Quantifying Lymphedema withlymphedema-research.org/Presentations-HNM/2014_Mayrovitz... ·...

Quantifying Lymphedema with Noninvasive Methodology

• Physical Principles• Practical Aspects• Potential Limitations

©Harvey N. Mayrovitz PhDProfessor of PhysiologyCollege of Medical SciencesNova Southeastern Universitymayrovit@nova.edu

NLN2014 – Washington D.C. 9/5/2014

HNM-NLN-2014

Why Measure/Quantify?

• Track at-risk patients• Early detection Early Tx• Severity stratification• Treatment outcomes• Documentation aspects• Research related

HNM-NLN-2014

Pre-Surgical Baseline

Threshold Change Detection

Therapy Initiation

Periodic Follow-ups

Measures and Criteria•Limb Volumes and Metrics•Limb Bioimpedance•Local Tissue Water

Dr. HN Mayrovitz

Early Detection of Lymphedema

HNM-NLN-2014

Methods Applicable to LIMBSLimb Girth (Circumference)

• Girth Limb Volume or Sum of Girths

Limb Volume• Water Displacement Limb Volume

Limb fluid content and its change• Bioimpedance BIA & BIS Whole Limb

• Tissue Dielectric Constant (TDC) Local

Physical and Structural Properties• Tonometry / Indentometry Various • Imaging: Ultrasound - MRI - Other

Methods Applicable to MOST Sites

• Head • Face • Neck • Breast • Trunk • Foot• Toe• etc

Fluid Content (TDC)Tissue Dielectric Constant

HNM-NLN-2014

Methods Applicable to MOST Sites

Force

Indentation(mm)

Physical Properties Tonometry/Identometry

04080

120160200240280320360400

1 2 3 4 5

F = 108 135 gr = 0.996, p < 0.001N = 24 legs

Forc

e (F

, g)

Indentation Depth (, mm)

Healthy legs measured 10 cm proximal to themedial malleolus

Tissue“Hardness”

Mayrovitz HN Lymphology

2009;42:88-98

Indent (mm)

Measure Force

HNM-NLN-2014

200

220

240

260

280

300

320

340

Pre-Treat Post LLLT

Forc

e (g

)Affected ArmOther Arm

200

250

300

350

400

450

Pre-Treat Post LLLT

Forc

e (g

)

Treated LegOther Leg

P<0.001

P<0.001

N = 38

P<0.001

P<0.001

Hardness Changes with LLLT

Data from: Mayrovitz HN & Davey S. Lymphology 2011;44:168-177

Force

Indentation(mm)

Arms

Legs

HN

M-N

LN-2014

Commercial Tonometers

ForceApplied

DisplacementDetermined

Pallota O. J Lymphoedema 2011;6:34-41

Modified from: Mellor et al. The Breast J. 2004;10:496-503

GelEntryEcho

Dermis

Subcutis

0.93 ± 0.13

Normal Lymphedematous

Ventral Forearm US-20 MHz

1.83 ± 1.28

Methods Applicable to MOST Sites Imaging Ultrasound MRI Other

mm

HNM-NLN-2014

Metric Measures for LIMBSTape Measure Girth at multiple points• Measure both limbs

Inter-limb differentials and sequential changes

Mark then Measure

Segment Length

• Measure one limb Sequential data but miss systemic changes

HNM-NLN-2014

Limb Girth Volume

Manual

Geometric Modelor Algorithm

Circumferences@ 4 – 12 cm

intervals

A

B

ab

L

C1

C2

General FrustumCalculationModel

V =L(A2B‐a2b)

3(A‐a)

Volume Tracking

Affected Limb

Contralateral Limb

Edema Volume

V = L/3 ( A1 + A2 + (A1 A2 )1/2

A1

A2

HNM-NLN-2014

Foot-Plate

Rail

Frame

Perometer: Girth Volume

IR d

iode

s

Mayrovitz HN et al. Advances in Wound Care 2000;113:272-276

D1

IR DiodeArray

IR Diode Array

D2

Area = KD1 D2

Frame

Perometer: Basic Principle

HNM-NLN-2014

Limb Girth & Volume LE Thresholds

Automated

Manual

GIRTHIf unilateral then lymphedema if•

inter-side differential > C1 cm or

if unilateral or bilateral then• change from pre-surgery > C2 cm

VOLUMEIf unilateral then lymphedema if• inter-side differential > V1 ml or• inter-side ratio > if unilateral or bilateral then• change from pre-surgery > V2 ml

HNM-NLN-2014

0

20

40

60

80

100

12 months 30 months

Lym

phed

ema

Rat

e (%

)

2 cm 200 ml 10% volume

Arm Lymphedema Metric Criteria

Differences • Between sides • or vs. baseline

Data from: Armer et al. J. Lymphoedema 2009;4:14-18

LE rate dependent on criteria used

HNM-NLN-2014

Mark in Relation To FLAT Surface

NOT along limb

Source of largeFollow‐up error

Practical Aspects of Limb Girth For Reproducibility: Mark along flat

HNM-NLN-2014

Volume Reduction (ml) (%)

1183 ± 778 17.2 ± 7.1 1180 ± 782 17.1 ± 7.2 1202 ± 781 17.4 ± 7.0

SegmentLength

4 cm8 cm

12 cm

4000

5000

6000

7000

8000

Leg

Volu

me

(ml)

4 cm 8 cm 12 cm

Pre-treatment Volumes

4000

5000

6000

7000

8000

Leg

Volu

me

(ml)

4 cm 8 cm 12 cm

Post-treatment Volumes

Bilateral lower extremity lymphedema

> = 10 MLD Tx

What Segment Length to Use?

N = 70

Mayrovitz et al. Physical Therapy 2007; 87: 1362-1368 HNM-NLN-2014

If you calculate on the basis of THIS

and its really more like THIS

Then you obtain a volume greater than the true value

Limb Shape as a Factor

HNM-NLN-2014

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2

Ratio (b/a)

Volu

me

Rat

io (V

e/Vc

)Limb Shape as a Factor

A

B

ab

L

C1

C2

C1 = 31.4 cmC2 = 28.3 cmL =10 cm

@ b/a 0.686.2% volume

deviation

Data from: Mayrovitz HN, Lymphology 2003;31:140-143

Ve

Vc

Ve = 0.938 Vc

HNM-NLN-2014

Photo from: K. Johansson & E Branje Acta Oncologica 2010;49:166-173Arm lymphoedema in a cohort of breast cancer survivors 10 years after diagnosis

LE if change in edema volume >= 5% from pre-surgery

Volumes via H2 O Displacement

Mostly usedas a so calledgold standardwhen comparing other methods and in research studies

0

1000

2000

3000

4000

0 1000 2000 3000 4000

Given LAV predict RAVLine of Identity95% +Interval

Left Arm Volume (LAV, ml by H2 O)

Rig

ht A

rm V

olum

e (R

AV,

ml)

Normal Arm Volume Differentials

Data from Gebruers N et al . Clin Physiol Funct Imaging 2007; 27:17-22

Line of Identity

95% +CI

100 RH FemalesRAV > LAV (2%)

0

50

100

150

200

250

300

ml

0-10 10 20 20-30 30-40 ARMArm Segment (cm)

DomNoDom

Normal Arm Volume Differentials

At-risk Arm Is:DominantNon-Dominant

3SD

Thr

esho

lds

(ml) If dominant = at-risk

Then Greater Threshold

Data from: Dylke ES et al Lymphatic Res Biology 2012;10:182-188

N =204HealthyWomenAge > 40

Girths via Perometer Volumes via frustum calculation

Hand Volume: H2 O Displacement

Ulna head

Z

Seg Vol = kZ [Ai + Ai+1 + (Ai Ai+1 )1/2]

Depth

From: Mayrovitz HN et al. Lymphology 2006;39:95-103

Metrics

HNM-NLN-2014

200

250

300

350

400

450

500

550

600

200 250 300 350 400 450 500 550 600

VM = 1.02 Vw – 12.0 mlr=0.985, p<0.001N=60 Hands

Algorithm vs. Water Displacement

Volume by water displacement (VW , ml)

Volu

me

by A

lgor

ithm

(V M

, ml)

LOA± 9.8%

HNM-NLN-2014

200

250

300

350

400

450

500

550

with arm LE (n=20) no arm LE (n=20)

Han

d Vo

lum

e (m

l) WaterPerometer

Hand Volume: H2 O vs. Perometer

Data from: Lee MJ et al. Lymphatic Research Biology 2011;9:13-18

Perometer values ~ 7.5% greater than H2 O values

r ~ 0.88

Figure-of-Eight: Hand volume Surrogate Pellecchia GL J Hand Therapy 2003;16:300-304

Maihafer GC J Hand Therapy 2003;16:305-310

cm (fig-8) vs. H2 O displacement (ml)

R = 0.94-0.95 but only normal handsTracking ability unproven

HNM-NLN-2014

WaterDisplacement

Compared toMetric

Measures

Foot Volume: H2 O Displacement

4

812 cm

0

12

L1

Y

X

A

B

CD E F

Foot in Water Filled Volumeter

Outflow Tube

L2

Mayrovitz HN et al. Lymphology 2005;38:20-27 HNM-NLN-2014

Algorithm vs. Water Displacement

N = 60 feetLOA = ± 9.3%

Mayrovitz HN et al. Lymphology 2005;38:20-27 HNM-NLN-2014

WaterDisplacement

PRO CON• Direct – Accurate

Limb/Hand/Foot volumes• Especially for irregularly shaped limbs

• Impractical for whole limbs• Bulky equipment • sterilization procedures• Patient mobility• Patient flexibility• Open wounds

Manual Girth • Low cost• Portable• Easy to use• Whole legs measureable• Hand & Foot algorithms• Limited ROM no issue• Wounds are not an issue

• Multiple measurements• Time factor• Volumes from calculations• Site repeatability

Optoelectronic(Perometer)

• Quick –Easy• Small segment lengths• Stored Measurements• Automatic processing• Selective processing

• Accuracy depends on proper positioning

• Patient mobility• Patient flexibility • Not portable• Space requirements• $$$ HNM-NLN-2014

Bioimpedance Analysis • Electrical Impedance of a limb depends

on the limb’s volume and constituents

• Lymphedema increase in low resistancefluid content of the limb

•Bioimpedance (BIOZ)•Bioimpedance Spectroscopy (BIS)•Bioimpedance Analysis (BIA)•Single Frequency BIA = SFBIA•Multi-Frequency BIA = MFBIA

HNM-NLN-2014

Muscle

Bone

Skin

Fat5-10%H2O 70-75% H2O

15-20% H2O

Limb Conducting Structures

1

20

2

Conductivity @ 5KHz

Relative to Bone

Bone

Muscle

Fat

HNM-NLN-2014

No cellsJust salineE

I

~

Z = E/I

LowFrequency

HighFrequency

SinusoidalVoltageExcitation Current increases

with frequency

Basic Operating Principle

HNM-NLN-2014

CellMembrane

Cell Interior

Cm

Ri

Re

Frequency Analysis Basis

Cell Exterior

~EI

Z = E / IECW

ICW

HNM-NLN-2014

Cole-Cole Plot: estimate parameters

increasing frequency

ReRi Re /(Ri + Re )

CellMembrane

Cell Interior

Cm

Ri

Re

Cell Exterior

~EI

Z = E / IECW

ICW

Low fHigh f

ECW + ICW ECW

MFBIA = BIS

R∞ R0

HNM-NLN-2014HNM-NLN-2014

Current Injecting ElectrodesVoltage Measuring

Electrodes

I

Basic Operating Principle

Mayrovitz HN Clinical Physiology 1998;18:234-242

Z = E/I

E

HNM-NLN-2014

Leg Volumes: Supine StandLe

g Vo

lum

e (m

l)

Time (minutes)

Blood volume shift to lower extremities

Supine Standing Supine

-20 -10 0 10 20 30 40 50

2100

2000

1900

1800

Girth-VolumeMeasurements

HNM-NLN-2014

Z Depends on Frequency & Volume

Time (minutes)

|Z| (

)

Supine Standing Supine-20 -10 0 10 20 30 40 50

130

120

110

100

090

080

070

Z ~ (1 / volume)Z ~ (1 / f)

5 KHz

500 KHz

HNM-NLN-2014

~

V

I

Z = V / I

Assessing Arm Lymphedema

Current Source

MeasuredVoltage

Current

HNM-NLN-2014

Single Frequency BIA ECW

HNM-NLN-2014

R∞

R0

Ri

Ri

R0

R0DOM

R0NONDOM

360.1± 45.8 354.8 ± 45.9

266.5 ± 39.2 257.8 ± 39.4

1052.3 ± 276.2 966.7 ± 264.9

2.988 ± 0.653 2.781 ± 0.595

0.986 ± 0.040

Multi-Frequency BIA

Nondomnant Dominant

Data from: Ward LC et al. Lymphatic Research Biology 2011;9:47-51

172 paired arms

3SD lymphedemathresholds

nondom/dom

dom = at-risk

1.134

nondom = at-risk

1.106

HNM-NLN-2014

SFBIA = MFBIA for estimating ECW

York SL et al. Breast Cancer Res Treat 2009;117:177-182

Inter-Limb Z RatiosArm LEArm controlsLeg LE

Single Frequency Bio-impedance (BIA)

Mul

tiple

Fre

quen

cy B

io-im

peda

nce

(BIS

) 2.4

2.2

2.0

1.8

0.8 1.2 1.6 2.0 2.4

Both estimateRe (Low f)

So …. Why use MFBIA (BIS)?

ECWICW

• If ICW relatively unchanged even with LE thenmay not have to depend on inter-arm ratios

• May be approximately true if muscle massdoes not significantly change since the largestfraction of ICW is associated with muscle

Proposed Concept

HNM-NLN-2014

0

1

2

3

4

5

6

Normal Arm At-Risk Arm

ECW

/ IC

W

Pre-SurgeryLymphedema Dx

ECW / ICW Ratios

Data from: Cornish BH et al. Angiology 2002;53:41-47

N = 20

HNM-NLN-2014

Tissue Dielectric Constant (TDC)Relative Permittivity (r )

PRINCIPLE

Local Tissue Water Assessment

HNM-NLN-2014

H

HO-

+ +

-

+

+ Dipole

What is Dielectric Constant?

H2 O Molecule Charge Separation

HNM-NLN-2014

What is Dielectric Constant?H

HO-

+

+

+

H

HO-

+

+

+

+

-+

-

Hydrogen bondingbetween water

molecules

2 molecules

“Hook-up”

HNM-NLN-2014

What is Dielectric Constant?

+

-

Time varying electric field of force - E

Dipole movement Displacement - Dof various types

D =

E

Dielectric Constant

= r 0 E

r = ratio /0 = TDC

H2 O @ 32oC

r = 76

HNM-NLN-2014

• Multi-Probe• Single Probe (compact)

Measurement Devices

HNM-NLN-2014

0.5 1.5 2.5 5.0 mmEffective Measurement Depth

Signal Generation and Processing

300 MHzSignal

Reflected Wave yields TDC

Control Unit

10 mm 15 mm 22 mm

p0.5 mm 1.5 mm 2.5 mm

TDCreadout

Multi-Probe

20 mm

Display has pressurebar indicator duringmeasurement

Effective measurement depth is between 1.5 & 2.5 mm Multi-Probe

Artery

Nerve

Capillary

Duct

Epidermis

Dermis

Arteriole

Hair

Pore

Fat

Hyp

o-D

erm

is

Single Probe (Compact)

HNM-NLN-2014

CenterConductor

OuterConductor

OuterConductor

ElectricField Lines

0102030405060708090

100

0 1 2 3 4 5 6 7 8 9 10

Elec

tric

Fie

ld

(% o

f Sur

face

fiel

d)

Depth (mm)

Effective Measurement Depth2.5 mm depth probe

Skin

Effective Measurement Depth

HNM-NLN-2014

GelEntryEcho

Dermis

Subcutis

0.93 ± 0.13mm

Low watercontent

High watercontent

Normal Lymphedematous

Ventral Forearm

InterrogationDepth (2.5 mm)

1.83 ± 1.28mm

5.0mm

1.5mm

subcutisskin

Modified from Mellor et al. The Breast J. 2004;10:496-503

Effective Measurement Depth

HNM-NLN-2014

0

10

20

30

40

50

60

70

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75

TDC

Calibration Example(2.5 mm probe)

Water (%)

Y = 0.632 X – 22.1r2 = 0.998, p < 0.001 Ethanol-Water

Mixture

Probe insertedInto Mixture

Cable to Control

Unit

TDC dependence on H2 O

HNM-NLN-2014

Skin Water Distribution

Skin Depth (m)

Skin

Wat

er C

onte

nt (%

)

aData: Nakagawi N et al. SRT, 2010:16:137-141; Confocal Raman Spectroscopy bData: Gniadecka et al. J Invest Dermatol 1998: 110:393-398 NIR-Raman Spec

20-24 yrs

60-68 yrs

aDermis 70-75% H2 O

N = 60

b70-90% Bound

0 50 100 150 200

Epidermis Dermis

Free and Bound Water

Protein (1 g)

Bound H2 O (0.2 – 0.5 g)

Limited Mobility H2 O

~ 20 g

Free Mobile

Free MobileFree Mobile

HNM-NLN-2014

Fluid

Protein

Blood Vessel

Lymphatic Dysfunction

LymphVessel

XBound and immobile water not readilymeasureable with

Standard BIA

Interstitium

HNM-NLN-2014

Data: Idy-Peretti et al. Int J Dermatol 1998;110:782-787 Hi-Res MRI N=21

Mobile water shows intense

Dermal Water in Lymphedema

1 mm 1 mm

Contralateral Leg Lymphedema calf

40% increasein Calf DermalWater in

Lymphedema11 primary LE10 secondary LE

TDC Features and Applications

Forearm

Biceps

Signal Generation

and Processing

TDCValue

Display

Foot DorsumTDC measurement

HNM-NLN-2014

20

25

30

35Fo

rehe

ad

Che

ek

Thum

b

A. F

orea

rm

D. F

orea

rm

Palm

-The

nar

Palm

-Cen

ter

D. H

and

Cnt

r1s

t Toe

Pla

ntar

A. G

aite

r

1stTo

e D

orsu

m

M. G

aite

r

L. G

aite

r

D. H

and

Web

D. F

oot

(1-2

)

D. F

oot

(4=5

)

M. M

alle

olus

TDC Site Variability32 females19 - 77 years1.5 mm Depth

Data From: Mayrovitz HN et al. Skin Research and Technology 2013;19:47–54

FL

CL

ML

Ophthalmic

Maxillary

Mandibular

42.1 ± 2.942.2 ± 2.5

SUPINE

39.0 ± 5.738.5 ± 5.2

43.7 ± 4.443.8 ± 4.0

N = 30 young adult males (25.0 ± 2.5 years) @ 1.5 mm depth

TDC Site Variability

Data from: Mayrovitz HN et al. Skin Research and Technology 2012;18:504–510

0

10

20

30

40

50

60

35 40 45 50 55 60 65 70 75

Total Body Water (%)

Fore

arm

TD

C @

5.0

mm

Dep

th N = 130 (50 females)Age 26.1 ± 3.0 (19-39)BMI 24.5 ± 4.0 (16-40)

Y = 0.929 X – 22.3r = 0.740, p < 0.001

Correlation with Total Body Water

HNM-NLN-2014

20

22

24

26

28

30

32

34

36

38

40

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5

Measurement Depth , mm)

TDC = 32.44 -0.185

r2 = 0.997, p<0.001N = 80 females

TDC Depth Dependence: ForearmTDC

Pattern of Depth DependenceMay Vary by Site0.5 mm

1.5 mm

2.5 mm

5.0 mm

Compact

HNM-NLN-2014

TDC Vascular Component

HNM-NLN-2014

Skin

Blo

od F

low

(a.u

.)TDC Vascular Component

From: Mayrovitz HN et al. Clinical Physiology and Functional Imaging 2013;33:55-61

Large vascular blood volume

& flow changes

Minor changesin TDC values

0.6

1.0

1.4

1.8

2.2TD

C in

ter-

arm

ratio

PatientsAffected/Control

1.64 ± 0.30N=18

Premenopausal Postmenopausal1.04 ± 0.04 1.04 ± 0.04

N=15 N=15

Mayrovitz HN  Lymphology 2007;40:87‐94

TDC Lymphedema Discriminations

HNM-NLN-2014

Pre-Surgery Reference TDC Ratios

Cancer Side Healthy Side

Axilla

Biceps

Forearm

1.029 ± 0.196(1.60)

1.012 ± 0.143(1.45)

1.003 ± 0.096(1.30)

N =103

N=80

Thorax0.999 ± 0.119

(1.35)

(3 SD Thresholds)

HNM-NLN-2014

0.900

0.950

1.000

1.050

1.100

1.150

1.200

pre-surgery 3 6 12 18 24

Thor

ax T

DC

Rat

io (A

t-Ris

k/C

ontr

ol S

ide)

0‐3‐6‐12‐18‐24 month (N=35)0‐3‐6‐12‐18 month (N=41)0‐3‐6‐9‐12 month (N=47)0‐3‐6  month (N=53)0‐3 month (N=60)

**** ** **

** *** * ****

Sequential TDC Ratio Changes

LateralThorax

HNM-NLN-2014

0.900

0.950

1.000

1.050

1.100

1.150

1.200

pre-surgery 3 6 12 18 24

Axi

lla T

DC

Rat

io (A

t-Ris

k/co

ntro

l)

0‐3‐6‐12‐18‐24 month (N=35)0‐3‐6‐12‐18 month (N=41)0‐3‐6‐9‐12 month (N=47)0‐3‐6  month (N=53)0‐3 month (N=60)

* *****

***

** ** ***

Sequential TDC Ratio Changes

Axilla

HNM-NLN-2014

TDC BIA/BIS (Delfin Technologies Ltd) (Impedimed Ltd) Operating principle Frequency applied EMF 300 MHz 4 - 1000 kHz Current flowing in the body Very Localized Much of the body Number of electrodes / probes 1 probe 4 electrodes Total single measurement time ~ 8 sec ~ 60 sec Measurement Depth 0.5 – 5 mm Undefined Measurement quantity Tissue dielectric constant Resistance Measurement parameter Skin-to-fat tissue fluid Parameter ~ to ECF Applicability Practically all body sites Limbs Patient preparation Patient position Any body position Supine

Arm-leg skin contact No effect Limbs must be abducted

Arm and hand position No restriction Palms flat on surface Shoe and socks removal Not needed to remove Must be removed Bladder emptying necessary No Yes Dominant side affects No Yes Measurement sites Hairy skin shaving Yes (very hairy) Yes Precautions for measurement Patient metal contact problem No Yes

Methods Features Comparison

HNM-NLN-2014