This cohort study examines the association between race and ethnicity and long-term quality of life after unilateral axillary lymph node dissection in women with breast cancer.
Key Points
Question
Are race and ethnicity associated with long-term quality of life (QOL) after axillary lymph node dissection (ALND) in women with breast cancer?
Findings
In this cohort study of 281 women with breast cancer followed up for a median of 2.97 years after unilateral ALND, QOL assessment using a validated lymphedema questionnaire revealed that racial and ethnic minority women and those with public medical insurance experienced worse physical QOL than White women or women with private insurance, even after adjusting for baseline questionnaire scores and clinicopathological variables.
Meaning
Findings of this study suggest that racial and ethnic minority status and public medical insurance are associated with worse physical QOL for women with breast cancer after unilateral ALND; understanding such disparities may enable targeted interventions to improve QOL for these patients.
Abstract
Importance
Higher lymphedema rates after axillary lymph node dissection (ALND) have been found in Black and Hispanic women; however, there is poor correlation between subjective symptoms, quality of life (QOL), and measured lymphedema. Additionally, racial and ethnic differences in QOL have been understudied.
Objective
To evaluate the association of race and ethnicity with long-term QOL in patients with breast cancer treated with ALND.
Design, Setting, and Participants
This cohort study enrolled women aged 18 years and older with breast cancer who underwent unilateral ALND at a tertiary cancer center between November 2016 and March 2020. Preoperatively and at 6-month intervals, arm volume was measured by perometer and QOL was assessed using the Upper Limb Lymphedema-27 (ULL-27) questionnaire, a validated tool for assessing lymphedema that evaluates how arm symptoms affect physical, psychological, and social functioning. Data were analyzed from November 2016 to October 2023.
Exposures
Breast surgery and unilateral ALND in the primary setting or after sentinel lymph node biopsy.
Main Outcomes and Measures
Scores in each domain of the ULL-27 were compared by race and ethnicity. Factors impacting QOL were identified using multivariable regression analyses.
Results
The study included 281 women (median [IQR] age, 48 [41-58] years) with breast cancer who underwent unilateral ALND and had at least 6 months of follow-up. Of these, 30 patients (11%) self-identified as Asian individuals, 57 (20%) as Black individuals, 23 (8%) as Hispanic individuals, and 162 (58%) as White individuals; 9 individuals (3%) who did not identify as part of a particular group or who were missing race and ethnicity data were categorized as having unknown race and ethnicity. Median (IQR) follow-up was 2.97 (1.96-3.67) years. The overall 2-year lymphedema rate was 20% and was higher among Black (31%) and Hispanic (27%) women compared with Asian (15%) and White (17%) women (P = .04). Subjective arm swelling was more common among Asian (57%), Black (70%), and Hispanic (87%) women than White (44%) women (P < .001), and lower physical QOL scores were reported by racial and ethnic minority women at nearly every follow-up. For example, at 24 months, median QOL scores were 87, 79, and 80 for Asian, Black, and Hispanic women compared with 92 for White women (P = .003). On multivariable analysis, Asian race (β = –5.7; 95% CI, −9.5 to −1.8), Hispanic ethnicity (β = –10.0; 95% CI, −15.0 to −5.2), and having Medicaid (β = −5.4; 95% CI, −9.2 to −1.7) or Medicare insurance (β = −6.9; 95% CI, −10.0 to −3.4) were independently associated with worse physical QOL (all P < .001).
Conclusions and Relevance
Findings of this cohort study suggest that Asian, Black, and Hispanic women experience more subjective arm swelling after unilateral ALND for breast cancer compared with White women. Black and Hispanic women had higher rates of objective lymphedema than their White counterparts. Both minority status and public medical insurance were associated with worse physical QOL. Understanding disparities in QOL after ALND is an unmet need and may enable targeted interventions to improve QOL for these patients.
Introduction
Despite the trend toward less extensive axillary surgery for breast cancer in the past quarter century, axillary lymph node dissection (ALND) remains necessary for a substantial proportion of patients with advanced nodal disease. Consequences of ALND include pain and paresthesias in the extremity, impaired strength, reduced mobility, and a long-term risk of breast cancer–related lymphedema that ranges from 20% to 30%.1,2,3,4,5 While the extent of nodal surgery remains the most important factor in lymphedema development, race and ethnicity are also associated with higher lymphedema incidence6; a recent prospective study from our institution found a more than 3-fold increased risk of lymphedema in Black and Hispanic women after ALND compared with White women.5
In addition to lymphedema risk, patients treated with ALND may have decreased quality of life (QOL), which may be independent of objective increases in arm volume.7,8,9,10,11 Multiple studies have shown poor correlation between subjective, or perceived, arm swelling and measured lymphedema, with a substantial proportion of patients reporting arm symptoms after ALND even in the absence of visible swelling.6,10,12 While a growing body of literature has revealed racial disparities in lymphedema development,5,13,14 there are limited data on racial differences in subjective arm symptoms and limb-specific QOL after ALND. Here we prospectively assess long-term health-related QOL and, specifically, racial disparities in QOL after unilateral ALND in a diverse cohort of women using a validated lymphedema-specific questionnaire (Upper Limb Lymphedema-27 [ULL-27] scale).15
Methods
Study Protocol and Data Source
Between November 2016 and March 2020, women aged 18 years or older with invasive breast cancer who underwent breast surgery and unilateral ALND in the primary setting or after sentinel lymph node biopsy at Memorial Sloan Kettering Cancer Center were enrolled into a prospective lymphedema screening study.16 This cohort study was approved by the Memorial Sloan Kettering Cancer Center Institutional Review Board; all patients gave written informed consent to participate. Prospective and retrospective data were entered into a Health Insurance Portability and Accountability Act–compliant database. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for reporting observational studies.
Volumetric arm measurements using a perometer (Perometer 350 NT; Pero-System), body mass index (BMI; defined as weight in kilograms divided by height in meters squared) measurements, and QOL assessments using the ULL-27 were obtained at preoperative baseline and longitudinally every 6 months after ALND for up to 36 months; this time frame was chosen because most cases of lymphedema are diagnosed within 3 years after surgery.3,17 Baseline surveys were administered before surgery in both the upfront surgery and neoadjuvant chemotherapy settings, with a median (IQR) time from survey completion to surgery of 8 (2-20) days.
Lymphedema was defined as a relative volume change (RVC) of 10% or greater from baseline using the following formula: RVC = [(A2U1)/(U2A1)] – 1, where A and U represent the volumes of affected and unaffected arms, respectively, at baseline (time 1) and at postoperative follow-up (time 2).18
Patient-reported QOL was measured using the ULL-27, a validated lymphedema-specific survey tool that assesses QOL in terms of physical, psychological, and social domains of functioning (eTable 1 in Supplement 1).15 The survey comprises 27 items across these 3 domains that evaluate the patient’s symptoms secondary to the affected extremity, with a 4-week look-back period. Each item is rated on a 5-point Likert scale; a score of 100 signifies no impairment, while a lower score reflects progressively worse QOL in a particular domain. The physical domain of the ULL-27 includes questions about arm symptoms (including swelling, heaviness, and tingling), skin texture (eg, whether the skin is tense or hard), and whether symptoms in the arm interfere with daily activities. The psychological and social domains assess the effect of arm symptoms on mood (including sadness, distress, and anger) and social activities (including going out to a restaurant or theater, or interactions with spouse or partner), respectively.
Data on subjective arm swelling were abstracted from the physical domain of the ULL-27 questions 3 and 4, which query the frequency of arm swelling reported by the patient (eTable 1 in Supplement 1). A response score of 3 or higher, corresponding to a symptom frequency of sometimes, often, or always, at any time point 6 or more months after surgery was recorded as subjective swelling. Lymphedema therapy referral occurred if patients had an RVC of 10% or greater or if the patient had subjective symptoms of arm swelling and requested a referral. Among those referred, treatment uptake in patients who attended at least 1 lymphedema therapy session was assessed.
Clinicopathological and demographic information were obtained by retrospective medical record review. Patients were grouped by self-reported race and ethnicity as Asian, Hispanic, non-Hispanic Black (hereafter, Black), and non-Hispanic White (hereafter, White). Individuals with undefined ethnicity were classified by race, while those who could not be categorized into a particular group or who were missing both race and ethnicity data were categorized as having other or unknown race and ethnicity.
Statistical Analysis
Clinicopathological and demographic characteristics were described for the overall cohort and compared by race and ethnicity using the Wilcoxon rank sum test for continuous variables and the Fisher exact test for categorical variables. Physical, psychological, and social QOL scores were reported at each time point using medians and IQRs and were compared by race and ethnicity after correcting for multiple comparisons using the Benjamini-Hochberg procedure, which controls for the false-discovery rate (rejecting the null hypothesis when it is in fact true) in the context of multiple hypothesis testing. Patterns of change in scores over time were graphed on box plots, constructed using the ggplot functionality in R, version 3.4.2 (R Project for Statistical Computing). Cumulative incidence of lymphedema was assessed using competing risk analysis, with death, locoregional recurrence, and contralateral breast surgery considered as competing events. The Gray test was used to compare differences in lymphedema rates by race and ethnicity.
Univariable and multivariable linear regression analyses were conducted in the generalized estimating equation framework to evaluate the association between clinicopathological characteristics and QOL scores. The QOL scores were assumed to follow a gaussian distribution; therefore, an identity link function was used in the generalized estimating equation model. The correlation structure was estimated to be that of independence, which was chosen based on assessment of the quasilikelihood information criterion. Variables that were significant in the univariable analysis at a type I error rate (α) of .05 were included in the multivariable analysis. Because QOL scores were analyzed in 3 separate domains, we set the permissible α = .0167 after Bonferroni correction in the multivariable analysis. Variation inflation factors were assessed to detect collinearity in the multivariable model. However, none of the variation inflation factors was significantly high (all <5). Additionally, given the small size of the other or unknown race and ethnicity category, we performed a sensitivity analysis by combining this category with the Asian race category. No changes in the univariable analyses or multivariable analyses for the 3 domains were observed, and therefore the original analyses were reported. All statistical analysis was done using R, version 3.4.2, and data were analyzed from November 2016 to October 2023. A P < .05 was considered statistically significant in all cases except for the multivariable analysis, which was conducted in 3 separate domains of QOL and for which a P < .0167 was considered significant after Bonferroni correction.
Results
Clinicopathological Characteristics
Between November 2016 and March 2020, 304 patients with breast cancer treated with unilateral ALND were enrolled in the study. Of these, 281 women with a baseline measurement and at least 6 months of follow-up were included. Median (IQR) patient age was 48 (41-58) years; median (IQR) BMI was 26 (23-31). Overall, 11% of patients (n = 30) self-identified as Asian, 20% (n = 57) as Black, 8% (n = 23) as Hispanic, 58% (n = 162) as White, and 3% (n = 9) as other or unknown race or ethnicity. Table 1 summarizes clinicopathological characteristics of the cohort stratified by race and ethnicity. Compared with Asian and White women, Black and Hispanic women had a higher baseline BMI (median [IQR], 30 [26-34] and 28 [23-32] vs 23 [21-29] and 26 [22-30], respectively; P < .001), higher clinical nodal stage at presentation (clinical nodal positivity, 83% and 87% vs 60% and 72%, respectively; P = .03), and higher rates of objective lymphedema (40% and 39% vs 17% and 21%; P = .02). Additionally, Black women were less likely than other racial or ethnic groups to have private insurance (46% vs ≥70% for the other 3 groups; P < .001). Otherwise, the type of surgery, nodal radiotherapy receipt, and neoadjuvant chemotherapy or adjuvant systemic therapy receipt did not differ by race and ethnicity.
Table 1. Clinicopathological Features of the Study Cohort, Stratified by Race and Ethnicity.
| Characteristic | No. (%) | P valuea | |||||
|---|---|---|---|---|---|---|---|
| All (N = 281) | Asian (n = 30) | Black (n = 57) | Hispanic (n = 23) | White (n = 162) | Other or unknown (n = 9) | ||
| Age, median (IQR), y | 48 (41-58) | 45 (36-52) | 49 (42-54) | 45 (37-51) | 49 (41-60) | 48 (41-58) | .03 |
| Baseline BMI, median (IQR) | 26 (23-31) | 23 (21-29) | 30 (26-34) | 28 (23-32) | 26 (22-30) | 25 (23-30) | <.001 |
| ALND affected dominant arm | 128 (46) | 12 (40) | 26 (46) | 12 (52) | 74 (46) | 4 (44) | .94 |
| Health insurance | |||||||
| Private | 186 (66) | 24 (80) | 26 (46) | 16 (70) | 115 (71) | 5 (56) | <.001 |
| Medicaid | 24 (9) | 4 (13) | 9 (16) | 4 (17) | 5 (3) | 2 (22) | |
| Medicare | 63 (22) | 2 (7) | 20 (35) | 1 (4) | 40 (25) | 0 | |
| Unknown or uninsured | 8 (3) | 0 | 2 (3) | 2 (9) | 2 (1) | 2 (22) | |
| Clinical T stage | |||||||
| 0 | 5 (2) | 0 | 2 (4) | 0 | 2 (1) | 1 (11) | .26 |
| 1 | 59 (21) | 7 (23) | 10 (18) | 7 (30) | 33 (20) | 2 (22) | |
| 2 | 127 (45) | 19 (63) | 28 (49) | 11 (48) | 67 (41) | 2 (22) | |
| 3 | 46 (16) | 1 (3) | 8 (14) | 2 (9) | 33 (20) | 2 (22) | |
| 4 | 44 (16) | 3 (10) | 9 (16) | 3 (13) | 27 (17) | 2 (22) | |
| Clinical N stage | |||||||
| 0 | 75 (27) | 12 (40) | 10 (18) | 3 (13) | 47 (29) | 3 (33) | .03 |
| 1 | 183 (65) | 15 (50) | 46 (81) | 16 (70) | 102 (63) | 4 (44) | |
| 2 | 6 (2) | 1 (3) | 0 | 1 (4) | 4 (3) | 0 | |
| 3 | 17 (6) | 2 (7) | 1 (2) | 3 (13) | 9 (6) | 2 (22) | |
| Histology | |||||||
| Ductal | 235 (84) | 28 (93) | 50 (88) | 18 (78) | 130 (80) | 9 (100) | .43 |
| Lobular or mixed | 38 (14) | 1 (3) | 6 (11) | 5 (22) | 26 (16) | 0 | |
| Otherb | 8 (3) | 1 (3) | 1 (2) | 0 | 6 (4) | 0 | |
| Receptor status | |||||||
| HR positive, ERBB2 negative | 190 (68) | 23 (77) | 35 (61) | 16 (70) | 109 (67) | 7 (78) | .25 |
| ERBB2 positive | 55 (20) | 5 (17) | 9 (16) | 5 (22) | 36 (22) | 0 | |
| HR negative, ERBB2 negative | 36 (13) | 2 (7) | 13 (23) | 2 (9) | 17 (10) | 2 (22) | |
| Surgery | |||||||
| BCS | 70 (25) | 4 (13) | 20 (35) | 9 (39) | 34 (21) | 3 (33) | .24 |
| Mastectomy | |||||||
| With reconstruction | 157 (56) | 18 (60) | 28 (49) | 11 (48) | 95 (59) | 5 (56) | |
| Without reconstruction | 54 (19) | 8 (27) | 9 (16) | 3 (13) | 33 (20) | 1 (11) | |
| No. of LNs removed, median (IQR) | 18 (14-23) | 15 (12-22) | 17 (13-22) | 20 (13-24) | 19 (14-23) | 15 (12-17) | .05 |
| No. of positive LNs, median (IQR) | 2 (1-5) | 3 (1-4) | 2 (1-3) | 2 (1-4) | 3 (1-7) | 2 (2-4) | .16 |
| Radiation, whole breast or chest wall | 265 (94) | 27 (90) | 56 (98) | 23 (100) | 151 (93) | 8 (89) | .22 |
| Regional nodal radiation | 260 (93) | 26 (87) | 56 (98) | 23 (100) | 147 (91) | 8 (89) | .09 |
| Neoadjuvant chemotherapy | 196 (70) | 18 (60) | 42 (74) | 19 (83) | 111 (69) | 6 (67) | .44 |
| Adjuvant capecitabine | 68 (24) | 8 (27) | 18 (32) | 5 (22) | 35 (22) | 2 (22) | .64 |
| Adjuvant ERBB2 therapy | 55 (20) | 5 (17) | 9 (16) | 5 (22) | 36 (22) | 0 | .52 |
| Endocrine therapy | 206 (73) | 27 (90) | 39 (68) | 19 (83) | 115 (71) | 6 (67) | .13 |
| Lymphedema diagnosis | 73 (26) | 5 (17) | 23 (40) | 9 (39) | 34 (21) | 2 (22) | .02 |
Abbreviations: ALND, axillary lymph node dissection; BCS, breast-conserving surgery; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); HR, hormone receptor; LN, lymph node.
P value reflects comparisons between the groups with significance set at .05; determined by Kruskal-Wallis rank sum or Fisher exact test.
Other histology includes mammary (n = 5), occult primary (n = 2), and high-grade carcinoma not otherwise specified (n = 1).
Lymphedema Incidence
At a median (IQR) follow-up of 2.97 (1.96-3.67) years, 73 patients (26%) developed lymphedema with a 2-year rate of 20% (95% CI, 15%-25%). Lymphedema incidence at 2 years was highest among Black (31% [95% CI, 20%-44%]) and Hispanic (27% [95% CI, 11%-47%]) women compared with Asian (15% [95% CI, 5%-31%]) and White women (17% [95% CI, 12%-24%]; P = .04) (Figure 1).
Figure 1. Cumulative Incidence of Lymphedema After Unilateral Axillary Lymph Node Dissection in Patients With Breast Cancer, Stratified by Race and Ethnicity.
Vertical line represents 2-year follow-up.
Subjective Arm Swelling and Lymphedema Therapy Referral
Overall, 154 of 281 women (55%) reported subjective symptoms of arm swelling at 6 months or more after surgery. Symptoms were more common among Asian (57%), Black (70%), and Hispanic (87%) women than White women (44%; P < .001). Lymphedema referral rates were high across the cohort (n = 196 [70%]) and did not differ by race or ethnicity. Among those referred, treatment uptake was highest among Black and Hispanic women (93% and 94% vs 83% for Asian women and 80% for White women; P = .03) (Table 2).
Table 2. Frequency of Subjective Arm Swelling, Referral for Lymphedema Therapy, and Uptake of Lymphedema Therapy Among Patients Referred, Stratified by Race and Ethnicity.
| Characteristic | No. (%) | P valuea | ||||
|---|---|---|---|---|---|---|
| Overall (N = 281) | Asian (n = 30) | Black (n = 57) | Hispanic (n = 23) | White (n = 162) | ||
| Subjective swelling | 154 (55) | 17 (57) | 40 (70) | 20 (87) | 72 (44) | <.001 |
| Referred for lymphedema therapy | 196 (70) | 23 (77) | 43 (75) | 17 (74) | 105 (65) | .32 |
| Received lymphedema therapyb | 163 (83) | 19 (83) | 40 (93) | 16 (94) | 84 (80) | .03 |
Fisher exact test; P value reflects comparisons between the groups with significance set at .05.
Denominator includes only those patients referred for therapy.
Trajectory of QOL Scores, Stratified by Race and Ethnicity
Physical Domain of the ULL-27
For the overall cohort, physical QOL decreased at 6 months after ALND and remained below baseline levels through 3 years of follow-up (Figure 2A). When stratified by race and ethnicity, Asian, Black, and Hispanic women reported lower physical QOL scores than White women at nearly all time points after ALND (Figure 2B). For example, at 24 months, median (IQR) QOL scores were 87 (75-95) for Asian, 79 (69-89) for Black, and 80 (61-84) for Hispanic women compared with 92 (77-98) for White women (P = .003). At 30 and 36 months, more than 80% of survey response data were missing due to attrition, limiting data interpretation.
Figure 2. Trajectory of Physical Quality of Life (QOL) Scores Over Time.
Horizontal lines represent median scores and dots represent outliers. The upper and lower bounds of the boxes represent the 75th and 25th percentiles of the IQR, respectively. ULL-27 indicates Upper Limb Lymphedema-27 questionnaire; whiskers, minimum and maximum scores (excluding outliers).
Psychological and Social Domains of the ULL-27
Psychological QOL for the cohort initially decreased at 6 months, then recovered almost to baseline levels by 2 years after ALND (eFigure 1A in Supplement 1) and did not differ among racial or ethnic groups at any time point (eFigure 1B in Supplement 1). Social QOL did not measurably decrease from baseline through 3 years of follow-up (eFigure 2A in Supplement 1) and, similarly, did not differ among racial or ethnic groups (eFigure 2B in Supplement 1).
Factors Associated With QOL After ALND
Physical Domain of the ULL-27
Factors associated with physical QOL on univariable analysis and multivariable analysis are listed in Table 3. On multivariable analysis, after adjusting for baseline scores, Asian race (β = –5.7; 95% CI, −9.5 to −1.8) and Hispanic ethnicity (β = –10.0; 95% CI, −15.0 to −5.2) were independently associated with worse QOL compared with White race (reference) (P < .001), while Black race was not (β = –3.2; 95% CI, −6.8 to 0.4). In addition, worse QOL was associated with having Medicaid (β = −5.4; 95% CI, −9.2 to −1.7) and Medicare health insurance (β = –6.9; 95% CI, −10.0 to −3.4) vs private insurance (reference; P < .001). Other factors associated with worse physical QOL included surgery on the dominant arm, clinical N3 disease, and receipt of adjuvant ERBB2 (formerly HER2)-targeted therapy (Table 3). While lymphedema therapy referral was also associated with worse QOL, this was likely a reflection of increased upper extremity symptoms that necessitated referral. A lymphedema diagnosis was not associated with worse physical QOL.
Table 3. Univariable and Multivariable Analyses of Factors Associated With Physical QOL.
| Characteristic | Univariable analysis | Multivariable analysis | ||
|---|---|---|---|---|
| β (95% CI) | P value | β (95% CI) | P valuea | |
| Baseline QOL score | 0.61 (0.48 to 0.74) | <.001 | 0.49 (0.36 to 0.63) | <.001 |
| Time point after ALND, mo | ||||
| 6 | 1 [Reference] | .23 | NA | NA |
| 12 | 3.7 (−0.2 to 7.7) | |||
| 18 | 3.1 (−1.2 to 7.4) | |||
| 24 | 4.6 (0.6 to 8.5) | |||
| 30 | −1.0 (−9.0 to 6.9) | |||
| 36 | 3.7 (−2.4 to 9.9) | |||
| Age | −0.04 (−0.16 to 0.08) | .52 | NA | NA |
| Race and ethnicity | ||||
| Asian | −5.1 (−9.5 to −0.7) | <.001 | −5.7 (−9.5 to −1.8) | <.001 |
| Black | −10.0 (−14.0 to −6.5) | −3.2 (−6.8 to 0.4) | ||
| Hispanic | −13.0 (−19.0 to −8.2) | −10.0 (−15.0 to −5.2) | ||
| White | 1 [Reference] | 1 [Reference] | ||
| Other or unknown | −8.9 (−16.0 to −1.6) | −4.3 (−11.0 to 2.0) | ||
| Health insurance status | ||||
| Private | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Medicaid | −10.0 (−15.0 to −5.7) | −5.4 (−9.2 to −1.7) | ||
| Medicare | −8.6 (−12.0 to −4.8) | −6.9 (−10.0 to −3.4) | ||
| Unknown or uninsured | 1.2 (−5.2 to 7.5) | −2.4 (−7.8 to 3.0) | ||
| Baseline BMI | −0.66 (−0.91 to −0.40) | <.001 | −0.22 (−0.46 to 0.01) | .06 |
| Change in BMI | 0.07 (−1.20 to 1.40) | .92 | NA | NA |
| Surgery on dominant arm | ||||
| No | 1 [Reference] | .002 | 1 [Reference] | .004 |
| Yes | −4.2 (−6.9 to −1.5) | −3.7 (−6.2 to −1.2) | ||
| Clinical T stage | ||||
| 0 | 2.1 (−5.3 to 9.5) | .38 | NA | NA |
| 1 | 1 [Reference] | |||
| 2 | 2.1 (−1.5 to 5.7) | |||
| 3 | 2.0 (−2.4 to 6.4) | |||
| 4 | −1.6 (−6.3 to 3.0) | |||
| Clinical N stage | ||||
| 0 | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| 1 | −5.8 (−8.6 to −2.9) | −2.9 (−6.6 to 0.8) | ||
| 2 | −11.0 (−21.0 to −1.6) | −7.0 (−17.0 to 2.6) | ||
| 3 | −17.0 (−24.0 to −10.0) | −15.0 (−21.0 to −7.8) | ||
| Type of breast surgery | ||||
| BCS | 1 [Reference] | <.001 | 1 [Reference] | .72 |
| Mastectomy | ||||
| With reconstruction | 6.7 (3.3 to 10.0) | 1.3 (−2.0 to 4.6) | ||
| Without reconstruction | 0.6 (−3.8 to 4.9) | 0.4 (−3.7 to 4.5) | ||
| No. of LNs removed | −0.09 (−0.25 to 0.07) | .29 | NA | NA |
| No. of positive LNs | 0.07 (−0.13 to 0.26) | .50 | NA | NA |
| Radiotherapy | ||||
| No | 1 [Reference] | .001 | 1 [Reference] | .26 |
| Yes | −6.8 (−11.0 to −2.9) | 3.9 (−2.8 to 11.0) | ||
| Regional nodal radiation | ||||
| No | 1 [Reference] | .001 | 1 [Reference] | .02 |
| Yes | −6.8 (−11.0 to −3.0) | −6.9 (−13.0 to −1.0) | ||
| Neoadjuvant chemotherapy | ||||
| Yes | 1 [Reference] | <.001 | 1 [Reference] | .39 |
| No | 5.3 (2.4 to 8.2) | −1.8 (−5.7 to 2.2) | ||
| Adjuvant capecitabine | ||||
| No | 1 [Reference] | .004 | 1 [Reference] | .54 |
| Yes | −4.8 (−8.1 to −1.6) | −1.2 (−4.9 to 2.6) | ||
| Adjuvant ERBB2 therapy | ||||
| No | 1 [Reference] | .008 | 1 [Reference] | .009 |
| Yes | −4.7 (−8.2 to −1.2) | −4.4 (−7.6 to −1.1) | ||
| Adjuvant endocrine therapy | ||||
| No | 1 [Reference] | .26 | NA | NA |
| Yes | 1.8 (−1.3 to 4.9) | |||
| Lymphedema | ||||
| No | 1 [Reference] | <.001 | 1 [Reference] | .82 |
| Yes | −6.2 (−9.3 to −3.2) | −0.4 (−3.6 to 2.8) | ||
| Lymphedema therapy referral | ||||
| No | 1 [Reference] | <.001 | 1 [Reference] | <.001 |
| Yes | −11.0 (−13.0 to −8.1) | −6.3 (−8.9 to −3.6) | ||
Abbreviations: ALND, axillary lymph node dissection; BCS, breast-conserving surgery; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); LN, lymph node; NA, not applicable; QOL, quality of life.
P value calculated using the Bonferroni correction for multiple comparison adjustment with significance set at <.0167.
Psychological and Social Domains of the ULL-27
In the psychological domain, after adjusting for baseline scores, a lymphedema diagnosis was associated with significantly worse QOL (β = –5.4 [95% CI, −8.5 to −2.3]; P < .001), while race and ethnicity and health insurance status were not (eTable 2 in Supplement 1). In the social domain, symptoms necessitating lymphedema therapy referral (β = –5.5; 95% CI, −8.1 to −2.8) and having Medicare health insurance (β = −5.7; 95% CI, −9.1 to −2.3) were associated with worse QOL (P < .001). No association with race or ethnicity and social QOL was observed (eTable 3 in Supplement 1).
Discussion
In this prospective cohort study assessing longitudinal changes in arm volume and health-related QOL in a diverse cohort of women with breast cancer treated with unilateral ALND, we observed several racial and ethnic differences. Similar to the findings of other contemporary studies,5,13,14,19 Black and Hispanic women in our cohort had higher rates of measured lymphedema than Asian and White women. Unlike prior studies, we also prospectively assessed QOL after unilateral ALND and found that Asian, Black, and Hispanic women experienced more subjective arm swelling and worse physical QOL scores at almost every time point compared with White women.
Subjective arm swelling was far more common than measured lymphedema in our cohort, reported in 55% vs 26% of patients throughout the follow-up period. McLaughlin et al20 similarly reported substantial discordance between perceived and measured lymphedema in their prospective cohort study of 336 patients treated with ALND, with 27% reporting arm swelling compared with 16% with measured lymphedema at 5 years of follow-up. Although their study did not specifically address differences in perceived lymphedema by race or ethnicity, they did report that higher BMI, infection, and injury to the ipsilateral arm were risk factors for perceived arm swelling.20
In our study, Black and Hispanic women reported the highest rates of arm swelling, 70% and 87%, respectively, compared with 44% among White women, which may be partially explained by the higher BMI observed among Black and Hispanic women enrolled on our study. Additionally, while not all women who report arm swelling have measured lymphedema, the higher observed rates of lymphedema in Black and Hispanic women undoubtedly also contributed to the higher reported subjective arm swelling. Interestingly, Asian women also reported higher rates of arm swelling (57%) than White women despite having the lowest BMI and lowest rates of measured lymphedema across racial and ethnic groups. A possible explanation may be related to the high rates of mastectomy observed in Asian women in our study, which may result in more postoperative sensory changes compared with lumpectomy, with the ensuing numbness contributing to a sensation of perceived swelling.4,6,10
In addition to higher rates of reported arm swelling, Asian, Black, and Hispanic women with breast cancer treated with unilateral ALND had lower physical QOL scores than White women at nearly every time point after surgery. In all patients, physical QOL decreased after 6 months and never returned to baseline levels, even after more than 2 years of follow-up. An earlier study from our group,11 which had a median follow-up of 1.2 years, left open the possibility that physical QOL would recover with longer follow-up11; however, findings of the current study with longer follow-up suggest that ALND leads to immediate and persistent reductions in QOL over time. Moreover, minority women in our study disproportionately experienced larger decreases in physical QOL after ALND, although the reason for this disparity is unclear. Studies have reported that racial minorities may experience more chemotherapy toxicity21,22 or more severe radiation fibrosis,23,24 which may also contribute to more troublesome arm symptoms and worse QOL. In addition, shoulder dysfunction and range-of-motion limitations after ALND, which were not directly captured in our study, may also contribute to QOL reductions that may vary by race and ethnicity.
After adjusting for baseline scores and other clinicopathological variables, Asian race and Hispanic ethnicity were independently associated with worse physical QOL, while Black race was not. While prior prospective and retrospective studies have reported worse lymphedema health-related QOL in Black women compared with White women,19,25 data on limb-specific QOL for Hispanic and Asian women are limited. To our knowledge, ours is the first prospective study to report on racial differences in QOL independent of other clinicopathological variables. Understanding this racial disparity provides an opportunity for prompt intervention. Strategies, such as an early exercise program, which was shown in the UK PROSPER (Prevention of Shoulder Problems) randomized clinical trial to significantly improve limb function, postoperative pain, and arm symptoms after axillary surgery,26 should be considered to mitigate symptoms in these high-risk groups. The impact of other targeted interventions on QOL, such as the use of prophylactic compression garments27 or surgical risk reduction with immediate lymphatic reconstruction, requires further study.
Health insurance status was also associated with QOL, with lower QOL in women with public health insurance, including Medicare and Medicaid. While health insurance status has been implicated in cancer stage at diagnosis28 and breast cancer outcomes,29 there are limited data on how insurance status mediates QOL after breast cancer treatment. While Black patients in our study were more likely to have public insurance, this did not appear to affect lymphedema therapy referrals or uptake in therapy in our cohort. However, the financial burden associated with lymphedema treatments, including direct costs and lost wages due to time away from work, may deter patients with financial hardships from seeking adequate treatment. Additionally, those with public insurance likely have higher out-of-pocket costs and lower insurance reimbursement for treatments,30 which may adversely affect QOL. Finding cost-effective strategies that can be offered to patients with limited financial resources is an important unmet need to improve QOL after ALND.
While understanding these disparities in QOL after axillary surgery is important to enable interventions, timely identification of symptoms and access to these interventions are equally important. In our study, 70% of patients were referred for lymphedema therapy, with the highest uptake of therapy among Black and Hispanic women, likely because a higher proportion of these women were symptomatic at the time of referral. A recent study examining variations in ancillary service referrals and use in patients with cancer found that, while Black and Hispanic patients were referred for rehabilitative services more often than White patients, use of services among Black women was not commensurately higher.31 Although our study did not specifically address the association between early lymphedema therapy referral and QOL overall or across racial and ethnic groups, future studies from this cohort will be able to answer whether early intervention can mitigate arm symptoms and provide benefit.
Unlike the physical domain of the ULL-27, the psychological and social domains revealed far less impairment after ALND across the cohort, with no difference among racial and ethnic groups. Questions pertaining to the physical domain are focused specifically on upper extremity symptoms and impairment, while the psychological and social domains focus on mood and social interactions, which the patient may not directly attribute to the extremity. Therefore, answers to questions in the psychological and social domains may instead reflect the implications of their overall cancer journey on their emotional state and relationships, rather than the impact of the extremity itself. Reassuringly, scores in these domains showed the most improvement after initial surgery and had the least implications for overall QOL.
Strengths and Limitations
Strengths of our study include its prospective design, diverse patient cohort, long-term follow-up, and use of a validated lymphedema-specific questionnaire to assess QOL. In addition, each patient completed multiple questionnaires over time, allowing for assessment of long-term QOL. The key limitations include attrition; while the initial study design anticipated an approximate 25% attrition rate, mandatory study closure during the COVID-19 pandemic resulted in higher-than-expected attrition for later study measurements. As a result, the extent of missing measurements exceeded 80% for the 30- and 36-month measurement time points, limiting data interpretation. Additionally, the other or unknown race category could not be further characterized and included a small number of patients (n = 9); therefore, analysis of associations in this group were limited.
Conclusions
In this study of a diverse cohort of women with breast cancer undergoing unilateral ALND who were prospectively followed up for lymphedema development and QOL changes, racial and ethnic disparities were identified. Black and Hispanic women had higher rates of objective lymphedema than their White counterparts. Moreover, Asian, Black, and Hispanic women experienced more subjective arm swelling and reported worse physical QOL than White women at nearly every time point after surgery. After adjusting for other clinicopathological features, both minority race and ethnicity and public insurance were associated with worse physical QOL. Understanding these disparities allows for identification of patients at high risk for impairment, which may enable prompt targeted interventions to improve QOL for these patients.
eTable 1. Upper Limb Lymphedema-27 Questionnaire, Distributed to All Study Patients at Baseline and at Regular Postoperative Intervals After Axillary Lymph Node Dissection
eTable 2. Univariable and Multivariable Analyses of Factors Associated With Psychological Quality of Life
eTable 3. Univariable and Multivariable Analyses of Factors Associated With Social Quality of Life
eFigure 1. Trajectory of Psychological Quality of Life Scores Over Time for (A) the Overall Cohort and (B) Stratified by Race and Ethnicity
eFigure 2. Trajectory of Social Quality of Life Scores Over Time for (A) the Overall Cohort and (B) Stratified by Race and Ethnicity
Data Sharing Statement
References
- 1.Ashikaga T, Krag DN, Land SR, et al. ; National Surgical Adjuvant Breast, Bowel Project . Morbidity results from the NSABP B-32 trial comparing sentinel lymph node dissection versus axillary dissection. J Surg Oncol. 2010;102(2):111-118. doi: 10.1002/jso.21535 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Che Bakri NA, Kwasnicki RM, Khan N, et al. Impact of axillary lymph node dissection and sentinel lymph node biopsy on upper limb morbidity in breast cancer patients: a systematic review and meta-analysis. Ann Surg. 2023;277(4):572-580. doi: 10.1097/SLA.0000000000005671 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.DiSipio T, Rye S, Newman B, Hayes S. Incidence of unilateral arm lymphoedema after breast cancer: a systematic review and meta-analysis. Lancet Oncol. 2013;14(6):500-515. doi: 10.1016/S1470-2045(13)70076-7 [DOI] [PubMed] [Google Scholar]
- 4.Hidding JT, Beurskens CH, van der Wees PJ, van Laarhoven HW, Nijhuis-van der Sanden MW. Treatment related impairments in arm and shoulder in patients with breast cancer: a systematic review. PLoS One. 2014;9(5):e96748. doi: 10.1371/journal.pone.0096748 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Montagna G, Zhang J, Sevilimedu V, et al. Risk factors and racial and ethnic disparities in patients with breast cancer–related lymphedema. JAMA Oncol. 2022;8(8):1195-1200. doi: 10.1001/jamaoncol.2022.1628 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McLaughlin SA, Brunelle CL, Taghian A. Breast cancer–related lymphedema: risk factors, screening, management, and the impact of locoregional treatment. J Clin Oncol. 2020;38(20):2341-2350. doi: 10.1200/JCO.19.02896 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Boyages J, Kalfa S, Xu Y, et al. Worse and worse off: the impact of lymphedema on work and career after breast cancer. Springerplus. 2016;5:657. doi: 10.1186/s40064-016-2300-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cornelissen AJM, Kool M, Keuter XHA, et al. Quality of life questionnaires in breast cancer–related lymphedema patients: review of the literature. Lymphat Res Biol. 2018;16(2):134-139. doi: 10.1089/lrb.2017.0046 [DOI] [PubMed] [Google Scholar]
- 9.Dominick SA, Natarajan L, Pierce JP, Madanat H, Madlensky L. The psychosocial impact of lymphedema-related distress among breast cancer survivors in the WHEL Study. Psychooncology. 2014;23(9):1049-1056. doi: 10.1002/pon.3510 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ridner SH. Quality of life and a symptom cluster associated with breast cancer treatment–related lymphedema. Support Care Cancer. 2005;13(11):904-911. doi: 10.1007/s00520-005-0810-y [DOI] [PubMed] [Google Scholar]
- 11.Zhang JQ, Montagna G, Sevilimedu V, et al. Longitudinal prospective evaluation of quality of life after axillary lymph node dissection. Ann Surg Oncol. 2022;29:4127-4136. doi: 10.1245/s10434-022-11623-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Armer JM, Ballman KV, McCall L, et al. Lymphedema symptoms and limb measurement changes in breast cancer survivors treated with neoadjuvant chemotherapy and axillary dissection: results of American College of Surgeons Oncology Group (ACOSOG) Z1071 (Alliance) substudy. Support Care Cancer. 2019;27(2):495-503. doi: 10.1007/s00520-018-4334-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kwan ML, Yao S, Lee VS, et al. Race/ethnicity, genetic ancestry, and breast cancer–related lymphedema in the Pathways Study. Breast Cancer Res Treat. 2016;159(1):119-129. doi: 10.1007/s10549-016-3913-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ren Y, Kebede MA, Ogunleye AA, et al. Burden of lymphedema in long-term breast cancer survivors by race and age. Cancer. 2022;128(23):4119-4128. doi: 10.1002/cncr.34489 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Launois R, Mègnigbêto A, Pocquet K, Alliot F. A specific quality of life scale in upper limb lymphedema: the ULL-27 questionnaire. Lymphology. 2002;35(suppl):181-187. doi: 10.1016/S1098-3015(11)71503-0 [DOI] [Google Scholar]
- 16.Memorial Sloan Kettering Cancer Center . A prospective surveillance program for assessment and treatment of breast cancer–related lymphedema after axillary lymph node dissection. Accessed December 18, 2023. https://classic.clinicaltrials.gov/ct2/show/NCT02743858?term=02743858&draw=2&rank=1.
- 17.McDuff SGR, Mina AI, Brunelle CL, et al. Timing of lymphedema after treatment for breast cancer: when are patients most at risk? Int J Radiat Oncol Biol Phys. 2019;103(1):62-70. doi: 10.1016/j.ijrobp.2018.08.036 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Ancukiewicz M, Russell TA, Otoole J, et al. Standardized method for quantification of developing lymphedema in patients treated for breast cancer. Int J Radiat Oncol Biol Phys. 2011;79(5):1436-1443. doi: 10.1016/j.ijrobp.2010.01.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Naughton MJ, Liu H, Seisler DK, et al. Health-related quality of life outcomes for the LEAP study-CALGB 70305 (Alliance): a lymphedema prevention intervention trial for newly diagnosed breast cancer patients. Cancer. 2021;127(2):300-309. doi: 10.1002/cncr.33184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.McLaughlin SA, Wright MJ, Morris KT, et al. Prevalence of lymphedema in women with breast cancer 5 years after sentinel lymph node biopsy or axillary dissection: objective measurements. J Clin Oncol. 2008;26(32):5213-5219. doi: 10.1200/JCO.2008.16.3725 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hu X, Kaplan CM, Martin MY, et al. Race differences in patient-reported symptoms during chemotherapy among women with early-stage hormone receptor–positive breast cancer. Cancer Epidemiol Biomarkers Prev. 2023;32(2):167-174. doi: 10.1158/1055-9965.EPI-22-0692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rosenzweig MQ, Mazanec SR. Racial differences in breast cancer therapeutic toxicity: implications for practice. Cancer Epidemiol Biomarkers Prev. 2023;32(2):157-158. doi: 10.1158/1055-9965.EPI-22-1111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Alam A, Mukhopadhyay ND, Ning Y, et al. A preliminary study on racial differences in HMOX1, NFE2L2, and TGFβ1 gene polymorphisms and radiation-induced late normal tissue toxicity. Int J Radiat Oncol Biol Phys. 2015;93(2):436-443. doi: 10.1016/j.ijrobp.2015.05.049 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tuamokumo NL, Haffty BG. Clinical outcome and cosmesis in African-American patients treated with conservative surgery and radiation therapy. Cancer J. 2003;9(4):313-320. doi: 10.1097/00130404-200307000-00014 [DOI] [PubMed] [Google Scholar]
- 25.Pusic AL, Cemal Y, Albornoz C, et al. Quality of life among breast cancer patients with lymphedema: a systematic review of patient-reported outcome instruments and outcomes. J Cancer Surviv. 2013;7(1):83-92. doi: 10.1007/s11764-012-0247-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Bruce J, Mazuquin B, Canaway A, et al. ; Prevention of Shoulder Problems Trial (PROSPER) Study Group . Exercise versus usual care after non-reconstructive breast cancer surgery (UK PROSPER): multicentre randomised controlled trial and economic evaluation. BMJ. 2021;375:e066542. doi: 10.1136/bmj-2021-066542 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Paramanandam VS, Dylke E, Clark GM, et al. Prophylactic use of compression sleeves reduces the incidence of arm swelling in women at high risk of breast cancer–related lymphedema: a randomized controlled trial. J Clin Oncol. 2022;40(18):2004-2012. doi: 10.1200/JCO.21.02567 [DOI] [PubMed] [Google Scholar]
- 28.Ko NY, Hong S, Winn RA, Calip GS. Association of insurance status and racial disparities with the detection of early-stage breast cancer. JAMA Oncol. 2020;6(3):385-392. doi: 10.1001/jamaoncol.2019.5672 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ayanian JZ, Kohler BA, Abe T, Epstein AM. The relation between health insurance coverage and clinical outcomes among women with breast cancer. N Engl J Med. 1993;329(5):326-331. doi: 10.1056/NEJM199307293290507 [DOI] [PubMed] [Google Scholar]
- 30.Bian J, Shen A, Yang W, Zhang L, Qiang W. Financial toxicity experienced by patients with breast cancer–related lymphedema: a systematic review. Support Care Cancer. 2023;31(6):354. doi: 10.1007/s00520-023-07800-9 [DOI] [PubMed] [Google Scholar]
- 31.Chan VHY, Heller DR, Berger ER, Capozza S, Greenup RA, Sanft TB. Ancillary treatment referrals and visits after breast cancer surgery in a sociodemographically diverse population. Ann Surg Oncol. 2023;30(9):5637-5648. doi: 10.1245/s10434-023-13431-5 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Upper Limb Lymphedema-27 Questionnaire, Distributed to All Study Patients at Baseline and at Regular Postoperative Intervals After Axillary Lymph Node Dissection
eTable 2. Univariable and Multivariable Analyses of Factors Associated With Psychological Quality of Life
eTable 3. Univariable and Multivariable Analyses of Factors Associated With Social Quality of Life
eFigure 1. Trajectory of Psychological Quality of Life Scores Over Time for (A) the Overall Cohort and (B) Stratified by Race and Ethnicity
eFigure 2. Trajectory of Social Quality of Life Scores Over Time for (A) the Overall Cohort and (B) Stratified by Race and Ethnicity
Data Sharing Statement
