Volume and Tissue Composition Changes Measured with Dual-Energy X-Ray Absorptiometry in Melanoma-Related Limb Lymphedema

Abstract

Abstracts

Background:

The aim of this cross-sectional study was to investigate the volume, fat mass, and lean mass in both upper and lower limbs measured with dual-energy X-ray absorptiometry (DXA) in melanoma patients with melanoma-related limb lymphedema.

Methods and Results:

Four hundred thirty-one patients who had undergone surgical treatment with wide local excision and unilateral axillary or inguinal sentinel lymph node biopsy and/or complete lymph node dissection participated in a survey, and they underwent clinical examination and measurements of their upper or lower limbs with DXA. Limb lymphedema was diagnosed on the basis of history and characteristic physical findings on the clinical examination. The inter-limb differences in volume, fat mass, and lean mass measured with DXA were categorized as none/mild, moderate, or severe according to reference values (taking handedness into account for the upper limbs). Of the 431 patients, 109 (25%) had clinical melanoma-related limb lymphedema corresponding to 23 (10%) who had undergone axillary nodal surgery and 86 (45%) who had undergone inguinal nodal surgery. The majority of patients developed lymphedema within the first year (90%) after surgery, and the majority of lymphedemas were categorized as mild. The increase in the volume of limbs with lymphedema was primarily due to an increase in fat mass.

Conclusion:

There is a high prevalence of melanoma-related limb lymphedema. The increase in volume in the limb with lymphedema is primarily due to an increase in fat mass. This increase in fat mass in limbs with lymphedema is important for the understanding of the pathophysiology and may be important in the treatment of lymphedema.

Introduction

Clinical limb lymphedema after lymph node surgery in cancer patients is a common problem that is often under-recognized and under-treated.¹ After lymph node surgery in melanoma patients, lymphedema has been reported to be present in 1%–35% of patients after sentinel lymph node biopsy (SLNB) of the axilla and groin and in 10%–83% of patients after axillary and inguinal complete lymph node dissection (CLND).^2–7 The lymphatic impairment caused by surgery results in an accumulation of protein-rich fluid that leads to edema and a chronic inflammatory reaction, which with time will lead to an excess of adipose tissue.^8,9 However, there is still an incomplete understanding of the pathophysiology involved in the development of lymphedema. Limb lymphedema is commonly diagnosed clinically and staged by using the International Society of Lymphology (ISL) staging system.¹⁰ Quantification of lymphedema includes volume measurement of the limb by water displacement, circumference measurements, and perometer.¹¹ However, these methods do not provide information on tissue composition in limbs with lymphedema. Dual-energy X-ray absorptiometry (DXA) provides information on both volume and tissue composition, and this method has high repeatability for measurements of both upper limb and lower limb.^11,12 The DXA scan provides data on body composition in terms of three compartments as defined by their density: bone mineral, fat mass, and lean mass. Limb volumes of the three compartments and, as such, the entire limb can be calculated by using the densities of each of the three compartments.^13,14

Treatment of lymphedema varies with the type of tissue affected. In stages with primarily excess fluid, decongestive treatment and compression garments can reduce the excess volume by promoting redirection of the lymph toward the functioning lymph nodes in unaffected areas.¹⁵ Super-microsurgical lymphaticovenular anastomosis aimed at draining the lymph to the venous system is also effective, especially early after the onset of lymphedema.¹⁶ However, in advanced stages of lymphedema where the excess in limb volume is dominated by an increase in adipose tissue, liposuction can be used to reduce limb volume.^17,18 It is, therefore, of importance to measure not only the volume but also the tissue composition and distribution in the affected limb compared with the normal limb.

After the treatment of cancer, for example, melanoma and breast cancer where lymph node surgery and/or radiotherapy are performed, lymphedema commonly arises unilaterally. A baseline measurement of the limb before the onset of lymphedema is rarely available. Unilateral limb lymphedema is, therefore, often described in terms of the volumetric difference between the affected and the unaffected limb.⁷ The measured differences between the limbs (at risk of or with lymphedema and the contralateral unaffected limb) should be compared with reference values in the population of interest by using the same measurement method. We have previously reported reference values of inter-limb differences measured with DXA.⁷ The objective of this study was, therefore, to investigate the inter-limb differences in volume, fat mass, and lean mass compared with reference values in patients who had undergone unilateral axillary or inguinal SLNB and/or CLND.

Methods

The study participants took part in a cross-sectional study assessing lymphedema and health-related quality of life in melanoma patients.

The inclusion criteria were: confirmed cutaneous melanoma, age ≥18 and ≤75 years, no current regional or metastatic disease, at least 1 year follow-up after wide local excision, and unilateral axillary or inguinal SLNB alone and/or CLND (Fig. 1). The exclusion criteria were: ilio-inguinal surgery, failure of SLNB technique, major surgery or previous lymph node surgery to the ipsilateral limb, lymphedema diagnosed before melanoma surgery or current treatment of any cancer or metastatic disease, inpatient hospital treatment within ±30 days, heart or kidney failure, or being mentally not fit for inclusion in the study.

FIG. 1.

Flow chart of inclusion of study participants. SLNB, sentinel lymph node biopsy.

Data were collected from and cross-checked between the electronic medical notes of collaborating hospitals, the prospectively registered The Danish National Pathology Register, and The Danish Melanoma Database. Participants completed an online survey before inclusion in the study.

The primary outcome variable in this study was the presence of clinical unilateral limb lymphedema.¹⁰ The diagnosis of lymphedema was based on a detailed clinical history and a thorough physical examination.^10,19,20 The history included onset, development of edema, and symptoms such as swelling, discomfort, feeling of tiredness, and pain in the limb. The physical examination was based on inspection and palpation of the limb, focusing on pitting and non-pitting edema and skin changes. If diagnosed with lymphedema, the stage was registered according to the ISL in stage I–III.¹⁰

All participants were examined clinically and scanned by the first author after training by and in accordance with the practice of the lymphedema physiotherapists and the institutional head of DXA (B.Z.).

The DXA scans of both the limb at risk of lymphedema and the contralateral limb were performed by using the same DXA scanner (Lunar Prodigy DXA mini fan beam scanner serial number DF +13189; GE Medical Systems, Madison, WI) as previously described.^7,11,12

The legs were scanned as part of a whole-body scan and analyzed with the software: Prodigy med enCore version 15 (GE Medical Systems). The arms were scanned separately and analyzed with the software: Small Animal Programme version 8.1027, individual analysis version (GE Medical Systems). Regions of interest (ROIs) were drawn manually by the examiner (first author): the upper arm, lower arm, and hand as well as the thigh, lower leg, and foot by using a simple mathematical algorithm and anatomical landmarks.⁷ Data from the DXA scans were exported to Microsoft Excel 2013 by using Prodigy med enCore version 15 Practical Management Tool and imported to SAS, version 9.4 (SAS Institute, Inc., Cary, NC). The data on fat mass and lean mass were given in grams for each ROI. The volume was calculated from the bone mineral content, fat mass, and lean mass by using the density for bone mineral content of 1.65 g/cm³, fat mass of 0.916 g/cm³, and lean mass of 1.04 g/cm³.¹⁴

The sample size calculation was based on the assumptions that ∼75% of patients undergoing SLNB procedure have negative lymph nodes.²¹ Among these, 15% are expected to develop lymphedema.⁶ The remaining 25% of melanoma patients will undergo CLND due to metastasis in the sentinel lymph node(s), of which 30% are expected to develop lymphedema.^6,22 To have a power of 0.90, a significance level of 0.05, and a two-sided 95% confidence interval adjusted for the uneven distribution of the risk of lymphedema in the SLNB-alone group and the CLND group, we planned to include 435 patients. Based on these numbers, it was expected to be possible to include ∼100 patients with lymphedema.

The patient-related variables were region of limb involved, self-reported time of onset, if a diagnosis of lymphedema had been given by a healthcare worker before the study, treatment with compression garment(s), stage of lymphedema using the ISL lymphedema staging system, and type of nodal surgery.¹⁰

The study was approved by the Danish Regional Committee on Biomedical Research Ethics (Journal No. H-4-2014-127), the Danish Data Protection Agency (2012-58-0004, Local Journal No. HEH-2015-003, I-Suite No. 03436), and the Danish Melanoma Group, and it was registered with ClinicalTrials.gov (Identifier No. NCT02352623). All patients gave written informed consent before their participation in the study.

Statistical analysis

The limb at risk of or with lymphedema (hereafter limb of interest) was compared with the contralateral limb, and the inter-limb percentage differences were calculated as: (limb-of-interest–contralateral)/contralateral * 100 and corrected for handedness (for the upper limbs) according to the reference value.⁷ The inter-limb percentage differences were compared with reference values and categorized as none/mild, moderate, or severe.⁷ Stratification was performed according to the limb involved and the location of lymphedema (upper arm, lower arm, hand, thigh, lower leg, and/or foot).

The ordinal association between the categorization of inter-limb percentage differences in the participants with and without lymphedema was tested by using the Kendall rank correlation coefficient, which compared the distribution between the two groups.

Associations between inter-limb percentage differences as dependent variables and independent variables were conducted with linear regression, Student's t-test, and one-way analysis of variance.

Multiple regression analyses with inter-limb percentage differences as dependent variables and significant variables in the univariate analysis were performed if n > 10.

The correlations between the independent variables were tested with Spearman's ρ correlation coefficient for the upper limb and lower limb, respectively.

p values below 0.05 were considered as indicating a significant difference. The statistical analyses were performed by using SAS, version 9.4 (SAS Institute, Inc.).

Results

The baseline characteristics of the study participants and lymphedema are summarized in Tables 1 and 2. Of the 431 patients, 109 (25%) in all had unilateral melanoma-related limb lymphedema corresponding to 23 (10%) who had undergone axillary nodal surgery and 86 (45%) who had undergone inguinal nodal surgery. Noteworthy is that the majority of patients only had a relatively mild stage of lymphedema and with onset within the first year after surgery. However, approximately one in three patients had not been diagnosed with lymphedema by a healthcare worker before the study, even though lymphedema had been reportedly present for an average of 3 years for patients with arm lymphedema and 4 years for patients with leg lymphedema. No patients had received radiotherapy treatment of melanoma or surgical treatment for lymphedema.

Table 1.

Characteristics of the Study Population

		Melanoma-related limb lymphedema
	Total (n = 431) n (%) or mean (SD), range	Upper limb (n = 23) n (%) or mean (SD), range	Lower limb (n = 86) n (%) or mean (SD), range
Gender
Female	224 (52)	13 (57)	57 (66)
Male	207 (48)	10 (43)	29 (34)
Age (years)	58 (12), 22–75	59 (11), 36–75	53 (12), 24–75
BMI (kg/m²)	27 (5), 14–47	29 (4), 22–35	26 (4), 17–37
Lymph node surgery
Axillary SLNB	198 (46)	10 (43)
Axillary CLND	42 (10)	13 (57)
Inguinal SLNB	151 (35)		53 (62)
Inguinal CLND	40 (9)		33 (38)

BMI, body mass index; CLND, complete lymph node dissection; SD, standard deviation; SLNB, sentinel lymph node biopsy.

Table 2.

Characteristics of Melanoma-Related Limb Lymphedema

	Upper limb lymphedema (n = 23) n (%) or mean (SD), range	Lower limb lymphedema (n = 86) n (%) or mean (SD), range
ISL lymphedema stage
1	12 (52)	36 (42)
2	11 (48)	48 (56)
3	0	2 (2)
Self-reported onset of lymphedema
≤1 year after surgery	21 (91)	77 (91)
2–3 years after surgery	1 (4)	6 (7)
4–5 years after surgery	0	1 (1)
>5 years after surgery	1 (4)	1 (1)
Duration of lymphedema, years	3 (3), 0–12	4 (4), 0–17
Diagnosed with lymphedema by a healthcare worker before the study
Yes	16 (70)	54 (63)
No	7 (30)	32 (37)
Use of compression garments
Daily	3 (13)	20 (23)
Sometimes	6 (26)	29 (34)
Never	14 (61)	37 (43)
Location of lymphedema
Upper arm	10 (43)
Lower arm	16 (70)
Hand	7 (30)
Thigh		39 (45)
Lower leg		79 (92)
Foot		32 (37)

ISL, International Society of Lymphology.

The degree of inter-limb percentage differences in the patients without lymphedema were close to expected when compared with our reference material, with the distribution of difference as none/mild (expected: 84%), moderate (expected: 13.5%), and severe (expected: 2.5%) (Table 3).⁷

Table 3.

Volume, Fat Mass, and Lean Mass in Upper (A) and Lower (B) Limbs Without and With Melanoma-Related Limb Lymphedema

	Categorization of the inter-limb percentage differences^a in patients without lymphedema of the upper limb (n = 217)			Categorization of the inter-limb percentage differences^a in patients with lymphedema of the upper limb (n = 23)
A	None/mild (%)	Moderate (%)	Severe (%)	N ^b	None/mild (%)	Moderate (%)	Severe (%)	p^c
Volume
Total arm	81	17	2	23	48	30	22	0.005
Upper arm	80	16	5	10	20	50	30	0.006
Lower arm	86	10	5	16	50	25	25	0.017
Hand	81	14	5	7	71	14	14	0.59
Fat mass
Total arm	80	16	4	23	52	22	26	0.014
Upper arm	77	17	6	10	40	50	10	0.055
Lower arm	81	15	4	16	31	13	56	0.002
Hand	82	12	6	7	57	29	14	0.24
Lean mass
Total arm	85	10	5	23	61	26	13	0.032
Upper arm	85	11	4	10	50	20	30	0.06
Lower arm	83	12	5	16	75	0	25	0.37
Hand	82	13	4	7	86	14	0	0.77

	Categorization of the inter-limb percentage differences^a in patients without lymphedema of the lower limb (n = 105)			Categorization of the inter-limb percentage differences^a in patients with lymphedema of the lower limb (n = 86)^a
B	None/mild (%)	Moderate (%)	Severe (%)	N ^b	None/mild (%)	Moderate (%)	Severe (%)	p^c
Volume
Total leg	90	9	1	86	59	28	13	<0.0001
Thigh	83	14	2	39	49	23	28	<0.0001
Lower leg	74	27	5	79	47	25	28	<0.0001
Foot	79	14	7	32	56	22	22	0.017
Fat mass
Total leg	85	15	0	86	67	31	1	0.005
Thigh	87	13	0	39	54	38	8	0.0004
Lower leg	88	12	0	79	67	32	1	0.001
Foot	89	11	0	32	84	16	0	0.56
Lean mass
Total leg	84	16	0	86	69	26	5	0.013
Thigh	91	8	1	39	80	11	9	0.11
Lower leg	83	17	0	79	67	23	10	0.009
Foot	84	16	0	32	78	22	0	0.49

The inter-limb percentage differences were calculated as (“Limb-of-interest”—contralateral)/(contralateral × 100) and categorized as none/mild, moderate, or severe according to reference value of limbs (taking handedness into account for the upper limbs) by Gjorup et al.⁷ Data are stratified by limb of interest (A, upper limb; B, lower limb) and lymphedema and if present by ^blocation hereof.

p Values were tested with the Kendall rank correlation coefficient.

The patients with upper limb lymphedema had a significantly increased volume of the affected limb, primarily due to a higher fat mass of both upper and lower arms (Table 3-A).

The patients with lower limb lymphedema had increased volumes of the affected limb, which were not only due to an increase in the fat mass of both the thigh and the lower leg but also due to an increase in the lean mass of the lower leg (Table 3-B). The correlations between the independent variables shown in Table 4 tested for the upper and lower limbs, respectively, were inspected for colinearity by using Spearman's ρ correlation. No strong correlation was observed (all ρ < 0.5) (data not shown).

Table 4.

Association Between Inter-Limb Percentage Differences in Volume, Fat Mass, and Lean Mass in Upper Limb Lymphedema (A) and Lower Limb Lymphedema (B) and Patient, Lymphedema, and Treatment Characteristics

	Mean (95% CI)	p	Mean (95% CI)	p	Mean (95% CI)	p	Mean (95% CI)	p
A A1. Upper limb volume differences^a (%)	Total arm (n = 23)		Upper arm (n = 10)		Lower arm (n = 16)		Hand (n = 7)
BMI^b							−2 (−4; −0.1)	0.041
Use of compression garments^c
Daily (reference; no)	8 (2; 15)	0.019	16 (3; 29)	0.026
Known lymphedema diagnosis^d
Yes (reference; no)	6 (0.6; 11)	0.031	15 (4; 26)	0.012			NA	NA
Duration of lymphedema	5 (1; 9)	0.008			10 (3; 17)	0.007
ISL lymphedema stage
Stages 2–3 (reference; stage 1)					11 (1; 22)	0.031
Lymph node surgery
CLND (reference; SLNB)	5 (0.3; 10)	0.037	12 (2; 23)	0.026	12 (1; 22)	0.03

A2. Upper limb fat mass differences (%)	Total arm (n = 23)		Upper arm (n = 10)		Lower arm (n = 16)		Hand (n = 7)
Use of compression garments^e
Sometimes (reference; no)							22 (7; 38)	0.017
Daily (reference; no)	15 (7; 23)	0.001	14 (1; 28)	0.04			28 (6; 50)	0.025
Lymphedema diagnosis
Yes (reference; no)	9 (2; 15)	0.012	15 (6; 25)	0.006	12 (−6; 31)	0.18	NA	NA
Duration of lymphedema	5 (0.2; 11)	0.042
ISL lymphedema stage
Stages 2–3 (reference; stage 1)							24 (10; 37)	0.006
Lymph node surgery	9 (3; 15)	0.004	13 (5; 22)	0.005	19 (7; 31)	0.004
CLND (reference; SLNB)

A3. Upper limb lean mass differences (%)	Total arm (n = 23)		Lower arm (n = 16)
Duration of lymphedema (years)	6 (0.9; 11)	0.022	11 (3; 19)	0.012
ISL lymphedema stage
Stages 2–3 (reference; stage 1)	6 (−0.2; 12)	0.059	13 (0.7; 25)	0.039

	Mean (95% CI)	p	Mean (95% CI)	p	Mean (95% CI)	p	Mean (95% CI)
B B1. Lower limb volume differences (%)	Total leg (n = 86)		Thigh (n = 39)		Lower leg (n = 79)		Foot (n = 32)
Duration of lymphedema					2 (0.6; 4)	0.007
Lymph node surgery
CLND (reference; SLNB)	3 (1; 5)	0.0003	5 (1; 8)	0.005

B2a. Female lower limb fat mass differences (%)	Total leg (n = 57)		Thigh (n = 32)		Lower leg (n = 52)
Lymphedema diagnosis
Yes (reference; no)	5 (0.2; 9)	0.04
Duration of lymphedema					4.9 (0.9; 9.0)	0.017
Lymph node surgery
CLND (reference; SLNB)	7 (4; 11)	0.0002	7 (1; 12)	0.015

B2b. Male lower limb fat mass differences (%)	Total leg (n = 29)		Thigh (n = 7)		Lower leg (n = 27)		Foot (n = 6)
BMI			−8 (−11; −4)	0.004
Lymphedema diagnosis
Yes (reference; no)			11 (2; 20)	0.022
ISL lymphedema stage
Stages 2–3 (reference; stage 1)							40 (8; 70)	0.024
Lymph node surgery
CLND (reference; SLNB)	9 (3; 15)	0.006	12 (3; 22)	0.022	15 (4; 25)	0.009

B3. Lower limb lean mass differences (%)	Total leg (n = 86)		Thigh (n = 39)		Lower leg (n = 79)		Foot (n = 32)
Use of compression garments
Daily	−3 (−5; −0.4)	0.024

The volume, fat mass, and lean mass of the lymphedema limbs are shown as inter-limb percentage differences, which were calculated as: (“Limb-of-interest”—contralateral)/(contralateral × 100) and categorized as none/mild, moderate, or severe according to reference values of limbs (taking handedness into account for the upper limbs) without lymphedema. The association between inter-limb percentage differences in total and regional volume, fat mass, and lean mass as dependent variables and independent variables was tested with linear regression (BMI and duration of lymphedema in years as continuous variables), Student's t-test (dichotomous variables), and one-way analysis of variance (>2 groups) and given as mean inter-limb percentage differences.

Calculated as weight in kilograms divided by height in meters squared.

Use of compression garments classified as none, sometimes, or daily.

Diagnosed with melanoma-related limb lymphedema by a healthcare worker before the study. Only significant results are shown.

CI, confidence interval.

Variables significantly associated with inter-limb percentage differences in volume and tissue composition of the limbs are shown in Table 4-A and -B. For the upper limbs with lymphedema, univariate analyses showed a significantly increased total volume and total fat mass, which were increased in patients who had been diagnosed with lymphedema before the study, used compression garments daily, with an increased duration of lymphedema and with CLND (Table 4-A1–A2). The lean mass was increased in lymphedema in upper limbs with an increased duration of lymphedema and an increase in the ISL stage (Table 4-A3). However, neither body mass index (BMI) nor age was associated with significant inter-limb differences of the volume or tissue composition of the limbs. All patients with lymphedema of the hand were diagnosed with lymphedema before the study (data not shown).

In patients with lower limb lymphedema, the duration of lymphedema was also significantly associated with an increase in total volume; in women also, an increase in total fat mass was observed (Table 4-B1a, B2a). Most noticeable was the significant decreasing fat mass in men with increasing BMI and the increase in fat mass in men with previous diagnosis of both lymphedema and CLND (Table 4-B2b).

Multiple analyses were performed for the upper limb total volume and lower arm volume, total fat mass and lower arm lean mass by using the inter-limb percentage differences as dependent variables (data not shown). The increase in total volume of the upper lymphedema limb was significantly associated with the use of compression garments (occasional 4.6%; 95% confidence interval [CI] 0.1% to 9.0%; p = 0.044 and daily 5.7%; 95% CI −0.2% to 11.7%; p = 0.057 [reference, none]), and with a longer duration of lymphedema (3.6%; 95% CI 0.4% to 6.7%; p = 0.029) and CLND (4.3%; 95% CI 0.4% to 8.3%; p = 0.034 [reference, SLNB]). The significant increase in the volume of the lower arm was only associated with the duration of lymphedema (7.2%; 95% CI 0.6% to 13.9%; p = 0.036) and CLND (8.8%; 95% CI 0.2% to 17.4%; p = 0.045 [reference, SLNB]).

Finally, multiple analysis showed that a significant increase in total fat mass of the upper limb was associated with daily use of compression garments (12.3%; 95% CI 5.9% to 18.6%; p = 0.0008 [reference, none]) and CLND (7.6%; 95% CI 3.4% to 11.9%; p = 0.002 [reference, SLNB]). For the lower limbs, too few variables in each subgroup were significant; hence, multiple analyses were not conducted.

Discussion

The clinical examination of lymphedema after nodal surgery is a straightforward method for identifying patients with limb lymphedema.¹⁰ This is supported by the present study, where volume and tissue composition of the limbs presented by inter-limb percentage differences measured with DXA were compared with reference values in which the categorization in patients without lymphedema was close to the expected.⁷

In contrast to most prior studies, the prevalence of limb lymphedema after nodal surgery was high. Melanoma-related limb lymphedema was present in 5% and 31% of patients after axillary SNLB and CLND, respectively, and in 35% and 83% of patients after inguinal SLNB and CLND, respectively. However, lymphedema was categorized as mild or moderate for the vast majority of patients. De Vries et al. defined upper limb lymphedema after axillary nodal surgery as an increase in volume of ≥10% of the affected limb compared with the contralateral unaffected limb measured with water displacement.²³ They similarly classified lower limb lymphedema after nodal surgery as an increase in volume of ≥6.5% of the affected limb compared with the contralateral unaffected limb measured with water displacement.²⁴ Starritt et al. suggested an objective definition for upper limb lymphedema after axillary dissection as a volume increase of >16% of the volume of the control limb, adjusted for handedness, measured with water displacement.⁴ Spillane et al.³ defined lower limb lymphedema after inguinal or ilio-inguinal dissection as a perometer volume percentage increase of ≥15% of the volume of the control limb.

The diagnosis of lymphedema is straightforward clinically, and volume and tissue measurements should be used in describing lymphedema. This is reflected by our findings: The volume of the upper limb with lymphedema was only 10% and 16% greater than the contralateral limb in 3 (13%) and 2 (9%) patients, respectively (data not shown); the volume of the lower limb with lymphedema was only 6.5% and 15% greater than the contralateral limb in 11 (13%) and 1 (1%) patients, respectively (data not shown). These findings support the higher sensitivity of the clinical diagnosis of lymphedema as compared with the diagnosis of lymphedema using volume differences. The use of cut-off point volume differences for the diagnosis of lymphedema is questionable and will, as shown in our study, underestimate the incidence of lymphedema. Since early detection of lymphedema is important, it is important to inform the patients about management and care of their affected limb, which can possibly delay progression, clinical diagnosis of lymphedema appears to be of more use than strict volume measurements. The low inter-limb volume differences in the lymphedema patients reported here could partly be explained by the treatment of lymphedema with compression garments, which were worn sometimes or daily by half of the patients with lymphedema. If worn on the day of the study, compression garments were removed ∼30 minutes before the DXA scan. However, in the patients with upper limb lymphedema who never used compression garments, no total volume increase reached the earlier mentioned definitions of upper limb lymphedema, respectively. In patients with lower limb lymphedema, 4 (11%) patients had total volumes >6.5% but no total volume increases were >15%.

The prevention of progression of lymphedema in the patients who received conservative treatment with compression is plausible. Our data showed the use of compression garments to be significantly associated with both the volume and the fat mass increase in the upper total limb as well as with the fat mass of the hand in multiple analyses. This is possibly due to a better patient compliance in patients with increased volume in their lymphedema limb. However, our data do not allow for a causal explanation hereof.

In the lower limb, the increase in fat mass was observed for both the thigh and the lower leg. In men, this was significantly associated with low BMI. This is possibly attributed to the relatively small amount of fat mass in the lower limbs in men with low BMI, in which a small increase in fat mass will represent a significant increase in the inter-limb difference expressed as a percentage.

In lymphedema, preventive measures such as exercise and avoiding aggravating factors could possibly alleviate symptoms.²⁵ Further, strict adherence to conservative treatment with compression garments might offset the progression of lymphedema.^15,25 A large proportion of patients with limb lymphedema report unmet needs regarding both information and support.^26–28

Information about lymphedema enables the patients to take responsibility for their condition as well as to communicate their needs. It is, therefore, of utmost importance to inform the patients of both the risk, and symptoms, and preventative measures of lymphedema.

Further studies are needed on these matters, evaluating the volume and tissue changes over time in patients undergoing surgery at risk of causing lymphedema, and examining the effect of preventive measures such as exercise and treatment, including use of conservative treatment as well as microsurgery.

Accurate measurement of changes in volume and tissue composition is fundamental to identify patients at the highest risk of development and progression of lymphedema as well as to select and monitor appropriate treatment of lymphedema.

The development of the treatment of melanoma with less invasive surgery and of lymphedema with microsurgery is evolving. The inflammatory process causing edema and increase in fat mass is evident in lymphedema.^9,18 Lymphedema can be diagnosed clinically, and the diagnosis of lymphedema should not be missed. Our data indicate that lymphedema affects patients in a non-uniform way, and this should be reflected in personalized information and treatment of patients with lymphedema.

Conclusion

The volume and tissue composition measured with DXA in both upper and lower melanoma-related limb lymphedema is significantly increased compared with the contralateral limb. The volumetric increases were, in particular, attributed to an increase in fat mass. However, the changes were for the majority mild compared with reference values. This study highlights the pathophysiology of lymphedema with an increase in both fat mass and lean mass. This knowledge is of importance in the research related to the prevention and treatment of lymphedema.

Footnotes

Acknowledgments

This study was supported by the Department of Plastic Surgery, the Department of Clinical Physiology and Nuclear Medicine, and The Research Council at Herlev and Gentofte Hospital, University of Copenhagen, by the private funds of Toemmerhandler Johannes Fog and Civilingenioer Bent Boegh og hustru Inge Boegh, and by The Danish Cancer Society, Denmark.

Author Disclosure Statement

No competing financial interests exist.

References

Keast

, Despatis

, Allen

, Brassard

. Chronic oedema/lymphoedema: Under-recognised and under-treated. Int Wound J, 2015; 12:328–333.

Moody

, Ali

, Carbone

, Singh

, Hardwicke

. Complications of sentinel lymph node biopsy for melanoma—A systematic review of the literature. Eur J Surg Oncol, 2017; 43:270–277.

Spillane

, Saw

RPM

, Tucker

, Byth

, Thompson

. Defining lower limb lymphedema after inguinal or ilio-inguinal dissection in patients with melanoma using classification and regression tree analysis. Ann Surg, 2008; 248:286–293.

Starritt

, Joseph

, McKinnon

, Lo

, de Wilt

JHW

, Thompson

. Lymphedema after complete axillary node dissection for melanoma. Ann Surg, 2004; 240:866–874.

Ul-Mulk

, Hölmich

. Lymph node dissection in patients with malignant melanoma is associated with high risk of morbidity. Dan Med J, 2012; 59:A4441.

Hyngstrom

, Chiang

Y-J

, Cromwell

, et al. Prospective assessment of lymphedema incidence and lymphedema-associated symptoms following lymph node surgery for melanoma. Melanoma Res, 2013; 23:1.

Gjorup

, Hendel

, Klausen

, Zerahn

, Hölmich

. Reference values for assessment of unilateral limb lymphedema with dual-energy X-ray absorptiometry. Lymphat Res Biol, 2017.

Warren

, Brorson

, Borud

, Slavin

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

, Kataru

, Mehrara

. Inflammatory manifestations of lymphedema. Int J Mol Sci, 2017; 18:1–13.

10.

Lymphoedema Framework.. Best Practice for the Management of Lymphoedema. International Consensus. London: MEP Ltd, 2006.

11.

Gjorup

, Zerahn

, Hendel

. Assessment of volume measurement of breast cancer-related lymphedema by three methods: Circumference measurement, water displacement, and dual energy X-ray absorptiometry. Lymphat Res Biol, 2010; 8:111–119.

12.

Gjorup

, Zerahn

, Juul

, Hendel

, Christensen

, Hölmich

. Repeatability of volume and regional body composition measurements of the lower limb using dual-energy X-ray absorptiometry. J Clin Densitom, 2016; 20:82–96.

13.

Fosbøl

MØ

, Zerahn

. Contemporary methods of body composition measurement. Clin Physiol Funct Imaging, 2015; 35:81–97.

14.

International Atomic Energy Agency.. Dual Energy X Ray Absorptiometry for Bone Mineral Density and Body Composition Assessment. IAEA Human Health Series. Austria: IAEA, 2010.

15.

Melam

, Buragadda

, Alhusaini

, Arora

. Effect of complete decongestive therapy and a home program for patients with post mastectomy lymphedema. J Phys Ther Sci, 2016; 27:2743–2748.

16.

Koshima

, Narushima

, Yamamoto

, Mihara

, Iida

. Recent advancement on surgical treatments for lymphedema. Ann Vasc Dis, 2012; 5:409–415.

17.

Boyages

, Kastanias

, Koelmeyer

, et al. Liposuction for advanced lymphedema: A multidisciplinary approach for complete reduction of arm and leg swelling. Ann Surg Oncol, 2015; 22(Suppl 3):S1263–S1270.

18.

Brorson

, Ohlin

, Olsson

, Karlsson

. Breast cancer-related chronic arm lymphedema is associated with excess adipose and muscle tissue. Lymphat Res Biol, 2009; 7:3–10.

19.

O'Donnell

, Rasmussen

, Sevick-Muraca

. New diagnostic modalities in the evaluation of lymphedema. J Vasc Surg Venous Lymphat Disord, 2017; 5:261–273.

20.

Lee

, Andrade

, Antignani

, et al. IUA-ISVI consensus for diagnosis guideline of chronic lymphedema of the limbs. Int Angiol, 2015; 34:311–332.

21.

The Danish Melanoma Database.. Annual quality report. 2014. The Danish Melanoma Group. 2015. www.melanoma.dk.

22.

Balch

, Gershenwald

, Soong

, et al. Final version of 2009 AJCC melanoma staging and classification. J Clin Oncol, 2009; 27:6199–6206.

23.

de Vries

, Vonkeman

, van Ginkel

, Hoekstra

. Morbidity after axillary sentinel lymph node biopsy in patients with cutaneous melanoma. Eur J Surg Oncol, 2005; 31:778–783.

24.

de Vries

, Vonkeman

, van Ginkel

, Hoekstra

. Morbidity after inguinal sentinel lymph node biopsy and completion lymph node dissection in patients with cutaneous melanoma. Eur J Surg Oncol, 2006; 32:785–789.

25.

Ammitzbøll

, Lanng

, Kroman

, et al. Progressive strength training to prevent lymphoedema in the first year after breast cancer—the LYCA feasibility study. Acta Oncol, 2017; 56:360–366.

26.

Girgis

, Stacey

, Lee

, Black

, Kilbreath

. Priorities for women with lymphoedema after treatment for breast cancer: Population based cohort study. BMJ, 2011; 342:d3442.

27.

Beesley

, Smithers

, Khosrotehrani

, et al. Supportive care needs, anxiety, depression and quality of life amongst newly diagnosed patients with localised invasive cutaneous melanoma in Queensland, Australia. Psychooncology, 2015; 24:763–770.

28.

Beesley

, Rowlands

, Hayes

, et al. Incidence, risk factors and estimates of a woman's risk of developing secondary lower limb lymphedema and lymphedema-specific supportive care needs in women treated for endometrial cancer. Gynecol Oncol, 2015; 136:87–93.