Systematic versus sentinel-lymph-node-driven axillary-lymph-node dissection in clinically node-negative patients with operable breast cancer. Results of the GF-GS01 randomized trial.

ABSTRACT

PURPOSE: Sentinel-lymph-node (SLN) resection seems to minimize systematic axillary-lymph-node dissection (sALND) side effects in operated breast cancer patients. We explored whether SLN resection achieves similar therapeutic outcomes as sALND but with fewer side effects. METHODS: A randomized, controlled, open-label trial with parallel-group design compared sALND restricted to cases with positive SLN biopsy (test arm, n = 774) versus SLN biopsy followed by sALND (control arm, n = 770). RESULTS: The five-year overall survivals in control and test arms were 96.42 and 95.64% (P = 0.2925). The estimated difference was nearly zero (precisely, - 0.79%, one-tailed 95% confidence interval (CI) limit - 2.44%). In a multivariate Cox model, the adjusted hazard ratio in the test arm was HR 0.81 (upper 95% CI limit 1.17). Advanced age (HR 1.05 per additional year, CI [1.03-1.08]), negative progesterone receptor (HR 2.17 [1.35-3.45]), SLN metastasis (HR 1.69 [1.03-2.79]), and only one SLN identification technique (HR 4.14 [1.21-14.18]) were associated with lower survival. Patients with ≥ 1 severe side effect at 1 month in control and test arms were 173/703 = 24.6% [21.5-28.0%] and 91/693 = 13.1% [10.7-15.9%] (P < 0.001). The estimated sensitivity of SLN biopsy (control arm) was 145/178 = 81.5% [74.8-86.7%]. CONCLUSIONS: Restricting ALND to cases with positive SLN biopsy does not affect the overall survival but reduces by 11.5% [7.5-15.6%] (P < 0.001) the risk of severe short-time side effects of sALND.