Post on 07-Sep-2018
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