Is it the Type of Axillary Surgery or the Number of Removed Lymph Nodes That Increases the Risk of Breast Cancer Related Lymphedema (BCRL)? Results from a Prospective Screening Trial

Abstract

Axillary surgery has been identified as the main risk factor for BCRL regardless the delivery of regional nodal radiation (RLNR). Yet it remains unknown if it is the type of axillary surgery or the number of removed lymph nodes (LN) that increases BCRL risks. Between 2008 and 2021, 3,350 patients (pts) who received surgery for breast cancer were enrolled in a lymphedema screening trial. Patients with bilateral breast cancer or without axillary surgery were excluded. Perometry was used to assess limb volume preoperatively in all patients. BCRL was defined as a ≥10% relative arm-volume increase arising >3 months postoperatively. The cohort was divided by axillary surgery type: axillary lymph node dissection (ALND) and sentinel lymph node biopsy (SLNB). Radiation was delivered using 3D conformal technique and RLNR was defined as the usage of anterior supraclavicular field. No hypofractionation was used and doses ranged between 50 and 50.4 Gy in 25-28 fractions. Multivariable Cox proportional hazard models compared the cumulative incidence of BCRL and local failure between different patient groups. After applying inclusion criteria, 2,623 pts were available with overall median follow-up s of 6.1 years. Of the entire cohort, 709 (27%) had ALND with a median of 16 LN removed, while 1,914 (73%) received SLNB only with a median of 2 LN removed. The median number of malignant LN and patients receiving RLNR was higher in ALND group compared to SLNB only group. Frequency distribution analysis showed that the main overlap between ALND and SLNB only groups happen in the range of 3-11 LN removed. Therefore, the primary analysis focused only on pts with 3-11 LN across both groups (n = 690: ALND n = 140, SLNB n = 550). The multivariable model adjusted for BMI, RLNR, age and breast surgery showed that in this group with 3-11 LN removed in both cohorts, ALND remained significantly associated with BCRL (HR: 4.2, p<0.0001). Separate analyses for the entire SLNB only group and ALND groups were conducted to evaluate if the BCRL risk increases per each removed LN within the same axillary surgery group. The multivariable analysis for SLNB only pts(N = 1,914) showed that for each LN removed the risk of BCRL did not increase significantly (HR:1.06, p = 0.3), similarly for ALND group (N = 709) for each LN removed (HR:1.02, p = 0.08). For pts with pathologic N2-N3 disease and clinical node negative without neoadjuvant chemotherapy receiving ALND, the number of LN removed did not significantly improve neither Local control (HR:1.02, p = 0.8) nor distant disease survival (HR:1.01, p = 0.6). ALND procedure per se is the main risk factor for BCRL not the number of LNs removed. For high-risk pts with >N2 disease, aggressive ALND did not improve tumor outcome. De-escalation with targeted axillary sampling followed by RLNR should be evaluated. Future lymphedema research should account for type of axillary surgery instead of number of LNs removed as a factor. (NCT01521741).