Segmental Bioimpedance Informs Diagnosis of Breast Cancer-Related Lymphedema

Abstract

Background:

Detection of lymphedema, particularly its mild stage, is clinically challenging. The aim of this study was to determine whether segmental bioimpedance spectroscopy (BIS) provided additional information to whole arm BIS in assessing women with or at risk of lymphedema following breast cancer.

Methods and Results:

Participants (n = 66), aged 61.6 ± 10.5 years (mean ± standard deviation [SD]), were grouped according to lymphedema status: (1) at-risk (n = 24) had no indicators of lymphedema and (2) lymphedema (n = 42) were suspected to be developing lymphedema or had previously met lymphedema diagnostic criteria and undergone treatment. For each upper limb, impedance was measured for the whole arm, hand and four 10 cm segments of the arm, commencing at the ulnar styloid. Interlimb impedance ratios for corresponding locations were calculated and compared to previously determined, normatively based thresholds based on 2SD and 3SD above the mean. Segmental BIS classified 19% more women with lymphedema than 3SD whole arm thresholds and the same number as 2SD whole arm thresholds. Segmental BIS identified localized lymphedema and patterns in lymphedema distribution that were undetectable by whole arm BIS. Neither 3SD whole arm nor segmental BIS thresholds found lymphedema where it was not present; however, 2SD whole arm thresholds alone classified one woman in the at-risk group as having lymphedema.

Conclusion:

Segmental BIS classified as many or more cases of lymphedema than whole arm BIS thresholds without finding lymphedema where it was likely not present while also providing additional information regarding the distribution of lymphedema within the limb.

Introduction

Lymphedema is a long-term swelling affecting regions of the body. Arm lymphedema resulting from the treatment of breast cancer is the most common cause of lymphedema in developed countries.^1–3 Early diagnosis and treatment of lymphedema likely results in improved response to treatment^4–6 and lessens the health burden of the condition in terms of cost, wellness, and management time.^7–9 However, localized or mild lymphedema can be difficult to identify¹⁰ and existing classification systems may not be sufficiently sensitive to detect it.¹¹

Bioimpedance spectroscopy (BIS) is a noninvasive quantitative measure of breast cancer-related lymphedema of the arm,^12–14 detecting lymphedema up to 10 months before other assessment methods.¹² During BIS measurement, a small electrical current is passed through the body region of interest and the impedance to the current determined. At low frequencies, this current does not enter the cells, rather passing through the extracellular fluid (ECF).¹² The impedance at low frequency is inversely related to ECF volume,¹⁵ and as lymph is a major component of ECF,¹⁶ BIS is particularly well-suited for detection of the higher levels of ECF found in lymphedema.

Current practice for breast cancer-related lymphedema uses interlimb BIS ratios calculated from the impedance of the arms to determine the likelihood of lymphedema being present. The use of both arms to calculate interlimb ratios removes the effect of weight and whole body fluid changes that are unrelated to lymphedema. To determine whether lymphedema is present, an individual's interlimb ratio is compared to BIS interlimb ratio thresholds previously determined for the dominant and nondominant limb from a normative population.^12,17 A BIS ratio that exceeds the BIS threshold is an indication that lymphedema is likely present. BIS ratios can also be monitored over time for progressive increases that are suggestive of lymphedema development,¹² and used to determine treatment effectiveness.^18,19

BIS thresholds, traditionally set at three standard deviations (3SDs) above a normative population mean,¹² have been used widely in clinical practice and research.^19–22 Its use has been challenged as it may be too high for identification of mild lymphedema.¹³ Fu et al.¹³ found 34% of women with previously treated lymphedema who still had a 200 mL volume difference between their arms (affected and unaffected) were not detected by 3SD whole arm BIS ratios. Similarly, Dylke et al.²³ found that the 2SD whole arm BIS threshold, rather than the 3SD threshold, had higher sensitivity and specificity for identifying women with mild to moderate lymphatic changes as determined by lymphoscintigraphy. One method, therefore, to improve lymphedema detection could be to use the more liberal 2SD BIS thresholds; however, another method to improve identification of localized or mild lymphedema may be to use segmental rather than whole arm measurements.¹¹

Previously, we demonstrated that BIS can reliably quantify impedance in smaller segments of the arm and the hand.^24,25 Normatively based BIS thresholds for four arm segments and the hand, accounting for limb dominance, in women without history of breast cancer or lymphedema were determined.²⁴ However, it is unknown whether segmental BIS assessment of the upper limb provides additional information to whole arm interlimb BIS ratios in women thought to be developing or currently living with lymphedema. Segmental BIS of the upper limb could potentially detect mild or localized cases of lymphedema and identify the distribution of lymphedematous changes within the limb. As a consequence, this knowledge may facilitate intervention at a very early stage and allow customization of lymphedema treatment for individual presentations, potentially improving treatment outcomes. The aims of this study, therefore, were to

1. Determine whether classification of lymphedema, based on segmental BIS ratios is in agreement with classification of lymphedema based on the 3SD and the more liberal 2SD whole-arm BIS ratios.

2. Determine whether additional cases of lymphedema are identified using segmental BIS ratios.

3. Describe the pattern of lymphedema distribution at a segmental level.

Materials and Methods

Participants

Sixty-six women who had completed treatment for unilateral breast cancer, including axillary node removal, were recruited through media advertisement and from health services in western New South Wales, Australia. The women were classified according to lymphedema status: (1) at risk for but currently no lymphedema ( at-risk ; n = 24), or (2) lymphedema ( lymphedema ; n = 42). The at-risk group comprised women who had never had a diagnosis of lymphedema and had no significant physical signs of lymphedema, such as pitting edema or skin tightness on assessment. The lymphedema group comprised women suspected to be developing lymphedema ( developing ; n = 20) or with established lymphedema for which they had undergone treatment ( treated ; n = 22). Participants within the developing group had been referred to a lymphedema service with suspected lymphedema. These participants reported symptoms of lymphedema such as swelling and heaviness, and presented with physical signs of lymphedema such as pitting edema and tissue changes; however, they had not been previously diagnosed nor had undergone any treatment for lymphedema. The treated group had medical record documentation of lymphedema diagnosis of exceeding normative-based thresholds for either interlimb circumference differences²⁶ or whole arm interlimb BIS ratios.¹² They had undergone treatment for lymphedema, which may have included compression garments, massage, exercises, skin care, or other treatments.

Women who were pregnant, had cardiac implants (pacemaker, defibrillator), metal implants in upper limbs, or medical conditions that could impact upper limb fluid volumes (e.g., cerebrovascular accident, limb deformity) were excluded from participating in the study. The participants' characteristics are reported in Table 1. The women in the at-risk group were, on average, younger than those in the treated and developing group (p = 0.05); however, there was no difference in age at time of breast cancer surgery (p = 0.10) or body mass index (BMI) (p = 0.51). The median (interquartile range) duration of lymphedema for the treated group was 4.6 (2.9–9.1) years.

Table 1.

Participant Characteristics by Diagnostic Group

	At-risk (n = 24)	Treated (n = 22)	Developing ( n = 20)	p
Age (years)^a	57.6 (10.7)	65.1 (10.7)	62.6 (8.9)	0.05
Age at time of surgery (years)^a	53.2 (7.5)	56.3 (10.1)	58.9 (8.7)	0.10
BMI (kg/m²)^a	29.3 (5.9)	29.5 (4.8)	33.1 (10.0)	0.51
Arm dominance (R:L) (n)	17:7	20:2	19:1
Dominance:nondominance affected (n)	11:13	9:13	12:8
Time between surgery and participation (years)^b	2.7 (1.1–6.3)	9.4 (4.9–12.4)	1.0 (0.5–3.9)	0.002
Duration of lymphedema (years)^b	—	4.6 (2.9–9.1)	—
Bioimpedance interlimb ratios (median and IQR)
Hand (MCP to wrist)	0.979 (0.945–1.002)	1.053 (0.998–1.138)	1.028 (0.992–1.170)	0.001
Segment A (wrist to 10 cm)	1.014 (0.974–1.037)	1.077 (1.013–1.308)	1.232 (1.100–1.390)	0.001
Segment B (10–20 cm)	1.016 (0.972–1.050)	1.199 (1.048–1.312)	1.150 (1.053–1.334)	0.001
Segment C (20–30 cm	1.015 (0.982–1.036)	1.078 (1.017–1.230)	1.160 (1.037–1.216)	0.004
Segment D (30–40 cm)	1.022 (0.951–1.055)	1.038 (0.988–1.216)	1.101 (1.069–1.260)	0.0006
Whole arm	1.018 (0.0974–1.038)	1.087 (0.032–1.229)	1.141 (1.093–1.254)	<0.0001

Values are mean (SD) by Kruskal–Wallis test.

Values are median (IQR) by Kruskal–Wallis test.

BMI, body mass index; IQR, interquartile range; MCP, metacarpalphalangeal; R:L, right:left; SD, standard deviation.

The BIS ratios for the whole arm in women in the at-risk group was significantly lower than for the women in both the treated and developing lymphedema groups (p ≤ 0.01; Table 1); however, there was no significant difference in BIS ratios for the whole arm between the treated and developing lymphedema groups (p = 0.33).

Ethics approval was obtained from the Human Research Ethics Committee of Catholic Healthcare (Lourdes Hospital and Community Health Service) and the University of Sydney's Human Research Ethics Committee. All participants provided written informed consent.

Assessments

All participants attended an assessment session of ∼30 minutes duration in which all measurements were completed. Height, to the nearest 5 mm and weight, to the nearest 0.1 kg, were measured for calculation of BMI (kg/m²). Whole arm and segmental BIS was used to quantify volume of ECF in the whole arm, hand, and arm segments using an SFB7 impedance spectrometer (Impedimed, Ltd., Brisbane, Australia) following previously described, reliable protocols.^24,25,27

In brief, sites for electrode placement were marked with an indelible pen with participants seated, shoulders extended to 90 degrees forward flexion with full elbow extension. After cleaning the sites with an alcohol wipe, electrodes were applied. The measurement electrodes were placed at the metacarpalphalangeal (MCP) joint of the third digit, mid-wrist (dorsum) in line with the ulnar styloid and on the dorsum of the arm at 10, 20, 30, and 40 cm proximal from the ulnar styloid process. The current drive electrodes were placed at the base of the third toe on the right foot, over the third MCP (for whole arm BIS) and the base of the nail bed of the third fingers (for hand and segmental measurements).

Participants were positioned in supine on a nonconductive surface for BIS measurement. The impedances, extrapolated to zero frequency (R₀), for the whole arm, hand, and arm segments for both upper limbs were obtained. The measurement electrodes for the whole arm measures were positioned based on the principles of equipotential.²⁸ In contrast, the measurement electrodes for the hand and segmental measurements spanned the region of interest. The hand and arm segments were defined as follows: hand segment from the third MCP joint to the ulnar styloid,²⁵ Segment A (Seg A): wrist to 10 cm; Segment B (Seg B): 10–20 cm; Segment C (Seg C): 20–30 cm; and Segment D (Seg D): 30–40 cm.^24,27 Calibration of the SFB7 was assessed daily using the test cell supplied by the manufacturer.

BIS files were analyzed using the Bioimp software (version 5.2.4.0). As R₀ is inversely proportional to ECF volume, BIS ratios were calculated by dividing the resistance of the unaffected limb/segment by the affected limb/segment (R_0unaffected/R_0affected) and then compared to thresholds according to limb dominance.^12,24

Participants were classified as having lymphedema if their interlimb BIS ratio exceeded previously established thresholds based on normative population.^12,24 The different criteria for classification of lymphedema status explored in this study were based on whole arm measurements, including exceeding ≥2SD and ≥3SD above the mean for the whole arm^12,23 and on segmental measurements, including one arm segment exceeding ≥3SD above the mean or ≥2 arm segments exceeding ≥2SD above the mean.²⁴

Statistical analysis

Participant characteristics, absolute impedance and BIS ratio data were checked for normality using the Shapiro–Wilk test. Differences between groups were assessed using analysis of variance and Kruskal–Wallis test. Due to lack of differences, the treated and developing groups were combined to form the lymphedema group, which was then compared to the at-risk group. Segments were classified as meeting 2SD threshold criteria if the segment BIS ratio was ≥2SD threshold but <3SD threshold. Segments were classified as meeting 3SD criteria if the segment BIS ratio was ≥3SD thresholds. Descriptive statistics were used to describe agreement between segmental and whole arm BIS, additional cases of lymphedema identified with segmental BIS and 2SD whole arm BIS, and patterns of lymphedema distribution. Data were analyzed using Prism 7 for Windows (version 7.01) and significance was set at p < 0.05.

Results

For the whole group, whole arm BIS criteria and segmental BIS criteria agreed in classification of the lymphedema status for 78.8% of all participants including 65.9% of those with lymphedema. Neither 3SD whole arm nor the segmental BIS criteria classified women in the at-risk group as having lymphedema. In contrast, in the at-risk group, the 2SD whole arm BIS criterion was exceeded in one woman at risk for lymphedema in the nondominant arm and was within <1% of the threshold for two other women, both of whom were at risk for lymphedema in their nondominant arm.

Segmental BIS and the conservative 3SD whole arm BIS criteria were in agreement in the classification of 32 women in the lymphedema group (n = 42). Twenty-one of the 32 women were classified as having lymphedema and 11 as not having lymphedema. Of the 10 women from the lymphedema group in whom there was not agreement, segmental BIS criteria classified nine as having lymphedema while the 3SD whole arm criteria classified one of the women with lymphedema. In women classified as having lymphedema by only segmental BIS, lymphedema was present in the dominant limb of six women and the nondominant limb of three women. The one woman identified only by the 3SD whole arm criteria had treated lymphedema in the nondominant limb. The overall agreement between segmental and 3SD whole arm BIS criteria was 85% in all participants and 76% in women with lymphedema.

Adoption of the more liberal 2SD whole arm criteria slightly improved overall agreement between segmental and 2SD whole arm criteria in the classification of lymphedema in the lymphedema group (n = 42). Overall, the 2SD diagnostic criteria agreed for 34 women of the women in the lymphedema group, with more women classified as having lymphedema (26 women) and less as not having lymphedema (8 women) than found with the 3SD whole arm thresholds. Disagreement in classification occurred for eight women: segmental BIS classified four women with lymphedema on the dominant side, 2SD whole arm classified four women, two of whom had lymphedema on the dominant side. One woman classified only by the 2SD whole arm threshold had one segment exceeding 2SD and another within 0.005 of meeting the segmental threshold. The overall agreement between segmental and 2SD whole arm BIS criteria was 86% in all participants and 81% in women with lymphedema.

Segmental BIS detected patterns in lymphedema distribution in the lymphedema group (n = 42; Fig. 1). Lymphedema was not detected equally throughout the arm segments. Twenty-seven women exceeded the 2SD segmental BIS threshold for Seg B (64%), 21 exceeded the Seg A (50%) and Seg C (50%) threshold, 15 exceeded Seg D (36%) threshold and 13 exceeded the hand (30%) threshold.

FIG. 1.

Individual participant data for women with (A) no lymphedema, (B) with lymphedema, or (C) developing lymphedema. Circles represent arm BIS ratio; letters represent the individual segments. Black filled circles: exceed 3SD thresholds; gray filled circles: exceed 2SD thresholds; white filled circles: indicate arm BIS ratio is within normal range. Letters are coded black if they exceed 3SD segment thresholds; dark gray if they exceed 2SD thresholds; light gray if the BIS ratio is within normal range. BIS, bioimpedance spectroscopy; SD, standard deviation.

Two different lymphedema distribution patterns were identified in the women categorized as having lymphedema by segmental BIS criteria: (1) generalized lymphedema that extended through at least half the limb, that is, at least two adjacent segments (n = 24); and (2) localized lymphedema, in which lymphedema was confined to only small areas of the limb (n = 6). The women who presented with generalized lymphedema met both segmental BIS criteria, that is, one segment exceeded 3SD thresholds and ≥2 segments exceeded 2SD thresholds. In addition, 21 of these 24 women also exceeded the 3SD whole arm threshold, two women exceeded only the 2SD whole arm thresholds and one woman was within 1.2% of the 2SD threshold but had two segments over 2SD and one over 3SD.

The six women with localized lymphedema met only one of the segmental BIS criterion, that is, one segment exceeding 3SD or two segments exceeding 2SD. Four of these six women had a single segment over a 3SD threshold; three were in the developing subgroup and had lymphedema in either Seg A or C whereas the fourth woman was in the treated subgroup and had lymphedema in only Seg B. The remaining two with localized lymphedema were in the treated subgroup, and lymphedema was identified in two nonadjacent segments: Seg B and Seg D for both. While three of the women with localized lymphedema also exceeded the 2SD whole arm criteria, none exceeded 3SD whole arm thresholds. In the 12 women who were not classified as having lymphedema by segmental BIS, only four had a segment over a 2SD threshold.

Conclusions

Segmental BIS criteria based on normatively determined thresholds for the hand and 10 cm segments of the arm provided information additional to that provided by whole arm BIS. Segmental BIS thresholds distinguished those with breast cancer-related lymphedema of the arm from those who displayed no signs or symptoms in women with mild to moderate swelling. In addition, segmental BIS quantified the often-uneven distribution²⁹ of lymphedema within the limb.

Segmental BIS classified more women with lymphedema than the conservative 3SD whole arm BIS criteria. Notably, the additional women classified by only segmental BIS had either localized or mild lymphedema that was unlikely sufficient to exceed the whole arm BIS threshold. The segments that exceeded the segmental criteria were typically in the forearm or around the elbow, with the proximal upper arm and hand least affected. Previous studies have similarly found this pattern, finding that it may be predictive of intractable lymphedema.²⁹ Furthermore, it is in agreement with current hypothesis of development of lymphedema in which there is localized failure of lymphatics in the mid-arm region.³⁰

Lowering the whole arm BIS threshold for classification of lymphedema to the 2SD whole arm criteria classified more women as having lymphedema than the 3SD criteria. However, as three women in the at-risk group, without any signs indicative of lymphedema, were either classified, or within 1% of classification, as having lymphedema, this threshold may be too liberal, introducing false positives. This may be specific to those at risk in their nondominant limb, as it was only in those whose nondominant limb was at risk that the 2SD whole arm thresholds classified as having lymphedema. Neither of the studies that advocated use of 2SD whole arm threshold^13,23 explored whether dominance of the affected arm impacted the superiority of 2SD or 3SD whole arm thresholds. It is unknown whether, similar to fat deposition in lymphedema,³¹ dominance of the limb affects preferentially the volume of ECF in developing lymphedema.

One participant in the treated subgroup was classified as having lymphedema by 3SD whole arm (and by default, the 2SD whole arm threshold) but not segmental BIS criteria. There are two possible explanations for this: measurement range and classification of a false positive. We hypothesize the lack of congruence is due to measurement range as whole arm BIS, using the principle of equipotentials,²⁸ measures impedance from the ulnar styloid to the shoulder, a distance that varies according to limb length. In contrast, segmental BIS measured only up to 40 cm from the ulnar styloid. Thus, whole arm BIS can potentially detect edema outside the measurement range of segmental BIS when the limb is longer than average or edema is found in regions proximal to 40 cm from the ulnar styloid.

Neither segmental nor the whole arm BIS criteria detected the presence of lymphedema in some women with or suspected of developing lymphedema. The most obvious reasons are that lymphedema was never present, it had resolved or it had reduced to below threshold levels. However, the use of absolute thresholds may be another explanation why some women with lymphedema may not exceed segmental or whole arm BIS criteria for identification of lymphedema. Absolute thresholds, such as those used in BIS, detect lymphedema when BIS ratios meet or exceed thresholds based on a set number of SDs above the mean of a normative population. When women have prebreast cancer surgery BIS ratios below the population mean they must experience an increase in their BIS ratios of more than 3SD to reach the absolute diagnostic thresholds (Fig. 2). Cornish et al.¹² proposed a change of 0.1, from a presurgical baseline, in whole arm BIS ratio as an alternate method of lymphedema detection, with 0.1 being the sum of 3SDs of the mean of a normative population. Thus, women with naturally low presurgery BIS ratios can develop lymphedema before achieving either of the absolute 2SD or 3SD whole arm thresholds. Segmental BIS ratio change values could also be calculated from Svensson et al's data²⁴; however, both whole arm and segmental change values require validation in a longitudinal study.

FIG. 2.

Normal curve distribution diagram. The arrow indicates the number of SDs required to exceed the 3SD absolute threshold from 2SD below the population mean. A woman with a healthy BIS ratio 2SD below the mean may develop lymphedema at 1SD above the population mean, this lymphedema would be detected by BIS ratio change but not absolute BIS thresholds.

Despite the small sample size, this study found that segmental BIS criteria classified additional women to the whole arm BIS criteria in both the new and known subgroups of the lymphedema group. These findings suggest that segmental BIS criteria provide another resource that can be used for early detection of lymphedema. In addition, the ability to measure objectively small localized changes with segmental BIS provides opportunity for further research into the etiology and progression of lymphedema, and the potential to better customize interventions to individual lymphedema presentations, thereby improving treatment outcomes, reducing lymphedema-associated costs and health burden, and improving quality of life.^11,32,33 Future research is needed to confirm whether segments that exceed the threshold are associated with dermal back flow identified on lymphoscintigraphy, a feature unique to lymphedema. Furthermore, the highlighted issues relating to dominance of the affected limb require further investigation with a larger sample. Regardless of the improved classification afforded by segmental BIS found in this study, lymphedema detection, due to the challenges presented by unique and individual lymphedema presentations and progressions, will likely need a combination of measures, both objective and subjective to improve detection rates; this could be achieved through a nomogram as suggested by Dylke et al.²³

Segmental BIS can supplement whole arm BIS in the detection of lymphedema and evaluation of treatment in women with and at risk of developing breast cancer related lymphedema. There was significant agreement between segmental BIS and whole arm BIS in the classification of lymphedema status; however, segmental BIS identified additional cases of lymphedema without falsely detecting lymphedema in the at-risk population. Segmental BIS provided information regarding the patterns of lymphedema distribution within the limb, including the presence of localized lymphedema in women with newly developing and previously treated lymphedema that was not detectable with whole arm measurements.

Footnotes

Acknowledgment

B.J.S. was supported by project funding from the NSW Health Education and Training Institute (HETI).

Author Disclosure Statement

B.J.S., E.S.D., and S.L.K. have no competing financial interests. L.C. Ward consults to Impedimed, Ltd. Impedimed had no involvement in the design, undertaking, or article preparation of this study.

References

Armer

, Heckathorn

. Post-breast cancer lymphedema in aging women: Self-management and implications for nursing. J Gerontol Nurs, 2005; 31:29–39.

Rockson

, Rivera

. Estimating the population burden of lymphedema. Ann N Y Acad Sci, 2008; 1131:147–154.

Warren

, Brorson

, Borud

, Slavin

. Lymphedema: A comprehensive review. Ann Plast Surg, 2007; 59:464–472.

Ramos

, O'Donnell

, Knight

. Edema volume, not timing, is the key to success in lymphedema treatment. Am J Surg, 1999; 178:311–315.

Forner-Cordero

, Munoz-Langa

, Forner-Cordero

, DeMiguel-Jimeno

. Predictive factors of response to decongestive therapy in patients with breast-cancer-related lymphedema. Ann Surg Oncol, 2010; 17:744–751.

Shah

, Arthur

, Wazer

, Khan

, Ridner

, Vicini

. The impact of early detection and intervention of breast cancer-related lymphedema: A systematic review. Cancer Med, 2016; 5:1154–1162.

Bilir

, DeKoven

, Munakata

. Economic benefits of BIS-aided assessment of post-BC lymphedema in the United States. Am J Manag Care, 2012; 18:234–241.

Shih

, Xu

, Cormier

, et al. Incidence, treatment costs, and complications of lymphedema after breast cancer among women of working age: A 2-year follow-up study. J Clin Oncol, 2009; 27:2007–2014.

Armer

, Henggeler

, Brooks

, Zagar

, Homan

, Stewart

. The health deviation of post-breast cancer lymphedema: Symptom assessment and impact on self-care agency. Self Care Depend Care Nurs, 2008; 16:14–21.

10.

Stanton

, Modi

, Mellor

, Levick

, Mortimer

. Diagnosing breast cancer-related lymphoedema in the arm. J Lymphoedema, 2006; 1:12–15.

11.

Stout Gergich

, Pfalzer

, McGarvey

, Springer

, Gerber

, Soballe

. Preoperative assessment enables the early diagnosis and successful treatment of lymphedema. Cancer, 2008; 112:2809–2819.

12.

Cornish

, Chapman

, Hirst

, et al. Early diagnosis of lymphedema using multiple frequency bioimpedance. Lymphology, 2001; 34:2–11.

13.

, Cleland

, Guth

, et al. L-dex ratio in detecting breast cancer-related lymphedema: Reliability, sensitivity, and specificity. Lymphology, 2013; 46:85–96.

14.

Czerniec

, Ward

, Refshauge

, et al. Assessment of breast cancer-related arm lymphedema—Comparison of physical measurement methods and self-report. Cancer Invest, 2010; 28:54–62.

15.

Warren

, Janz

, Slavin

, Borud

. The use of bioimpedance analysis to evaluate lymphedema. Ann Plast Surg, 2007; 58:541–543.

16.

Cornish

, Chapman

, Thomas

, Ward

, Bunce

, Hirst

. Early diagnosis of lymphedema in postsurgery breast cancer patients. Ann N Y Acad Sci, 2000; 904:571–575.

17.

Ward

, Dylke

, Czerniec

, Isenring

, Kilbreath

. Confirmation of the reference impedance ratios used for assessment of breast cancer-related lymphedema by bioelectrical impedance spectroscopy. Lymphat Res Biol, 2011; 9:47–51.

18.

Cornish

, Bunce

, Ward

, Jones

, Thomas

. Bioelectrical impedance for monitoring the efficacy of lymphoedema treatment programmes. Breast Cancer Res Treat, 1996; 38:169–176.

19.

Shah

, Vicini

, Beitsch

, et al. The use of bioimpedance spectroscopy to monitor therapeutic intervention in patients treated for breast cancer related lymphedema. Lymphology, 2013; 46:184–192.

20.

Vicini

, Shah

, Lyden

, Whitworth

. Bioelectrical impedance for detecting and monitoring patients for the development of upper limb lymphedema in the clinic. Clin Breast Cancer, 2012; 12:133–137.

21.

Bundred

, Stockton

, Keeley

, et al. Comparison of multi-frequency bioimpedance with perometry for the early detection and intervention of lymphoedema after axillary node clearance for breast cancer. Breast Cancer Res Treat, 2015; 151:121–129.

22.

Kim

, Jeon

, Sung

, Jeong

, Do

, Kim

. Prediction of treatment outcome with bioimpedance measurements in breast cancer related lymphedema patients. Ann, 2011; 35:687–693.

23.

Dylke

, Schembri

, Bailey

, et al. Diagnosis of upper limb lymphedema: Development of an evidence-based approach. Acta Oncologica, 2016; 55:1477–1483.

24.

Svensson

, Dylke

, Ward

, Kilbreath

. Segmental impedance thresholds for early detection of unilateral upper limb swelling. Lymphat Res Biol, 2015; 13:253–259.

25.

Dylke

, Alsobayel

, Ward

, Liu

, Webb

, Kilbreath

. Use of impedance ratios to assess hand swelling in lymphoedema. Phlebology, 2014; 29:83–89.

26.

Dylke

, Yee

, Ward

, Foroughi

, Kilbreath

. Normative volume difference between the dominant and nondominant upper limbs in healthy older women. Lymphat Res Biol, 2012; 10:182–188.

27.

Czerniec

, Ward

, Lee

, Refshauge

, Beith

, Kilbreath

. Segmental measurement of breast cancer-related arm lymphoedema using perometry and bioimpedance spectroscopy. Support Care Cancer, 2011; 19:703–710.

28.

Cornish

, Jacobs

, Thomas

, Ward

. Optimizing electrode sites for segmental bioimpedance measurements. Physiol Measurement, 1999; 20:241–250.

29.

Stout

, Pfalzer

, Levy

, et al. Segmental limb volume change as a predictor of the onset of lymphedema in women with early breast cancer. PM R, 2011; 3:1098–1105.

30.

Stanton

AWB

, Svensson

, Mellor

, Peters

, Levick

, Mortimer

. Differences in lymph drainage between swollen and non-swollen regions in arms with breast-cancer-related lymphoedema. Clin Sci, 2001; 101:131–140.

31.

Dylke

, Ward

, Meerkin

, Nery

, Kilbreath

. Tissue composition changes and secondary lymphedema. Lymphat Res Biol, 2013; 11:211–218.

32.

Cormier

, Xing

, Zaniletti

, Askew

, Stewart

, Armer

. Minimal limb volume change has a significant impact on breast cancer survivors. Lymphology, 2009; 42:161–175.

33.

Soran

, Ozmen

, McGuire

, et al. The importance of detection of subclinical lymphedema for the prevention of breast cancer-related clinical lymphedema after axillary lymph node dissection; a prospective observational study. Lymphat Res Biol, 2014; 12:289–294.