Abstract
PURPOSE: Axillary lymph node dissection (ALND) has traditionally been recommended for breast cancer patients with positive sentinel lymph nodes (SLNs). Although omitting ALND is now widely accepted for patients undergoing breast-conserving surgery (BCS) with limited sentinel lymph node biopsy (SLNB) involvement, based on trials such as Z0011, evidence for patients undergoing total mastectomy (TM) remains limited and conflicting. This meta-analysis aimed to evaluate whether ALND can be safely omitted in TM patients with 1-2 positive SLNs by comparing survival outcomes between ALND and SLNB alone groups. METHODS: A systematic search of PubMed, Embase, Cochrane Library, and Web of Science (up to December 2024) identified 29 studies (6 randomized controlled trials and 23 observational) involving a total of 146 407 patients. Survival outcomes, including overall survival (OS), disease-free survival (DFS), and recurrence-free survival (RFS), were pooled using random- or fixed-effects models. Heterogeneity and publication bias were assessed using I2 statistic, Begg's test, and Egger's test. Subgroup analyses were performed based on study design, type of surgical (TM vs. BCS), and pathological T-stage. RESULTS: Overall, no significant differences in OS (OR = 0.93, 95% confidence interval [CI]: 0.83-1.04), DFS (OR = 1.02, 95% CI: 0.87-1.20), or RFS (OR = 1.08, 95% CI: 0.89-1.30) were observed between ALND and SLNB alone groups. However, in the TM subgroup, ALND was associate with improved OS (OR = 0.75, 95% CI: 0.62-0.90). Similarly, patients with T3-4 tumors demonstrate better OS outcomes with ALND. No significant differences in DFS or RFS were observed across subgroups. CONCLUSION: SLNB alone provides comparable survival outcomes to ALND in early breast cancer (EBC) patients with 1-2 positive SLNs, supporting its safety in those undergoing BCS. However, while our analysis suggests a potential survival advantage with ALND in TM patients and those with advanced T-stage (T3-4), this observation requires cautious interpretation due to potential selection bias, residual confounding from unmeasured variables and confounding by indication where ALND may reflect more intensive multimodal therapy rather than causally improving OS. Therefore, these subgroup findings should not be considered practice-changing at this stage. Further high-quality prospective studies are warranted to validate these associations and optimize patient selection criteria.