Incidence and risk factors of breast cancer-related lymphedema in Korea: a nationwide retrospective cohort study

Abstract

Background:

Breast cancer-related lymphedema (BCRL) is a secondary lymphedema that occurs after breast cancer related treatments. BCRL develops from damage or dysfunction of the normally functioning lymphatic system due to surgery, radiation therapy, and rarely due to cancer recurrence. This nationwide, retrospective study was aimed at investigating the incidence and risk factors of BCRL using the database of the Korean National Health Insurance Service (NHIS).

Methods:

Patients with newly diagnosed breast cancer who underwent breast surgery from 1 January 2017 to 31 December 2020, were recruited. The incidence was compared by four groups according to the operation type of breast cancer [breast conserving surgery (BCS) with sentinel lymph node biopsy (S), BCS with axillary lymph node dissection (A), total mastectomy (TM) with S, modified radical mastectomy (MRM)]. The incidence rates of lymphedema were calculated by the number of incident events by the total follow-up period. Cox proportional hazard regression was used to calculate the risk of incidence of lymphedema based on a patients’ characteristics, breast cancer treatment, and comorbidities.

Results:

The final cohort of operation subjects that satisfied the inclusion criteria was 34 676. BCRL occurred in 4242 patients (12.2%), and the median follow-up period was 695.4 days. The BCRL was diagnosed in the BCS with S (8.0%), BCS with A (23.5%), TM with S (10.7%), and MRM (28.5%) with an incidence of 40.8, 132.2, 55.8, and 171.8 per 1000 person-years, respectively. Young age, obesity, chemotherapy, radiotherapy, residence in metropolitan areas, and hyperlipidemia were identified as risk factors.

Conclusion:

In Korea, the incidence of BCRL was found to be 12.2%, with the highest risk observed among patients who underwent MRM. Therefore, surgical oncologists should meticulously assess the appropriate surgical approach and consider providing education to patients with risk factors for BCRL, aiming to ensure effective prevention strategies.

Introduction

Highlights

• Breast cancer-related lymphedema (BCRL) occurred frequently within 2 years after breast cancer diagnosis or surgery.

• In this study, a total of 34 676 subjects were included, in whom BCRL occurred in 4242 (12.2%).

• Patients who underwent modified radical mastectomy had the highest risk of lymphedema (aHR, 3.73; 95% CI: 3.40–4.08).

• Young age, obesity, chemotherapy, radiotherapy, residence in metropolitan areas, and hyperlipidemia were identified as risk factors.

According to the GLOBOCAN 2020 data, breast cancer is the most commonly diagnosed cancer and the fifth most common cause of cancer related death worldwide¹. With improvements in early detection and the development of treatment modalities for breast cancer, the survival rate in breast cancer patients is increasing, leading to increased importance being placed on patient quality of life (QoL). Lymphedema is a chronic disease characterized by an accumulation of lymphatic fluid, resulting in skin and tissue changes². Breast cancer-related lymphedema (BCRL) develops as a result of damage or dysfunction of the normally functioning lymphatic system due to surgery, radiotherapy (RT), and rarely due to cancer recurrence^2,3. BCRL significantly impacts patient QoL^4–6.

A systemic review analyzing 30 prospective cohort studies showed that the incidence of BCRL was 21.4%, and it frequently occurred within 2 years of diagnosis or surgery for breast cancer⁷. BCRL is characterized by symptoms such as swelling, pain, or heaviness in the affected arm, and affects the function, social activities, and psychological well-being of patients^8–11. The risk factors for BCRL include young age, obesity, smoking, comorbidities such as collagen disease and hypertension, advanced cancer stage, total mastectomy, axillary dissection, higher levels of axillary node excision, presence of surgical complications, such as postoperative bleeding and surgical site infection, chemotherapy, and RT^5,7,12–14. BCRL cannot be cured; however, the symptoms can be managed. Thus, efforts are required to improve the prevention, early detection, and management of BCRL⁶. Surgical oncologists should educate patients with BCRL risk factors to promote precaution and prevention.

Most previous studies investigating the risk factors for BRCL included relatively small sample sizes, and insufficient consideration of factors such as patient comorbidities and socioeconomic status. The Korean National Health Insurance Service (NHIS) provides mandatory healthcare to nearly all Korean citizens. The NHIS collects health data from nearly 50 million insured subjects, including admission and outpatient visit records, diagnoses, drug prescriptions, and health examination data. This nationwide, retrospective cohort study identified the incidence and risk factors of BCRL using data extracted from the NHIS database.

Methods

Data source and study population

This large, nationwide, retrospective cohort study was conducted using the National Health Information Database (NHID) of the NHIS. The NHIS is a governmental institution that provides medical information extracted from the health care insurance data of 97% of all Korean citizens. The NHID includes information such as the patients’ basic characteristics, socioeconomic status, disease diagnoses based on the 10th revision of the International Classification of Disease (ICD-10) codes, medical treatment and procedures, and drug prescriptions. The NHID used in this study included 344 061 patients diagnosed with breast cancer (NHIS-1-004) between 1 January 2002 and 31 December 2020¹⁵.

From 1 January 2017 to 31 December 2020, 61 857 patients aged 18 years or older were identified as being newly diagnosed with breast cancer under the ICD-10 code C50 (indicating invasive breast cancer), and the co-payment policy code V193 (the expanding benefit coverage for cancer patients in Korea). The V193 code is a unique system in which registration is possible after a definitive diagnosis of breast cancer via biopsy, and increases the cancer diagnosis accuracy of the NHID¹⁶. The diagnosis for BCRL was defined as 1 or more instances of the ICD-10 codes I972, I890, and I891 identified after surgical treatment for breast cancer. The index date for the patients was the date of the breast cancer surgery.

The exclusion criteria were as follows: 1) patients who had other cancer(s) prior to the diagnosis of breast cancer; 2) patients with an unclear breast surgery code; 3) patients who underwent two or more surgeries for breast cancer; 4) patients lacking study data due to not undergoing a health examination within the 3 years prior to breast cancer diagnosis; 5) patients who underwent breast surgery prior to the breast cancer diagnosis; 6) patients diagnosed with lymphedema prior to breast cancer treatments; 7) patients with the same dates for breast cancer surgery and chemotherapy treatment (Fig. 1). Patients were followed-up until the date of lymphedema diagnosis or 31 December 2020. Participants were categorized into four groups according to the type of surgery for breast cancer: breast conserving surgery (BCS) with sentinel lymph node biopsy (S), BCS with axillary lymph node dissection (A), total mastectomy (TM) with S, and modified radical mastectomy (MRM). The incidence of BRCL was subsequently compared among the four groups.

Figure 1.

Flow chart of patients who met the inclusion/exclusion criteria for the study.

The work has been reported with the strengthening the reporting of cohort, cross-sectional and case–control studies in surgery (STROCSS) criteria¹⁷ (Supplemental Digital Content 1, http://links.lww.com/JS9/C98)

Variables and operational definitions

We collected basic information, including age, residential area, economic status, and life style factors (e.g. smoking, alcohol consumption, and obesity), for all patients. To identify risk factors for BCRL, we assessed patient comorbidities, including a history of hypertension, diabetes, chronic kidney disease, and dyslipidemia according to the ICD-10 codes (hypertension, I10 and I15; diabetes mellitus, E10-E14; chronic kidney disease, N18; and dyslipidemia, E78), in combination with the electronic data interchange code of related medications. Comorbidities were deemed present if they were diagnosed at least once during the previous 3 years. ICD-10 codes were also used to identify the type of surgery, as follows: BCS with S, N7137; BCS with A, N7136; TM with S, N7139; MRM, N7138. Chemotherapy, RT, hormonal therapy, and target therapy for the study population were also investigated to identify the impact of treatment options for breast cancer on BCRL. Detailed operational definitions used in this study are as shown in Supplementary Table 1 (Supplemental Digital Content 2, http://links.lww.com/JS9/C99).

Statistical analysis

Descriptive statistics are presented as the mean and SD for continuous variables and as numbers with percentages for categorical variables. Kaplan–Meier analysis and log-rank test were used to assess the cumulative incidence rates of BCRL during the entire follow-up period, and the results were described as 1000 person-years. A cumulative incidence plot was used to compare the incidence of BCRL among the four groups.

Cox regression analyses (adjusted for age, obesity, smoking, economic status, residential area, operation type, treatment modalities, and comorbidities), were performed to evaluate risk factors for BCRL during the study period. Cox proportional hazards regression analysis was used to calculate the adjusted hazard ratio (aHR) and 95% CI for the risk of BCRL.

Statistical analyses were performed with SAS software (version 9.4; SAS Institute) and R (version 4.0.3, R Foundation for Statistical Computing).

Ethics approval and consent to participate

The study protocol was approved by the Institutional Review Board (IRB) of the OO National University Research Council (IRB no.: 2023-09-014, date of approval: 19 September 2023). The IRB waived the need for informed consent.

Results

Baseline characteristics of the study participants

A total of 34 676 patients diagnosed with breast cancer between 1 January 2017 and 31 December 2020 were included in this analysis (Fig. 1). BCRL developed in 4242 patients (12.2%), and the median follow-up period was 695.4 days. The patients enrolled in this study were mostly aged between 40 and 59 years (64.9%). Among patients diagnosed with BCRL, 6.0% were smokers and 23.3% were alcohol drinkers. In the BCRL group, 72.6% had low economic status. Of the entire cohort, 58.4% underwent BCS with S for breast cancer. Patients who received neoadjuvant chemotherapy accounted for 19.5%, whereas those who received adjuvant chemotherapy accounted for 34.5%. The percentage of patients receiving RT, hormonal therapy, and targeted therapy was 76.8, 76.8, and 16.1%, respectively. The prevalence rates of hypertension, diabetes mellitus, chronic kidney disease, and dyslipidemia are shown in Table 1.

Table 1.

Baseline characteristics.

	Total (N=34 676)	BCRL (-) (n=30 434)	BCRL (+) (n=4242)
Age at surgery, years	53.73±10.61	53.86±10.66	52.82±10.25
Age, years
≤39	2291 (6.6)	1961 (6.4)	330 (7.7)
40–59	22 488 (64.9)	19 652 (64.6)	2836 (66.9)
≥60	9897 (28.5)	8821 (29.0)	1076 (25.4)
BMI	23.57±3.51	23.53±3.49	23.89±3.60
High (≥23 kg/m2)	18 035 (52.0)	15 677 (51.5)	2358 (55.6)
Smoking status, yes	1931 (5.6)	1677 (5.5)	254 (6.0)
Alcohol consumption, yes	8954 (25.8)	7965 (26.2)	989 (23.3)
Economic status, lowa	24 664 (71.1)	21 584 (70.9)	3080 (72.6)
Residential area
Metropolitan cities	25 044 (72.2)	21 932 (72.1)	3112 (73.4)
Mid-size and small cities	7627 (22.0)	6732 (22.1)	895 (21.1)
Rural areas	2005 (5.8)	1770 (5.8)	235 (5.5)
Type of surgery
BCS with S	20 255 (58.4)	18 632 (61.2)	1623 (38.3)
BCS with A	3359 (9.7)	2569 (8.4)	790 (18.6)
TM with S	7443 (21.5)	6645 (21.8)	798 (18.8)
MRM	3619 (10.4)	2588 (8.6)	1031 (24.3)
Use of chemotherapy
No	15 962 (46.0)	14 649 (48.1)	1313 (31.0)
Yes	18 714 (54.0)	15 785 (51.9)	2929 (69.0)
Neoadjuvant	6753 (19.5)	5412 (17.8)	1341 (31.6)
Adjuvant	11 961 (34.5)	10 373 (34.1)	1588 (37.4)
Radiotherapy, yes	26 620 (76.8)	23 287 (76.5)	3333 (78.6)
Hormonal therapy, yes	26 643 (76.8)	23 534 (77.3)	3109 (73.3)
Target therapy, yes	5583 (16.1)	4714 (15.5)	869 (20.5)
Comorbidities
Hypertension	8441 (24.3)	7444 (24.5)	997 (23.5)
Diabetes mellitus	4350 (12.5)	3840 (12.6)	510 (12.0)
Chronic kidney disease	311 (0.9)	272 (0.9)	39 (0.9)
Dyslipidemia	17 865 (51.5)	15 686 (51.5)	2179 (51.4)

BCRL, breast cancer related lymphedema; BCS, breast conserving surgery; S, sentinel lymph node biopsy; A, axillary lymph node dissection; TM, total mastectomy; MRM, modified radical mastectomy.

^aPeople whose health insurance payments were less than 30% of the national population.

Incidence of BCRL

The incidence rate of BCRL was 64.2 per 1000 person-years across the entire cohort. The slope of the cumulative BCRL events became smaller over time after the breast cancer surgery (Fig. 2A). The incidence rates of BCRL according to the breast cancer surgery type are shown in Figure 2B. During the follow up period, BCRL occurred in 8.0% (1623/20 255) of the patients who underwent BCS with S, 23.5% (790/3359) of BCS with A, 10.7% (798/7443) of TM with S, and 28.5% (1,031/3619) of MRM. In patients who underwent BCS with S, the incidence of BCRL was 40.8 per 1000 person-years, which was lower than that of BCS with A (132.2 per 1000 person-years). In patients who had TM and S, the incidence of BCRL was 55.8 per 1000 person-years, which was lower than that of MRM (171.8, per 1000 person-years).

Figure 2.

A: Incidence rate of BCRL in the entire cohort; B: Specific incidence rates of BCRL in patients treated with BCS with S, BCS with A, TM with S, and MRM.

Risk factors of breast cancer-related lymphedema

The results of the univariate and multivariate Cox regression analysis are presented in Table 2. The risk of BRCL was higher in those under the age of 39 years (HR, 1.15; 95% CI: 1.03–1.29) than in those between ages 40 and 59 years, while the risk was lower in those aged 60 years or above (HR, 0.87; 95% CI: 0.81–0.93) than in those aged between 40 to 59 years. Obesity (HR, 1.17; 95% CI: 1.10–1.24) and low economic status (HR, 1.08; 95% CI: 1.01–1.15) were identified as factors that increased the risk of BCRL. Compared to BCS with S, MRM (HR, 4.06; 95% CI: 3.751–4.39), TM with S (HR, 1.38; 95% CI: 1.26–1.50), and BCS with A (HR, 3.16; 95% CI: 2.90–3.44) showed a higher risk for BCRL. Patients who received neoadjuvant chemotherapy (HR, 2.86; 95% CI: 2.65–3.09), adjuvant chemotherapy (HR, 1.60; 95% CI: 1.48–1.72), RT (HR, 1.09; 95% CI: 1.01–1.17), and targeted therapy (HR, 1.39; 95% CI: 1.29–1.50) had a higher risk of BCRL.

Table 2.

Factors that predict the development of BCRL according to Cox Regression Analysis

	Univariate analysis		Multivariate analysis
	HR (95% CI)	P	aHR (95% CI)	P
Age (≤39 vs 40–59)	1.15 (1.03–1.29)	0.017	1.06 (0.94–1.19)	0.347
Age (≥60 vs 40–59)	0.87 (0.81–0.93)	<0.001	0.88 (0.81–0.95)	0.002
BMI (≥23 kg/m2 vs <22.9 kg/m2)	1.17 (1.10–1.24)	<0.001	1.16 (1.09–1.24)	<0.001
Smoking status (yes vs no)	1.11 (0.98–1.26)	0.114	1.05 (0.92–1.20)	0.436
Alcohol consumption (yes vs no)	0.93 (0.87–1.00)	0.060	0.98 (0.90–1.05)	0.501
Economic status (Low vs High)	1.08 (1.01–1.15)	0.027	0.99 (0.93–1.06)	0.783
Residential area (Metro vs Small and medium)	1.07 (0.99–1.15)	0.090	1.10 (1.02–1.18)	0.016
Residential area (Rural vs Small and medium)	1.00 (0.86–1.15)	0.980	1.02 (0.8–1.17)	0.8215
MRM vs BCS with S	4.06 (3.75–4.39)	<0.001	3.73 (3.40–4.08)	<0.001
TM with S vs BCS with S	1.38 (1.26–1.50)	< 0.001	1.59 (1.43–1.77)	<0.001
BCS with A vs BCS with S	3.16 (2.90–3.44)	<0.001	2.77 (2.53–3.03)	<0.001
Neoadjuvant chemotherapy (yes vs no)	2.86 (2.65–3.09)	<0.001	1.71 (1.56–1.88)	<0.001
Adjuvant chemotherapy (yes vs no)	1.60 (1.48–1.72)	<0.001	1.16 (1.07–1.26)	<0.001
Radiotherapy (yes vs no)	1.09 (1.01–1.17)	0.027	1.29 (1.17–1.42)	<0.001
Hormonal therapy (yes vs no)	0.78 (0.73–0.83)	<0.001	0.91 (0.85–0.98)	0.009
Target therapy (yes vs no)	1.39 (1.29–1.50)	<0.001	0.97 (0.89–1.05)	0.412
Hypertension (yes vs no)	0.96 (0.90–1.03)	0.270	0.97 (0.89–1.05)	0.439
Diabetes mellitus (yes vs no)	0.96 (0.87–1.05)	0.369	0.93 (0.84–1.03)	0.163
Chronic kidney disease (yes vs no)	1.08 (0.79–1.48)	0.619	1.13 (0.82–1.55)	0.468
Hyperlipidemia (yes vs no)	1.01 (0.95–1.08)	0.666	1.07 (1.00–1.14)	0.049

A, axillary lymph node dissection; BCRL, breast cancer related lymphedema; BCS, breast conserving surgery; HR, hazard ratio; MRM, modified radical mastectomy; S, sentinel lymph node biopsy; TM, total mastectomy.

In multivariate analysis, age 60 years or above (aHR, 0.88; 95% CI: 0.81–0.95) was associated with a lower risk of BCRL than ages 40 to 59 years. Obesity (aHR, 1.16; 95% CI: 1.09–1.24) was a factor that increase the risk of BCRL. Regarding nontreatment-related risk factors, residence in a metropolitan city (aHR, 1.10; 95% CI: 1.02–1.18) was associated with an increased risk of BRCL compared with living in small and medium cities. Patients who underwent MRM (aHR, 3.73; 95% CI: 3.40–4.08), TM with S (aHR, 1.59; 95% CI: 1.43–1.77), and BCS with A (aHR, 2.77; 95% CI: 2.53–3.03) also had an increased risk of BCRL compared to those who underwent BCS with S. Neoadjuvant chemotherapy (aHR, 1.71; 95% CI: 1.56–1.88), adjuvant chemotherapy (aHR, 1.16; 95% CI: 1.07–1.26), and RT (aHR, 1.29; 95% CI: 1.17–1.42), were also all associated with an increased risk of BCRL. Hyperlipidemia (aHR, 1.07; 95% CI: 1.00–1.14) was associated with an increased risk of BCRL. A forest plot based on the results of multivariate analysis is shown in Figure 3.

Figure 3.

The forest plot based on results of multivariate analysis.

Discussion

BCRL is a potential adverse event that can occur following breast cancer treatment. This chronic condition is associated with functional disability and psychosocial impacts in patients. BCRL affects about 3–5 million patients per year worldwide¹⁸. The incidence of BCRL in patients who underwent breast cancer treatment has been reported to vary widely from 1 to 65% according to the duration of follow-up, therapeutic modality and extent, and diagnostic criteria of BCRL¹⁹. According to the extent of breast cancer surgery, the incidence of BCRL in previous studies was higher in those who underwent total mastectomy compared to those treated with BCS^8,14,20. In this study, BCRL incidence and risk were higher for total mastectomy (10.7%) than for BCS (8.0%) in SLNB (sentinel lymph node biopsy) patients (aHR: 1.59, 95% CI: 1.43–1.77). According to the results of a prior meta-analysis of 72 studies, ALND (axillary lymph node dissection, 19.9%) increases the incidence of BCRL more than SLNB (5.6%)⁷. When more than five axillary LNs are removed (18.2%), BCRL incidence is increased compared with the removal of less than five LNs (3.3%)²⁰. In this study, BCRL incidence and risk were higher in those who underwent ALND (23.5%) compared to those who underwent SLNB (8%) in patients with BCS (2.77, 95% CI: 2.53–3.03).

After the publication of the NSABP B-32 trial, SLNB has replaced ALND as the primary strategy for management of the axilla in the past two decades²¹. The ACOSOG Z0011 trial and the AMAROS trial provided the basis for omitting ALND in patients with clinically node-negative disease who had positive sentinel nodes^22,23. Recently, real-world practice is changing to de-escalate axillary surgery extents and omit SLNB in selected patients, such as SLNB alone or targeted axillary dissection in clinically node positive patients who have undergone neoadjuvant chemotherapy^24–26. Following these changes, according to the nationwide cohort study from Denmark published in 2023, the incidence of BCRL, which was 8.16 per 100 women in 2007, decreased to 4.55 in 2017¹⁴.

In the last decade, the number of women undergoing breast reconstruction following total mastectomy has increased. Furthermore, cosmetic outcomes following cancer treatment have improved, while diffuse distribution of breast cancer cells has become detectable due to improvements in preoperative evaluation methods. In Korea, Since April 2015, the Korean National Health Insurance has reimbursed breast cancer patients who undergo reconstruction, and breast reconstruction patients subsequently increased from 19.4% in 2015 to 53.4% in 2018²⁷. As the incidence of breast reconstruction has increased, studies on the effects of breast reconstruction on BCRL risk have also been conducted. Patients who undergo reconstruction typically undergo skin-sparing mastectomy, with healthy lymphatic channels in the spared skin, leading to a lower incidence of BCRL compared with patients who undergo extensive resection of the breast and axillary skin without reconstruction²⁸. Furthermore, patients who undergo mastectomy without reconstruction may experience tissue adhesion, fibrosis, and contracture of the breast skin, chest wall, and axilla, resulting in obstructed lymphatic flow and possibly lymphedema development²⁹. According to a prior prospective study published in 2016, the 2 year cumulative incidence of BCRL in a cohort of patients who underwent immediate breast reconstruction was 5.13% compared to 26.66% for no reconstruction²⁸. In several retrospective cohort studies, immediate breast reconstruction reduced the risk of lymphedema^29–31. In the present cohort, 45.3% of the patients who underwent mastectomy had immediate or delayed breast reconstruction. In this study, BCRL occurred in 15.7% of the patients who underwent breast reconstruction, which was significantly lower than the incidence in patients who did not undergo breast reconstruction (17.0%) (P=0.035).

Nontreatment-related risk factors for BCRL have been reported to include age, obesity, smoking, subclinical edema, cellulitis, comorbidity, low educational level, low income, etc^7,12,14,32. Studies reported that age as a risk factor for BCRL had different age criteria and showed heterogeneous results^8,33–37. Pezner et al.³⁴ found that 25% of patients older than 60 years developed BCRL, compared with 7% of those aged <60 years, and concluded that patients older than 60 years were at an increased risk. Hayes et al.³⁵ also reported that age 50 years or older was associated with a threefold increased odds of developing lymphedema. In contrast, patients aged between 40 and 59 years in the present study had a lower BCRL risk compared to the younger age group, but had a higher BCRL risk when compared to those aged 60 years or older. In a previous Danish nationwide follow-up study, younger age was associated with an increased risk of BCRL 5–7 years after surgery⁸.

A higher BMI at diagnosis is a well-established risk factor for the development of lymphedema. In our cohort, patients with a BMI >23 kg/m² had a significantly higher aHR for BCRL than those with a BMI <23 kg/m². Obesity is known to induce BCRL by exerting a negative impact on lymphatic density in the subcutaneous tissue, lymphatic endothelial cell proliferation, lymphatic leakiness, collecting-vessel pumping capacity, and clearance of macromolecules³⁸. Furthermore, some reports have indicated that the risk of BCRL increases in patients who experience a loss or gain of postoperative weight of 10 pounds or more over a period of 1 month³⁹. Therefore, education and management strategies are necessary for weight management in breast cancer patients. However, the relationship between lymphedema incidence and personal characteristics has received relatively little attention in previous research. In this study, the risk was increased in patients with a low economic status and those who resided in a metropolitan city. Several studies have reported that the odds of developing BCRL increased in patients with a low education level, low income, and those without a significant relationship (unmarried)^36,37.

In this study, adjuvant and neoadjuvant chemotherapies were found to be risk factors for BCRL. However, conclusive evidence regarding whether chemotherapy is a risk factor for BCRL is currently lacking, with previous studies reporting contrasting results^{5,7,9,14,20,37,40,41}. In one study, anthracycline-based chemotherapy (HR 1.46; 95% CI: 1.04–2.04) was significantly associated with an increased BCRL risk³⁷. Furthermore, increased extracellular fluid, which induces fluid retention in the peripheral edema of the extremities, is a common side effect of taxane-based chemotherapy^42,43. According to the results of a prospective cohort study by Kilbreath et al.²⁰, taxane-based chemotherapy was reported to increase the risk of arm swelling, which was identified as a risk factor for BCRL development within the first 18-months postoperatively (OR: 13.5; 95% CI: 4.8–38.1). However, Swaroop et al.⁴¹ compared patients who received adjuvant taxane-based chemotherapy (HR: 1.14; P=0.62) with those who received nontaxane chemotherapy (HR: 1.56; P=0.40), and found no differences in BCRL incidence. Although not conclusive, our results suggest that prospective BCRL surveillance is necessary in patients who are scheduled to receive anthracycline or docetaxel based chemotherapy. However, it is unclear whether neoadjuvant chemotherapy affects the risk of BCRL. According to a retrospective study of patients who received neoadjuvant chemotherapy and ALND for clinically node positive breast cancer, neoadjuvant chemotherapy was not a risk factor for BCRL (P=0.61)⁴⁴. In a prospective study by Specht et al.⁴⁵, the BCRL incidence in patients who received neoadjuvant chemotherapy (23%) was not statistically significant (HR: 0.76; P=0.39) compared to that in patients who received adjuvant chemotherapy (15%). However, BCRL risk increased ninefold (P=0.038) when residual lymph node disease was present postchemotherapy in patients who received neoadjuvant chemotherapy⁴⁵. These results suggest that, rather than chemotherapy itself, the disease severity and the resulting surgery extent have a greater impact on BCRL.

RT is a postoperative adjuvant treatment for breast cancer that reduces local recurrence and improves survival rate. According to a prior prospective study conducted in 627 patients, over the 22.8 months follow-up, BCRL incidence was higher in those who had RT compared to those who did not amongst a cohort of patients treated with mastectomy and SLNB (10 vs. 2.19%)⁴⁶. Furthermore, in patients who underwent mastectomy and ALND, BCRL incidence was higher in those who had RT compared to those who did not (30.1 vs. 19.3%)⁴⁶. In a prospective cohort study conducted on 1501 breast cancer patients who underwent BCS or mastectomy, BCRL occurred in only 3.1% of the patients when RT was received on the breast or chest wall only⁴⁷. However, BCRL occurred in 21.1% of patients when regional lymph node radiation (RLNR) was performed, while RLNR significantly increased the risk of BCRL compared with breast/chest wall radiation alone (HR: 1.70; 95% CI: 1.07–2.70)⁴⁷. According to the results of a meta-analysis of 21 studies, BCRL risk increased even more when both breast/chest wall radiation and RLNR were performed compared to breast/chest wall radiation alone in patients who had ALND (OR: 8.70; 95% CI: 4.19–18.09)⁴⁸. The extent of axillary surgery is thought to have a greater impact on the incidence of BCRL compared to RT. Therefore, in patients with advanced breast cancer expected to have RLNR, it would be helpful to de-escalate the axillary stage prior to axillary surgery through neoadjuvant chemotherapy to prevent BCRL.

BCRL is an incurable disease that exerts adverse physical and emotional effects on patients; therefore, prevention and early detection are critical. Surgical oncologists should provide an education for patients on preventive measures of BCRL, such as avoiding excessive heat on an ‘at risk’ limb, not having chemotherapy administered into the limb unless medically necessary, and trying to avoid infections⁴⁹. Also, surgical oncologists must provide patient education in clinical practice in patients who have the risk factors for BCRL identified in this study. Furthermore, introducing prospective BCRL screening programs involving multidisciplinary teams in breast cancer clinics using standard objective tools, such as perometry, will enable diagnosis at an earlier stage and earlier intervention of BCRL⁵⁰.

The primary strength of this study is that it included a large cohort representative of Korea’s general population. However, this study has several limitations. First, the NHID did not allow us to search for information such as laboratory results, imaging, cancer stage, or pathological results, including the number of LNs excised, details of treatment, such as the radiation field, or the radiation technique. Second, we were unable to fully address selection bias as we could not include patients with asymptomatic BCRL or those who did not receive health care. Third, recurrent or metastatic breast cancer was not identified in this study. In this study, those who received more than two breast cancer surgeries or who did not undergo any surgery at all were excluded to prevent the inclusion of patients with late-stage, recurrent, or metastatic breast cancer. Fourth, the relatively short follow-up period of this study is a limitation, although it should be noted that 70 and 90% of BCRL cases occur within 1 and 2 years after surgery, respectively^7,51,52. Fifth, patients who received hormonal therapy had reduced HR for BCRL compared to patients who did not. Since such hormonal therapy is performed for 5 years in most cases, a longer follow-up period may be needed to clearly demonstrate its effect on BCRL.

Conclusion

In the present study, we found that the incidence of BCRL among patients with breast cancer in Korea was 12.2%, with the highest risk observed among patients who underwent MRM. More aggressive surgery, young age, obesity, neoadjuvant or adjuvant chemotherapy, RT, and hyperlipidemia were identified as risk factors for BCRL. Therefore, surgical oncologists should meticulously assess the appropriate surgical approach and consider providing education to patients with risk factors for BCRL to ensure effective prevention strategies.

Ethical approval

The study protocol was approved by the Institutional Review Board (IRB) of the Jeonbuk National University Research Council (IRB no.: CUH 2023-09-014, date of approval: 19 September 2023).

Consent

The IRB waived the need for informed consent.

Sources of funding

None.

Author contribution

H.J.Y. and J.S.K.: concept and design; H.R.A., H.E.J., and H.J.Y.: drafting of the manuscript; S.Y.K., S.H.J., and J.S.K.: critical review of the manuscript for important intellectual content; C.Y.J. and J.S.K.: statistical analysis; H.J.Y.: administrative, technical, or material support; S.H.J.: supervision. All authors contributed in acquisition, analysis, or interpretation of data.

Conflicts of interest disclosure

The authors declare that they have no financial conflicts of interest with regard to the content of this report.

Research registration unique identifying number (UIN)

1. Name of the registry: not applicable.

2. Unique identifying number or registration ID: not applicable.

3. Hyperlink to your specific registration (must be publicly accessible and will be checked): not applicable.

Guarantor

Hyun Jo Youn.

Data availability statement

The study data are available from the corresponding author upon reasonable request.

Provenance and peer review

Not commissioned, externally peer-reviewed.

Supplementary Material

js9-110-3518-s001.docx ^{(31.6KB, docx)}

js9-110-3518-s002.docx ^{(22.8KB, docx)}