with local physicists.

The BRachytherapy Applicator Dosimetry (BRAD) phantom, described in Section

6.4.1, provided an audit tool for the comparison of planned and delivered dose distributions

in 2D planes around clinical gynaecological applicators with HDR or PDR brachytherapy

treatment equipment using Ir-192 or Co-60 sources. The process constitutes a full system

‘end to end’ audit which provides a combined performance assessment of CT scanning,

reconstruction of the applicator, dose prescribing, dose calculation and source data in the

treatment planning system (TPS), plan export consistency, delivery performance at the

treatment unit with clinical treatment applicators, and inherent confirmation of source

calibration data in the TPS. The advanced film dosimetry methods described in Chapter 5 are

utilised for the audit dose measurement.

8.2. Methodology

8.2.1. Establishment of a Working Party of the Institute of Physics and Engineering in

Medicine (IPEM)

The author submitted a request to the Institute of Physics and Engineering in

Medicine (IPEM) for funding to establish a Working Party for brachytherapy audit, and when

this was approved, chaired the group. The working party was established to primarily bring

in funding, but also to coordinate brachytherapy audit activity and ensure no duplication of

effort between three separate audits that were coincidentally being conducted in 2013-2014

in the UK: BRAD ‘end to end’ audit (the current work); well-chamber intercomparison audit

(source strength verification); INTERLACE clinical trial audit (dose to a point from a short

active length in a straight plastic catheter). The funding request had been made to cover the

cost of construction of BRAD phantom and expenses for conducting the UK audits. The author

obtained separate funding from Ashland, USA, to cover the cost of Gafchromic EBT3 film for

the audits.

8.2.2. Brachytherapy Applicator Dosimetry ‘End to End’ Audit Methodology

The audit methodology proposed in Section 6.4.2 was used, with the BRAD

phantom described in Section 6.4.1. The full audit protocol is provided in Appendix D, and a

risk assessment for conducting the audit summarised in Appendix C. The audit process was

designed to match as closely as possible to that for a patient treatment. A typical gynaecology

cervix treatment applicator was CT scanned in approximate clinical orientation using local

brachytherapy acquisition protocols. The data was exported to the TPS, applicator

reconstructed using normal clinical methods, and a treatment planned using normal clinical

145

procedure and typical dose prescription. The DICOM RTDose file was exported for later

analysis, and the treatment data sent to the HDR or PDR unit and treatment delivered exactly

as for a patient. Gafchromic EBT3 films were placed in the BRAD phantom prior to treatment

delivery to measure the delivered dose.

During the audit at each radiotherapy centre, the local brachytherapy physics

practice was informally discussed. The intention was to discover any significant departures

from accepted standard practice across the UK, any opportunities for improvement, or any

particular safety risks that may be identified. Following the audit, feedback was elicited via

email to evaluate the process and any suggestions for development or further work.

Any quantitative results of the audit which deviated significantly from mean

values or any notable discrepancies in physics brachytherapy procedures from accepted

standard practice across the UK were discussed with the lead physicist for brachytherapy or

the head of radiotherapy physics at the host centre. The results of the audits were

confidential and only anonymised data is reported here and in publications. The limitations

of the scope of the audit were also discussed (see Appendices C and D).

8.2.3. Data Analysis

Audit films were analysed using the methodology described in Section 6.4.3:

triple-channel dosimetry and linear dose-scaling with reference dose films scanned

simultaneously with the audit films. The DICOM RTDose file was exported from the audited

centre’s TPS and analysed in FilmQAPro software against the film-measured dose maps. The

percentage dose difference between film-measurement and TPS-calculation was evaluated

at prescription Point A (Viswanathan and Thomadsen 2012) along with a distance to dose

agreement on the film. The difference in 2D dose distribution planes around the clinical

treatment applicator between film-measured and TPS-calculated doses were evaluated using

isodose overlay and gamma analysis (Low et al. 1998, Low and Dempsey 2003).

To investigate if there was any correlation between the results of the percentage

difference in planned and measured Point A doses and other defined variables such as type

of applicator and manufacturer, unpaired two-tailed t-tests were applied to the data. This is

a measure of the strength of a linear association between two variables. However, such

analysis is problematic as there are a multitude of influences on the overall accuracy of dose

delivery in the brachytherapy audit. A perceived relationship may be incorrectly estimated

due to the failure to account for confounding factors, i.e. omitted-variable bias. The

difference in Point A agreement between metal and plastic applicators is assessed. Any

difference in Point A agreement with other variables is tested for metal and plastic

applicators separately such that the material type doesn’t confound the analysis.

146

8.2.4. Schedule of Audits

Figure 8.1 provides a map of all audited centres in the UK. Two pilot audits were

undertaken in March and April 2013, with the remaining audits being conducted from May

2013 to August 2014. A total of 46 of the 47 brachytherapy centres in the UK were audited.

The majority of audits were personally conducted by the author (83%, n=38), with three

other physicists auditing several centres (n=8 total), having received prior training from the

author. The data analysis for every audit was performed by the author.

Figure 8.1. Map of 46 radiotherapy centres that participated in the BRAD national audit of clinical

brachytherapy dosimetry in the United Kingdom during 2013 and 2014. Marker colours

indicate different auditors.

147

8.3. Results

8.3.1. Point A Dose Measurement for Cervix Treatment Applicators

Figures 8.2 and 8.3 show results of the differences in Point A dose between film-

measured and TPS-calculated, for the 46 audited brachytherapy centres. The standard

uncertainty in this parameter was estimated at 3.0% (Section 6.4.5), represented as vertical

lines on the Figures. Point A is defined laterally to both the left and to the right of the

treatment applicator, and Figure 8.2 shows the percentage dose difference values for both

the left and right films at Point A. There is a mean difference of 1.6% between the left and

right Point A values, range 0.0 to 6.0% across all audits. All left and right Point A doses are

consistent within the estimated standard uncertainty. Figure 8.3 shows the mean of the left

and right Point A percentage differences for each audited centre, with the data categorised

by applicator type; plastic, metal, or metal with dose attenuation correction in the TPS. The

mean Point A difference for the plastic applicators was -0.6% (1.3% sd), the mean for the

metal applicators was -3.0% (1.3% sd), and the mean for the metal applicators with

attenuation corrected in TPS was -0.4% (two data values, 0.5 and -1.2).

Table 8.1 provides data on the statistical significance of differences in Point A

dose agreement values with various parameters of the audit data, including construction

material of applicator, ring or ovoid type, manual or TPS-model based applicator

reconstruction, and manufacturer, the latter groups being further subdivided by material

type to avoid omitted-variable bias. There is a high statistical significance of the difference

between metal and plastic applicators for the percentage dose difference at Point A, p<0.001

in unpaired two-tailed t-test. A statistical significance, p=0.02, was seen for the mean

percentage difference at Point A between plastic ring and plastic ovoid applicators. All other

tests showed no statistical significance, comparing metal ring and metal ovoid applicators,

reconstruction method, or manufacturer.

Figures 8.4 and 8.5 show results of the ‘distance to agreement’ (DTA) between

TPS-calculated and film-measured doses at Point A, for the 46 audited brachytherapy

centres. An uncertainty in the measured DTA has been estimated using a typical maximum

dose gradient at Point A of 6% mm-1 and the 3.0% standard uncertainty in dose difference at

Point A (Section 6.4.5), deriving a standard uncertainty in distance to agreement of 0.5 mm

(represented as vertical lines on the figures). Figure 8.4 shows the DTA values for both the

left side and right side Point A, for each audited centre. The mean difference in DTA value

between left and right side for each centre is 0.4 mm, with a range 0.0 to 1.0 mm. Figure 8.5

shows the mean DTA for each centre, with the data categorised by applicator type; plastic,

metal, or metal with dose attenuation correction in the TPS. The mean DTA for plastic

applicators was 0.5 mm (0.2 mm sd) and for metal applicators was 0.7 mm (0.3 mm sd), with

a combined mean DTA for all data values of 0.6 mm, range 0.1 to 1.2 mm.

Pearson’s product-moment correlation coefficient was used to assess

correlation between CT slice width and percentage dose difference at Point A and DTA at

Point A. No correlation was found in either case, r = -0.09 and -0.05 respectively.

148

Figure 8.2. Percentage dose difference, film-measured compared to TPS-planned, at Point A left

and Point A right, for cervix treatment applicators, at 46 audited centres.

Figure 8.3. Percentage dose difference, film-measured compared to TPS-planned, mean Point A for

cervix treatment applicator, by applicator type, at 46 audited centres. (Mean of each

data set shown as horizontal lines)

149

Table 8.1. p-values and statistical significance of unpaired two-tailed t-test comparing dose

difference at Point A (film-measured to TPS-planned) for various grouped data from the

46 audits. Table includes number (n), mean, standard deviation (sd), and standard error

of the mean (sem) for each of the grouped data sets.

Material:

Plastic v.

Metal

Type:

ring v. ovoid

(plastic)

Type:

ring v. ovoid

(metal)

All

pla

stic

app

licat

ors

All

met

al

app

licat

ors

Pla

stic

rin

g

app

licat

ors

Pla

stic

ovo

id

app

licat

ors

Met

al r

ing

app

licat

ors

Met

al o

void

app

licat

ors

n 26 18 19 7 8 9

Mean -0.63 -2.96 -0.27 -1.57 -3.00 -3.09

sd 1.32 1.18 1.13 1.39 1.10 1.28

sem 0.27 0.28 0.27 0.53 0.39 0.43

p-value <0.001 0.023 0.881

Statistical

significance

Highly

significant Significant

Not

significant

Reconstruction:

manual v. model

(plastic)

Reconstruction:

manual v. model

(metal)

Manufacturer:

Varian v. Elekta

(plastic)

Manufacturer:

Varian v. Elekta

(metal)

Pla

stic

man

ual

-

reco

nst

ruct

ion

Pla

stic

TP

S-

reco

nst

ruct

ion

Met

al m

anu

al-

reco

nst

ruct

ion

Met

al T

PS-

reco

nst

ruct

ion

Pla

stic

Var

ian

app

licat

or

Pla

stic

Ele

kta/

Nu

clet

ron

app

licat

or

Met

al V

aria

n

app

licat

or

Met

al E

lekt

a/

Nu

clet

ron

app

licat

or

n 12 14 12 6 0 25 15 3

Mean -0.79 -0.49 -3.03 -2.83 -3.11 -2.20

sd 1.39 1.30 1.25 1.14 1.23 0.46

sem 0.40 0.36 0.36 0.47 0.32 0.27

p-value 0.573 0.76 (no plastic Varian

applicators) 0.23

Statistical

significance

Not

significant

Not

significant

Not

significant

150

Figure 8.4. Distance to agreement, film-measured compared to TPS-planned, at Point A left and

Point A right, for cervix treatment applicators, at 46 audited centres.

Figure 8.5. Distance to agreement, film-measured compared to TPS-planned, mean Point A for

cervix treatment applicator, by applicator type, at 46 audited centres. (Mean of each

data set shown as horizontal lines)

151

8.3.2. Dose Distribution Measurement for Cervix Treatment Applicators

Figure 8.6 shows isodose comparisons between film-measured and TPS-

calculated dose distributions from three audits using the BRAD phantom, typical of the

results of the national brachytherapy audit. Two of the four films from each audit are shown

in the Figure; planes left lateral and anterior to the cervix treatment applicator. (The opposite

planes were similar and are not reproduced here). Inset in the Figure are 2D planes through

the DICOM RTDose data, with the location of the evaluated film region shown; the bright

areas are high dose regions in the RTDose data, corresponding to the position of the

treatment applicator. Good agreement is seen between the film and TPS dose distributions,

generally better than 1 mm deviation for isodose lines in the range 2 to 16 Gy, with deviations

increasing at lower doses further from the applicator, up to 2 mm for isodoses 0.7 to 1.5 Gy.

Figure 8.7 shows gamma passing rates of the film-measured and TPS-calculated

dose distributions, for the 46 audited brachytherapy centres. The data are the mean passing

rate of all four films for each audit, with the data categorised by applicator material and

gamma evaluation criteria (all locally normalised percentage difference and with 2 Gy lower

cut-off threshold). The film dose distributions were resampled to the lower TPS RTDose grid

resolution and interpolated as necessary for analysis (via FilmQAPro software), to prevent

film noise artificially increasing the passing rate. The gamma passing rates for centre 33 on

Figure 8.7 are below the typical range of other centres, likely the result of library applicator

misalignment, discussed below in Section 8.3.4.1. With this outlier omitted, the gamma

passing rate for all audits exceed 98% at 5% (local) 3 mm criteria, and exceeded 95% at 3%

(local) 2 mm criteria, except for centre 16, at 92%. The TPS at this centre (Oncentra GYN) had

a grid resolution limitation of 2.1 mm rather than the usual 1 mm for audit data, hence it is

unsurprising the results for gamma with small distance criteria may be compromised. The

gamma passing rate at 2% (local) 1.5 mm are more variable, within a range 82% to 99%. Table

8.2 provides the mean and standard deviation gamma passing rates, at the three criteria, for

all 46 audited centres.

152

Figure 8.6. BRAD dose distribution comparison between TPS-calculated (thick lines) and film-

measured (thin lines), isodose lines in the range 0.7 Gy to 16 Gy. The graphs to the left

are planes lateral to a cervix applicator, to the right are planes anterior to a cervix

applicator, on the axis of the intrauterine (IU) tube. Inset images are the RTDose planes

showing the evaluated region of interest. For (a) 70 mm IU and ovoid applicator, (b) 60

mm IU and ring applicator, and (c) 40 mm IU and ring applicator.

153

Figure 8.7. Mean gamma passing rate, at three different criteria, comparing film-measured to TPS-

calculated 2D dose distributions in four planes around cervix treatment applicators, at

46 audited centres.

Table 8.2. Mean and standard deviation (sd) gamma passing rate, at three different criteria, for 46

audited centres

n Gamma passing rate at:

5% (local), 3 mm 3% (local), 2 mm 2% (local), 1.5 mm

mean (sd) mean (sd) mean (sd)

All applicator types 46 99.7 (1.0) 97.8 (3.2) 90.0 (4.8)

Plastic applicators 26 99.6 (1.3) 97.6 (3.7) 89.3 (4.6)

Metal applicators 18 99.9 (0.2) 98.4 (1.9) 91.0 (4.9)

Attenuation corrected

metal applicators (data

values only)

2 99.7, 100.0 95.3, 99.2 85.2, 93.8

154

8.3.3. Review of Physics Procedures for HDR Brachytherapy Treatment

The national brachytherapy audit afforded the opportunity to review local

brachytherapy physics practice with host physicists and compare to accepted standards

across UK. This was a valuable outcome of the audit visits. A summary of findings which

provided opportunities for development or further review, that were reported to audited

centres, is provided below.

Brachytherapy equipment

Occasionally the local brachytherapy physicist had difficulty assembling the treatment

applicator, and was unsure of correct alignment. In one audit, the applicator was initially

assembled with the ovoid curvature in the wrong direction with respect to the IU

curvature. In two cases the physicist was unsure of the correct ring cap dimension to be

used, or whether this made a difference. The cervical stopper was omitted or located in

the wrong position with respect to the ovoids in two audits.

Imaging

The majority of centres use CT reconstruction slice width of between 1.0 and 2.0 mm

for brachytherapy treatment planning. In 4 of the 46 centres, 3.0 mm slice width was

used, which has potential to increase uncertainties in applicator reconstruction.

TPS – applicator reconstruction and use

There were several occasions when the TPS applicator library did not match as well as

would be expected to the CT-image of the applicator, in terms of length of IU, curvature

of IU, tilt of ring with respect to IU, or relative height of ovoids. At two centres, it was

not clear which applicator library should be used to match the physical applicator used

for the audit. In one case, the difference between library applicator and imaged

applicator was sufficiently great to seek further investigation by the manufacturer.

(Section 8.3.4.1).

A few centres appeared to lack confidence, and QC evidence, of the physical position of

the first distal dwell position in an applicator in relation to the CT image of that

applicator on the TPS. Several different techniques were in use to confirm the dwell

positions were correctly positioned on the CT data along the IU. In two audits the locally

expected distance from tip to first dwell was not reproduced when the auditor asked for

a measurement to be made on the TPS.

In 1 audit, the treatment planner placed the ring source path at the wrong level on the

CT image of the applicator, due to sub-optimal image windowing, and the auditor had

to intervene prior to plan finalisation.

Depending on the TPS, Point A may be automatically imported along with the applicator

library, and subsequent movement of the applicator can change position of Point A if

they are linked. In one audit this was not appreciated by the planner and the auditor

had to intervene to correct the position of Point A.

155

2 of the 20 centres using metal applicators make a 2% correction in the TPS for

attenuation, the other centres do not. A few physicists were unaware whether their TPS

made a correction, and the magnitude of attenuation.

TPS - planning procedures

There were several centres that had different definitions of prescription Point A

compared to consensus view, or used an alternative prescription point. For ring

applicators, the vast majority define Point A as 20 mm up from the surface of the ring,

often measured as 20 mm + cap dimension from level of sources in ring. One centre

measured 20 mm from the level of the sources to define their Point A, a few others used

a cervical stopper rather than the ring surface. One centre used their own locally defined

point C for prescription, 7 mm lateral to the ring plastic edge at the level of the sources.

Another centre defined Point A laterally from a specific dwell point in the IU. For ovoids,

some use the centre of the cervical stopper (flange) while others use the distal surface.

In a few centres, the standard loading pattern, used as a starting point for manual

optimisation, had unexplained rotation of dwell positions in the ring to be not laterally

symmetric, and one had dwell positions in the IU giving an active length approximately

2 cm proximal to the level of the ring (noted to be an error during the audit). In several

audits, the planner stated the ‘standard loading pattern’ is always amended for clinical

use, such as reducing the library plan loading of 6 dwells each side of the ring to 4, or

rotating the dwells in the ring.

In one centre, a planner did not appreciate a difference between the TPS allowed step

size and the actual step size limitations at the HDR treatment unit, causing a systematic

difference between planned and delivered doses of 0.5 mm.

Variations in the level of optimisation of HDR brachytherapy treatment planning across

the UK is significant. The process for individual planning of patients’ treatments ranged

from delivery of standard plans with no optimisation, indeed one centre only performed

retrospective dosimetry on the TPS after treatment delivery had taken place, to fully

inverse-planned treatments to 3D target volumes (HR-CTV) based on MRI imaging. The

majority of centres (approximately three-quarters) undertook a reasonable level of

manual optimisation based on CT-derived 3D target volumes.

Data transfer

1 centre did not use network transfer of plan data from the TPS to the treatment unit,

instead relying on CD data transfer. Two centres planned the treatment at the TPS using

standard loading, but then did not transfer any data to the treatment unit and instead

recalled the standard plan at the treatment unit.

Quality control

3 of the 46 audited centres had no independent check method for TPS calculations,

although these were all in centres where standard plans were delivered and instead had

consistency checks against local data tables.

156

At one centre the local physicist stated that film measurements “several months ago”

had shown dwell positions in the ring were in error by 5 mm on one side, but that a

correction was awaiting independent verification prior to clinical implementation. At the

request of the local physicist the audit was conducted with the dwell position correction

implemented since “the current clinical technique was known to have source position

errors”. (This was followed up with the Head of Radiotherapy Physics at the centre, who

stated their TPS studies had shown no significant change in clinically relevant plan

dosimetry).

There is a variation in QC measurements performed, such as whether to routinely assess

dwell position accuracy in clinical applicators including ovoid channels. (In 2013 a

significant error was publicised in one centre in Japan in which dwell positions in ovoids

had significant error “about 3 cm”, affecting 100 patients. In 2014, a different

manufacturer issued a field safety notice indicating dwell positions in ovoids may be 2-

3 mm from expected positions. It is recommended to periodically check dwell positions

in all applicators).

8.3.4. Results Requiring Further Investigation

8.3.4.1. Library Applicator Misalignment

There was only one audit in which a clear discrepancy was seen between the

TPS-calculated and film-measured dose distribution on isodose overlay analysis, as illustrated

in Figure 8.8. A good agreement is seen between TPS-calculated (thick lines) and film-

measured (thin lines) isodoses in the audit results of Figure 8.8(a), obtained from a different

audit for comparison. By contrast, in Figure 8.8(b), there is a clear discrepancy between the

film and TPS isodose lines at the level of the distal tip of the IU applicator. The Figure shows

the TPS applicator library overlay on the CT image, which is in good agreement in 8.8(a) but

there is a clear discrepancy in 8.8(b) resulting in shift in TPS modelled dwell positions of

approximately 3 mm. In the audit of Figure 8.8(b), the TPS applicator library was correctly

aligned with the CT image at the level of the ring, but there was a difference in the apparent

length of the IU applicator of approximately 3 mm. With retrospective analysis, there were

only two other centres using the same specific applicator design in the UK audit: in one case

a similar discrepancy was seen but of a lower magnitude, isodose shift around 1 to 2 mm,

and in the other this discrepancy was mitigated in routine clinical practice by decoupling the

IU and ring applicators and manually aligning them with the CT image. The results were

reported to the manufacturer who conducted an investigation and stated “applicator

libraries should be considered as an ideal representation but may deviate due to production

tolerances, mechanical forces within a patient, or incorrect sterilization and handling”

(personal email to author, September 2014). The manufacturer concluded that “The

difference in tip position between Applicator Library model and imaged applicator can be

caused by product variations. Recommendation is to use the Applicator Library in

combination with the appropriate CT markers to ensure accurate positioning of the model”

(personal email to author, October 2014).

157

Figure 8.8. BRAD audit isodose comparison between TPS-calculated (thick lines) and film-measured

(thin lines) for (a) audit with good agreement, and (b) audit with discrepancy at the distal

end of the IU applicator. Inset enlargements of CT images of distal tip of IU with TPS

library applicator overlay and source dwell positions.

158

8.3.4.2. Prescription Dose Error

There was only one audit in which the initial film results indicated a potentially

significant dose error. Subsequent analysis was able to find ‘human factors’ as a cause and

there was no systematic error in dosimetry at the audited centre. However, this served as a

reminder that ‘human error’ may be a significant cause of incorrect treatment delivery.

Figure 8.9 shows excerpts of the TPS planning information hard copy. The

intention of the audit was to deliver 7 Gy prescription dose, and this was entered in the TPS

and is shown on the front page of the TPS data sheet, Figure 8.9(a). Analysis of the audited

film dose at Point A gave a measured dose value of 10.2 Gy, a 46% increase on the

prescription value. Subsequent analysis confirmed the TPS had calculated the Point A doses

correctly, shown in Figure 8.9(b), but the point of normalisation used in the TPS had been

incorrectly set by the planner. (The actual TPS dose to Point A and the film measured doses

are used in the earlier sections of this Chapter to derive percentage difference and DTA

results).

Figure 8.9. BRAD audit with incorrect dose normalisation. Section of plan parameters hard copy

from TPS, showing (a) prescription dose of 7 Gy and (b) calculated dose to Point A

average 10.2 Gy

159

8.3.5. Feedback from Audited Centres

Of the 46 radiotherapy centres audited, 25 provided feedback after the audit via

email. All responses were positive and found the audit to be of value. A representative

selection of comments is provided below:

Administration and set-up

“Thank you very much for the quick response with the detailed report.”

“Thank you for the comprehensive report, a valuable exercise. It’s reassuring to know

that the dose distribution is good and that there weren’t any glaring discrepancies.”

“The process was well organised, including communications pre and post audit,

equipment delivery, documentation.”

“Brachytherapy physics dosimetry audit has lagged behind external beam so it's great

that the process has at last started.”

“I think the audit was very good. It was quick, required minimal preparation from us,

and measured something that was clinically relevant. We are very satisfied with the

whole process.”

“I thought the audit was quite painless, especially as you came and did the

measurements yourself and we weren't battling with unfamiliar kit and instructions. It

is always reassuring to know that a more clinical situation has been verified rather than

just the source output.”

“We found the audit process very helpful and reassuring. External audit is a very

important process to ensure consistency across the UK, it is a valuable source of

assurance and gives confidence to the clinical team.”

“I definitely feel that the audit was useful; as far as I can recall it is the first time we have

had an audit on the ring and tandem and got the results back, so it’s reassuring that

we’re giving the doses we think we are.”

“This is the first time we have been audited in this way and it complements the internal

QA we do.”

“I was really impressed with the efficiency and comprehensiveness of the audit process.

The format of the report is similarly comprehensive but succinct, and ideal for sharing

with clinical/radiographer colleagues as a confidence-booster. We also really

appreciated the knowledge-sharing and interaction during the planning process that

your involvement brought to the day.”

“I think the audit was valuable in confirming the whole process from scanning, planning

through to treatment. The process seemed quite efficient as well.”

“The E2E [end to end] tests you’ve carried out here are of a much higher standard than

we could have hoped to achieve in the time available, and we have the added benefit of

being benchmarked against other centres.”

160

Methodology

“You have demonstrated film dosimetry is a good option to measure brachytherapy

dose.”

“We tried to get something sensible with TLD during commissioning but due to the large

dose gradients we had to accept errors of around 15%.”

“It tested the whole of the patient pathway very smoothly.”

“I was very impressed with the audit, in particular the design of a phantom that enabled

such accurate measurements close to a brachytherapy source.”

“The phantom is impressive and a very useful tool.”

Results

“We will certainly have a review of our procedure based on your recommendations set

out in the report.”

“I will now make changes to the CT reconstruction.”

“I am very pleased (and relieved) that the results are good, that is reassuring.”

“Very pleased to see that the results give no cause for concern.”

“General discussions about brachytherapy procedures was very interesting and useful.

The report was prompt, detailed and informative.”

“I take on board your comments and agree… I am reviewing Point A location for our new

applicators.”

“It is reassuring that the results are good despite the issues raised during the audit. I

found the audit very useful and your comments/suggestions will certainly be taken on

board.”

“The audit highlighted the fact that we need to review our whole process to bring us

into line with what other centres are doing and to have confidence in our own working

practices.”

Criticism and suggestions for future audits

“The phantom is somewhat fragile.”

“The only part that was tricky was setting up the phantom, holding down the ring whilst

the clamp was tightened.”

“I got the feeling that it was more designed around the ring system than our ovoids and

that is perhaps why we had a bit of a gap between the flange and the phantom surface

maybe an interchangeable central piece might fix that.”

“Extending the audit to suit other applicators would be great for future work.”

“Possible future audits – maybe a skin type treatment with Frieburg flap and tissue

equivalent phantom rather than just water? Also, possibly the current phantom could

be adapted to take needles as well”

161

“When digitising the applicators, they weren’t as easy to visualise as when they are in a

patient, so that might have introduced a small positional error.”

“My only suggestion would be to develop it further for commercialisation, perhaps by

reworking so that the pieces interlock into a more robust solid cube which wouldn’t

require submersion in water.”

“Voluming and the interpretation of the HR-CTV is one huge area of confusion. Would

be interesting to send an image set round the country and see if the patient would get

the same treatment from different centres.”

“It's too early to audit in vivo dosimetry in brachy yet but that would be good in the

future. Also ... in a few years, non-TG43-based dosimetry.”

“Adapting the phantom to provide for instance a ‘prostate hdr’ audit, would be as useful

as this ‘gynae hdr’ audit. GAF would also appear well suited to test curved ‘brachy

moulds’.”


The BRAD ‘end to end’ brachytherapy dosimetry audit system has been

successfully used to audit 46 radiotherapy centres in UK, with generally good results. The

mean percentage dose difference of the film-measured dose from the TPS-calculated dose

at Point A was -0.6% for plastic applicators and -3.0% for metal applicators, Figure 8.3, with

a high statistical significance of the difference, Table 8.1. The attenuation due to metal

applicators has been calculated using Monte Carlo techniques in Section 7.3.4, giving 2.0%

as an estimate for Ir-192 in the vicinity of Point A, and increasing as path-length obliquity

through the applicator increases. The standard uncertainty of the dose difference between

TPS-calculated and film-measured dose at Point A has been estimated as 3.0% (k=1) in

Section 6.4.5. It follows that all of the audit determined doses at Point A (mean of the Point

A doses for the left and right side films) were within one standard uncertainty from the

planned value for all of the audited centres, if the attenuation of the metal applicators not

modelled in the TPS, is taken into account. The accuracy of dose delivery agrees with the

planned dose within the uncertainty of the audit methodology, which is therefore deemed

to be within clinically acceptable levels.

Differences between film-measured Point A for the left and right side films could

indicate lateral displacement of the source within the applicator, movement of the applicator

intrauterine tube from the central axis of the phantom, offset of applicator reconstruction in

the treatment planning system, unintentional movement of either of the films, or be the

result of inherent uncertainty in the film dosimetry process. All of the left to right side

differences were within the estimated standard uncertainty.

There was a high statistical significance for the difference in agreement between

film-measured and TPS-calculated dose to Point A for plastic and metal applicators, as would

be expected due to the lack of compensation for metal attenuation in TPS. Table 8.1 also

162

showed a moderate statistical significance between ring and ovoid plastic applicators, but

not for ring and ovoid metal applicators. The uncertainty in the physical audit set-up was

considered to be less for ring applicators than for ovoids, especially where the design of the

applicator meant the ovoids were not parallel with the top surface of the BRAD phantom, or

for those plastic ovoid applicators with curved IU tubes. Also the ovoid applicator geometry

is in general not as fixed as for ring applicator, which may have increased uncertainty

between imaging and delivery. The plastic ovoid applicators may be more susceptible to

movement than the metal ovoids between the time of CT and treatment delivery due to the

fixing mechanisms. No statistically significant difference was seen between manual or TPS-

model-assisted applicator reconstructions for either metal or plastic applicators. The overall

accuracy is likely to be primarily a function of the skill, experience and knowledge/training of

the operator rather than whether an applicator-model library is available in the TPS.

However, it was seen that an applicator model library is a more rapid method of applicator

reconstruction, and could be expected to improve efficiency. However, this does rely on the

applicator library model being an accurate representation of the true physical situation, and

care must be taken for each patient to ensure the applicator library and actual imaged

applicator are in geometric agreement, otherwise errors are possible as discussed in Section

8.3.3 and 8.3.4.1. There was an insufficient sample size for reliable statistics to compare

Varian and Nucletron treatment systems without omitted-variable bias: There were no

plastic Varian applicators in use, and a very small number of metal Nucletron applicators.

However, the mean value for Nucletron and Varian metal applicators is in agreement within

one standard deviation.

The mean distance to agreement, DTA, for all audits was 0.6 mm, which is of the

same order as the derived standard uncertainty estimate of 0.5 mm. The DTA for metal

applicators was higher than for plastic applicators, 0.7 mm compared to 0.5 mm. This is most

likely a result of the metal attenuation not accounted for in the TPS calculation. The DTA is

always a positive value for both left and right Point A films, hence any displacement of the IU

applicator from the central axis of the phantom will increase the DTA on both films, and the

mean DTA, in contrast to the mean percentage dose difference at Point A in which applicator

shifts may increase dose in one film compensated by a decrease in the other film’s Point A

location.

Excellent agreement was seen between film-measured and TPS-calculated

isodose distributions over a wide dose range, generally 0.7 to 16 Gy analysed film, shown in

Figure 8.6. The gamma evaluation of the two dose distributions for each audit had passing

rates exceeding 95% at 3% (local) 2 mm for all audits, except one centre in which a

misalignment was evident between the TPS applicator library and the CT-imaged applicator,

discussed in Section 8.3.4.1. The mean gamma passing rate over all audits was 97.8% (3.2%

sd) at 3% (local) 2 mm and 99.7% (1.0% sd) at 5% (local) 3 mm. It may be concluded that no

errors were detected in the basic TG-43 source data in use within the TPS at each of the

audited centres, and that treatment equipment performance was sufficient to faithfully

delivery the calculated dose distribution within clinically significant parameters, for the

situations and locations measured.

163

25 of the 46 audited centres provided feedback; all finding the process to be a

positive experience. In particular the process was perceived as being novel for brachytherapy

and valuable to cover the ‘end to end’ dosimetry process. The associated discussion of

physics brachytherapy processes was seen to be valuable and host physicists were happy to

receive suggestions for development.

164

165

CHAPTER 9

Summary Conclusions

and Future Work

166

9.1. Summary Conclusions

This section provides an overview of principal conclusions from the distinct

sections of this thesis, and cites relevant publications by the author. Further discussion and

conclusions are provided at the end of each relevant preceding chapter. The research work

also included an initial publication reviewing physics-aspects of dose accuracy in HDR

brachytherapy (Palmer et al. 2012b), and concluded with an editorial on brachytherapy

dosimetry (Palmer 2014a) and a publication reviewing dosimetric audit in brachytherapy

(Palmer et al. 2014b). Several oral and poster presentations were also given at major national

and international conferences, including in the UK (Brighton, Glasgow, London and

Portsmouth), mainland Europe (Barcelona, Geneva and Vienna), North America (Austin), and

Australia (Sydney), details provided in Section 1.4.

9.1.1. Quality Control of HDR Brachytherapy Equipment

9.1.1.1. UK Survey and Analysis

Chapter 3 provided data and analysis on a comprehensive survey of quality

control practice for HDR/PDR brachytherapy, which collated data from 31 clinically active

brachytherapy centres in the UK. This was the only systematic analysis of its kind to be

undertaken in recent years and provides a valuable benchmark resource. While there was

agreement on the need for QC testing of basic performance parameters, such as source

strength, QC schedules differed significantly in other areas, such as the need for routine

testing of dwell position accuracy in clinical treatment applicators. There was no evidence of

‘system tests’ or confirmation of delivered dose distributions with clinical treatment

applicators, and this was identified as an objective for further study in the thesis. Decisions

on QC testing should be based on risk assessment and local factors, and it is recommended

centres should review their practice against the UK survey benchmark data presented in this

thesis and associated publication (Palmer et al. 2012a).

9.1.1.2. Development of QC and Commissioning Tests and Performance Requirements

From results of the UK survey, dwell position accuracy, transit dose, and end to

end process dosimetry were identified as QC tests that would benefit from procedure

development. A proposed method for the assessment of source dwell position, movement

profile, and transit dose using a video camera was described and tested, in Chapter 3. It was

shown that dwell positions and transit dose corrections can be sub-optimal even with

modern HDR treatment systems. A source positioning error due to transfer tube curvature

and a movement profile with inappropriate transit corrections were presented, and following

discussion with the manufacturer, a new control system software version was released which

improved accuracy (Palmer 2013a).

167

A planning study evaluation of the impact of systematic dwell position errors

showed that an accuracy of 1.0 mm is an optimum tolerance to limit changes in typical

planning quality parameters (D90, D2cc) to within 3%. The tolerance limit for random dwell

positioning errors may be larger, if it is assumed a level of compensation from adjacent dwells

due to the random nature. It was also shown the dose at prescription Point A, for

gynaecological treatment applicators, is relatively insensitive to dwell position errors

(typically 0.7% change for a systematic 4 mm error). The implication is that dose

measurements at Point A, for QC or audit, while valuable for point-dose and source strength

confirmation are insufficient to confirm dwell position accuracy and overall (dose

distribution) treatment quality.

9.1.2. HDR Brachytherapy Audit Dosimetry

9.1.2.1. Evaluation of Dosimeters

Chapter 4 presented an evaluation of three dosimeters specifically chosen as

candidates for the measurement of dose distributions around brachytherapy treatment

applicators in the context of an external audit. One dosimetry system was chosen for

investigation in each of the following three potential measurement modes: multiple point

detectors, one or several 2D detectors, full 3D detector. Doped silica glass optical fibres,

radiochromic film, and solid radiochromic polymer were investigated in terms of (a)

preparation, processing and calibration, (b) measurement of radial dose from an isolated

HDR source, and (c) measurement of dose distribution from multiple HDR source dwell

positions.

The manual preparation process of doped silica glass dosimeters from optical

fibres was inefficient, with a 75% rejection rate for batch sensitivity consistency of the fibre

used in this study, and the process was labour intensive. With further refinement the process

may be improved, but fundamentally fibre dosimeters were abandoned as the spatial

resolution in their long dimension was unsatisfactory for brachytherapy audit

measurements, and challenges with handling and accurate positioning were significant.

The use of Presage, a solid radiochromic polymer, gave encouraging results for

full 3D measurement of brachytherapy dosimetry, albeit with cylindrical averaging required

to reduce noise in measurements of radial dose from a brachytherapy source. It was likely

that further research of the optical CT readout system would be beneficial but that the

development required was too great to implement for a brachytherapy audit in a reasonable

timescale. The limited physical dimensions, availability and cost of the dosimeter were also

negative features, and it was decided full 3D measurement was unnecessary to fulfil the

brachytherapy audit objectives.

Initial testing of radiochromic film, Gafchromic EBT3, as a dosimeter for

brachytherapy audit was successful, with robust calibration achievable over a wide dose

range (up to 50 Gy) and both radial dose measurement and dose distribution measurement

agreeing with Monte Carlo and TPS calculations within 5% over a range of clinical relevant

conditions. It was expected that with further research and development of techniques,

168

reduction of uncertainties and overall accuracy may be improved. At the current stage of

development of the three dosimeters, only Gafchromic film was able to be calibrated in terms

of absolute dose, the others requiring individual normalisation of each detector. It was

decided to progress with film dosimetry as the most promising dosimeter for brachytherapy

audit measurements. Results and further discussion are presented in Chapter 4 (Palmer et

al. 2013d).

9.1.2.2. Optimisation of Radiochromic Film Dosimetry

Having decided to proceed with a radiochromic film dosimetry system for

brachytherapy audit, further research and development was required on the technique to

ensure an accurate, robust methodology with minimised uncertainties, described in Chapter

5. Radiochromic film is widely used in radiotherapy but its application in brachytherapy is less

well documented, especially using advanced film techniques.

Post-irradiation darkening of film was investigated as this was important in the

context of a dosimetry audit, since films were exposed at the audited centre and analysis

performed at a later time at a different location. Darkening of film continues following

irradiation as a logarithmic function with time. For the audit, a methodology was derived in

which reference dose films were exposed at a similar time to audit films and both were

scanned simultaneously. Scanning would take place at a time interval after exposure at least

a multiple of four of the time difference between reference and audit film exposures for

reliable dosimetry. Scanner and film related uncertainties were comprehensively

investigated. The lateral scanner effect could be significant at brachytherapy dose levels, with

errors up to 23% at a 9 cm lateral position off axis at 14 Gy for the particular equipment used.

It was essential to limit analysis regions to close to the axis of the scanner and with triple-

channel dosimetry, errors within 1% at the edges of an 8 cm scan width were achievable. The

potential for natural film curvature during the scanning processes was also found to be a

potentially significant error, up to 4% for 2 mm vertical displacement, which had not

previously been reported in the literature. However, this could be completely mitigated using

a glass compression plate on top of the film at scanning.

The application of triple-channel film dosimetry with linear calibration scaling

using simultaneously scanned reference dose films was investigated and found to be an

appropriate technique for the brachytherapy audit. Comparison of triple-channel and single-

channel dosimetry showed significantly improved results with the former in controlled test

cases and for typical brachytherapy dose distribution measurements. The triple-channel

approach was also able to correct for surface contamination of the film, particularly useful in

the context of a brachytherapy audit, and appeared to correct for any potential of

manufacturing variations in film active layer thickness. Results and further discussion are

presented in Chapter 5 (Palmer et al. 2013b, Palmer et al. 2014c, and Palmer et al. 2015b).

169

9.1.3. ‘End to End’ Audit of Brachytherapy Dosimetry

9.1.3.1. Design of an Audit Phantom and Audit Methodology

There have been several HDR brachytherapy audits reported in the literature,

see Section 6.1, but no fully ‘end to end’ approaches nor any UK-wide external audit for

brachytherapy, prior to the current work. The value of such audit, concerned with the full

treatment process rather than being limited to a more conventional audit of only source

strength, was discussed in Section 6.2. Chapter 6 presented the development of a unique

film-based ‘end to end’ dosimetric audit system for gynaecological brachytherapy. A novel

phantom was designed to enable the measurement of point dose and dose distributions

around clinical brachytherapy applicators in the context of a full system dosimetry audit;

including CT, treatment planning and treatment delivery. The BRachytherapy Applicator

Dosimetry phantom (BRAD) was designed and successfully tested (Palmer et al. 2013c). The

design concept was to rigidly hold four film dosimeters at accurately known positions with

respect to a clinical treatment applicator, and sample the dose distribution in clinically

relevant regions with comparison to intended doses from the TPS. The phantom essentially

consists of a Solid Water (Gammex RMI457) frame, clamps for the applicator and film, and

accompanying full-scatter water tank.

The film dosimetry methodology for the audit, described in Section 6.4.2, was

specified to address and mitigate many of the characteristics and uncertainties associated

with the use of radiochromic film, discussed in Chapter 5. This included attention to film

batch, orientation, time since exposure, position on scanner, forced physical flatness,

reference dose films in the same scanner image for calibration scaling, and use of triple-

channel dosimetry calibration. It was decided to use percentage difference and distance to

agreement (separately) at prescription Point A, as well as gamma analysis and visual isodose

overlay inspection comparing planned and measured dose distributions across the films. The

BRAD system was tested for sensitivity to simulated errors of dwell position calibration error

and systematic dwell time/prescription dose error. Results clearly discriminated between

normal delivery and the simulated error cases. Finally, an uncertainty budget was evaluated

for the percentage dose difference at prescription Point A between TPS-calculated and film-

measured dose, at standard uncertainty of 3.0% (k=1). The audit procedure also importantly

included scope for a discussion between auditor and host physicist of the brachytherapy

procedures in use at the centre.

9.1.3.2. Monte Carlo Simulations

Monte Carlo code calculations were used to support the research work in this

thesis, discussed in Chapter 7. The local implementation of MCNP5 and the Ir-192 and Co-60

HDR source model approximations were validated in terms of dose with radial distance

compared to published reference data, all agreement was within 1%. MCNP5 was used to

determine the required physical dimensions of the water tank for the BRAD audit system to

ensure full scatter at the measurement points. A cube of side length 400 mm was selected to

provide full-scatter at the prescription Point A, approximately 20 mm from the centre of the

170

phantom. The Monte Carlo code was used to confirm the water equivalence of the EBT3 film

and that its presence did not perturb dose deposition, in situations resembling the BRAD

audit. The percentage dose change with and without film at prescription Point A in the BRAD

phantom was ≤ 0.1% for Ir-192 and Co-60 sources. An estimate of the dose attenuation by

metallic applicators in the BRAD measurement situation was calculated, in order to interpret

results from the UK audit. For Ir-192 the dose reduction was at least 2.0% (increasing with

longitudinal distance from sources) and for Co-60 was 1.5%, for typical clinical brachytherapy

applicator dimensions.

9.1.3.3. UK Audit

The BRAD system was successfully used to undertake ‘end to end’ audits at 46

of the 47 brachytherapy centres in the UK, demonstrating such an audit is worthwhile and

feasible. This was funded and organised via a Working Party of the IPEM, which the author

proposed and chaired. Full results and discussion of the UK audit are provided in Chapter 8

(and in Palmer et al. 2015a). It may be concluded that no errors were detected in the basic

TG-43 source data in use within the TPS at each of the audited centres, and that treatment

equipment performance was sufficient to faithfully deliver the calculated dose distribution

within clinically significant parameters, for the situations and locations measured. In all cases,

the measured dose at prescription Point A was in agreement with the TPS calculated dose

within experimental uncertainty, and within clinically relevant tolerance. The mean

difference between measured and calculated dose was 0.6% for plastic applicators and 3.0%

for metal applicators, with high statistical significance of the difference (p<0.001). The higher

percentage difference with metal applicators being due to a lack of compensation for

applicator attenuation in TPS, which in Section 7.3.4 has been calculated to be in excess of

2%. The mean distance to agreement at Point A between measured and calculated dose was

0.6 mm, within clinically relevant tolerance. Excellent agreement was seen between film-

measured and TPS-calculated isodose distributions, typically over a film-measured range of

0.7 to 16 Gy, with a mean gamma passing rate over all audits of 97.8% (3.2% sd) at 3% (local)

2 mm criteria, with 2 Gy cut-off. Two audits required follow-up to investigate the cause of

sub-optimal audit results. Both were resolved, the first being most likely due to applicator

overlay misalignment at the TPS (subsequently also investigated by the manufacturer) and

the second due to ‘human error’ in dose normalisation at the TPS.

The brachytherapy audit afforded the opportunity to review local brachytherapy

physics practice with host physicists and make comparisons to accepted standards across the

UK. Suggestions for improvement were reported to the local centres, which included HDR

equipment issues, imaging protocols, applicator reconstruction procedure, planning

procedures, data transfer methods, and quality control tests. This was a valuable outcome of

the audit visits. Variations in the level of patient-specific optimisation of HDR brachytherapy

treatment planning across the UK was significant. The process for individual planning of

patients’ treatments ranged from delivery of standard plans with no optimisation, indeed

one centre only performed retrospective dosimetry on the TPS after treatment delivery had

171

taken place, to fully inverse-planned treatments to 3D target volumes (HR-CTV) based on

magnetic resonance imaging.

Around half of the audited centres provided feedback after the audit; all

reporting the process had provided valuable reassurance of dosimetric accuracy. Particularly

the ‘end to end’ approach was perceived as being of particular value, and that their

brachytherapy processes had not previously been subject to comprehensive audit.

9.2. Future Work

9.2.1. Quality Control and Commissioning Testing for HDR Brachytherapy

While the fundamental tests of source strength, linear position accuracy and

dwell time were included in all UK centres’ QC schedules, Chapter 3, there was scope for

improved consistency and consideration of other QC tests. This will be included and

discussed in a forthcoming revision of IPEM Report 81, for which the author is a joint-editor.

The BRAD system measurement was conducted as a one-off national audit in

the UK during 2013-2014. However, the phantom is now available via the IPEM virtual

phantom library and proposed future work is to continue to provide the equipment and

dosimetry service as may be requested by centres commissioning new treatment equipment.

The ‘end to end’ audit approach used in this research work was well received by

host centres, as the ‘system test’ is a valuable addition to equipment-centric traditional QC

methods. It would be valuable if a simple QC-test could be developed for brachytherapy to

provide an ‘end to end’ test that could be conveniently implemented at centres without the

need for a more complex BRAD-type measurement system. This would be valuable for both

QC and equipment commissioning, rather than limiting the technique to essentially a ‘one-

off’ national audit, which has been the basis of the current research. It may be possible to

use point detectors (such as diodes) at selected locations informed by the results of this

study, if this is more convenient for routine QC and if centres do not have sufficient

equipment or experience for film dosimetry.

A new film phantom design incorporating a patient surface may be valuable for

QC or commissioning, with film an ideally suited dosimeter for 2D surface measurements,

particularly if curved surfaces were incorporated. Such a phantom could be utilised for

validation of interstitial breast brachytherapy as well as for skin surface treatments. Both of

these scenarios challenge current TG-43 TPS systems, and would be valuable for the

commissioning of advanced algorithm-based TPS.

A possible application of film dosimetry in brachytherapy is for the confirmation

of TG-43 model data for each individual HDR or PDR source, which may be performed when

new sources are first received at radiotherapy centres. There are no publications on the

potential variability of individual brachytherapy sources due to manufacturing tolerances and

distribution of radioactive material, and film could be applied to undertake new-source

commissioning tests. A strip of film could be positioned within a cylinder with the source on

172

the axis, and the resulting dose profile around the film would inform of any radial anisotropy

of the source, currently assumed to be negligible. The source could also be positioned with

its long axis in the plane of the film, to confirm TG-43 away-along dose rate tables, or

variations in dose rate with polar angle.

9.2.2. Brachytherapy Film Dosimetry Development

The use of Gafchromic EBT3 film for brachytherapy dose distribution

measurement in 2D planes has been demonstrated and successfully applied in this work. It

would be possible to extend the application of radiochromic film in brachytherapy to a

pseudo-3D measurement system using a stack of film dosimeters, perhaps with Solid Water

spacers, as has been achieved by McCaw et al. (2013) in external beam radiotherapy. Further

Monte Carlo simulation should be undertaken to confirm water-equivalence over a larger

volume and any perturbation compared to the equivalent water medium.

9.2.3. Brachytherapy Dosimetric Audit

9.2.3.1. Development of the BRAD System for ‘End to End’ Dosimetry Audit

The concept of ‘end to end’ external audit of brachytherapy dosimetry was well

received by all audited centres in UK. The feedback received from centres, discussed in

Section 8.3.5, and further reflection based on practical experience conducting the audits, has

been used to propose the following possible modifications to the BRAD system.

The concept of CT scanning the treatment applicator in the BRAD frame alone

and then inserting into the water tank for treatment delivery was designed to allow

applicators to be in near clinical orientation for imaging (without their open ends being

submerged in water) and then providing full scatter for irradiation. However, the appearance

of the applicator in the phantom (with large surrounding air volume) was reported by a few

physicists as being different to the image usually obtained when surrounded by the patient.

The BRAD system could be modified to recreate more closely the patient CT image with

additional water equivalent material.

The BRAD phantom could be further developed to incorporate CT-imaged

structures, or at least CT-visible outlines, to enable contouring processes to be included in

the ‘end to end’ system audit. This would then provide opportunity for volume-based

prescribing and evaluation in treatment planning, including evaluating the ability to cover the

target volume and avoid organs at risk, perhaps correlating results with the level of

optimisation of the plan (standard plan, manual optimisation, TPS-based inverse plan

optimisation). Structures of different density, modelling true patient anatomy, and realistic

patient-based phantom size (rather than nominal full scatter) could be included to test

calculation accuracy for non-TG43 based TPS algorithms (Papagiannis et al. 2014). This may

provide a valuable commissioning system as these algorithms become adopted in routine

clinical practice. A brachytherapy dosimetry audit developed in this way would then fully

173

satisfy the Level III audit definition proposed by the Australian Clinical Dosimetry Service for

an “anthropomorphic phantom end to end” audit (Kron et al. 2013).

The BRAD phantom could be modified with a replacement top-plate that would

allow cervix treatment applicators incorporating both intracavitary and interstitial needles to

be tested. Dose distribution shaping by the addition of low-weighted needles is becoming

accepted practice and may see widespread adoption. A more radical redesign of the BRAD

concept could be pursued for ‘end to end’ dosimetry testing of non-gynae applicators, such

as skin brachytherapy (with realistic contours and ‘missing scatter’), and oesophageal and

bronchial brachytherapy (with associated air volumes).

A significant potential source of uncertainty in the BRAD system was the

alignment of films. It would be possible to redesign the phantom to use a continuous single

film that sampled both left and right Point A as well as regions distal to the tip of the

intrauterine applicator. This would limit the uncertainty to movement of a single film with

respect to the applicator, rather than the potential of independent movement in two films

about the applicator. This would aid confirmation of alignment between film-measured and

TPS-calculated dose distributions.

9.2.3.2. Future Directions for Audit

There is no doubt that clinical and physics aspects of brachytherapy will continue

to develop at a significant rate, including the realisation of advanced and functional imaging

in brachytherapy, a new era of dose calculation algorithms for brachytherapy, and an

evolution away from traditional template planning to fully patient-specific optimisation. As a

result, the scope of dosimetric audit in brachytherapy will continue to be challenged to be fit

for purpose. Future directions will require more complexity of the BRAD-system.

Film dosimetry has been successfully applied for the first UK ‘end to end’ audit

described in this thesis. However, a limitation of film dosimetry is that it only provides a 2D

dose measurement. While this is preferable to point-dosimetry, true 3D measurements of

dose distributions in brachytherapy may be desirable and further research and development

work is advocated for high precision measurements in 3D, such as with Presage, or perhaps

an equivalent 3D EBT3 material or film-stack (as discussed above).

174

175

References

176

Acar H., Chiu-Tsao S.-T., Ozbay I., Kemikler G., Tuncer S. Evaluation of material heterogeneity

dosimetric effects using radiochromic film for COMS eye plaques loaded with 125I seeds (model

I25.S16). Med. Phys. 2013;40:011708.

Adamson J., Newton J., Yang Y., Steffey B., Cai J., Adamovics J., Oldham M., Chino J., Craciunescu O.

Commissioning a CT-compatible LDR tandem and ovoid applicator using Monte Carlo calculation and

3D dosimetry. Med. Phys. 2012;39:4515-4523.

Aird E.G.A., Jones C.H., Joslin C.A.F., Klevenhagen S.C., Rossiter M.J., Welsh A.D., Wilkinson J.M.,

Woods M.J., Wright S.J. Recommendations for brachytherapy dosimetry. Report of a joint working

party of the BIR and the IPSM. London, UK: British Institute of Radiology (BIR), 1993.

Aldelaijan S., Mohammed H., Tomic N., Liang L.H., DeBlois F., Sarfehnia A., Abdel-Rahman W.,

Seuntjens J., Devic S. Radiochromic film dosimetry of HDR 192Ir source radiation fields. Med. Phys.

2011;38:6074-6083.

Amols H.I., Klein E.E. QA procedures in radiation therapy are outdated and negatively impact the

reduction of errors. Med. Phys. 2011;38:5835-5837.

Anderson C., Lowe G., Wills R. Critical structure movement in cervix brachytherapy. Radiother. Oncol.

2013;107:39-45.

Andrassy M., Niatsetsky Y., Perez-Calatayud J. Co-60 versus Ir-192 in HDR brachytherapy: scientific and

technological comparison. Rev. Fis. Med. 2012;13:125-30.

Andres C., del Castillo A., Tortosa R., Alonso D., Barquero R. A comprehensive study of the Gafchromic

EBT2 radiochromic film. A comparison with EBT. Med. Phys. 2010;37(12):6271-6278.

Arjomandy B., Tailor R., Anand A., Sahoo N., Gillin M., Prado K., Vici M. 2010 Energy dependence and

dose response of Gafchromic EBT2 film over a wide range of photon, electron and proton beam

energies. Med. Phys. 2010;37:1942-7.

Arsenault C., Bissonnette J.-P., Dunscombe P., Johnson H., Mawko G., Seuntjens J. Standards for quality

control at Canadian radiation treatment centres: Brachytherapy remote afterloaders. Canadian

Association of Provincial Cancer Agencies (CAPCA), December 2006.

ASAHI SHIMBUN, Asia & Japan Watch internet media site. Hospital delivers radiation to wrong spots

in 100 cancer treatment cases, 26 December 2013. [cited 2014 Jan 5, 1 p.]. Available from

http://ajw.asahi.com/article/behind_news/social_affairs/AJ201312260063.

ASTRO. Safety is no accident. A framework for quality radiation oncology and care. Sponsored by

ASTRO. Developed and endorsed by AAMD, AAPM, ABR, ABS, ACR, ACRO, ARS, ASTRO, ASRT, AFROC,

SCAROP, SROA. Fairfax, VA, USA: American Society for Radiation Oncology (ASTRO), 2012.

Aukett R.J., Ratcliffe A.J., Mills J.A., Bonnett D.E. Audit techniques for LDR remote afterloading

brachytherapy. IPEM Meeting on current developments in brachytherapy, London, 1999.

Austerlitz C., Mota H., Almeida C.E., Allison R., Sibata C. Quality assurance of HDR 192Ir sources using

a Fricke dosimeter. Med. Phys. 2007;34:1348-1353.

Awunor O.A., Dixon B., Walker C. Direct reconstruction and associated uncertainties of 192Ir source

dwell positions in ring applicators using Gafchromic film in the treatment planning of HDR

brachytherapy cervix patients. Phys. Med. Biol. 2013;58:3207.

Baldock C., De Deene Y., Doran S., Ibbott G., Jirasek A., Lepage M., McAuley K.B., Oldham M., Schreiner

L.J. Polymer gel dosimetry. Phys. Med. Biol. 2010;55:R1-R63.

177

Baltas D., Geramani K., Ioannidis G.T., Hierholz K., Rogge B., Kolotas C., Muller-Sievers K., Milickovic

N., Kober B., Zamboglou N. Comparison of calibration procedures for 192Ir high-dose-rate

brachytherapy sources. Int. J. Radiat. Oncol. Biol. Phys. 1999;43(3):653-61.

Baras P., Seimenis I., Kipouros P. Polymer gel dosimetry using a three-dimensional MRI acquisition

technique. Med. Phys. 2002;29:2506-16.

Bastin K.T., Podgorsak M.B., Thomadsen B.R. The transit dose component of high dose rate

brachytherapy: direct measurements and clinical implications. Int. J. Radiat. Oncol. Biol. Phys.

1993;26:695-702.

Beaulieu L., Carlsson Tedgren A., Carrier J.-F., Davis S.T., Mourtada F., Rivard M.J., Thomson R.M.,

Verhaegen F., Wareing T.A., Williamson J.F. Report of the Task Group 186 on model-based dose

calculation methods in brachytherapy beyond the TG-43 formalism: Current status and

recommendations for clinical implementation. Med. Phys. 2012;39(10):6208-36.

Bekerat H. Improving the energy response of external beam therapy (EBT) Gafchromic dosimetry films

at low energies (≤100 keV). Master of Science Thesis, McGill University, Montreal, Canada. October

2012.

Bernstein M., Mehta K.J., Yaparpalvi R., Kuo H., Kalnicki S. Results of the hybrid interstitial-intracavitary

Utrecht applicator for cervical cancer in an outpatient setting. Radiother. Oncol. 2012;103:S116.

Bidmead A.M., Sander T., Locks S.M., Lee C.D., Aird E.G., Nutbrown R.F., Flynn A. The IPEM code of

practice for determination of the reference air kerma rate for HDR (192)Ir brachytherapy sources

based on the NPL air kerma standard. Phys. Med. Biol. 2010;55:3145-59.

Bolton S. Radiotherapy Physics Audit. Scope 2009;18(3):8-16.

Borca C.V., Pasquino M., Russo G., Grosso P., Cante D., Sciacero P., Girelli G., La Porta M.R., Tofani S.

Dosimetric characterization and use of GAFCHROMIC EBT3 film for IMRT dose verification. J. Appl. Clin.

Med. Phys. 2013;14(2):158–71.

Bouchard H., Lacroix F., Beaudoin G., Carrier J.-F., Kawrakow I. On the characterization and uncertainty

analysis of radiochromic film dosimetry. Med. Phys. 2009;36:1931-1946.

Bradley D.A., Hugtenburg R.P., Nisbet A., Rahman A.T.A., Issa F., Noor N.M., Alalawi A. Review of doped

silica glass optical fibre: Their TL properties and potential applications in radiation therapy dosimetry.

Appl. Radiat. Isotopes. 2012;71:2-11.

Brahme A. Dosimetric precision requirements in radiation therapy. Acta. Radiologica. Oncology.

1984;23:379-91.

Brown T.A.D., Hogstrom K.R., Alvarez D., Matthews K.L., Ham K., Dugas J.P. Dose-response curve of

EBT, EBT2, and EBT3 radiochromic films to synchrotron-produced monochromatic x-ray beams. Med.

Phys. 2012;39:7412-7417.

BRAPHYQS. Working Group of GEC-ESTRO. [cited 2014 Jan 21, 11p.]. Available from

http://www.estro.org/about/governance-organisation/committees-activities/gec-estro-braphyqs.

Butson M.J., Yu P.K.M., Cheung T., Metcalfe P. Radiochromic film for medical radiation dosimetry.

Mater. Sci. Eng., R. 2003;41:61-120.

Casanova Borca V., Pasquino M., Russo G., Grosso P., Cante D., Sciacero P., Girelli G., La Porta M.R.,

Tofani S. Dosimetric characterization and use of GAFCHROMIC EBT3 film for IMRT dose verification. J.

Appl. Clin. Med. Phys. 2013;14(2):158-171.

178

Carlsson Tedgren A., Grindborg J.-E. Audit on source strength determination for HDR and PDR 192Ir

brachytherapy in Sweden. Radiother. Oncol. 2008;86(1):126-30.

Carrasco M.A., Perucha M., Luis F.J., Baeza M., Herrador M. A comparison between radiochromic EBT2

film model and its predecessor EBT film model. Phys. Med.2013;29(4):412-422.

Casey K.E., Alvarez P., Kry S.F., Howell R.M., Lawyer A., Followill D. Development and implementation

of a remote audit tool for high dose rate (HDR) Ir-192 brachytherapy using optically stimulated

luminescence dosimetry. Med. Phys. 2013;40:112102.

Chassagne D., Dutreix A., Almond P., Burgers J., Busch M., Joslin C. Dose and volume specification for

reporting intracavitary therapy in gynaecology. International Commission on Radiation Units and

Measurements, ICRU Report 38, 1985.

Chavel P., Lowenthal S. Noise and coherence in optical image processing. I. The Callier effect and its

influence on image contrast. J. Opt. Soc. Am. 1978;68(5):559-568.

Crijns W., Maes F., van der Heide U.A., van der Heuvel F. Calibrating page sized Gafchromic EBT3 films.

Med. Phys. 2013;40:012102.

Daigh JD. Healthcare Inspection Review of Brachytherapy Treatment of Prostate Cancer, Philadelphia,

Pennsylvania and Other VA Medical Centers. Department of Veterans Affairs, Office of Inspector

General, Washington, USA 2010. [cited 2013 Dec 23. 110 p.]. Available from:

http://www.va.gov/oig/54/reports/VAOIG-09-02815-143.pdf.

de Almeida C.E., Pereira A.J., Marechal M.H., Pereira G., Cruz J.C., Ferraz J.C., Giordani A.J., Khalil C.M.,

Martins R.H., Menegussi G., Moreira D., Rocha J.R., Pinto M.A. Intercomparison of calibration

procedures for 192Ir HDR sources in Brazil. Phys. Med. Biol. 1999;44:N31-8.

De Deene Y., Reynaert N., De Wagter C. On the accuracy of monomer/polymer gel dosimetry in the

proximity of a high-dose-rate 192Ir source. Phys. Med. Biol. 2001;46:2801-2825.

De Deene Y., Hurley C., Venning A., Vergote K., Mather M., Healy B.J., Baldock C. A basic study of some

normoxic polymer gel dosimeters. Phys. Med. Biol. 2002;47:3441-3463.

Dempsey C. Lessons learned from a HDR brachytherapy well ionisation chamber calibration error.

Australas. Phys. Eng. Sci. Med. 2011;34(4):529-33.

Dempsey C., Smith R., Nyathi T., Ceylan A., Howard L., Patel V., Das R., Haworth A. ACPSEM

brachytherapy working group recommendations for quality assurance in brachytherapy. Australas.

Phys. Eng. Sci. Med. 2013;36:387-396.

Devic S., Tomic N., Soares C.G., Podgorsak E.B. Optimizing the dynamic range extension of a

radiochromic film dosimetry system. Med. Phys. 2009;36(2):429-437.

Devic S., Aldelaijan S., Mohammed H., Tomic N., Liang L.-H., DeBlois F., Seuntjens J. Absorption spectra

time evolution of EBT-2 model GAFCHROMICTM film. Med. Phys. 2010;37(5):2207-2214.

Devic S. Radiochromic film dosimetry: past, present, and future. Phys Med. 2011;27(3):122–134.

DeWerd L.A., Ibbott G.S., Meigooni A.S., Mitch M.G., Rivard M.J., Stump K.E., Thomadsen B.R.,

Venselaar J.L.M. A dosimetric uncertainty analysis for photon-emitting brachytherapy sources: Report

of AAPM Task Group No. 138 and GEC-ESTRO. Med. Phys. 2011;38:782-801.

Doran S.J. The history and principles of chemical dosimetry for 3-D radiation fields: Gels, polymers and

plastics. Appl. Radiat. Isotopes. 2009;67:393-398.

179

Dunn W.L., Shultis, J.K. Exploring Monte Carlo methods. Academic Press, 2012.

Dunscombe P. Recommendations for safer radiotherapy: what’s the message? Front. Oncol.

2012;2:129-1-6.

Dunscombe P., Lau H., Silverthorne S. The Ottawa Orthovoltage Incident. Report of the Panel of

Experts convened by Cancer Care Ontario. October 2008. [cited 2014 Jan 21, 48p.]. Available from

https://www.cancercare.on.ca/search/default.aspx?q=orthovoltage&type=0,6-76,6-40484|-1,1377-

78.

Eaton D.J., Earner B., Faulkner P., Dancer N. A national dosimetry audit of intraoperative radiotherapy.

Brit. J. Radiol. 2013;86:20130447.

Eckert & Ziegler BEBIG GmbH. Urgent Field Safety Notice reference 2014LCT82. Mispositioning [sic]

of source. Eckert & Ziegler Bebig GmbH, Berlin, Germany, 17 March 2014.

Elfrink R.J.M., Kolkman-Deurloo I.K.K., van Kleffens H.J., Rijnders A., Schaeken B., Aalbers T.H.L., Dries

W.J.F., Venselaar J.L.M. Quality control of brachytherapy equipment in the Netherlands and Belgium;

current practice and minimum requirements. Radiother. Oncol. 2001a;62:95-102.

Elfrink R.J.M., Kolkman-Deurloo I.-K.K., van Kleffens H.J., Rijnders A., Schaeken B., Aalbers T.H.L., Dries

W.J.F., Venselaar J.L.M. Determination of the accuracy of implant reconstruction and dose delivery in

brachytherapy in The Netherlands and Belgium. Radiother. Oncol. 2001b;59:297-306.

EORTC. Global Clinical Trials Quality Assurance of Radiation Therapy Harmonisation Group. [cited 2014

Jan 21.]. Available from http://rtqaharmonisation.org/.

Epson. Personal communication with e-Service Support Technician at Epson Customer Inter@ction

Centre, email received 07/01/2014.

Evans M.D.C., Devic S., Podgorsak E.B. High dose-rate brachytherapy source position quality assurance

using radiochromic film. Med. Dosimetry 2007;32:13-5.

Ferreira I.H., Duterix A., Bridier A., Chavaudra J., Svensson H. The ESTRO-QUALity assurance network

(EQUAL). Radiother. Oncol. 2000;55:273-84.

Fonseca G.P., Rubo R.A., Minamisawa R.A., dos Santos G.R., Antunes P.C.G., Yoriyaz H. Determination

of transit dose profile for a 192Ir HDR source. Med. Phys. 2013;40(5):051717-1-8.

Fraass B., Doppke K., Hunt M., Kutcher G., Starkschall G., Stern R., van Dyke J. American association of

physicists in medicine radiation therapy committee task group 53: quality assurance for clinical

radiotherapy treatment planning. Med. Phys. 1998;25:1773-829.

Gerardy I., Rodenas J., van Dycke M., Gallardo S., Ceccolini E. Dosimetric characterization of a

brachytherapy applicator using MCNP5 modelisation and in-phantom measurements. Appl. Radiat.

Isotopes 2010;68:735-737.

Gerbaulet A., Potter R., Mazeron J.-J., Meertens H., Van Limbergen E. (Editors). The GEC ESTRO

handbook of brachytherapy. Brussels, Belgium: ESTRO, 2002.

Glasgow G.P., Bourland J.D., Grigsby P.W., Meli J.A., Weaver K.A. A report of AAPM task group no. 41

remote afterloading technology. New York, USA; AAPM, 1993.

Goetsch S.J., Attix F.H., DeWerd L.A., Thomadsen B.R. A new re-entrant ionization chamber for the

calibration of iridium-192 high dose rate sources. Int. J. Radiat. Oncol. Biol. Phys. 1992;24(1):167-70.

180

Goorley, T. Using MCNP for medical physics applications. Computational Medical Physics Working

Group Workshop II, Gainesville FL, October 2007. Los Alamos National Laboratory, 2007;LA-UR-07-

6618. [cited 25 June 2014, 107p]. Available from

https://laws.lanl.gov/vhosts/mcnp.lanl.gov/pdf_files/la-ur-07-6618.pdf

Gorjiara T., Kuncic Z., Doran S., Adamovics J., Baldock C. Water and tissue equivalence of a new

PRESAGE® formulation for 3D proton beam dosimetry: A Monte Carlo study. Med. Phys. 2012;39:7071-

7079.

Granero D., Perez-Calatayud J., Ballester F. Technical note; Dosimetric study of a new Co-60 source

used in brachytherapy. Med. Phys. 2007;34:3485-3488.

Granero D., Perez-Calatayud J., Pujades-Claumarchirant M., Ballester F., Melhus C.S., Rivard J.

Equivalent phantom sizes and shapes for brachytherapy dosimetric studies of 192Ir and 137Cs. Med.

Phys. 2008;35(11):4872-4877.

Grochowska P., Izewska J. An IAEA survey of dosimetry audit networks for radiotherapy. SSDL

Newsletter No. 61. IAEA, June 2013. [cited 2014 Jan 21, 36p.]. Available from http://www-

pub.iaea.org/MTCD/Publications/PDF/Newsletters/SSDL-61.pdf.

Guo P.Y., Adamovics J.A., Oldham M. Characterization of a new radiochromic three-dimensional

dosimeter. Med. Phys. 2006;33:1338-1345.

Haie-Meder C., Siebert F.A., Potter R. Image guided, adaptive, accelerated, high dose brachytherapy

as model for advanced small volume radiotherapy. Radiother. Oncol. 2011;100:333-43.

Haworth A., Wilfert L., Butler D., Ebert M.A., Todd S., Bucci J., Duchesne G.M., Joseph D., Kron T.

Australasian brachytherapy audit: Results of the ‘end-to-end’ dosimetry pilot study. J. Med. Imag.

Radiat. On. 2013;57:490-498.

Hayashi N., Watanabe Y., Malmin R., Kato H. Evaluation of triple channel correction acquisition method

for radiochromic film dosimetry. J. Radiat. Res. 2012;53(6):930-935.

Heeney C., McClean B., Kelly C. A dosimetric intercomparison of brachytherapy facilities in Ireland,

Scotland and the North of England. Radiother. Oncol. 2005;74:149-56.

Hellebust T.P., Kirisits C., Berger D., Perez-Calatayud J., De Brabandere M., De Leeuw A., Dumas I.,

Hudej R., Lowe G., Wills R. and Tanderup K. Recommendations from Gynaecological (GYN) GEC-ESTRO

Working Group: Considerations and pitfalls in commissioning and applicator reconstruction in 3D

image-based treatment planning of cervix cancer brachytherapy. Radiother. Oncol. 2010;96:153-60.

Hellebust T.P., Tanderup K., Lervag C. Dosimetric impact of interobserver variability in MRI-based

delineation for cervical cancer brachytherapy. Radiother. Oncol. 2013;107:13-9.

Hepel J.T., Wazer D.E. A comparison of brachytherapy techniques for partial breast irradiation.

Brachytherapy 2012;11:163-75.

HSE. Equipment used in connection with medical exposure. Guidance Note PM77 (Third edition).

Health and Safety Executive, UK, 2006. [cited 18 Oct 2012]. Available from:

www.hse.gov.uk/pubns/guidance/pm77.pdf.

Hussein M., Tsang Y., Thomas R.A.S., Gouldstone C., Maughan D., Snaith J.A.D., Bolton S.C., Nisbet A.,

Clark C.H. A methodology for dosimetric audit of rotational radiotherapy using a commercial detector

array. Radiother. Oncol. 2013;108(1):78-85.

181

IAEA. International Atomic Energy Agency, Dosimetry and Medical Radiation Physics Section.

Calibration of photon and beta ray sources used in brachytherapy. IAEA-TECDOC-1274. Austria,

Vienna: IAEA, 2002.

IAEA. Commissioning and quality assurance of computerized planning systems for radiation treatment

of cancer. International Atomic Energy Agency, Technical Reports Series No. 430. Vienna, Austria;

IAEA, 2004.

IAEA. Live chart of nuclides. Online resource only. Nuclear Data Section, International Atomic Energy

Agency, Vienna, Austria. [cited 25 June 2014]. Available from:

https://www-nds.iaea.org/relnsd/vcharthtml/VChartHTML.html

IAEA QUATRO (Quality Assurance Team for Radiation Oncology). Comprehensive Audits of

Radiotherapy Practices: A Tool for Quality Improvement. Vienna, Austria: IAEA, 2007.

IAEA. Standards, Applications and Quality Assurance in Medical Radiation Dosimetry (IDOS). External

Quality Audits, Session 7. Proceedings of an International Symposium, Vol. 2; 2010 November 9-12,

Vienna, Austria. Vienna, Austria: International Atomic Energy Agency, 2011.

Ibbott G.S., Thwaites D.I. Audits for advanced treatment dosimetry. J. Phys.: Conf. Ser.

2015;573:012002.

ICRP Publication 86. Prevention of accidents to patients undergoing radiation therapy. Ann. ICRP.

2000;30(3).

ICRP Publication 97. Prevention of high-dose-rate brachytherapy accidents. Ann. ICRP. 2005a;35(2).

ICRP Publication 98. Radiation safety aspects of brachytherapy for prostate cancer using permanently

implanted sources. Ann. ICRP. 2005b;35(3).

ICRU. Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Rays in

Radiotherapy Procedures. International Commission on Radiation Units and Measurement, Report 24,

1976.

IPEM Radiotherapy Physics Topic Group. Survey of Quality Control Practice in UK Hospitals. Scope

1992;1:49-61. As cited in Mayles W.P.M., Lake R., McKenzie A., Macaulay E.M., Morgan H.M., Jordan

T.J., Powley S.K., Editors. Physics aspects of quality control in radiotherapy. Report no. 81. York, UK:

IPEM, 1999.

Issa F., Rahman A.T.A., Hugtenburg R.P., Bradley D.A., Nisbet A. Establishment of Ge-doped optical

fibres as thermoluminescence dosimeters for brachytherapy. Appl. Radiat. Isotopes. 2012;70:1158-

1161.

Ibbott G., Ma C.M., Rogers D.W.O., Seltzer A.M., Williamson J.F. Anniversary Paper: Fifty years of

AAPM involvement in radiation dosiemtry. Med. Phys. 2008;35:1418-27.

Izewska J., Grochowska P., Meghzifene A. IAEA survey of national dosimetry audit networks for

radiotherapy. Radiother. Oncol. 2012;103:S548.

Jordan K. Review of recent advances in radiochromic materials for 3D dosimetry. J. Phys. Conf. Ser.

2010;250:012043.

Karaiskos P., Angelopoulos A., Sakelliou L. Monte Carlo and TLD dosimetry of an Ir-192 high dose-rate

brachytherapy source. Med. Phys. 1998;25:1975-84.

Kirisits C., Rivard M.J., Baltas D., Ballester F., De Brabandere M., van der Laarse R., Niatsetski Y.,

Papagiannis P., Hellebust T.P., Perez-Calatayud J., Tanderup K., Venselaar J.L.M., Siebert F.-A. Review

182

of clinical brachytherapy uncertainties: Analysis guidelines of GEC-ESTRO and the AAPM. Radiother.

Oncol. 2014;110(1):199-212.

Kirov A.S., Williamson J.F., Meigooni A.S. TLD, diode and Monte Carlo dosimetry of an Ir-192 source

for high dose-rate brachytherapy. Phys. Med. Biol. 1995;40:2015-36.

Kron T., Haworth A., Williams I. Dosimetry for audit and clinical trials: challenges and requirements. J.

Phys. Conf. Ser. 2013;444:012014.

Krstajic N., Doran S.J. Focusing optics of a parallel beam CCD optical tomography apparatus for 3D

radiation gel dosimetry. Phys. Med. Biol. 2006;51:2055-2075.

Krstajic N., Doran S.J. Characterization of a parallel-beam CCD optical-CT apparatus for 3D radiation

dosimetry. Phys. Med. Biol. 2007a;52:3696-3713.

Krstajic N., Doran S.J. Fast laser scanning optical-CT apparatus for 3D radiation dosimetry. Phys. Med.

Biol. 2007b;52:N257-N263.

Kubo H.D., Glasgow G.P., Pethel T.D., Thomadsen B.R., Williamson J.F. High dose-rate brachytherapy

treatment delivery; report of the AAPM radiation therapy committee task group no. 59. Med. Phys.

1998;25:375-403.

Lee C. Personal communication with Chris Lee, Brachytherapy Physicist, The Clatterbridge Cancer

Centre NHS Foundation Trust. 2012 May 17.

Lee C. Recent developments and best practice in brachytherapy treatment planning. Br. J. Radiol.

2014;87:20140146.

Lewis D., Micke A., Yu X., Chan M. An efficient protocol for radiochromic film dosimetry combining

calibration and measurement in a single scan. Med. Phys. 2012;39:6339-6350.

Li G., Zheng H., Sun G., Wu Y. Photon energy spectrum reconstruction based on Monte Carlo and

measured percentage depth dose in accurate radiotherapy. Prog. Nuc. Sci Tech Conf. Ser. 2011;2:160-

164. The Joint International Conference of the 7th Supercomputing in Nuclear Applicaton and the 3rd

Monte Carlo, October 2010, Tokoyo, Japan. [cited 25 June 2014] Available from

http://www.aesj.or.jp/publication/pnst002/data/160-164.pdf

Lillicrap S.C., Owen B., Williams J.R., Williams P.C. Code of practice for high-energy photon therapy

dosimetry based on the NPL absorbed dose calibration service. Phys. Med. Biol. 1990;35:1355.

Low D.A., Harms W.B., Mutic S., Purdy J.A. A technique for the quantitative evaluation of dose

distributions. Med. Phys. 1998;25:656-661.

Low D.A., Dempsey J.F. 2003 Evaluation of the gamma dose distribution comparison method

Med.Phys. 2003;30:2455–2464.

Low D.A. Gamma dose distribution evaluation tool. J. Phys.: Conf. Ser. 2010;250:012071.

Lowenstein J.R., Roll J., Hanson H.W., Davis D.S., Lanciano R., Calkins A., Petereit D., Varia M., Ibbott

G.S. Radiotherapy quality assurance of gynecologic oncology group (COG) protocol 165, a

cooperative group study of carcinoma of the cervix. RPC, 2002. [cited 2014 Jan 21, 11p.]. Available

from

http://rpc.mdanderson.org/RPC/Publications/RPC_Presentations/2002%20AAPM%20Posters/ASTRO

%20Poster%202002%20JL.pdf.

Majumdar B., Patel N.P., Vijayan V. The non-uniformity correction factor for the cylindrical ionization

chambers in dosimetry of an HDR 192Ir brachytherapy source. J. Med. Phys. 2006;31:83-88.

183

Martisikova M., Ackermann B., Jakel O. Analysis of uncertainties in Gafchromic EBT film dosimetry of

photon beams. Phys. Med. Biol. 2008;53(24):7013–27.

Massillon-JL G., Minniti R., Mitch M.G., Soares C.G. Measurement of the absorbed dose distribution

near an 192Ir intravascular brachytherapy seed using a high-spatial-resolution gel dosimetry system.

Phys. Med. Biol. 2012;57:3407-3418.

Masillon-JL G., Chiu-Tsao S.-T., Domingo-Munoz I., Chan M.F. Energy dependence of the new

Gafchromic EBT3 film: dose response curves for 50 kV, 6 and 15 MV x-ray beams. Int. J. Med. Phys.

Clin. Eng. Radiat. Oncol. 2012;1:60-65.

Mayer R.R., Ma F., Chen Y., Miller R.I., Belard A., McDonough J., O’Connell J.J. Enhanced dosimetry

procedures and assessment for EBT2 radiochromic film. Med. Phys. 2012;39(4):2147-2155.

Mayles W.P.M., Lake R., McKenzie A., Macaulay E.M., Morgan H.M., Jordan T.J., Powley S.K., editors.

Physics aspects of quality control in radiotherapy. Report no. 81. York, UK: IPEM, 1999.

McCaw T.J., Micka J.A., DeWerd L.A. Development and characterization of a three-dimensional film

stack dosimeter. J. Phys.: Conf. Ser. 2013;444:012018.

McConn Jr R.J., Gesh C.J., Pagh R.T., Rucker R.A., Williams III R.G. Radiation portal monitor project:

Compendium of material composition data for radiation transport modelling. Pacific Northwest

National Laboratory, Richland, Washington, 2011.

McDermott P.N., Somnay A.R. and Alecu R. Acceptance testing and commissioning of a new model

HDR afterloader. Radiother. Oncol. 1996;39:S25.

MNCP5 Development Team. Using MCNP5 for medical physics applications. Document LA-UR-05-

2755. American Nuclear Society Topical Meeting – Monte Carlo, Chatanooga, TN, USA. April 2005.

McJury M., Tapper P.D., Cosgrove V.P., Murphy P.S., Griffin S., Leach M.O., Webb S., Oldham M.

Experimental 3D dosimetry around a high-dose-rate clinical 192Ir source using a polyacrylamide gel

(PAG) dosimeter. Phys. Med. Biol. 1999;44:2431-2444.

Meertens H. In-phantom calibration of Selectron-LDR sources. Radiother. Oncol. 1990;17:369-78.

Melhus C.S., Rivard J. Approaches to calculating AAPM TG-43 brachytherapy dosimetry parameters for

137CS, 125I, 192Ir, 103Pd, and 169Yb sources. Med. Phys. 2006;33(6):1729-1737.

Mendez I., Hartman V., Hudej R., Strojnik A., Casar B. Gafchromic EBT2 film dosimetry in reflection

mode with a novel plan-based calibration method. Med. Phys. 2013;40(1):011720-1-9.

Mendez I.,Peterlin P., Hudej R., Strojnik A., Casar B. On multichannel film dosimetry with channel-

independent perturbations. Med. Phys. 2014;41(1):011705

Menegotti L., Delana A., Martignano A. Radiochromic film dosimetry with flatbed scanners: A fast and

accurate method for dose calibration and uniformity correction with single film exposure. Med. Phys.

2008;35(7):3078–3085.

Micke A., Lewis D.F., Yu X. Multichannel film dosimetry with nonuniformity correction. Med. Phys.

2011;38(5):2523-2534.

Mijnheer B.J., Battermann J.J., Wambersie A. What degree of accuracy is required and can be achieved

in photon and neutron therapy? Radiother. Oncol. 1987;8:237-52.

Mishra V., Waterman F.M., Suntharalingam N. Anisotropy of an 192iridium high dose rate source

measured with a miniature ionization chamber. Med. Phys. 1997;24:751-5.

184

Molineu A., Hernandez N., Nguyen T., Ibbott G., Followill D. Credentialing results from IMRT

irradiations of an anthropomorphic head and neck phantom. Med. Phys. 2013;40:022101.

Mosleh-Shirazi M.A., Hadad K., Faghihi R., Baradaran-Ghahfarokhi M., Naghshnezhad Z., Meigooni A.S.

Echoseed model 6733 Iodine-125 brachytherapy source: Improved dosimetric characterization using

the MCNP5 Monte Carlo code. Med. Phys. 2012;39(8):4653-4659.

Nath R., Anderson L.L., Meli J.A., Olch A.J., Stitt J.A., Williamson J.F. Code of practice for brachytherapy

physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56. Med. Phys.

1997;24:1557-98.

Nesvacil N., Tanderup K., Hellebust T.P. A multicentre comparison of the dosimetric impact of inter-

and intra-fractional anatomical variations in fractionated cervix cancer brachytherapy. Radiother.

Oncol. 2013;107:20-5.

NHS DH. National Health Service Department of Health. Manual for Cancer Standards version 5.0.

London, UK: DH, 2013. [cited 2014 Jan 21, 63p.]. Available from http://www.cquins.nhs.uk/.

NHS England. 2013/14 NHS Standard Contract for Brachytherapy and Molecular Radiotherapy (All

Ages). Section B Part 1 – Service Specification No. B01/S/b. NHS Commissioning Board 2013 [cited

2013 Dec 23. 21 p.]. Available from: http://www.england.nhs.uk/wp-content/uploads/2013/06/b01-

brachytpy-molecular-radio.pdf.

NHS England. Clinical audit. [cited 2014 Jan 21. 1 p.]. Available from:

http://www.england.nhs.uk/ourwork/qual-clin-lead/clinaudit/

Niroomand-Rad A., Blackwell C.R., Coursey B.M., Gall K.P., Galvin J.M., McLaughlin W.L., Meigooni A.S.,

Nath R., Rodgers J.E., Soares C.G. Radiochromic film dosimetry: recommendations of AAPM Radiation

Therapy Committee Task Group 55. American Association of Physicists in Medicine Report no. 63. Med.

Phys. 1998;25(11):2093-2115.

Nisbet A., Thwaites D.I. A dosimetric intercomparison of electron beams in UK radiotherapy centres.

Phys. Med. Biol. 1997;42:2393-409.

NNDC. Chart of nuclides. Online resource only. National Nuclear Data Centre, Brookhaven National

Laboratory, NY, USA. [cited 25 June 2014]. Available from http://www.nndc.bnl.gov/chart/

NRC. United States Nuclear Regulatory Commission. NRC Information Notice 2009-17: Reportable

medical events involving treatment delivery errors caused by confusion of units for the specification

of brachytherapy sources. Washington, USA: NRC; 2009. [cited 31st Aug 2012]. Available from:

http://pbadupws.nrc.gov/docs/ML0807/ML080710054.pdf.

Nucletron. Urgent field safety notice: Flexitron HDR. Reference 555.00267FSN. 2014 [cited 17 Sept

2014, 2p]. Available from:

http://www.mhra.gov.uk/home/groups/fsn/documents/fieldsafetynotice/con249618.pdf.

Paelinck L., De Neve W., De Wagter C. Precautions and strategies in using a commercial flatbed scanner

for radiochromic film dosimetry. Phys. Med. Biol. 2007;52:

Palmer A., Mzenda B. Performance assessment of the BEBIG Multisource® high dose rate

brachytherapy treatment unit. Phys. Med. Biol. 2009;54:7417-7434.

Palmer A., Mzenda B., Kearton J., Wills R. Analysis of regional radiotherapy dosimetry audit data and

recommendations for future audits. Br. J. Radiol. 2011;84:733-42.

185

Palmer A.L., Bidmead M., Nisbet A. A survey of quality control practices for high dose rate (HDR) and

pulsed dose rate (PDR) brachytherapy in the United Kingdom. J. Contemp. Brachyther. 2012a;4(4):232-

40.

Palmer A., Bradley D., Nisbet A. Physics-aspects of dose accuracy in high dose rate (HDR)

brachytherapy; source dosimetry, treatment planning, equipment performance and in vivo

verification techniques. J. Contemp. Brachy. 2012b;4:81-91.

Palmer A., Hayman O., Ioannou L., Nagar Y.S. Is HDR equipment performance suitable for modern

brachytherapy? Positional errors, dosimetric impact & case study. Radiother. Oncol. 2012c;103:S132-

3.

Palmer A., Hayman O., Muscat S. Treatment planning study of the 3D dosimetric differences between

Co-60 and Ir-192 sources in high dose rate (HDR) brachytherapy for cervix cancer. J. Contemp.

Brachyther. 2012d;4:52-9.

Palmer A. Impact of software changes: transit dose and source position accuracy of the Eckert and

Ziegler BEBIG GmbH multisource high dose rate (HDR) brachytherapy treatment unit. J. Radiother.

Pract. 2013a;12:80–87.

Palmer A.L., Nisbet A., Bradley D. Verification of high dose rate brachytherapy dose distributions with

EBT3 Gafchromic film quality control techniques. Phys. Med. Biol. 2013b;58:497-511.

Palmer A.L., Lee C., Ratcliffe A.J., Bradley D., Nisbet A. Design and implementation of a film dosimetry

audit tool for comparison of planned and delivered dose distributions in high dose rate (HDR)

brachytherapy. Phys. Med. Biol. 2013c;58:6623-6640.

Palmer A.L., Di Pietro P., Alobaidli S., Issa F., Doran S., Bradley D., Nisbet A. Comparison of methods

for the measurement of radiation dose distributions in high dose rate (HDR) brachytherapy: Ge-doped

optical fiber, EBT3 Gafchromic film, and PRESAGE® radiochromic plastic. Med. Phys.

2013d;40(6):061707-1-11.

Palmer A.L. BJR brachytherapy dosimetry special feature. Br. J. Radiol. 2014a;87:20140506.

Palmer A.L., Bradley D.A., Nisbet A. Dosimetric audit in brachytherapy. Br. J. Radiol.

2014b;87:20140105.

Palmer A.L., Bradley D., Nisbet A. Evaluation and implementation of triple-channel radiochromic film

dosimetry in brachytherapy. J. Appl. Clin. Med. Phys. 2014c;14(4):280-296.

Palmer A.L., Diez P., Gandon L., Wynn-Jones A., Bownes P., Lee C., Aird E., Bidmead M., Lowe G.,

Bradley D., Nisbet A. A multicentre ‘end to end’ dosimetry audit for cervix HDR brachytherapy

treatment. Radiother. Oncol. 2015a (in press, http://dx.doi.org/10.1016/j.radonc.2014.12.006).

Palmer A.L., Bradley D., Nisbet A. Evaluation and mitigation of potential errors in radiochromic film

dosimetry due to film curvature at scanning. J. Appl. Clin. Med. Phys. 2015b (in press).

Papagiannis P., Pantelis E., Karaiskos P. Current state of the art brachytherapy treatment planning

dosimetry algorithms. Br. J. Radiol. 2014;87:20140163.

Park C.C., Yom S.S., Podgorsak M.B., Harris E., Price R.A. Jr, Bevan A., Pouliot J., Konski A.A., Wallner

P.E. American Society for Therapeutic Radiology and Oncology (ASTRO) Emerging Technology

Committee report on electronic brachytherapy. Int. J. Radiat. Oncol. Biol. Phys. 2010;76(4):963-72.

Pattersen M.N., Aird E., Olsen D.R. Quality assurance of dosimetry and the impact on sample size in

randomized clinical trials. Radiother. Oncol. 2008;86(2):195-9.

186

Perez-Calatayud J., Granero D., Ballester F. Phantom size in brachytherapy source dosimetric studies.

Med. Phys. 2004;31(7):2075-2081.

Perez-Calatayud J., Ballester F., Das R.K., DeWard L.A., Ibbott G.S., Meigooni A.S., Ouhib Z., Rivard

M.J., Sloboda R.S., Wiliamson J.F. Dose calculation for photon-emitting brachytherapy sources with

average energy higher than 50 keV: Report of the AAPM and ESTRO. Med. Phys. 2012a;39:2904-2929.

Perez-Calatayud J., Ballester F., Das R.K., DeWerd L.A., Ibbott G.S., Meigooni A.S., Ouhib Z., Rivard M.J.,

Sloboda R.S., Williamson J.F. Dose calculation for photon-emitting brachytherapy sources with average

energy higher than 50 keV: Full report of the AAPM and ESTRO. Report of the High Energy

Brachytherapy Source Dosimetry (HEBD) Working Group. American Association of Physicists in

Medicine, MD, USA, 2012b.

Petric P., Hudej R., Rogelj P. Uncertainties of target volume delineation in MRI guided adaptive

brachytherapy of cervix cancer: A multi-institutional study. Radiother. Oncol. 2013;107:6-12.

Pierquet M., Craciunescu O., Steffey B., Song H., Oldham M. On the feasibility of verification of 3D

dosimetry near brachytherapy sources using Presage/Optical-CT. J. Phys. Conf. Ser. 2010;250:012091.

Pi Medical Research. 3D Matrix Phantom (Baltas Phantom). [cited 2014 Jan 21. 1 p.]. Available from

http://www.pi-medical.gr/products/tools/matrix-phantom

Potter R., Haie-Meder C., Van Limbergen E., Barillot I., De Brabandere M., Dimopoulos J., Dumas I.,

Erickson B., Lang S., Nulens A., Petrow P., Rownd J., Kirisits C. Recommendations from gynaecological

(GYN) GEC ESTRO working group (II): concepts and terms in 3D image-based treatment planning in

cervix cancer brachytherapy-3D dose-volume parameters and aspects of 3D image-based anatomy,

radiation physics, radiobiology. Radiother. Oncol. 2006;78:67-77.

Potter R., Kirisits C. Upcoming ICRU/GEC ESTRO recommendations for brachytherapy in cancer of the

cervix (1). Radiother. Oncol. 2012;103:S42.

Ratcliffe A.J., Bonnett D.E., Farajollahi A.R., Aukett R.J., Mills JA. Quality assurance audit methods for

low dose rate afterloading machines. GEC-ESTRO Brachytherapy Meeting, Stockholm, 1997.

RCR. The Royal College of Radiologists, Society and College of Radiographers, Institute of Physics and

Engineering in Medicine, National Patient Safety Agency, British Institute of Radiology. Towards Safer

Radiotherapy. London, UK: The Royal College of Radiologists, 2008. [cited 24 Sept 2012, 85p].

https://www.rcr.ac.uk/docs/oncology/pdf/Towards_saferRT_final.pdf.

Reda S.M. Comparative study on homogeneous and inhomogeneous photon dose distribution from a

high dose rate 192Ir brachytherapy source using MCNP5. Arab J. Nuc. Sci. App. 2013;46(4):165-175.

Reinhardt S., Hillbrand M., Wilkens J.J., Assmann W. Comparison of Gafchromic EBT2 and EBT3 films

for clinical photon and proton beams. Med. Phys. 2012;39:5257-5262.

Rivard M.J., Coursey B.M., DeWerd L.A., Hanson W.F., Huq M.S., Ibbott G.S., Mitch M.G., Nath R.,

Williamson J.F. Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for

brachytherapy dose calculations. Med. Phys. 2004;31:633-674.

Rivard M.J., Venselaar J.L.M., Beaulieu L. The evolution of brachytherapy treatment planning. Med.

Phys. 2009;36:2136-53.

Rivard M.J., Granero D., Perez-Calatayud J., Ballester F. Influence of photon energy spectra from

brachytherapy sources on Monte Carlo simulations of kerma and dose rates in water and air. Med.

Phys. 2010;37(2):869-876.

187

Rivard M.J., Beaulieu L., Mourtada F. Enhancements to commissioning techniques and quality

assurance of brachytherapy treatment planning systems that use model-based dose calculation

algorithms. Med. Phys. 2010;37(6):2645-58.

Rivard M.J., Reed J.L., DeWerd L.A. 102Pd strings: Monte Carlo assessment of a new approach to

brachytherapy source design. Med. Phys. 2014;41:011716-1-11.

Roue A., Ferreira I.H., Van Dam J., Svensson H., Venselaar J.L.M. The EQUAL-ESTRO audit on geometric

reconstruction techniques in brachytherapy. Radiother. Oncol. 2006;78(1):78-83.

Roue A., Venselaar J.L.M., Ferreira I.H., Bridier A., Van Dam J. Development of a TLD mailed system for

remote dosimetry audit for 192IR HDR and PDR sources. Radiother. Oncol. 2007;83(1):86-93.

RPC. M.D. Anderson Dosimetry Laboratory. Radiological Physics Centre. [cited 21 Jan 2014]. Available

from: http://rpc.mdanderson.org/RPC/home.htm.

Sahoo N. Measurement of transit time of a remote after-loading high dose rate brachytherapy source.

Med. Phys. 2001;28:1786-1790.

Sander T. Air kerma and absorbed dose standards for reference dosimetry in brachytherapy. Br. J.

Radiol. 2014;87:20140176.

Schoenfeld A.A., Poppinga D., Harder D., Doerner K.-J., Poppe B. The artefacts of radiochromic film

dosimetry with flatbed scanners and their causation by light scattering from radiation-induced

polymers. Phys. Med. Biol. 2014;59:3575-97.

Sellakumar P., Kumar A.S., Supe S.S., Anand M.R., Nithya K., Sajitha S. Evaluation of dosimetric

functions for Ir-192 source using radiochromic film. Nucl. Instrum. Methods Phys. Res., Sect. B

2009;267:1862-1866.

Sharma S.D., Bianchi C., Conte L. Radiochromic film measurement of anisotropy function for high-dose-

rate Ir-192 brachytherapy source. Phys. Med. Biol. 2004;49:4065-72.

Shultis J.K., Faw R.E. An MCNP primer. Dept. of Mechanical and Nuclear Engineering, Kansas State

University, Manhattan, KS 66506: December 2011. [cited 21 Sept 2014, 45p]. Available from:

http://krex.k-state.edu/dspace/handle/2097/15754.

Siauw T., Cunha A., Atamturk A., Hsu I.C., Pouliot J., Goldberg K. IPIP: a new approach to inverse

planning for HDR brachytherapy by directly optimizing dosimetric indices. Med. Phys. 2011;38:4045-

51.

Sim G.S., Wong J.H.D., Ng K.H. The use of radiochromic EBT2 film for the quality assurance and

dosimetric verification of 3D conformal radiotherapy using Microtek ScanMaker 9800XL flatbed

scanner. J. Appl. Clin. Med. Phys. 2013;14(4):85-95.

Song W, Webster M., Han D., Einck J., Scanderbeg D., Vuong T., Devic S. Dynamic modulated

brachytherapy (DMBT): concept, design, and simulations. Radiother. Oncol. 2012;103:S44-5.

Spiller A., on behalf of Eckert & Ziegler Bebig GmbH. Personal communication with the author via

email, July 2012.

Stern R.L., Liu T. Dwell position inaccuracy in the Varian GammaMed HDR ring applicator. J. Appl. Clin.

Med. Phys. 2010;11:291-6.

Suchowerska N., Jackson M., Lambert J., Yin Y.B., Hruby G., McKenzie D.R. Clinical trials of a urethral

dose measurement system in brachytherapy using scintillation detectors. Int. J. Radiat. Oncol. Biol.

Phys. 2011;79:609-15.

188

Sudom T.S., Ratcliffe A.J. Radiotherapy physics audit – A technique for Iridium-192 hairpins. IPEM

Biennial Radiotherapy Physics Meeting, York, 1998.

Sureka C.S., Sunny C.S., Subbaiah K.V. Dose distribution for endovascular brachytherapy using Ir-192

sources: comparison of Monte Carlo calculations with radiochromic film measurements. Phys. Med.

Biol. 2007;52:525-37.

Sutherland J.G.H., Rogers D.W.O. Monte Carlo calculated absorbed-dose energy dependence of EBT

and EBT2 film. Med. Phys. 2010;37:1110-1116.

Sydney Morning Herald. Cancer therapy missed tumour sites. 21 June 2003a. [cited 2014 Jan 21,

1p.]. Available from http://www.smh.com.au/articles/2003/06/20/1055828490830.html.

Sydney Morning Herald. Critically ill received wrong radiation therapy due to human errors: review.

12 September 2003b. [cited 2014 Jan 21, 1p.]. Available from

http://www.smh.com.au/articles/2003/09/11/1063268518775.html?from=storyrhs.

Tamponi M., Bona R., Poggiu A., Marini P. A practical tool to evaluate dose distributions using

radiochromic film in radiation oncology. Physica Medica 2014 (in press, available online ahead of print,

August 2014).

Tanderup K., Hellebust T.P., Lang S., Granfeldt J., Potter R., Lindegaard J.C. Kirisits C. Consequences of

random and systematic reconstruction uncertainties in 3D image based brachytherapy in cervical

cancer Radiother. Oncol. 2008;89:156-63.

Tanderup K., Nesvacil N., Potter R., Kirisits C. Uncertainties in image guided adaptive cervix cancer

brachytherapy: Impact on planning and prescription. Radiother. Oncol. 2013;107:1-5.

Thomadsen BR. Achieving Quality in Brachytherapy. Madison, USA; Taylor & Francis, 1999.

Thomadsen B., Lin S.-W., Laemmrich P. Waller T., Cheng A., Caldwell B., Rankin R., Stitt J. Analysis of

treatment delivery errors in brachytherapy using formal risk analysis techniques. Int. J. Radiat. Oncol.

Biol. Phys. 2003; 57: 1492-1508.

Thomadsen B.R., Williamson J.F., Rivard M.J., Meigooni A.S. Anniversary paper: past and current

issues, and trends in brachytherapy physics. Med. Phys. 2008;35:4708-23.

Tolli H., Johansson K.-A. Quality assurance in brachytherapy: the displacement effect in the vicinity of

60Co and 192Ir brachytherapy sources. Phys. Med. Biol. 1993;38:1485-1492.

Thwaites D.I., Williams J.R., Aird E.G., Klevenhagen S.C., Williams P.C. A dosimetric intercomparison of

megavoltage photon beams in UK radiotherapy centres. Phys. Med. Biol. 1991;37(2):445-461.

Thwiates D.I., Nisbet A. The Scottish+ radiotherapy dosimetry audit system – an example of the UK

network. Proceedings of the World Congress of Medical Physics, Sydney, Australia, 2003a. Int. Fed.

Med. Phys. Biol. Eng. ISBN 1 877040 14 2: A03.13.4200.

Thwaites D.I., Powley S., Nisbet A., Allahverdi M. The United Kingdom’s radiotherapy dosimetry audit

network. In: Standards and Codes of Practice for Medical Radiation Dosimetry, IAEA:Vienna (2003b),

IAEA-STI-PUB-1153, vol 2, 183-190.

Thwaites D. Accuracy required and achievable in radiotherapy dosimetry: have modern technology

and techniques changed our views? J. Phys.: Conf. Ser. 2013;444:012006.

Toye W.C., Das K.R., Todd S.P. An experimental MOSFET approach to characterize (192)Ir HDR source

anisotropy. Phys. Med. Biol. 2007;52:5329-39.

189

USNRC, United States Nuclear Regulatory Commission. NRC Information Notice 2009-17: Reportable

medical events involving treatment delivery errors caused by confusion of units for the specification

of brachytherapy sources. NRC, Washington 2009. [cited 2012 Aug 31. 5 p.]. Available from

http://pbadupws.nrc.gov/docs/ML0807/ML080710054.pdf.

Van Dyk J., Barnett R.B., Cygler J.E., Shragge P.C. Commissioning and quality assurance of treatment

planning computers. Int. J. Radiat. Oncol. Biol. Phys. 1993;20:261-73.

Van Hoof S.J., Granton P.V., Landry G., Podesta M., Verhaegen F. Evaluation of a novel triple-channel

radiochromic film analysis procedure using EBT2. Phys. Med. Biol. 2012;57(13):4353-4368.

Varian Medical Systems. Urgent field safety notice: Varisource iX console incorrect dwell times under

special circumstances. Varian document PNL-VS-02.2010. April 2010. [cited 17 Sept 2014, 5p].

Available from: http://www.mhra.gov.uk/home/groups/dts-iac/documents/fieldsafetynotice/

con081794.pdf.

Venselaar J.L.M., Brouwer W.F.M., van Straaten B.H.M., Aalbers A.H.L. Intercomparison of calibration

procedures for Ir-192 HDR sources in The Netherlands and Belgium. Radiother. Oncol. 1994;30(2):155-

61.

Vanselaar J, Perez-Calatayud J. (Editors). A Practical guide to quality control of brachytherapy

equipment. ESTRO Booklet No. 8. Brussels, Belgium: ESTRO, 2004.

Venselaar J.L.M., Baltas D., Meigooni A.S., Hoskin P.J. (Editors). Comprehensive brachytherapy physical

and clinical aspects. Boca Raton, FL, USA: CRC Press, 2013.

Veres, A. Personal communication via email with Attila Veres, Physicist in Charge, Equal-ESTRO

Laboratory, Paris-Villejuif, France. 2013 Dec 6.

Viswanathan A.N., Thomadsen B. American Brachytherapy Society consensus guidelines for locally

advanced carcinoma of the cervix. Part I: General principles. Brachytherapy 2012;11: 33-46.

Wai P., Adamovics J., Krstajic N., Ismail A., Nisbet A., Doran S. Dosimetry of the microSelectron-HDR

Ir-192 source using Presage™ and optical CT. Appl. Radiat. Isotopes 2009;67:419-42.

Wallace A.B. Acceptance testing, commissioning and quality assurance for a 370GBq Ir-192 HDR

brachytherapy afterloader. Australas. Phys. Eng. Sci. Med. 1997;20:112-6.

Wanderas A.D., Sundset M., Langdal I., Danielsen S., Frykholm G., Marthinsen A.B.L. Adaptive

brachytherapy of cervical cancer, comparison of conventional point A and CT based individual

treatment planning. Acta. Oncol. 2012;51:345-54.

Wilkinson D.A. High dose rate (HDR) brachytherapy quality assurance: a practical guide. Biomed.

Imaging Interv. J. 2006;2:2-7.

X-5 Monte Carlo Team. MCNP - A general Monte Carlo N-Particle Transport Code, version 5. Volume

I: Overview and theory. Los Alamos National Laboratory, 2008a;LA-UR-03-1987. [cited 25 June

2014, 416p]. Available from https://laws.lanl.gov/vhosts/mcnp.lanl.gov/pdf_files/la-ur-03-1987.pdf

X-5 Monte Carlo Team. MCNP - A general Monte Carlo N-Particle Transport Code, version 5. Volume

II: User’s guide. Los Alamos National Laboratory, 2008b;LA-UR-03-1987.

Yang Y., Rivard M.J. Monte Carlo simulations and radiation dosimetry measurements of peripherally

applied HDR Ir192 breast brachytherapy D-shaped applicators. Med. Phys. 2009;36(3):809-815

Ye S-J., Brezovich I.A., Shen S., Duan J., Popple R.A., Pareek P.N. Attenuation of intracavitary

applicators in 192Ir-HDR brachytherapy. Med. Phys. 2004;31(7):2097-2106.

190

Zehtabian M., Faghihi R., Sina S. A Review on Main Defects of TG-43, Brachytherapy, Dr. Kazushi Kishi

(Ed.), ISBN: 978-953-51-0602-9, InTech, 2012. http://www.intechopen.com/books/brachytherapy/a-

review-on-the-main-defects-of-aapm-task-group-43

Zhang M., Zhang L., Yuan S. A study of experimental measurements of dosimetric parameters in HDR

Ir-192 source. Med. Dosim. 2010;35:250-4.

Zilio V.O., Joneja O.P., Popowski Y. Absolute depth-dose-rate measurements for an Ir-192 HDR

brachytherapy source in water using MOSFET detectors. Med. Phys. 2006;33:1532-9.

191

Appendices

192

A. Abstracts of Peer Reviewed Journal Papers Resulting from

this Research

Palmer A.L., Bidmead M., Nisbet A. A survey of quality control practices for high dose rate

(HDR) and pulsed dose rate (PDR) brachytherapy in the United Kingdom. J. Contemp.

Brachyther. 2012;4:232-240.

Purpose: A survey of quality control (QC) currently undertaken in UK radiotherapy centres for high dose

rate (HDR) and pulsed dose rate (PDR) brachytherapy has been conducted. The purpose was to benchmark

current accepted practice of tests, frequencies and tolerances to assure acceptable HDR/PDR equipment

performance. It is 20 years since a similar survey was conducted in the UK and the current review is timed

to coincide with a revision of the IPEM Report 81 guidelines for quality control in radiotherapy.

Material and Methods: All radiotherapy centres in the UK were invited by email to complete a

comprehensive questionnaire on their current brachytherapy QC practice, including: equipment type,

patient workload, source calibration method, level of image guidance for planning, prescribing practices,

QC tests, method used, staff involved, test frequencies, and acceptable tolerance limits.

Results: Survey data was acquired between June and August 2012. Of the 64 centres invited, 47 (73%)

responded, with 31 centres having brachytherapy equipment (3 PDR) and fully completing the survey, 13

reporting no HDR/PDR brachytherapy, and 3 intending to commence HDR brachytherapy in the near

future. All centres had comprehensive QC schedules in place and there was general agreement on key

test frequencies and tolerances. Greatest discord was whether source strength for treatment planning

should be derived from measurement, as at 58% of centres, or from the certified value, at 42%. IPEM

Report 81 continues to be the most frequently cited source of QC guidance, followed by ESTRO Booklet

No.8.

Conclusions: A comprehensive survey of QC practices for HDR/PDR brachytherapy in UK has been

conducted. This is a useful reference to which centres may benchmark their own practice. However

individuals should take a risk-assessment based approach, employing full knowledge of local equipment,

clinical procedures and available test equipment in order to determine individual QC needs.

Palmer A., Bradley D., Nisbet A. Physics-aspects of dose accuracy in high dose rate (HDR)

brachytherapy: source dosimetry, treatment planning, equipment performance and in-vivo

verification techniques. J. Contemp. Brachyther. 2012;4(2):81-91.

This study provides a review of recent publications concerned with dosimetric accuracy in high dose rate

(HDR) brachytherapy treatments. This includes work on the determination of dose rate fields around

brachytherapy sources, the capability of treatment planning systems, the performance of treatment units,

and methods to verify dose delivery. The main aims of the paper is to review the current techniques and

methods employed for HDR brachytherapy dosimetry, to discuss the current achievable accuracy in HDR

brachytherapy and to identify future areas for research in this field.

A systematic review was conducted of journal articles discussing concepts, research, and practices of

dosimetry in HDR brachytherapy published from 2002 to 2012. Medline and Embase databases were

searched with keywords “high dose rate”, “brachytherapy”, “dosimetry”, and alternative forms, limited

to English language 116 articles were found of which 84 were directly relevant for the purposes of this

review. This information was supplemented by reviewing reference lists and personal files. No other

systematic reviews of HDR brachytherapy dosimetry methods were located. This work highlights the

determinants of accuracy in HDR dosimetry and treatment delivery and presents a selection of papers,

focusing on articles from the last five years, to reflect active areas of research and development.

193

The majority of publications are related to Ir-192, this being the most common isotope in clinical use, with

a lack of work for newer HDR sources such as Co-60. With the exception of the ESTRO mailed dosimetry

service, there is little dosimetric audit activity reported in the literature, when compared with external

beam radiotherapy verification. Apart from Monte Carlo modelling of source dosimetry, there is no clear

consensus on the optimum techniques to be used to assure dosimetric accuracy through all the processes

involved in HDR brachytherapy treatment.

Palmer A. Impact of software changes: Transit dose and source position accuracy of the

Eckert & Ziegler BEBIG GmbH Multisource high dose rate (HDR) brachytherapy treatment

unit. J. Radiother. Pract. 2013:12;80-87.

Purpose: Medical device performance checks are essential following changes to control system software.

This work investigates the effects of new software on the performance of a high dose rate (HDR)

brachytherapy treatment unit.

Methods and Materials: A performance assessment was undertaken of the Eckert & Ziegler BEBIG GmbH

MultiSource® HDR treatment unit following software upgrade. Video recordings of source transits were

used to calculate transit doses, and autoradiography used to measure source dwell positions. Results were

compared to a previous study.

Results: All results showed improved performance with the new compared to old control software.

Optimal source movement profiles were observed with maximum transit speeds of 63 (+/-4) mm s-1

between dwells of 5.0 mm separation. The maximum error in transit dose correction with the new

software was 2.5 % at 10.0 mm perpendicular from the source axis, compared to 5.6 % previously. The

new software eliminated a causal relationship between curvature of the source transfer tubes and dwell

position uncertainty.

Conclusions: This work demonstrates the need for comprehensive medical device system checks following

software changes. Technical improvements in HDR device performance have been achieved with the new

software; reducing transit doses, improving transit dose correction, and improving source positioning

accuracy.

Palmer A.L., Di Pietro P., Alobaidli S., Issa F., Doran S., Bradley D., Nisbet A. Comparison of

methods for the measurement of radiation dose distributions in high dose rate (HDR)

brachytherapy: Ge-doped optical fibre, EBT3 Gafchromic film, and PRESAGE® radiochromic

plastic. Med. Phys. 2013;40(6):061707-1-11.

The measurement of dose distributions in clinical brachytherapy, for the purpose of quality control,

commissioning or dosimetric audit, is challenging and requires development. Radiochromic film dosimetry

with a commercial flatbed scanner may be suitable but careful methodologies are required to control

various sources of uncertainty. Triple-channel dosimetry has recently been utilised in external beam

radiotherapy to improve the accuracy of film dosimetry, but its use in brachytherapy, with characteristic

high maximum doses, steep dose gradients and small scales, has been less well researched. We investigate

the use of advanced film dosimetry techniques for brachytherapy dosimetry evaluating uncertainties and

assessing the mitigation afforded by triple-channel dosimetry. We present results on post-irradiation film

darkening, lateral scanner effect, film surface perturbation, film active layer thickness, film curling, and

examples of the measurement of clinical brachytherapy dose distributions. The lateral scanner effect in

brachytherapy film dosimetry can be very significant, up to 23% dose increase at 14 Gy, at ±9 cm lateral

from the scanner axis for simple single-channel dosimetry. Triple-channel dosimetry mitigates the effect,

but still limits the useable width of a typical scanner to less than 8 cm at high dose levels to give dose

uncertainty to within 1%. Triple-channel dosimetry separates dose and dose-independent signal

components and effectively removes disturbances caused by film thickness variation and surface

194

perturbations in the examples considered in this work. The use of reference dose films scanned

simultaneously with brachytherapy test films is recommended to account for scanner variations from

calibration conditions. Post-irradiation darkening, which is a continual logarithmic function with time,

must be taken into account between the reference and test films. Finally, films must be flat when scanned

to avoid Callier-type effect and to provide reliable dosimetric results. We have demonstrated that

radiochromic film dosimetry with Gafchromic EBT3 film and a commercial flatbed scanner is a viable

method for brachytherapy dose distribution measurement, and uncertainties may be reduced with triple-

channel dosimetry and specific film scan and evaluation methodologies.

Palmer A.L., Nisbet A., Bradley D. Verification of high dose rate brachytherapy dose

distributions with EBT3 Gafchromic film quality control techniques. Phys. Med. Biol.

2013;58:497-511.

It is essential that quality control (QC) techniques are developed to keep pace with modern high dose rate

(HDR) brachytherapy. Current QC methods may be insufficient to fully assure the accuracy of 3D-

optimised dose delivery. This work presents an evaluation of Gafchromic EBT3 film, with multi-channel

analysis, in HDR dose environments for advanced QC and commissioning. ‘Film-array in water’ and ‘three-

channel Solid Water block’ purpose-designed phantoms are utilised. Dose and dose-rate dependency and

practical film usage has been evaluated. EBT3 measurements of dose with radial distance from a HDR

source are compared to Monte Carlo data. Semi-3D dose distributions around clinical HDR applicators are

compared to treatment plans. The measurement of delivery accuracy for inverse-planned pseudo-clinical

test cases, with correct delivery and simulated treatment errors, has also been investigated. Local gamma

criteria of 3%, 3 mm is recommended with passing rates of at least 96% typically achieved. The system is

sensitive to simulated errors in HDR delivery, with significant reductions of passing rate. It has been

demonstrated that EBT3 Gafchromic film, in combination with multi-channel analysis, is appropriate for

applicator, treatment unit and planning system commissioning measurements as well as practical routine

QC to confirm agreement of planned and delivered complex HDR brachytherapy dose distributions.

Palmer A.L., Lee C., Ratcliffe A.J., Bradley D., Nisbet A. Design and implementation of a film

dosimetry audit tool for comparison of planned and delivered dose distributions in high

dose rate (HDR) brachytherapy. Phys. Med. Biol. 2013;58:6623-6640.

A novel phantom is presented for ‘full system’ dosimetric audit comparing planned and delivered dose

distributions in HDR gynaecological brachytherapy, using clinical treatment applicators. The

BRachytherapy Applicator Dosimetry (‘BRAD’) test object consists of a near full-scatter water tank with

applicator and film supports constructed of Solid Water, accommodating any typical cervix applicator.

Film dosimeters are precisely held in four orthogonal planes bisecting the intrauterine (IU) tube, sampling

dose distributions in the high risk clinical target volume (HR-CTV), Points A and B, bladder, rectum and

sigmoid. The applicator position is fixed prior to CT scanning and through treatment planning and

irradiation. The CT data is acquired with the applicator in a near clinical orientation to include applicator

reconstruction in the system test. Gamma analysis is used to compare treatment planning system

exported RTDose grid with measured multi-channel film dose maps. Results from two pilot audits are

presented, using Ir-192 and Co-60 HDR sources, with mean gamma passing rates of 98.6% using criteria

of 3% local normalisation and 3 mm distance to agreement (DTA). The mean DTA between prescribed

dose and measured film dose at Point A was 1.2 mm. The phantom was funded by IPEM and will be used

for a UK national brachytherapy dosimetry audit.

195

Palmer A.L., Bradley D., Nisbet A. Evaluation and implementation of triple-channel

radiochromic film dosimetry in brachytherapy. J. Appl. Clin. Med. Phys. 2014;15(4):280-296.

The measurement of dose distributions in clinical brachytherapy, for the purpose of quality control,

commissioning or dosimetric audit, is challenging and requires development. Radiochromic film dosimetry

with a commercial flatbed scanner may be suitable but careful methodologies are required to control

various sources of uncertainty. Triple-channel dosimetry has recently been utilised in external beam

radiotherapy to improve the accuracy of film dosimetry, but its use in brachytherapy, with characteristic

high maximum doses, steep dose gradients and small scales, has been less well researched. We investigate

the use of advanced film dosimetry techniques for brachytherapy dosimetry evaluating uncertainties and

assessing the mitigation afforded by triple-channel dosimetry. We present results on post-irradiation film

darkening, lateral scanner effect, film surface perturbation, film active layer thickness, film curling, and

examples of the measurement of clinical brachytherapy dose distributions. The lateral scanner effect in

brachytherapy film dosimetry can be very significant, up to 23% dose increase at 14 Gy, at ±9 cm lateral

from the scanner axis for simple single-channel dosimetry. Triple-channel dosimetry mitigates the effect,

but still limits the useable width of a typical scanner to less than 8 cm at high dose levels to give dose

uncertainty to within 1%. Triple-channel dosimetry separates dose and dose-independent signal

components and effectively removes disturbances caused by film thickness variation and surface

perturbations in the examples considered in this work. The use of reference dose films scanned

simultaneously with brachytherapy test films is recommended to account for scanner variations from

calibration conditions. Post-irradiation darkening, which is a continual logarithmic function with time,

must be taken into account between the reference and test films. Finally, films must be flat when scanned

to avoid Callier-type effect and to provide reliable dosimetric results. We have demonstrated that

radiochromic film dosimetry with Gafchromic EBT3 film and a commercial flatbed scanner is a viable

method for brachytherapy dose distribution measurement, and uncertainties may be reduced with triple-

channel dosimetry and specific film scan and evaluation methodologies.

Palmer A.L., Bradley D., Nisbet A. Dosimetric audit in brachytherapy. Br. J. Radiol.

2014;87:20140105.

Dosimetric audit is required for the improvement of patient safety in radiotherapy and to aid optimisation

of treatment. The reassurance that treatment is being delivered in line with accepted standards, that

delivered doses are as prescribed, and that quality improvement is enabled, is as essential for

brachytherapy as it is for the more commonly audited external beam radiotherapy. Dose measurement

in brachytherapy is challenging due to steep dose gradients and small scales, especially in the context of

an audit. Several different approaches have been taken for audit measurement to date; thimble and well-

type ionisation chambers, thermoluminescent detectors, optically stimulated luminescence detectors,

radiochromic film, and alanine. In this work, we review all of the dosimetric brachytherapy audits that

have been conducted in recent years, look at current audits in progress, and propose required directions

for brachytherapy dosimetric audit in the future. The concern over accurate source strength

measurement may be essentially resolved with modern equipment and calibration methods, but

brachytherapy is a rapidly developing field and dosimetric audit must keep pace.

Palmer A.L. BJR brachytherapy dosimetry special feature. Br. J. Radiol. 2014;87:20140506.

Editorial publication that did not contain an abstract.

196

Palmer A.L., Bradley D., Nisbet A. Evaluation and mitigation of potential errors in

radiochromic film dosimetry due to film curvature at scanning. J. Appl. Clin. Med. Phys. 2015

(in press)

This work considers a previously overlooked uncertainty present in film dosimetry which results from

moderate curvature of films during the scanning process. Small film samples are particularly susceptible

to film curling which may be undetected or deemed insignificant. In this study we consider test cases with

controlled induced curvature of film and with film raised horizontally above the scanner plate. We also

evaluate the difference in scans of a film irradiated with a typical brachytherapy dose distribution with

the film naturally curved and with the film held flat on the scanner. Typical naturally occurring curvature

of film at scanning, giving rise to a maximum height 1 to 2 mm above the scan plane, may introduce dose

errors of 1 to 4%, and considerably reduce gamma evaluation passing rates when comparing film-

measured doses with treatment planning system calculated dose distributions, a common application of

film dosimetry in radiotherapy. The use of a triple-channel dosimetry algorithm appeared to mitigate the

error due to film curvature compared to conventional single-channel film dosimetry. The change in pixel

value, and calibrated reported dose, with film curling or height above the scanner plate, may be due to

variations in illumination characteristics, optical disturbances, or a Callier-type effect. There is a clear

requirement for physically-flat films at scanning to avoid the introduction of a substantial error source in

film dosimetry. Particularly for small film samples, a compression glass plate above the film is

recommended to ensure flat-film scanning. This effect has been overlooked to date in the literature.

Palmer A.L., Diez P., Gandon L., Wynn-Jones A., Bownes P., Lee C., Aird E., Bidmead M., Lowe

G., Bradley D., Nisbet A. A multicentre ‘end to end’ dosimetry audit for cervix HDR

brachytherapy treatment. Radiother. Oncol. 2015 (in press, http://dx.doi.org/

10.1016/j.radonc.2014.12.006).

Purpose: To undertake the first multicentre fully ‘end to end’ dosimetry audit for HDR cervix

brachytherapy, comparing planned and delivered dose distributions around clinical treatment applicators,

with review of local procedures.

Materials and Methods: A film-dosimetry audit was performed at 46 centres, including imaging, applicator

reconstruction, treatment planning and delivery. Film dose maps were calculated using triple-channel

dosimetry and compared to RTDose data from treatment planning systems. Deviations between plan and

measurement were quantified at prescription Point A and using gamma analysis. Local procedures were

also discussed.

Results: The mean difference between planned and measured dose at Point A was -0.6% for plastic

applicators and -3.0% for metal applicators, at standard uncertainty 3.0% (k=1). Isodose distributions

agreed within 1 mm over a dose range 2-16 Gy. Mean gamma passing rates exceeded 97% for plastic and

metal applicators at 3% (local) 2 mm criteria. Two errors were found: one dose normalisation error and

one applicator library misaligned with the imaged applicator. Suggestions for quality improvement were

also made.

Conclusions: The concept of ‘end to end’ dosimetry audit for HDR brachytherapy has been successfully

implemented in a multicentre environment, providing evidence that a high level of accuracy in

brachytherapy dosimetry can be achieved.

197

B. Abstracts of Conference Presentations and Posters Resulting

from this Research

Palmer A., Ioannou L., Hayman O. Nagar Y.S. Is HDR equipment performance suitable for

modern brachytherapy? Positional errors, dosimetric impact & case study. Proceedings of

the World Congress of Brachytherapy, Barcelona, Spain. Radiother. Oncol. 2012;103:S132-3.

Purpose: To study the effect of simulated HDR source dwell position errors on the dose to the clinical

target volume (HR-CTV) and organs at risk in cervix cancer. Determine the clinically relevant positional

accuracy requirement; inform QC needs and treatment uncertainty estimates. Interpret the performance

of an HDR unit using this data.

Method: Simulated dwell position errors, 0.2 to 10.0 mm, were introduced in eight HDR cervix plans and

the dosimetric effect on ICRU and GEC-ESTRO DVH parameters calculated. CT plans on the Eckert & Ziegler

BEBIG (EZ) HDRplus planning system, with Co-60 source and IUT/split-ring applicator were used

throughout. Two error modes were simulated: (a) systematic proximal shift (away from HDR unit) position

calibration error; (b) source cable take-up lag, possible with older EZ software, first dwell unaffected and

others shifted distally. Video source position tests for the EZ HDR MultiSource were used with the above

simulation data to predict effect on clinical dose delivery.

Results: Figure. 1 shows the relationship between systematic errors in source dwell positions, mode (a),

and the % change in DVH metrics, D90 and D2cc. The mean and interquartile range from eight plans is

provided. There is an approximate linear relationship; errors of 1.0 mm result in 2.0 % change, 4.0 mm

error 10.0 % change. ICRU A pt. exhibits less change than DVH metrics, 0.4% and 0.7% respectively.

Bladder doses increase while rectum, sigmoid and HR-CTV decrease due to the proximal shift direction of

dwells. This would be reversed for distal shifts. Table 1 gives additional detailed DVH data for 0.2 to 2.0

mm position errors for both modes. Thirty QC test results of source position were collated for the EZ HDR

unit. 13% were recorded within 0.25 mm error, 64% within 0.5 mm, and 23% within 1.0 mm. Previous

software version of the EZ unit had the potential for small dwell position errors due to curvature of the

transfer tube: This effect on dose delivery uncertainty is given in Table 1(b). This is eliminated with new

software.

Conclusion: An accuracy of at least 1.0 mm in dwell positions is required to limit the effect on clinically

relevant dose delivery parameters to within an acceptable level of 3.0%. Only at 0.2 mm accuracy are the

effects all within 1.0%. A QC action level of 0.5 mm is proposed, with 1.0 mm maximum tolerance. All

recorded QC position results for the EZ unit were within the clinically relevant accuracy level. Errors due

to source cable take-up lag, with potential DVH uncertainty up to 6%, have been eliminated with the new

software.

Palmer A.L., Nisbet A., Bradley D.A. Semi-3D dosimetry of high dose rate brachytherapy

using a novel Gafchromic EBT3 film-array water phantom. J. Phys.: Conf. Ser.

2013;444:012101. Published abstract, poster and oral presentation at the conference:

Proceedings of the International Conference on 3D Dosimetry (IC3DDose), Sydney, Australia,

November 2012.

There is a need to modernise clinical brachytherapy dosimetry measurement beyond traditional point

dose verification to enable appropriate quality control within 3D treatment environments. This is to keep

pace with the 3D clinical and planning approaches which often include significant patient-specific

optimisation away from ‘standard loading patterns’. A multi-dimension measurement system is required

to provide assurance of the complex 3D dose distributions, to verify equipment performance, and to

enable quality audits. However, true 3D dose measurements around brachytherapy applicators are often

198

impractical due to their complex shapes and the requirement for close measurement distances. A solution

utilising an array of radiochromic film (Gafchromic EBT3) positioned within a water filled phantom is

presented. A calibration function for the film has been determined over 0 to 90Gy dose range using three

colour channel analysis (FilmQAPro software). Film measurements of the radial dose from a single HDR

source agree with TPS and Monte Carlo calculations within 5 % up to 50 mm from the source. Film array

measurements of the dose distribution around a cervix applicator agree with TPS calculations generally

within 4 mm distance to agreement. The feasibility of film array measurements for semi-3D dosimetry in

clinical HDR applications is demonstrated.

Palmer A.L., Diez P., Aird E., Lee C., Radcliffe A., Gouldstone C., Sander T., Bradley D., Nisbet

A. Development of a UK dosimetry audit for HDR/PDR brachytherapy. Med. Phys. Int.

2013;1(2):233 (oral presentation at ICMP2013, Brighton, UK)

Purpose: To report on the development of the first comprehensive dosimetry audit of HDR/PDR

brachytherapy, part-funded by IPEM and RTTQA Interlace trial, supported by NPL. External audit in

brachytherapy is an underdeveloped quality assurance process. Worldwide, only seven audits have been

reported, limited to point dose or source strength, and have omitted treatment applicator dosimetry.

Method: The audit uses two complementary phantoms: (1) accurate measurement of point dose using

alanine from a series of source dwells in a straight plastic catheter, (2) measurement of dose distribution

around clinical cervix applicators using film and comparison to treatment planning system calculations.

Phantom (1) is constructed from Solid Water, within a Perspex scatter box, and contains a central HDR

source catheter with three concentric measuring points at 20 mm, spaced by 120°, which can

accommodate both alanine pellets and Farmer chambers. Phantom (2) consists of a Solid Water frame

which precisely and rigidly holds the applicator and four films, contained within a full scatter water tank.

The latter constitutes a ‘system test’ including applicator CT reconstruction.

Results: Alanine and film dosimetry systems have been calibrated and validated for use in brachytherapy

applications for both Ir-192 and Co-60 sources. Pilot audits have been completed demonstrating the

suitability of the phantom designs. Comparison of planned and measured dose distribution had a mean

gamma passing rates of 98.6% using criteria of 3% local normalisation and 3 mm distance to agreement.

Conclusion: A brachytherapy dosimetry audit has been developed that can provide a unique quality

assurance assessment.

Patel I., Palmer A. Revision of IPEM guidance on quality control of radiotherapy equipment.

Med. Phys. Int. 2013;1(2):179 (invited oral presentation at ICMP2013, Brighton, UK).

Invited presentation, no abstract requested.

Palmer A.L., Nisbet A., Bradley D.A. A new standard for HDR brachytherapy quality control:

practical and advanced film dosimetry for treatment applicators and sources. Proceedings

of the 2nd Estro Forum, Geneva, Switzerland. Radiother. Oncol. 2013;106:S368-9.

Purpose: Robust and relevant quality control (QC) is essential to ensure accuracy and safety in

radiotherapy. QC in HDR brachytherapy has not kept pace with clinical modernisation e.g. volume-based

prescriptions and patient-specific inverse-optimisation. New QC standards are required to confirm dose

distribution accuracy and avoid errors. Typical QC is currently limited to straight catheters, point doses

and well chambers. A 'state of the art' practical implementation of advanced multichannel film dosimetry

is presented to verify delivered dose distributions around treatment applicators and sources. 3 cases are

presented; a cervix ring applicator, a shielded vaginal cylinder, and an isolated source catheter. The

application of gamma assessment criteria in brachytherapy is discussed.

199

Materials and Methods: A film-array water phantom was used to accurately position Gafchromic EBT3

film, cut to 12 x 12 cm. Multichannel film dosimetry separating dose-dependent and dose-independent

components was used (FimQAPro). Dose-maps were compared to treatment plans using isodoses and

gamma. Absolute film dose values were used with no re-normalisation of any data. Isolated source doses

were compared to TG-43 source model values. The value and sensitivity of gamma for brachytherapy

applications was assessed by multivariate analysis of area/position and calculation parameters. Eckert &

Ziegler Bebig GmbH HDR multiSource treatment unit, with Co-60 source, and HDRplus treatment planning

system (TPS) were used throughout.

Results: Figure 1 shows dose maps from 2 films positioned adjacent to and bisecting the cervix applicator

intrauterine (IU) channel, overlaid on TPS isodoses. Agreement in isodoses, 75 cGy to 2000 cGy, is

generally within 1.0 mm. A comparison of the 2 symmetric films confirms sufficient reproducibility.Table

1 provides example gamma results for the cervix and shielded vaginal applicators. The passing rate in

brachytherapy is sensitive to the defined interest region, in the cervix example ranging from 95% at typical

HR-CTV to 100% at a bladder position, for 3% (local) / 2 mm criteria, evaluated over 9 cm2 regions. The

full-region, 144 cm2, passing rate was ~ 98%.The validity of the TG-43 general source model for individual

supplied HDR sources was successfully verified. Gamma passing rates > 95% at 3% (local) / 2 mm between

5 mm and 50 mm from the source.

Conclusions: There is an absence of clinically-relevant QC for modern brachytherapy. We have presented

a practical, robust method of advanced film dosimetry of treatment applicators, which is more closely

aligned to clinical treatments than current QC. Planned and measured isodoses agreed closely, with high

gamma passing rates. Film dosimetry is advocated to confirm validity of the general TG-43 model for

individual supplied sources. The use and sensitivity of gamma for brachytherapy must be carefully

considered; we propose separate calculations in a number of clinically relevant regions.

Nisbet A., Palmer A.L.*, Bradley D.A. Available guidance, current UK practice, and future

directions for HDR brachytherapy quality control. Proceedings of the 2nd Estro Forum,

Geneva, Switzerland. Radiother. Oncol. 2013;106:S370. (*main and presenting author).

Purpose: A survey of high dose rate (HDR) brachytherapy quality control (QC) procedures undertaken at

radiotherapy centres in the United Kingdom (UK) is reported. Published recommendations and guidance

for HDR QC are also reviewed and compared to current UK practice. Recent changes in clinical

brachytherapy techniques and the impact on required QC is discussed. Modern methods to determine

optimum quality checking processes are indicated. This work is conducted in the context of the recent

‘point/counterpoint’ debate in Medical Physics that “QA procedures in radiation therapy are outdated

and negatively impact the reduction of errors” and a review of the dosimetric accuracy in HDR.

Materials and methods: All UK radiotherapy centres were asked to participate in a survey of their

approach and practice for HDR brachytherapy QC. This included guidance used, frequencies and tolerance

values for individual QC tests. A comprehensive evaluation of responses was conducted detailing

popularity of tests, and the average and range values of testing and tolerance. A literature search was

conducted on general guidance, specific QC techniques in both brachytherapy and teletherapy, and on

risk-based systems for quality assurance.

Results: Survey data was acquired from 31 UK radiotherapy centres and statistical analysis of responses

performed. 45 possible individual QC tests were identified. There was general agreement on

measurement frequency and tolerance for key QC tests, e.g. measurement of source position in a straight

catheter, checked daily and with a 1.0mm tolerance in most centres. There was disagreement on a

number of tests, e.g. the need for regular x-ray imaging of applicators. There was absence of tests that

may be deemed necessary for modern brachytherapy practice, e.g. confirmation of planned and delivered

dose distributions. There is likely a need to move from a device-centred to a system-centred approach,

using risk-based assessment methods to determine required QC testing, with emphasis on clinical

processes rather than simple device operation. Table 1 provides sample key results from the work.

200

Conclusion: The only contemporary benchmark survey of HDR QC practice has been undertaken. The

outcome of this work is a review of current practice against available recommendations, relevant recent

changes in clinical brachytherapy techniques, and the use of modern quality process assessments.

Recommendations for appropriate, optimised QC for HDR brachytherapy are made.

Palmer A.L. UK Brachytherapy Audit: origin, funding and collaboration. Presented at IPEM,

NPL, CTRad meeting on “Reaching a Consensus on Verification of Radiotherapy Delivery”,

held at NPL 13th December 2013. (Invited oral presentation).


Palmer A.L. Group hugs in brachytherapy physics. Presented at Medical Physics Research

Away Day, Portsmouth Hospitals NHS Trust, 26/02/14. Iinvited oral presentation).


Palmer A.L., Bradley D.A., Nisbet A. Comprehensive audit of brachytherapy dose

distributions: A methodology and UK audit results. Poster at ESTRO33 conference, Vienna,

Austria, April 2014. (Radiother. Oncol. supplement in press).

Purpose: To describe a methodology for comprehensive dosimetric audit in HDR/PDR brachytherapy and

present results of audits undertaken in the United Kingdom (UK). This work is conducted under the

auspices of the Institute of Physics and Engineering and Medicine (IPEM). Complementary to routine QA

measurements, an independent system-test provides assurance of the dosimetric accuracy of the

combination of all elements of the brachytherapy-physics process and equipment performance.

Dosimetric audit for external beam radiotherapy is common, but is much less developed for

brachytherapy, albeit equally valid; this deficit is addressed by present work. We describe a unique

brachytherapy ‘end-to-end’ quality assurance assessment involving CT scan, applicator reconstruction,

treatment planning system (TPS) calculation, multiple-2D dose measurement around clinical treatment

applicators, and comparison with intended dose distribution and prescription dose.

Materials and methods: A novel water and Solid-WaterTM phantom was used for the audits, conducted

by a visiting-physicist at 10 radiotherapy centres in the UK, June to October 2013, with a further 25 audits

scheduled until April 2014, to be presented. A clinical treatment applicator, ring or ovoids and intrauterine

(IU) tube, was CT scanned in the phantom, reconstructed on CT, treatment planned and the planned dose

delivered to the phantom. Dose was measured using Gafchromic EBT3 film in two planes bisecting the IU,

with 3-channel analysis. The measured dose distribution was compared to the TPS exported RTDose using

isodose overlay and gamma analysis, and compared to the prescribed dose at Point A using distance to

agreement and point percentage difference, shown in the figure. A full uncertainty analysis was

performed. The standard operating procedures at each centre were also reviewed during the audit.

Results: The table provides results comparing TPS planned and film-measured dosimetry from 10 audits

in the UK, including Ir-192, Co-60, HDR and PDR units. Isodose comparison showed mean gamma passing

rate of 99.0% at 3% (local norm.) 3 mm, and mean ‘distance to agreement’ of prescribed and measured

dose at prescription Point A of 1.0 mm (0.6 mm standard uncertainty, k=1).

Conclusion: A brachytherapy audit methodology has been established to provide assurance of dose

delivery: planned dose distribution and prescription dose. The first 10 audits in the UK have shown

clinically acceptable results. Processes were reviewed and comment given, including clarity on dwell

positions within applicators, and use of optimised compared to ‘standard-plans’. The audit system has

also been used during new equipment commissioning and to investigate a reported high toxicity series of

treatments.

201

Palmer A.L., Bradley D.A., Nisbet A. Improving quality assurance of HDR brachytherapy:

Verifying agreement between planned and delivered dose distributions using DICOM

RTDose and advanced film dosimetry. Poster at AAPM 2014 annual meeting, Austin, Texas,

July 2014.

Purpose: HDR brachytherapy is undergoing significant development, and quality assurance (QA) checks

must keep pace. Current recommendations do not adequately verify delivered against planned dose

distributions: This is particularly relevant for new treatment planning system (TPS) calculation algorithms

(non TG-43 based), and an era of significant patient-specific plan optimisation. ‘Full system checks’ are

desirable in modern QA recommendations, complementary to device-centric individual tests. We present

a QA system incorporating TPS calculation, dose distribution export, HDR unit performance, and dose

distribution measurement. Such an approach, more common in external beam radiotherapy, has not

previously been reported in the literature for brachytherapy.

Methods: Our QA method was tested at 24 UK brachytherapy centres. As a novel approach, we used the

TPS DICOM RTDose file export to compare planned dose distribution with that measured using

Gafchromic EBT3 films placed around clinical brachytherapy treatment applicators. Gamma analysis was

used to compare the dose distributions. Dose difference and distance to agreement were determined at

prescription Point A. Accurate film dosimetry was achieved using a glass compression plate at scanning to

ensure physically-flat films, simultaneous scanning of known dose films with measurement films, and

triple-channel dosimetric analysis.

Results: The mean gamma pass rate of RTDose compared to film measured dose distributions was 98.1%

at 3%(local), 2 mm criteria. The mean dose difference, measured to planned, at Point A was -0.5% for

plastic treatment applicators and -2.4% for metal applicators, due to shielding not accounted for in TPS.

The mean distance to agreement was 0.6 mm.

Conclusions: It is recommended to develop brachytherapy QA to include full-system verification of

agreement between planned and delivered dose distributions. This is a novel approach for HDR

brachytherapy QA. A methodology using advanced film dosimetry and gamma comparison to DICOM

RTDose files has been demonstrated as suitable to fulfil this need.

Palmer A.L. Advanced Film Dosimetry for Brachytherapy. Invited oral presentation as

Guest Speaker at Gafchromc Symposium, Ashland Inc, held during AAPM 2014 annual

conference, Austin, Texas, July 2014. Available at:

http://www.filmqapro.com/documents/Palmer_Brachy%20Audit_Gafchromic_AAPM_2014

.pdf


Palmer A.L. Auditing HDR/PDR Brachytherapy Physics in UK. Invited oral presentation at

IPEM Biennial Radiotherapy Meeting, Sept 2014, Glasgow.

The reassurance that treatment is being delivered in line with accepted standards, that delivered doses

are as prescribed, that opportunities for quality improvement are taken, and that quality control practices

are optimised, is a fundamental requirement in radiotherapy. This is as essential for brachytherapy as it

is for the more commonly audited and reviewed external beam radiotherapy [1]. Unfortunately, dose

measurement in brachytherapy is challenging due to steep dose gradients and small scales, especially in

the context of an external audit. After a brief review of errors in brachytherapy, to set the scene, this

presentation will discuss the concept of dosimetric audit and the recent resurgence of brachytherapy

audit activity in the UK, benefiting from 4 complementary audits being conducted in the UK during 2014.

The presentation will focus on the recently completed IPEM-sponsored BRachytherapy Applicator

http://www.filmqapro.com/documents/Palmer_Brachy%20Audit_Gafchromic_AAPM_2014.pdf

http://www.filmqapro.com/documents/Palmer_Brachy%20Audit_Gafchromic_AAPM_2014.pdf

202

Dosimetry (BRAD) audit, which comprised an end-to-end system check of planned and delivered dose

distributions around clinical brachytherapy applicators at every brachytherapy centre in UK [2]. The test

object is now available via the IPEM virtual phantom library. Recommendations for film dosimetry

techniques as used for the audit will be discussed [3]. Observations on the diversity of brachytherapy-

physics practice encounter during the audits will be mentioned, with a few suggestions for improvement.

In line with the brachytherapy audit theme, the presentation will also include a brief review of an audit of

brachytherapy quality control practice that was conducted recently in the UK, identifying areas of

consensus and variance.

Palmer A.L., Bradley D.A., Nisbet A. Monte Carlo derived correction factors for

brachytherapy film dosimetry for audit and QC. Poster presentation at ESTRO 3rd Forum

conference, April 2015, Barcelona.

Purpose: Radiochromic film is a valuable dosimeter in brachytherapy used for dosimetry audit or QC.

However, the water-equivalence of film and any resulting dose perturbation when used with Ir-192 and

Co-60 HDR sources has not been sufficiently studied. The need for sufficiently large phantoms for full-

scatter is understood, but an optimum phantom size for practical transport for audit measurements for

both Ir-192 and Co-60, with appropriate correction factors, has not been determined. We address both

issues using Monte Carlo derived correction factors. This work was undertaken for an IPEM brachytherapy

audit in the UK, and will be of value for others conducting brachytherapy film dosimetry measurements

or conducting Monte Carlo calculations in brachytherapy.

Methods: MCNP5 code and Ir-192 and Co-60 HDR source models were initially validated by comparing

radial dose with consensus data. Due to small scale geometries, *F8 tally with full photon and electron

transport was used, with 1E9 particle histories. Gafchromic EBT3 film chemical composition and geometry

were modelled and calculations made to determine any dose perturbation due to the film compared to a

uniform water case. A separate geometry was constructed to study scatter corrections at distances 20 to

75 mm from Ir-192 and Co-60 sources within water spheres of 100 to 600 mm, and equivalent cubes. An

optimum phantom size for brachytherapy dosimetry audit was proposed.

Results: For both Ir-192 and Co-60, there was negligible percentage change in dose due to the presence

of a 15 mm strip of Gafchromic EBT3 film, shown in Table 1, confirming its ‘water equivalence’ for HDR

brachytherapy dosimetry measurements. MCNP5 derived phantom scatter correction factors for Ir-192

and Co-60 are provided in Figure 1. There is agreement within 0.1% with data for Ir-192 calculated by

Perez-Calatayud et al. (Med. Phys. 2004). A water phantom cube of side length 400 mm was considered

optimum for brachytherapy dosimetry audit measurements, requiring scatter correction factors at 20 mm

and 60 mm from the source of 0.0% and 0.4% for Ir-192 and 0.0% and 0.2% for Co-60, respectively. This

is a practical maximum size for convenient transportation.

Conclusions: Monte Carlo calculations using MCNP5 were conducted to support HDR brachytherapy

dosimetry audit and QC measurements. It was confirmed Gafchromic EBT3 film does not disturb dose

distributions compared to water-only cases. An optimum phantom size of 400 mm side cube was

proposed with a maximum 0.4% scatter correction at 60 mm from the source. This data has been utilised

in an IPEM dosimetry audit in the UK and may be applied in other brachytherapy dosimetry situations.

203

C. Risk Assessment for Brachytherapy Dosimetric Audit with the

BRAD Phantom

Table C.1. Risk assessment for BRAD audit.

Assessed by: Tony Palmer Date: August 2013 M

itig

atio

n o

r fu

rth

er

acti

on

s re

qu

ired

No

fu

rth

er a

ctio

n r

equ

ired

No

fu

rth

er a

ctio

n r

equ

ired

Ensu

re g

oo

d c

om

mu

nic

atio

n

of

limit

atio

ns

of

aud

it. N

o

furt

her

act

ion

. Err

or

wo

uld

be

ou

tsid

e sc

op

e o

f th

is

rese

arch

/au

dit

.

Res

earc

h t

o e

nab

le f

irst

use

of

film

fo

r b

rach

yth

erap

y

do

sim

etry

sys

tem

au

dit

is

app

rop

riat

e an

d s

afe

Res

earc

h t

o e

nab

le f

irst

use

of

film

fo

r b

rach

yth

erap

y

do

sim

etry

sys

tem

au

dit

is

app

rop

riat

e an

d s

afe

Ris

k ra

tin

g*

Grade (1-25)

3 3 9 4 3

Consequence (1-5 3 3 3 4 3

Likelihood (1-5) 1 1 3 1 1

Cu

rren

t co

ntr

ol m

easu

res

Sco

pe

of

aud

it a

nd

lim

itat

ion

s cl

earl

y

stat

ed o

n p

re-a

ud

it p

roce

du

re a

nd

aud

it r

epo

rt

Au

dit

ors

rec

eive

dir

ect

trai

nin

g fr

om

lead

sci

enti

st. D

etai

led

an

d r

ob

ust

test

ing

of

do

sim

etry

sys

tem

per

form

ance

Au

dit

is n

eces

sari

ly li

mit

ed t

o

mea

sure

men

t o

f sp

ecif

ic p

aram

eter

s.

Pu

rpo

se o

f au

dit

is a

‘sys

tem

’ tes

t to

be

sen

siti

ve t

o s

ever

al e

rro

rs.

Lim

itat

ion

s o

f sc

op

e cl

earl

y st

ated

.

Det

aile

d a

nd

ro

bu

st t

esti

ng

of

do

sim

etry

sys

tem

per

form

ance

un

der

take

n. P

ilot

aud

its

and

ch

eck

on

err

or

det

ecti

on

un

der

take

n.

Det

aile

d a

nd

ro

bu

st t

esti

ng

of

do

sim

etry

sys

tem

pe

rfo

rman

ce

un

der

take

n. P

ilot

aud

its

and

ch

eck

on

err

or

det

ecti

on

un

der

take

n.

Po

ssib

le o

utc

om

e /

per

son

s at

ris

k

Au

dit

ou

tpu

t n

ot

suff

icie

nt

or

asp

ects

of

qu

alit

y n

ot

suff

icie

ntl

y ve

rifi

ed. H

ost

cen

tre

no

t sa

tisf

ied

.

Au

dit

res

ult

s n

ot

avai

lab

le o

r in

corr

ect

.

Was

ted

tim

e an

d r

eso

urc

e o

r fu

rth

er w

ork

to

reso

lve

inco

rre

ct r

esu

lts

Do

sim

etri

c er

ror

rem

ain

s u

nd

etec

ted

.

Au

dit

or/

aud

it s

po

nso

r m

ay b

e ac

cuse

d o

f

neg

ligen

ce.

Inap

pro

pri

ate

reas

sura

nce

giv

en t

o h

ost

ph

ysic

ist

and

rad

ioth

erap

y ce

ntr

e. E

rro

r

con

tin

ues

un

det

ecte

d. P

ote

nti

al s

ign

ific

ant

har

m t

o p

atie

nts

co

nti

nu

es.

Au

dit

or/

aud

it s

po

nso

r m

ay b

e ac

cuse

d o

f

neg

ligen

ce.

Ho

st p

hys

icis

t an

d r

adio

ther

apy

dep

artm

ent

incu

r ex

pen

se a

nd

res

ou

rces

un

nec

ess

arily

.

Trea

tmen

t eq

uip

men

t ta

ken

ou

t o

f cl

inic

al

use

un

nec

ess

arily

del

ayin

g p

atie

nts

,

trea

tmen

t.

Haz

ard

Sco

pe

of

aud

it n

ot

un

der

sto

od

bet

wee

n a

ud

ito

r

and

ho

st c

entr

e

Au

dit

is n

ot

con

du

cted

corr

ect

ly d

ue

to e

qu

ipm

ent

or

aud

ito

r er

ror

Erro

r ex

ists

in h

ost

cen

tre

bra

chyt

her

apy

do

sim

etry

bu

t

aud

it d

oes

no

t m

easu

re t

his

spec

ific

par

amet

er a

nd

rem

ain

s u

nd

etec

ted

Au

dit

mea

sure

s a

par

amet

er

wh

ich

co

nta

ins

an e

rro

r b

ut

do

es n

ot

det

ect

actu

al e

rro

r

pre

sen

t in

bra

chyt

her

apy

do

sim

etry

(fa

lse

neg

ativ

e)

Au

dit

inco

rre

ctly

rep

ort

s an

erro

r in

bra

chyt

her

apy

do

sim

etry

th

at d

oes

no

t

actu

ally

exi

st (

fals

e p

osi

tive

)

* Likelihood: 1 rare, 2 unlikely, 3 possible, 4 likely, 5 almost certain Consequence: 1 negligible, 2 minor, 3 moderate, 4 major, 5 catastrophic

204

D. Protocol for Brachytherapy Dosimetric Audit with the BRAD

Phantom

BRachytherapy Applicator Dosimetry Audit Protocol v2. Jan 2014 Written by Tony Palmer on behalf of an IPEM RT-SIG Working Party on Brachytherapy Audit

Introduction The BRachytherapy Applicator Dosimetry (BRAD) phantom provides an audit tool for the comparison of planned and delivered dose distributions in several 2D planes around clinical gynaecological applicators with HDR or PDR brachytherapy treatment equipment using Ir-192 or Co-60 sources. (A complementary audit by RTTQA is being conducted in collaboration, aimed at accurate measurement of point dose from a straight catheter). The process provides a ‘full system’ end-to-end audit which provides a combined performance assessment of CT scanning and reconstruction of the applicator, dose prescribing, dose calculation and source data in the treatment planning system (TPS), plan export consistency, and delivery at the treatment unit measuring dose distribution from clinical treatment applicators, and inherent confirmation of source calibration value in the TPS. The audit does not include clinical or technical assessment of outlining accuracy. A ‘standard plan’ or recent/typical treatment plan is used. Local CT scanning protocol and dose prescription levels can be used for the audit. Advanced three-channel dosimetry is used for film dose analysis. The TPS-planned and film-measured dose distributions are compared using isodose overlay and gamma analysis. The percentage difference at Point A and the distance to agreement between the dose measured at Point A and position of the measured prescription isodose is also provided. Further details on the phantom design, testing and typical results are available in Palmer et al. 2013a and on general film dosimetry for brachytherapy dose distribution assessment in Palmer et al. 2013b. 1. Scope, Responsibilities, and Preparations for the Audit The determination of a ‘clinically relevant’ departure between intended and delivered radiation dose distributions may be more difficult to define in brachytherapy than conventional accepted practice in external beam radiotherapy, and may be dependent on local factors for each radiotherapy centre. It is likely to be in the region of a dose difference greater than 5.0%, or a shift in clinically relevant isodose lines of greater than 3.0 mm that is considered significant in brachytherapy. The film dosimetry method has an increased uncertainty compared to conventional primary dosimetric methods such as ionisation chamber dosimetry, but is used here to provide a spatial dose distribution assessment. Films are calibrated in MV radiation against an ionisation chamber with calibration traceable to NPL primary standard. The brachytherapy audit is conducted as a ‘spot check’ only and is not a comprehensive assessment of all possible treatment modes or equipment. This necessarily constitutes an assessment of one specific aspect of physics dosimetry alone, not any clinical aspects of treatment. The result is of course valid only at the time of measurement. The audit is conducted under the auspices of a Working Party of IPEM RT-SIG, and is therefore underwritten by professional indemnity insurance of IPEM, covering activities of members while acting on behalf of IPEM. A general disclaimer is made that it is the responsibility of the local physicist for accuracy of all treatments and caution should be exercised in interpreting results from any single audit process. This must be done in conjunction with comprehensive local commissioning and quality control measurement schedules. Preparing for the audit: The CT scan acquired on the day is utilised hence no pre-planning is required. It is essential that a clinical cervix treatment applicator is available with central IUT and ovoids or ring, that a CT session has

205

been booked (single scan required) and access to brachytherapy planning system and treatment equipment. It is advisable to confirm ability to export and retrieve RTDose files from the TPS prior to the audit. 2. Review of Local Brachytherapy Process During the audit, the local brachytherapy physics practice will be informally discussed. The following information must be recorded:

CT scan reconstructed image width for clinical brachytherapy and audit.

Method of applicator reconstruction on CT, e.g. manual or library applicators, use of marker wire

Any offset of dwell positions for curvature of source wire.

QC/commissioning done to verify positions of actual dwells compared to library or manual placement of dwells

Dose and fractionation in clinical use for cervix treatment.

Prescription method used, e.g. Point A, D90 HR-CTV.

Exact local definition of Point A.

Prescribed dose to Point A for the audit plan (normally same as clinical practice)

Plan creation method: e.g. Standard library plan, optimisation methods used.

Whether nominal or actual activity is used at TPS, and method of correction for source activity between TPS and treatment unit.

Independent method used to verify dwell position and times, e.g. software, spreadsheet, check table.

Applicator used for the audit; specific type, and dimensions of ring/ovoids and IUT.

Print out of dwell positions, times and dose values from TPS and comparison to print out from treatment unit after delivery. Hard copies to be retained.

3. Audit Methodology 3.1 Equipment set-up Figure 1 shows the BRAD phantom with gynaecological cervix applicator in place, all positioned within the water tank (prior to filling). Importantly, the ring or ovoids must be in contact with the upper surface of the phantom, the applicator must be rotated such that the source tubes are parallel with a film plane indicated by scribed marks, and the clamp must be ‘finger tight’ to prevent any movement of the applicator during the audit procedure. Please note the BRAD phantom is fragile and is likely to be damaged if dropped.

Figure 1. BRAD phantom in water tank with applicator in position. Ring is in contact with upper surface of phantom, scribe marks used to align rotation of applicator with film plane, held in place by applicator clamp.

206

To ensure the intrauterine (IU) tube is aligned with the axis of the phantom, an appropriate sized collar must be selected and inserted into the central cavity. Collars for IU external diameters of 3, 4, 5 and 6 mm are available. Spacers may be used in the central channel prior to insertion of collars to adjust the height of the collar to securely hold the IU while not preventing ring/ovoids being in contact with upper surface of the phantom. The number of spacers and collars, shown in figure 2, should be chosen to best-fit the clinical applicator provided. For a 60 mm IU length, one 10 mm spacer should be inserted followed by the appropriate collar. For a 50 mm IU length, two 10 mm spacers should be inserted followed by the appropriate collar. For a 40 mm IU length, three 10 mm spacers should be inserted, followed by the appropriate collar. If a curved IUT or inclined ovoids are used, it is possible the ring or ovoids will not be parallel with the surface of the phantom. In this case, the distance offset at the level of the IUT must be measured (usually easier to measure the displacement and the maximum edge and take central position as half of this value). This is important to correctly interpret the position of Point A on the films. A metal rod is supplied to push out any collars or spaces at the end of the measurement (and may be required initially if collars/spacers have been left in place from a previous measurement), using the access hole under the phantom. (The spacers and collars are necessarily a snug fit in the channel and pliers may be required to push the metal rod to avoid bending).

Figure 2. Collars, 35 mm total length with 30 mm internal cavity of 3, 4, 5 or 6 mm diameter, used to hold IU on axis of phantom. Spacers each 10 mm length to adjust height of collar in central cavity to permit ring/ovoids in contact with upper surface of the phantom for different IU lengths.

3.2 CT scan Once the BRAD phantom is set-up with the test applicator securely located, the CT scan can be performed with the phantom on its side, as shown in figure 3. This orientation is used to image the applicator in a near clinical orientation. It is important to not adjust the applicator position or handle the phantom by the clamp after CT scanning until completion of the procedure. The usual local CT scan acquisition parameters, slice width etc, must be used to properly include assessment of clinical applicator reconstruction/alignment accuracy in the audit. (Film must not be inserted in the phantom during the CT scan). The acquired CT data set must then be exported to the treatment planning computer (TPS) via the usual clinical process.

207

Figure 3. BRAD phantom on side for CT scanning, with brachytherapy applicator in position approximating clinical use.

3.3 Treatment planning and DICOM RTDose export The CT images should be imported into the treatment planning system (TPS) and the treatment applicator reconstructed and aligned using normal clinical procedure. A typical cervix treatment plan should be created, using ‘standard plan’ or ‘usual loading’ or copied from a recent clinical example from the local centre, either prescribing to Point A or noting its dose from the TPS. (A dose of at least 7 Gy to Point A should be used for film dosimetry). Confirmation that the locally used definition of Point A is as defined by the 2012 ABS recommendations (Viswanathan and Thomadsen 2012) or the dose at this point determined. A print-out of intended dwell positions and times must be retained and the plan then sent to the HDR treatment unit using the usual clinical process. The treatment plan should be exported to the treatment unit using the normal clinical procedure, after performing any local independent check process. A 3D dose calculation must be performed to produce the required RTDose data. It is essential the dose grid is aligned orthogonally to the two film planes in the phantom to enable comparison between film-measured and TPS-calculated doses, when creating RTDose file. An inclined axis may have been used for curved IUT or inclined ovoids during treatment planning and this must be re-set normal to the phantom surface for RTDose calculation. The RTDose DICOM file must be obtained for the treatment plan and retained for later comparison of the planned dose distribution with the film-measured distribution. It is essential that the dose calculation grid is aligned orthogonal in both planes with respect to the film holders in the BRAD phantom. It may necessary to re-orient the axes for creation of the 3D dose grid if this alignment is not achieved in normal clinical practice (e.g. for curved IUT or inclined ovoids). The calculation grid must be centred on the applicator and be of dimensions around 18 x 18 x 18 cm, with 0.1 cm resolution. The resulting file size should be of the order 10 to 20 MB. Please see the supplementary information in Appendix A1 to export of RTDose files from various TPS.

The RTDose file should be transferred to an encrypted USB memory stick on the day of the audit to be taken away for analysis. If this is not possible, the file may be sent electronically via Secure File Transfer (SFT) on NHS mail or via a ‘large file transfer service’ to Tony Palmer.

208

3.4 HDR treatment unit delivery The treatment plan must be imported to the HDR or PDR treatment unit using normal clinical procedure and prepared for treatment. The water tank should be filled with room temperature tap water, a yellow plastic tub is supplied to transfer water into and out of the water tank. Gafchromic film has some sensitivity to ambient light and should be kept in the dark at all times other than for HDR treatment irradiation. One film must be placed on each holder in the BRAD phantom, with the long axis vertical. The films should be labelled in the lower corner away from the IU tube with their position (‘N’ under the applicator transfer tubes, and ‘E’, ‘S’ and ‘W’ moving clockwise looking down on the phantom). It is essential that each film is pushed inwards and upwards, in the BRAD phantom against the end stops (behind the holding plates) see figure 4. Extremely light pressure with the supplied screwdriver is required on the screws to hold the film in place – Take extra care not to over tighten and shear off the screws!

Figure 4. Films positioned on all four applicator vanes. Each film is pushed inwards and upwards against end-stops (see arrows) and held in place behind screwed plates.

The BRAD phantom with films should then be carefully lowered into the full water tank and the applicator connected to the treatment unit, as shown in figure 5. The treatment plan should then be delivered. Once complete, a print out of the delivered dwell position and times should be retained. After disconnecting the applicator transfer tubes the BRAD phantom should be removed from the water tank and the films removed and dried before storing in the dark.

Figure 5. Tank filled with tap water and BRAD phantom in use with brachytherapy treatment unit.

At the same date and approximate time as the HDR/PDR irradiation, two reference films from the same batch will be exposed to known MV radiation dose at Portsmouth. These films act as references for subsequent film dosimetry analysis.

209

3.5 Film analysis and results Films must be returned immediately, using light-tight packaging and appropriate protection to prevent bending, to Tony Palmer at Portsmouth for analysis. Postal details are provided in the appendix A4. An envelope is supplied for this purpose. Triple-channel analysis is used to generate film dose maps which are compared to the RTDose data using isodose overlay, shown in figure 6, and gamma analysis.

Figure 6. Comparison of TPS and film-measured dose distribution analysed with isodose overlay and gamma evaluation.

A report of the results will be provided approximately one week following the audit visit. Appendices A1. Supplementary Information for the Export of RTDose file from Treatment Planning Systems Export of DICOM RTDose from brachytherapy treatment planning systems is not normally performed in routine clinical practice, hence supplementary information is provided here which may be useful. It is essential that the dose calculation grid is aligned orthogonally in both planes with respect to the film holders in the BRAD phantom. It may necessary to re-orient the axes for creation of the 3D dose grid if this alignment is not achieved in normal clinical practice (e.g. for curved IUT or inclined ovoids). The calculation grid must be centred on the applicator and ideally be of dimensions around 18 x 18 x 18 cm, with 0.1 cm resolution. It may be necessary to change from default settings to achieve this. The resulting file size should be of the order 10 to 20 MB. Varian BrachyVision. With the patient plan open, go to File>Export>Wizard, and follow the prompts selecting an appropriate destination. Nucletron Oncentra Brachy / Masterplan. With the patient plan open, go to Plan>Calculate 3D Dose Grid, and then using the CM Export window, expand the study and select the RTDOSE file in the left pane and highlight ‘1’ in the right Dicom Objects pane. With the file selected, click Export button and select an appropriate destination. If a suitable destination peer has not been set-up for DICOM export, it should be possible to use the Multicase export function to save the file to desktop. The resulting RTDose file may need a Dicom preamble adding (e.g. via PowerDicom) for use in FilmQAPro. Nucletron Oncentra GYN. There is less flexibility in calculating and exporting RTDose from Oncentra GYN, and it may not be possible to undertake full audit analysis. Please contact Tony Palmer to discuss alternatives if the following procedure is unsuccessful.

210

The dose grid resolution cannot be independently set and will depend on the margin set around the active source dwell positions. A margin of 5 cm will result in a grid resolution around 0.2 cm. It is preferable to calculate and export RTDose at several resolutions for analysis, e.g. margins of 7, 5, 3 cm. Create a folder on the desktop and set this as the DICOM destination: Use Nucletron Smoothbase, configure Dicom, add destination and name it. Select Dicom file and navigate to the created folder. To amend the grid: Dose settings/dose grid position/enter margin size, click apply ASDP (active source dwell position – must be done after creating the plan). Save the dose after saving the plan. To export the RTDose file: Use Nucletron Smoothbase, export to Dicom, complete study, select patient and study, click load, select the Dicom destination named above. It is possible to have Smoothbase and Oncentra GYN open at the same time to facilitate ease of exporting several files with different grid sizes. Eckert & Ziegler Bebig HDRPlus. With the patient plan open, select Patient>Export>Dicom and select an appropriate destination using the windows explorer style interface. A2. Equipment Checklist Please use the checklist below to ensure all equipment is replaced into the travel case at the end of the audit.

BRAD solid water phantom with applicator support arm

White lid tub containing 3 x 10 mm length spacers and 4 x 35 mm length collars (3, 4, 5, 6 mm internal diameter cavity)

Red lid tube containing 2 x metal rods to remove spacers and collars

Clear Perspex water tank

Yellow plastic box for transfer of water into and out of water tank (not shown in photo).

Screwdriver (green/black handle)

Pliers (black handle, not shown in photo)

Figure 7. Transport case showing properly loaded equipment. (Yellow box for ‘water transfer’ should be placed inside the water tank).

A3. Uncertainty Analysis and Typical Results with Suggested Tolerance The combined standard uncertainty in the determination of dose difference at a point between film-measured dose and TPS dose is 3.0% (k=1), and the standard uncertainty in distance to agreement is estimated as 0.5 mm (k=1). Both of these values apply to the dose gradient around Point A, which is approximately 6% mm-1. Multiple film-dosimetry measurements have demonstrated repeatability of film dose distributions assessed by gamma comparison exceeding 95% passing rate at 1% (local), 1.5 mm criteria. Gamma criteria of 5% local normalisation and 3 mm distance to agreement and 3% 2 mm, are used in comparing film-measured and TPS calculated dose distributions. Pilot audits have achieved passing rates exceeding 95% at 5%, 3 mm criteria, and an acceptable limit of 90% is therefore suggested, although local physics must make a final acceptability decision based on local factors. The mean distance between the prescription isodose and the position of Point A in pilot audits was 1.2 mm, and an acceptable limit of 3.0 mm is suggested. The local physicist must interpret results and decide acceptability based on local brachytherapy clinical practice.

211

E. BRAD Phantom Availability in IPEM Virtual Equipment Library

(Internet Site)

Figure E.1 is a screenshot of the IPEM webpage advertising the BRAD phantom, developed

through this project.

Figure E.1. IPEM webpage advertising BRAD phantom in Virtual Phantom Library (only available to

IPEM members hence internet address not given) [cited 28/03/2014].

212

F. MCNP5 Input File for Validation Test

The following two input files were used to validate the implementation of

MCNP5 software code for the brachytherapy applications presented in this research thesis,

as described in Section 7.2.2. The first file is for the Nucletron mHDR-v2 Ir-192 source and

the second for the Eckert & Ziegler BEBIG MultiSource Co0.A86 Co-60 HDR source.

Validation MCNP5 implementation Ir192 dose rate with radial distance c c Input file by Tony Palmer, 2013 c c ----------------------- c cell definitions c ----------------------- c c Source and capsule c 101 3 -22.42 -1001 1004 -1005 imp:p,e 1 $ Source 102 2 -8.000 1001 -1002 1004 -1005 imp:p,e 1 $ Capsule cylinder shell 103 2 -8.000 -1002 1003 -1004 imp:p,e 1 $ Capsule lower end cap 104 2 -8.000 -1002 1005 -1006 imp:p,e 1 $ Capsule upper end cap c c Boundaries of problem c 105 1 -0.998 (-1009 1006 -1007):(-1003 1008 -1007): (-1006 1003 1002 -1007) #107 #108 #109 #110 #111 #112 #113 #114 #115 imp:p,e 1 $ Fill with water 106 0 1007:-1008:1009 imp:p,e 0 $ Outside known universe c c Tally cells for radial dose c 107 1 -0.998 1010 -1011 1012 -1013 imp:p,e 1 $ tally r=0.75cm 108 1 -0.998 1010 -1011 1014 -1015 imp:p,e 1 $ tally r=1.0cm 109 1 -0.998 1010 -1011 1016 -1017 imp:p,e 1 $ tally r=1.5cm 110 1 -0.998 1010 -1011 1018 -1019 imp:p,e 1 $ tally r=2.0cm 111 1 -0.998 1010 -1011 1020 -1021 imp:p,e 1 $ tally r=3.0cm 112 1 -0.998 1010 -1011 1022 -1023 imp:p,e 1 $ tally r=4.0cm 113 1 -0.998 1010 -1011 1024 -1025 imp:p,e 1 $ tally r=5.0cm 114 1 -0.998 1010 -1011 1026 -1027 imp:p,e 1 $ tally r=6.0cm 115 1 -0.998 1010 -1011 1028 -1029 imp:p,e 1 $ tally r=7.0cm c ----------------------- c surface cards c ----------------------- c c Nucletron HDR v2 simplified Ir-192 source in steel capsule c 1001 cx 0.0325 $ Outside radius source, inside radius steel capsule 1002 cx 0.0425 $ Outside radius steel capsule 1003 px -0.02 $ Lower edge of lower end cap of capsule 1004 px 0.001 $ Upper edge of lower end cap of capsule, lower edge source 1005 px 0.36 $ Lower edge of upper end cap of capsule, upper edge source 1006 px 0.38 $ Upper edge of upper end cap of capsule c c Boundaries of problem c 1007 cx 60 $ Radius of edge of known universe 1008 px -60 $ Lower edge of known universe 1009 px 60 $ Upper edge of known universe c c Tally cell surfaces c 1010 px 0.155 $ Lower plane of tally cells 1011 px 0.205 $ Upper plane of tally cells 1012 cx 0.725 $ Radii of cylinders for tally cells 1013 cx 0.775 1014 cx 0.975 1015 cx 1.025 1016 cx 1.475 1017 cx 1.525 1018 cx 1.975 1019 cx 2.025 1020 cx 2.975 1021 cx 3.025 1022 cx 3.975 1023 cx 4.025 1024 cx 4.975 1025 cx 5.025 1026 cx 5.975 1027 cx 6.025 1028 cx 6.975 1029 cx 7.025

213

c ----------------------- c Data cards c ----------------------- c Material cards c ----------------------- c Pacific Northwest National Laboratory, Compendium Material Comp, March 2011 c c Water, liquid c density = 0.998 g/cm3 m1 1000 0.666657 $ H (element and atomic fraction) 8000 0.333343 $ O c c Stainless steel (316L) c density = 8.000 g/cm3 m2 6000 0.001384 $ C 14000 0.019722 $ Si 15000 0.000805 $ P 16000 0.000518 $ S 24000 0.181098 $ Cr 25000 0.020165 $ Mn 26000 0.648628 $ Fe 28000 0.113247 $ Ni 42000 0.014434 $ Mo c c Iridium c density = 22.42 g/cm3 m3 77000 1.0000 $ Ir c c c ----------------------- c Source card c ----------------------- c Isotropic cylindrical Ir-192 volume source c sdef POS=0 0 0 axs=1 0 0 ext=d1 rad=d2 erg=d3 par=2 $ Ir-192 source c si1 0.001 0.36 $ axial source extent, sampling range sp1 -21 0 $ density constant with x c si2 0 0.0325 $ radial source extent, sampling range sp2 -21 1 $ density proportional to radius c c Emission data from www.nndc.bnl.gov Nuclear Data Sheet 113, 1871 (2012) c not included very low emission prob (<0.03%) or very low energies (<10keV) c Discrete Ir-192 energies in MeV and probability of emission si3 L 0.0615 0.0630 0.0633 0.0650 0.0651 0.0668 0.0711 0.0714 0.0714 0.736 0.0754 0.0757 0.0778 0.1364 0.2013 0.2058 0.2833 0.2960 0.3085 0.3165 0.3745 0.4165 0.4205 0.4681 0.4846 0.4891 0.5886 0.6044 0.6125 0.8845 1.0615 sp3 1.190 2.020 0.176 0.300 2.620 4.440 0.238 0.460 0.161 0.681 0.531 1.021 0.364 0.199 0.471 3.310 0.266 28.712 29.700 82.869 0.727 0.670 0.069 47.840 3.190 0.438 4.522 8.216 5.340 0.292 0.053 c c c ----------------------- c Tally cards c ----------------------- c f6:p 107 108 109 110 111 112 113 114 115 fm6 1.6e-10 $ Convert MeV/g to Gy per source particle fc6 f6 Energy deposition (Gy) per source particle c c *f8:p,e 107 108 109 110 111 112 113 114 115 fc8 *f8 Energy deposition (MeV) per particle history c c c ----------------------- c Other data c ----------------------- mode p e $ Transport photons and electrons c $ Phys card defaults detailed physics cut:p j 0.01 $ Cut photon transport at 10keV nps 1000000000 $ Number of particle histories cut card print 110 $ Print first 50 histories End of File

214

Validation MCNP5 implementation Co60 dose rate with radial distance c c Input file by Tony Palmer, 2013 c c ----------------------- c cell definitions c ----------------------- c c Source and capsule c 101 3 -8.900 -1001 1004 -1005 imp:p,e 1 $ Source 102 2 -8.000 1001 -1002 1004 -1005 imp:p,e 1 $ Capsule cylinder shell 103 2 -8.000 -1002 1003 -1004 imp:p,e 1 $ Capsule lower end cap 104 2 -8.000 -1002 1005 -1006 imp:p,e 1 $ Capsule upper end cap c c Boundaries of problem c 105 1 -0.998 (-1009 1006 -1007):(-1003 1008 -1007): (-1006 1003 1002 -1007) #107 #108 #109 #110 #111 #112 #113 #114 #115 imp:p,e 1 $ Fill with water 106 0 1007:-1008:1009 imp:p,e 0 $ Outside known universe c c Tally cells for radial dose c 107 1 -0.998 1010 -1011 1012 -1013 imp:p,e 1 $ tally r=0.75cm 108 1 -0.998 1010 -1011 1014 -1015 imp:p,e 1 $ tally r=1.0cm 109 1 -0.998 1010 -1011 1016 -1017 imp:p,e 1 $ tally r=1.5cm 110 1 -0.998 1010 -1011 1018 -1019 imp:p,e 1 $ tally r=2.0cm 111 1 -0.998 1010 -1011 1020 -1021 imp:p,e 1 $ tally r=3.0cm 112 1 -0.998 1010 -1011 1022 -1023 imp:p,e 1 $ tally r=4.0cm 113 1 -0.998 1010 -1011 1024 -1025 imp:p,e 1 $ tally r=5.0cm 114 1 -0.998 1010 -1011 1026 -1027 imp:p,e 1 $ tally r=6.0cm 115 1 -0.998 1010 -1011 1028 -1029 imp:p,e 1 $ tally r=7.0cm c ----------------------- c surface cards c ----------------------- c c EZ BEBIG HDR simplified Co-60 Co0.A86 source in steel capsule c 1001 cx 0.030 $ Outside radius source, inside radius steel capsule 1002 cx 0.045 $ Outside radius steel capsule 1003 px -0.055 $ Lower edge of lower end cap of capsule 1004 px 0.001 $ Upper edge of lower end cap of capsule, lower edge source 1005 px 0.35 $ Lower edge of upper end cap of capsule, upper edge source 1006 px 0.405 $ Upper edge of upper end cap of capsule c c Boundaries of problem c 1007 cx 60 $ Radius of edge of known universe 1008 px -60 $ Lower edge of known universe 1009 px 60 $ Upper edge of known universe c c Tally cell surfaces c 1010 px 0.155 $ Lower plane of tally cells 1011 px 0.205 $ Upper plane of tally cells 1012 cx 0.725 $ Radii of cylinders for tally cells 1013 cx 0.775 1014 cx 0.975 1015 cx 1.025 1016 cx 1.475 1017 cx 1.525 1018 cx 1.975 1019 cx 2.025 1020 cx 2.975 1021 cx 3.025 1022 cx 3.975 1023 cx 4.025 1024 cx 4.975 1025 cx 5.025 1026 cx 5.975 1027 cx 6.025 1028 cx 6.975 1029 cx 7.025 c ----------------------- c Data cards c ----------------------- c Material cards c ----------------------- c Pacific Northwest National Laboratory, Compendium Material Comp, March 2011 c c Water, liquid c density = 0.998 g/cm3 m1 1000 0.666657 $ H (element and atomic fraction) 8000 0.333343 $ O

215

c c Stainless steel (316L) c density = 8.000 g/cm3 m2 6000 0.001384 $ C 14000 0.019722 $ Si 15000 0.000805 $ P 16000 0.000518 $ S 24000 0.181098 $ Cr 25000 0.020165 $ Mn 26000 0.648628 $ Fe 28000 0.113247 $ Ni 42000 0.014434 $ Mo c c Cobalt c density = 8.900 g/cm3 m3 27000 1.0000 $ Co c c c ----------------------- c Source card c ----------------------- c Isotropic cylindrical Co-60 volume source c sdef POS=0 0 0 axs=1 0 0 ext=d1 rad=d2 erg=d3 par=2 $ Co-60 source c si1 0.001 0.35 $ axial source extent, sampling range sp1 -21 0 $ density constant with x c si2 0 0.03 $ radial source extent, sampling range sp2 -21 1 $ density proportional to radius c c Emission data from www.nndc.bnl.gov Nuclear Data Sheet 113, 1871 (2012) c not included very low emission prob (<0.03%) or very low energies (<10keV) si3 L 1.1732 1.3325 $ Discrete Co-60 energies in MeV sp3 99.87 100.00 $ Probability (normalised) c c c ----------------------- c Tally cards c ----------------------- c f6:p 107 108 109 110 111 112 113 114 115 fm6 1.6e-10 $ Convert MeV/g to Gy per source particle fc6 f6 Energy deposition (Gy) per source particle c c *f8:p,e 107 108 109 110 111 112 113 114 115 fc8 *f8 Energy deposition (MeV) per particle history c c c ----------------------- c Other data c ----------------------- mode p e $ Transport photons and electrons c $ Phys card defaults detailed physics cut:p j 0.01 $ Cut photon transport at 10keV nps 1000000000 $ Number of particle histories cut card print 110 $ Print first 50 histories End of File

216

G. MCNP5 Input File for Phantom Scatter Evaluation

The following input file was used to determine scatter correction factors for

phantoms of various size, for use in brachytherapy dosimetric audit, as described in Section

7.2.3. The file was appropriately modified to independently vary the water phantom and tally

cell radii. The input file included here is for an Ir-192 point source. The source definition was

amended for calculations with Co-60.

Effect of phantom size scatter on point Ir192 source c c Input file by Tony Palmer, 2013 c c ----------------------- c cell definitions c ----------------------- c 100 1 -0.998207 -1001 imp:p,e 1 $ Water sphere (material density surfaces) 101 1 -0.998207 1001 -1002 imp:p,e 1 $ Spherical shell for tally 102 1 -0.998207 1002 -1003 imp:p,e 1 $ Rest of water sphere 103 2 -0.001205 1003 -1004 imp:p,e 1 $ Air 104 0 1004 imp:p,e 0 $ Outside known universe c ----------------------- c surface cards c ----------------------- c 1001 sph 0 0 0 5.9 $ Sphere at origin, rad of inner edge sphere tally 1002 sph 0 0 0 6.1 $ Sphere at origin, rad of outer edge sphere tally 1003 sph 0 0 0 30 $ Sphere at origin, rad to edge of water phantom 1004 sph 0 0 0 80 $ Sphere at origin, rad to outside world c ----------------------- c Data cards c ----------------------- c Material cards c ----------------------- c Pacific Northwest National Laboratory, Compendium Material Comp, March 2011 c c Water, liquid c density = 0.998 g/cm3 m1 1000 0.666657 $ H (element and atomic fraction) 8000 0.333343 $ O c Air, dry, near sea level m2 6000 0.000150 $ C 7000 0.784431 $ N 8000 0.210748 $ O 18000 0.004671 $ Ar c c ----------------------- c Source card c ----------------------- c Point isotropic Ir-192 source at origin sdef POS 0 0 0 erg=d1 par=2 $ Ir-192 source c Emission data from www.nndc.bnl.gov Nuclear Data Sheet 113, 1871 (2012) c not included very low emission prob (<0.03%) or very low energies (<10keV) c Discrete Ir-192 energies in MeV and probability of emission si1 L 0.0615 0.0630 0.0633 0.0650 0.0651 0.0668 0.0711 0.0714 0.0714 0.736 0.0754 0.0757 0.0778 0.1364 0.2013 0.2058 0.2833 0.2960 0.3085 0.3165 0.3745 0.4165 0.4205 0.4681 0.4846 0.4891 0.5886 0.6044 0.6125 0.8845 1.0615 sp1 1.190 2.020 0.176 0.300 2.620 4.440 0.238 0.460 0.161 0.681 0.531 1.021 0.364 0.199 0.471 3.310 0.266 28.712 29.700 82.869 0.727 0.670 0.069 47.840 3.190 0.438 4.522 8.216 5.340 0.292 0.053 c c ----------------------- c Tally cards c ----------------------- c f6:p 101 $ in MeV/g per particle history fm6 1.6e-10 $ Convert MeV/g to Gy fc6 Energy deposition (Gy) per source particle $ Comment string for tally c c ----------------------- c Other data c ----------------------- mode p $ Transport photons 1000000000 $ Number of particle histories cut card print 110 $ Print first 50 histories End of File

217

H. MCNP5 Input File for the Effect of the Presence of EBT3 Film

on Measured Dose

The following input file was used to estimate the perturbation in delivered dose

due to the presence of EBT3 film from a uniform water medium case, for the geometries

involved in the BRAD phantom and dose to prescription Point A, as described in Section 7.2.4.

The input file below was modified to replace the film with water to calculate the undisturbed

case.

Effect of film side on to Co60 10mm source using 15mm film ring c c Input file by Tony Palmer, 2013 c c Simulation of source dwells in ring applicator c Source train 10mm length, 10mm from film edge, c Use film disc around source geometry to improve tally efficiency c Radiation traverse 15mm film to Point A c c ----------------------- c cell definitions c ----------------------- c c Source and capsule c 101 3 -8.900 -1001 1004 -1005 imp:p,e 1 $ Source 102 2 -8.000 1001 -1002 1004 -1005 imp:p,e 1 $ Capsule cylinder shell 103 2 -8.000 -1002 1003 -1004 imp:p,e 1 $ Capsule lower end cap 104 2 -8.000 -1002 1005 -1006 imp:p,e 1 $ Capsule upper end cap c c Boundaries of problem c 105 1 -0.998 (-1009 1006 -1007):(-1003 1008 -1007): (-1006 1003 1002 -1007) #107 #108 #109 #110 #111 #112 imp:p,e 1 $ Fill with water 106 0 1007:-1008:1009 imp:p,e 0 $ Outside known universe c c Tally cells c 107 1 -0.998 1011 -1007 1012 -1013 imp:p,e 1 $ tally lower layer 108 1 -0.998 1011 -1007 1013 -1014 imp:p,e 1 $ tally central layer 109 1 -0.998 1011 -1007 1014 -1015 imp:p,e 1 $ tally upper layer c c Film c 110 4 -1.35 1010 -1011 1012 -1013 imp:p,e 1 $ film lower layer 111 5 -1.20 1010 -1011 1013 -1014 imp:p,e 1 $ film central layer 112 4 -1.35 1010 -1011 1014 -1015 imp:p,e 1 $ film upper layer c ----------------------- c surface cards c ----------------------- c c EZ Bebig Co0.A86 simplified Co-60 1cm length source in steel capsule c 1001 cx 0.030 $ Outside radius source, inside radius steel capsule 1002 cx 0.045 $ Outside radius steel capsule 1003 px -0.02 $ Lower edge of lower end cap of capsule 1004 px 0.001 $ Upper edge of lower end cap of capsule, lower edge source 1005 px 1 $ Lower edge of upper end cap of capsule, upper edge source 1006 px 1.02 $ Upper edge of upper end cap of capsule c c Boundaries of problem c 1007 cx 5 $ Radius of edge of known universe 1008 px -5 $ Lower edge of known universe 1009 px 5 $ Upper edge of known universe c c Tally and film cell surfaces c 1010 cx 1 1011 cx 2.5 $ 1.5cm width ring 1012 px 0.4861 1013 px 0.4986 1014 px 0.5014 1015 px 0.5139 c ----------------------- c Data cards

218

c ----------------------- c Material cards c ----------------------- c Pacific Northwest National Laboratory, Compendium Material Comp, March 2011 c c Water, liquid c density = 0.998 g/cm3 m1 1000 0.666657 $ H (element and atomic fraction) 8000 0.333343 $ O c c Stainless steel (316L) c density = 8.000 g/cm3 m2 6000 0.001384 $ C 14000 0.019722 $ Si 15000 0.000805 $ P 16000 0.000518 $ S 24000 0.181098 $ Cr 25000 0.020165 $ Mn 26000 0.648628 $ Fe 28000 0.113247 $ Ni 42000 0.014434 $ Mo c c Cobalt c density = 8.900 g/cm3 m3 27000 1.0000 $ Co c c c EBT3 matte polyester film base c density = 1.35 g/cm3 m4 1000 0.36364 $ H 6000 0.45455 $ C 8000 0.18182 $ O c c EBT3 active film layer c density = 1.20 g/cm3 m5 1000 0.56800 $ H 3000 0.00600 $ Li 6000 0.27600 $ C 8000 0.13300 $ O 13000 0.01600 $ Al c c ----------------------- c Source card c ----------------------- c Isotropic cylindrical Co-60 volume source c sdef POS=0 0 0 axs=1 0 0 ext=d1 rad=d2 erg=d3 par=2 $ Co-60 source c simulating length of four source dwells in ring or ovoid applicator c si1 0.001 1 $ axial source extent, sampling range sp1 -21 0 $ density constant with x c si2 0 0.0325 $ radial source extent, sampling range sp2 -21 1 $ density proportional to radius c c Emission data from www.nndc.bnl.gov Nuclear Data Sheet 113, 1871 (2012) c not included very low emission prob (<0.03%) or very low energies (<10keV) si3 L 1.1732 1.3325 $ Discrete Co-60 energies in MeV sp3 99.87 100.00 $ Probability (normalised) c c c ----------------------- c Tally cards c ----------------------- c f6:p 107 108 109 T fm6 1.6e-10 $ Convert MeV/g to Gy per source particle fc6 f6 Energy deposition (Gy) per source particle c c *f8:p,e 107 108 109 T fc8 *f8 Energy deposition (MeV) per particle history c c c ----------------------- c Other data c ----------------------- mode p e $ Transport photons and electrons c $ Phys card defaults detailed physics cut:p j 0.01 $ Cut photon transport at 10keV nps 1000000000 $ Number of particle histories cut card ctme 0.1 $ 2880 $ Time cut card print 110 $ Print first 50 histories End of File

219

I. MCNP5 Input File for the Evaluation of Treatment Applicator

Attenuation

The following input file was used to calculate the dose attenuation due to the

presence of metallic treatment applicators, as described in Section 7.2.5. The file below is for

an Ir-192 source with a Varian titanium cervix applicator. The input file was modified for other

manufactures’ titanium and steel applicator dimensions with both Ir-192 and Co-60 sources.

Attenuation Varian Titanium Applicator, long 192Ir source c c Input file by Tony Palmer, 2014 c c ----------------------- c cell definitions c ----------------------- c 101 4 -22.42 -1001 1014 -1017 imp:p,e 1 $ Source 102 2 -7.860 1001 -1002 1014 -1017 imp:p,e 1 $ Capsule cylinder shell 103 2 -7.860 -1002 1013 -1014 imp:p,e 1 $ Capsule lower end cap 104 2 -7.860 -1002 1017 -1018 imp:p,e 1 $ Capsule upper end cap 105 3 -4.510 1003 -1004 1012 -1021 imp:p,e 1 $ Applicator cylinder shell 106 1 -0.998 1005 -1006 1015 -1016 imp:p,e 1 $ Tally cell, mid source, inner 107 1 -0.998 1007 -1008 1015 -1016 imp:p,e 1 $ Tally cell, mid source, middle 108 1 -0.998 1009 -1010 1015 -1016 imp:p,e 1 $ Tally cell, mid source, outer 109 1 -0.998 1005 -1006 1019 -1020 imp:p,e 1 $ Tally cell, above source, inner 110 1 -0.998 1007 -1008 1019 -1020 imp:p,e 1 $ Tally cell, above source, middle 111 1 -0.998 1009 -1010 1019 -1020 imp:p,e 1 $ Tally cell, above source, outer 112 1 -0.998 -1003 1012 -1013 imp:p,e 1 $ Fill inside applicator, below 113 1 -0.998 -1003 1018 -1021 imp:p,e 1 $ Fill inside applicator, above 114 1 -0.998 1002 -1003 1013 -1018 imp:p,e 1 $ Fill from capsule to applicator 115 1 -0.998 1004 -1011 1012 -1021 #106 #107 #108 #109 #110 #111 #117 #118 #119 imp:p,e 1 $ Fill water phantom 116 0 1011:-1012:1021 imp:p,e 0 $ Outside known universe 117 1 -0.998 1005 -1006 1022 -1023 imp:p,e 1 $ Tally cell, end source, inner 118 1 -0.998 1007 -1008 1022 -1023 imp:p,e 1 $ Tally cell, end source, middle 119 1 -0.998 1009 -1010 1022 -1023 imp:p,e 1 $ Tally cell, end source, outer c ----------------------- c surface cards c ----------------------- c 1001 cx 0.0325 $ Outside radius source, inside radius steel capsule 1002 cx 0.0425 $ Outside radius steel capsule 1003 cx 0.1 $ Inside radius metal applicator 1004 cx 0.15 $ Outside radius metal applicator 1005 cx 1.9 $ Inside radius first set tally cells 1006 cx 2.1 $ Outside radius first set tally cells 1007 cx 3.9 $ Inside radius second set tally cells 1008 cx 4.1 $ Outside radius second set tally cells 1009 cx 5.9 $ Inside radius third set tally cells 1010 cx 6.1 $ Outside radius third set tally cells 1011 cx 8 $ Radius of edge of known universe 1012 px -2 $ Lower edge of known universe 1013 px -0.1 $ Lower edge of lower end cap of capsule 1014 px 0.001 $ Upper edge of lower end cap of capsule, lower edge source 1015 px 2.5 $ Lower edge mid source tally cells 1016 px 3.5 $ Upper edge mid source tally cells 1017 px 6 $ Lower edge of upper end cap of capsule, upper edge source 1018 px 6.1 $ Upper edge of upper end cap of capsule 1019 px 8.5 $ Lower edge above source tally cells 1020 px 9.5 $ Upper edge above source tally cells 1021 px 11 $ Upper edge of known universe 1022 px 4.5 $ Lower edge end source tally cells 1023 px 5.5 $ Lower edge end source tally cells c ----------------------- c Data cards c ----------------------- c Material cards c ----------------------- c Pacific Northwest National Laboratory, Compendium Material Comp, March 2011 c c Water, liquid c density = 0.998 g/cm3 m1 1000 0.666657 $ H (element and atomic fraction) 8000 0.333343 $ O c c Stainless steel (202)

220

c density = 7.860 g/cm3 m2 6000 0.003405 $ C 7000 0.004866 $ N 14000 0.009708 $ Si 15000 0.000528 $ P 16000 0.000255 $ S 24000 0.188773 $ Cr 25000 0.086851 $ Mn 26000 0.659160 $ Fe 28000 0.046454 $ Ni c c Titanium c Material data from Elekta for Cervix Rotterdam Applicator c density = 4.510 g/cm3 m3 22000 1.0000 $ Ti c c Iridium c density = 22.42 g/cm3 m4 77000 1.0000 $ Ir c c ----------------------- c Source card c ----------------------- c Elongated source (6 cm) to simulate series of dwells in a straight channel c Isotropic cylindrical volume source c sdef POS=0 0 0 axs=1 0 0 ext=d1 rad=d2 erg=d3 par=2 $ Ir-192 source c si1 0 6 $ axial source extent, sampling range sp1 -21 0 $ density constant with x c si2 0 0.035 $ radial source extent, sampling range sp2 -21 1 $ density proportional to radius c c Emission data from www.nndc.bnl.gov Nuclear Data Sheet 113, 1871 (2012) c not included very low emission prob (<0.03%) or very low energies (<10keV) c Discrete Ir-192 energies in MeV and probability of emission si3 L 0.0615 0.0630 0.0633 0.0650 0.0651 0.0668 0.0711 0.0714 0.0714 0.736 0.0754 0.0757 0.0778 0.1364 0.2013 0.2058 0.2833 0.2960 0.3085 0.3165 0.3745 0.4165 0.4205 0.4681 0.4846 0.4891 0.5886 0.6044 0.6125 0.8845 1.0615 sp3 1.190 2.020 0.176 0.300 2.620 4.440 0.238 0.460 0.161 0.681 0.531 1.021 0.364 0.199 0.471 3.310 0.266 28.712 29.700 82.869 0.727 0.670 0.069 47.840 3.190 0.438 4.522 8.216 5.340 0.292 0.053 c c ----------------------- c Tally cards c ----------------------- c f6:p 106 107 108 $ in MeV/g per particle history fm6 1.6e-10 $ Convert MeV/g to Gy fc6 Energy deposition (Gy) per source particle, tallies at mid source c f16:p 117 118 119 fm16 1.6e-10 fc16 Energy deposition (Gy) per source particle, tallies end of source c f26:p 109 110 111 fm26 1.6e-10 fc26 Energy deposition (Gy) per source particle, tallies above source c c ----------------------- c Other data c ----------------------- mode p $ Transport photons c $ Phys card defaults detailed physics cut:p j 0.01 $ Cut photon transport at 10keV nps 1000000000 $ Number of particle histories cut card ctme 2880 $ Time cut card print 110 $ Print first 50 histories End of File

221

J. Questionnaire Used for the Survey of HDR/PDR Quality

Control Practice in the UK

The following questionnaire was emailed to all radiotherapy centres in the UK to survey HDR/PDR QC practice, see Section 3.2.1.

Survey of Quality Control Practices for HDR and PDR Brachytherapy in UK Please detail in the table below the HDR/PDR brachytherapy QC tests performed at your centre, their frequency and tolerance, and additional supporting information. By completing this survey you are providing consent for all of the data to be retained indefinitely and to be collated for the purposes of publication, however all data will be annonomised prior to any dissemination. Name and centre information is held purely for contact purposes. If you do not have either HDR or PDR equipment, can you please email to state this, and also whether you intend to purchase HDR/PDR in the future and anticipated timescale. Please return the completed document to [email protected]. Thank you, Tony Palmer. Name of person completing form: Email address: Date of completion of survey: Radiotherapy centre name and city: HDR or PDR: Equipment make and model: Source change frequency: Isotope: If applicable, 2nd HDR or PDR unit in Dept: Equipment make and model: Source change frequency: Isotope: Number of HDR and PDR fractions delivered per year: Session(s) (time of day) in which QC is generally performed: Use of measured or certificate source stength for treatment planning: Source strength first measurement method: Source strength second independent method: Origin of TG-43 source model data (manufacturer, journal or consensus data set): Inependent check method for treatment planning system generated plans: Imaging method for each clinical site/treatment technique: Prescription to point A or HR-CTV volume for cervix: Do you record Point A dose if prescribing to volume: Do you record ICRU OAR point doses or GEC-ESTRO DVH parameters or both? Do you perform treatment plan dwell optimisation, for which sites? Optimisation method for each clinical site/treatment technique: Please list the primary sources of guidance used in establishing your QC schedule: When was the content of your QC schedule last reviewed? Any further comments or additional information:

222

A

spec

t

QC

tes

t

Incl

ud

ed in

sch

ed

ule

d r

ou

tin

e Q

C

test

s at

yo

ur

cen

tre?

Met

ho

d u

sed

?

Per

form

ed b

y?

Targ

et m

easu

rem

ent

freq

uen

cy?

Typ

ical

ly a

chie

ved

mea

sure

men

t

freq

uen

cy?

Tole

ran

ce li

mit

(fo

r cl

inic

al u

se)?

Co

mm

ents

Source strength Initial measurement on receipt of source

Independent measurement on receipt

Repeated measurements during life of source

Leak testing of source

Treatment unit function Confirmation of accuracy of source data at treatment unit (calibration date and time, isotope)

Confirmation of decay correction accuracy performed by treatment unit for treatment plans

Plan data transfer check from TPS

Simulated treatment functionality test

Verification of machine timer accuracy

Operating system display and print-out as expected and in agreement

Test of function with mains power loss

Test of Uninterruptable Power Supply (UPS)

Applicators, source position and source movement

Visual inspection of applicators and transfer tubes for damage

Measurement of dimensions and angles of applicators and transfer tubes

X-ray imaging of applicators

Measurement verification of source dwell positions in straight catheter (not clinical applicator)

Measurement verification of source dwell positions within clinical applicators

Measurement verification of actual source dwell positions against TPS planned positions in complex geometry e.g. within applicator ring

Source position relative to dummy source or guide wires

Applicator/transfer tube connection interlock and simulated error

Verification of expected position of internal applicator shielding

Measurement of source transit times

Confirm that error code 'meanings and actions' are available at the treatment unit

Radiation monitor of applicators after use

Treatment planning system (TPS)

Check accuracy/consistency of source model data used by TPS i.e. TG-43 data, against reference values

Check accuracy of specific source data used by TPS e.g. source strength at calibration date, and decay corrected strength, against reference values

Calculation of standard plans compared to reference data

Calculation of standard plans using independent check calculation system compared to reference data and to TPS

223

Repeat of tests performed at commissioning, e.g. DVH accuracy, geometric tests, etc

Imaging 2D kV imaging: Test of TPS reconstruction algorithm using scan of physical reconstruction phantom and/or virtual image phantom

2D kV imaging: Please list any other brachytherapy-specific QC tests performed

CT imaging: Test of TPS reconstruction algorithm using scan of physical reconstruction phantom and/or virtual image phantom

CT imaging: Please list any other brachytherapy-specific QC tests performed

MR imaging: Test of TPS reconstruction algorithm using scan of physical reconstruction phantom and/or virtual image phantom

MR imaging: Please list any other brachytherapy-specific QC tests performed

Ultrasound imaging: Test of TPS reconstruction algorithm using scan of physical reconstruction phantom and/or virtual image phantom

Ultrasound imaging: Please list any other brachytherapy-specific QC tests performed

Image data transfer accuracy to TPS for all imaging modalities used

In-vivo dosimetry Type of in-vivo measurement system, if available

Calibration of in-vivo dosimetry system

Test of in-vivo measurement system against expected or planned values for typical/simulated treatment measured in a phantom

Facilities Visual (CCTV) and audible (intercom) patient monitoring system

Radiation warning lights

Independent radiation monitor

Interlocks (door, timer delay, etc)

Emergency stop

Contingencies Practice of simulated emergency procedures, e.g. source stick

Presence of emergency source containers, forceps, etc

Review of responsibilities: who removes applicators in source stick situation

Other Please list any additional tests that do not appear above

224

225

226