Point-of Care Technology: Perspective from South

Africa

Prof W. StevensWits University & NHLS

Content

� General Considerations

� South African perspective

� Three feasibility studies• CD4 PIMA

– Validation steps– Sampling issues– Impact on clinical outcome

• GeneXpert TB• Future Viral load

Definition

� Point of Care testing (POCT): refers to testing that is performed near or at the site of the patient with the result leading to a possible change in the care of the patient.

� Outpatient clinic, ER, theatre, mobile clinics, PHC clinics, or even small laboratories

� Small bench top analysers such as blood gas machines or full blood count analyzers, Portable hand held devices such as glucometers, strip based assays

(Warsinke, 2009; Plebani,20009)

Trends in POC

• Fastest growing segment in diagnostics : 10% annually

• 31000 different instruments, 25% of total lab revenue

• Growth areas : cardiac, coagulation, HIV testing , drugs of abuse, tight glycaemic control, molecular diagnostics for infections

• Expanded test menus

• Consolidated platforms e.g. glucose, lactate, creatinine

• Self-contained cartridges

• Improved analytical capability

• Improved data management systems, multi-vendor enterprise solutions

4

Common themes in Research and Development facilitate new assays …

� Microfluidics for fluid delivery

� Minituarization/ low-cost microfabrication

� Magnetism/magnetoresistance

� Efficient, inexpensive light sources (semi-conductor based lighting)

� Affordable microelectronics and digital imaging hardware

� Simplified power sources

� Electromagnetic Actuation of Fluids using Micro/NanoParticles

5

Desirable Specifications of a POC test

� Rapid, results within time-frame of a consultation/clinic visit

� Require minimal training, easy to perform� Simple equipment: Hand held detection

device/strip� Whole blood� Battery operated: an advantage� No specialised laboratory set-up� Stable, temp independent, no cold chain� Affordable, commercially available� Quality control� Ability to transfer data

Advantages Disadvantages� Quality and efficiency of care

can be improved in certain scenarios

� Improved accessibility

� Improve patient compliance and LTFU

� Improved turnaround time

� Smaller sample volumes

� Economic benefits –

� reduced length of stay

� reduced complications and readmission

� Improved patient and clinician satisfaction

� Difficulties with quality control/documentation

� Higher unit of cost/reagent

� Greater personnel requirements at clinic

� Longer patient wait times

� Poor regulatory control

� Data management/audit issues

� Slower sequential processing time/throughput in high clinics

� Over-servicing

7

Current Activities in South Africa

• Component of new National laboratory policy (Maputo declaration)

• Establishment of an NHLS POC working group

• Proposed role in investigation of:– Technologies, equipment– Method validation– Development of SOP’s, training programs and manuals,

checklist– Supportive internal QC and EQA programs– Ethics clearance for clinical studies and pilot sites– CDC support

Recent analysis of rapid HIV VCT testing in South Africa: concerns over performance

Pilot multi-disciplinary POC HIV clinic site

Immediate CD4 count result and retention in care

� Hypothesis : Patients who receive and immediate CD4 count result will more likely present to a facility for ART i nitiationor for wellness.

� Design� Prospective randomized study

� Patients receiving standard of care + immediate CD4� Patients receiving standard of care + information leaflet� Patients receiving standard of care only

� Appointment in one week � No take-home message

� All patients get questionnaire on loss to follow-up reasons � Patients were recruited before their HIV test:

� negative patients are asked about VCT experience (Evaluation of post-test counseling)

� Tests were performed on a BD FACS count machine by lay personnel after safety and technical training

� (R. Osih, 2010)

Patient Characteristics

All Standard of Care

+ Leaflet Immediate CD4 test

Total number 344 119 116 119

Female (%) 223 (64.8) 76 (63.8) 69 (59.5) 78 (65.5)

Age (Range) 31.8 (19-59) 31.1 (19-56) 31.7 (20-50) 32.5 (20-59)

Mean CD4 count (cells/mm3) 346.4 335.4 336.9 364.7

> 215 (%) 232 (67.4) 78 (67.2) 74 (63.7) 80 (67.8)

< 215 (%) 112(32.6) 28 (32.8) 35 (36.3) 39 (32.2)

There were no significant differences in age, gender, employment status and education between the groups

Results

� 107/225 (47.4%) of patients who did not receive an immediate CD4count collected their results after a week

� Patients with CD4 counts < 215 were more likely to report for care if they got an immediate result when testing

Standard

of care

+Leaflet Immediate CD4

count

Reported for care 10 13 23

Did not report for care 18 22 16

Total 28 35 39

35.7% 37.1% 58.9%

P value=0.012

CD4 testing:

� C

13

Beckman Coulter XL

PIMA

Centralised flow cytometry

PIMA CD4

Overview of POC CD4 Technologies1. Rapid CD4 tests = laminar flow strips

BC ZyomyxTM First SIGN™

2. Low volume Fluorescent Cartridge Analyzers

PIMA™ LabNow™ Daktari

3. Low volume Flow Analyzers

FACSCount™ PointCare™

•Screening for ART only•Positive/Negative answer•Cut-off 350cells/µl

•Imperial Cancer funded•Largely prototype•No peer reviewed validation/ publication

•Absolute CD4 count (no CD4%)•10 samples/instrument/day•Biohazard waste-disposal (can be disposed with sharps

waste)•No cold-chain•No peer reviewed validation/ publication•Battery operated but needs mains for charging

•Absolute CD4 count and CD4%, •20-40 samples/instrument/day•Cold chain required for some reagents•Needs special biohazardous waste management•FACSCount is “open” system whereas PointCare is a

‘closed-tube’ system (less biohazardous risk)•“Mini”-laboratory required (only for FACSCount)•Peer reviewed validation/ publication FACSCount

onlyFlow Cytometry/ National CD4 Reference Laboratory, Department of Molecular Medicine and Haematology, CMJAH NHLS

Low volume Fluorescent Cartridge Analyzer systemsPIMATM

• 20min/day internal QC procedure

• Sample analyses 20 minutes

• 10 sample/instrument/8 hour day

• Open blood system/ some risk to operator

PIMATM - PHASE I Laboratory validation:

Fixed volume EDTA venous blood

•Bland-Altman analyses revealed a bias of -16±36 (95%LOA -88 to 55.2)

•%Similarity indicated tight reproducibility agreement of 98%±4.5precision to laboratory based testing(CV) of

4.6% Flow Cytometry/ National CD4 Reference Laboratory, Department ofMolecular Medicine and Haematology, CMJAH NHLS

• Bland-Altman analyses revealed a bias of -38±138 (-16±36 )95%LOA: -388 to 311 , (-88 to 55.2)

• %Similarity showed poorer reproducibility agreement of 99% ±22.7 (98%±4.5) with precision to laboratory based testing

(CV) of 23.0% (4.6%)

PIMATM - PHASE II Well resourced clinic (CMJH)

Finger prick - Capillary bloodN= 95

PIMATM - PHASE III Rural Clinic evaluation: Polokwane

Finger prick - Capillary blood

Phase III clinic data

0 20 40 60 800

200

400

600

800

1000

MeanStd. Deviation

%Sim135.3100.5

Coefficient of variation 74.28%

Sample

%S

imila

rity

Phase III clinic data(1 outlier removed= possible typing error)

0 20 40 60 800

100

200

300

400

MeanStd. Deviation

%Sim123.845.53

Coefficient of variation 36.78%

Sample

%S

imila

rity

Bland-Altman(n=68)

0

500

1000

1500

2000

-800

-600

-400

-200

0

200

400

600

800

Bias

SD of bias

95% Limits of Agreement

From

To

116.4

187.0

-250.2

482.9

Average

Diff

eren

ce

Linear regression

0 500 1000 1500 20000

500

1000

1500

2000 Best-fit values

Slope 0.9610 ± 0.07974Goodness of Fit

R square 0.7111

NHLS

PIM

A

Flow Cytometry/ National CD4 Reference Laboratory, Department of Molecular Medicine and Haematology, CMJAH NHLS

• Bland-Altman analyses revealed

bias of 116.4±187 (-38±138) (-16±36 )

95%LOA: -250 to 483 (-388 to 311) (-88 to 55.2)

• %Similarity showed poorer reproducibility / agreement

123.8% ±45.5 (99 % ±22.7) (98% ±4.5)

with precision to laboratory based testing

(CV) of 36.8% (23.0%) (4.6%)

N=61

PIMA™ Summary ARY

Excellent potential but finger prick sampling introduces error/ CV 5% ↑ >30%

Waste disposal of closed cartridges into biohazard ‘sharps” containers

Instruments able to analyze stabilized blood for external quality control programmes

Internal quality control material available and results indicate very good reproducibility of the system with CV’s <2%

Not for paediatric samples as this require a CD4% of lymphocytes

Nucleic Acid Testing20

GeneXpert for TB

Centralized automated testing500ul processed, 1ml raw sputum

60 days

Evaluation of TB diagnostics in TB suspects compared to culture and clinical outcome

� Sputum from visit 2 � HIV+ (69.6%)

Laboratory comparison to culture (single sputum)

Total

n=286

Smear

microscopy

n=282

MDRPlus

(Hain)

n=251

LightCycler

MTB

n=259

Gene Xpert

n=187

Sensitivity 54.4% (45.2;63.5) 72.1% (62;80) 70.9%(61;79) 84.8%(75;92)

Specificity 99.4%(96;99.9) 97.3%(93;99) 96.6%(92;99) 97%(92;99)

PPV* 98.5%(91;99.9) 94.9%(87;98) 94%(86;98) 95.7%(87;99)

NPV* 74.4%(68;80) 83%(77;88) 82%(75;87) 89.7%(82;94)

Performance in HIV positive individuals

HIV +

n=200

Smear

microscopy

n=196

MDRPlus

(Hain)

n=182

LightCycler

MTB

n=189

Gene Xpert

n=133

Sensitivity 48.7%(37;60) 65.2%(53;76) 63.5%(51;74) 83% (70;91)

Specificity 99.2%(95;99.9) 97%(92;99) 95.6%(90;98) 96%(89;99)

PPV 97.4%(85;99) 93.7%(82;98) 90.4%(78;96) 93.6%(82;98)

NPV 74.5%(66;81) 82%(74;88) 80%(72;86) 89.5%(81;95)

Continued..

Smear + Culture +MDRPlus (Hain) n=42/45 (93.3%) indeterminate

LightCycler MTB (Roche) n=46/46 (100%)

Gene Xpert (Cepheid) n=34/34 (100%)

Smear - Culture +MDRPlus (Hain) n=10/29 (34.4%) 2

indeterminate

LightCycler MTB (Roche) n=12/44 (27.3%)

Gene Xpert (Cepheid) n=17/27 (62.9%)

LightCycler detected n=2 M.avium

72.5% in Boehme, NEJM

Comparison to clinical outcome: Visit 4 and Any TB

Case definitions:

� Definite TB: Any culture positive in the presence of symptoms of pulmonary TB with or without positive smears for AFB.

� Probable TB: Culture negative, and at least one smear positive for AFB in the presence of symptoms of pulmonary TB.

� Possible TB: Negative smears, negative cultures for TB in the presence of symptoms of pulmonary TB with compatible chest X ray or clinical response to TB treatment (weight gain).

� No TB: Negative bacteriology and culture for TB and resolution of symptoms without TB treatment with or without treatment for another condition or an alternative diagnosis made.

Smearn=139

MDRPlus (Hain)n=121

LightCycler MTB

(Roche)n=130

Gene Xpert (Cepheid)

n=116

MGIT Culture

n=10745.2%

(34.4%;56.4%)53.5%

(41.4%; 65.3%)54.4%

(42.8; 65.5%)62.3%

(49.7%; 73.5%)73.8%

(62.8%; 82.5%)

100%(91.8%; 100%)

94.0%(82.4%; 98.4%)

94.1%(82.7%; 98.4%)

93.6%(81.4%; 98.3%)

100%(91.8%; 100%)

sensitivity

specificity

Available technology for Viral load in Tiered Laboratory Network

Tertiary Reference Laboratory Nucleic acid Testing strategies

Molecular methods: target or signal amplification

Roche Amplicor (manual and COBAS Amplicor, COBAS Ampliprep/ Amplicor)**

Roche Taqman, vs.1** and 2Abbott RealTime** (LCx obselete)

bDNA (K-PCR)

Nuclisens EASYQ (mini-MAG or EASYMAG)* version 2

In-house assays*

Biocentric assayRNA qualitative screening test

Second Tier May do NAT

Exavir Load

Primary Care POC*Sample collection: DBS

Technology Landscape for HIV viral load POCT

� Sudden and dramatic expansion of options� Specimen Processing has been the bottle-

neck previously� Technologies can be divided:

• In clinical validation• Close to clinical validation • Early development• Recognized Technologies that may in future be

applied to this use

Department of Molecular Medicine and Haematology

Category Technology General Principle

In early clinical validation LIAT (IQUUM) Multiplex real-time PCR

SAMBA (DRW) Isothermal amplification, Dipstick readout, plan for automation and visual read-out

HDA (Biohelix) Helicase dependent amplification

NAT technology that could be adapted

GeneXpert Cepheid Multiple targets, real-timeTB, MRSA

BD Max (Handylab) Real-time microfluidic PCRInfluenza

Enigma (Enigma diagnostics)

Twist Dx Recombinase polymerase amplification (RPA)

Early development EO-NAT HIV (wave 80)

Semi-Bio

North-western University p24

Eiken chemical

Inverness, Abbott etc

© IQuum, Inc. 2010

LIAT analyzerLiat TubeLiat Tube

1. Collect sample1. Collect sample

2. Scan barcode2. Scan barcode

3. Insert the tube3. Insert the tube

Get resultsGet results

Liat AnalyzerAssay chemistry: silica magnetic bead extraction & multiplex real-time PCR detection

Internal control strategy: RNA process control to minimize likelihood of a false negative readout

Time to result: 1hour

200ul plasma

Department of Molecular

Medicine and Haematology

Liat HIV Quant Assay (59 mins) compared to Abbott m2000 HIV Assay

� 93 HIV clinical samples tested

� 1 sample is Abbott assay low positive, Liat Assay negative

� 1 sample is Liat assay low positive, Abbott Assay negative

� For concordant samples, the viral load result had a 96.5% correlation coefficients between the 2 assays

� US based studies ongoing

y = 0.946x + 0.8302

R2 = 0.9652

1

2

3

4

5

6

7

8

1 2 3 4 5 6 7

Abbott HIV Viral Load [log10]

Liat

Vira

l Loa

d [lo

g10]

Early Field validation studies

Shugi Chen, 2010, IQUUM

LIAT

y = 0.9817x + 0.1187R2 = 0.9157

0

1

2

3

4

5

6

7

0 1 2 3 4 5 6 7

Reference Assay VL(log10)

Lia

t A

ssay

VL

(lo

g10)

�92% correlation with Abbott m2000 with 75 clinical specimens (clades A, B, C, and D) �Training took 5 minutes�Easy to use�Assay takes 60 minutes

�Fiscus, unpublished data 2010 (IAS, Vienna)

Department of Molecular

Medicine and Haematology

LIAT Summary

•• Turnaround time Turnaround time of 1 hour compared to several hours to days for existing assays

•• Full automation from sample to result for POCFull automation from sample to result for POC with no sample manipulation, reagent handling, or use of supporting laboratory equipment such as centrifuges, vortexes, pipettes, or computers

•• Able to accept POC compatible sample matricesAble to accept POC compatible sample matrices such as saliva and whole blood

•• Portable and battery operation Portable and battery operation to enable bedside or field use

•• Maintains sensitivity at low sample volumeMaintains sensitivity at low sample volume , ideal for pediatric patients

• Meets requirements of MSF round-table discussion document

In early clinical validation: SAMBA technology

• isothermal signal amplification• 2 hours from collection to result• Automation of front end extraction underway• Preliminary data: sensitivity <400 RNA copies/ml

Combination Strategy likely in SA

• Total Assay numbers in April 09-March 2010:• 3,190,668 CD4 assays (69 sites)

• 1,233,462 viral load assays (18 labs)

• 254,738 HIV DNA PCR assays

• Point –of- Care throughput:• CD4: 20 minutes per assay : 8-10 per day

• Viral load: 1-2 hours per assay: 4-8 per day

• Assuming 22 days per month X 12 months

• Total numbers of instruments for POC alone:• 1510 analyzers for CD4

• 704-1408 analyzers for viral load and DNA PCR

Likely Challenges for Implementation….

• Equipment Selection• Volume basi s

• Appropriate method validation• Study design• Phase I: Review of technology• Phase II: Laboratory based validation against gold

standard methodology : accuracy, precision etc• Phase III: Limited and expanded field validation

(small pilot sites)• Phase IV: Evaluation of clinical outcome of

interventionDepartment of Molecular Medicine and

Haematology

Continued…

� Appropriate internal quality control� External quality assurance programs� Instrument maintenance strategy� Good supply chain� Data retention and consolidation� Interpretation at clinic sites� Biohazard disposal� Regulatory environment

Acknowledgements

� Debbie Glencross and Lindi Coetzee� Lesley Scott: GeneXpert data� NHLS staff� POCWG� POCWG from ACTG� Clinic sites: R.Osih, F.Venter, I.Sanne� CDC in SA