The presence and extent of extracapsular extension (ECE) and the need for axillary lymph node dissection (ALND) in patients who meet ACOSOG Z11 eligibility criteria.

Abstract

1019 Background: Whether ECE mandates ALND in patients with ≤2 positive sentinel nodes (SN) is controversial. ACOSOG Z11 excluded patients with matted nodes, but did not comment on microscopic ECE. In a prospective, consecutive series of patients, we sought to determine if ECE correlates with the number of positive axillary lymph nodes (LN) and if ECE ≤2mm clinically differs from ECE >2mm. Methods: In 8/2010 an institutional treatment algorithm based on the Z11 results was prospectively applied to consecutive patients having BCS. ALND was performed for ≥3 +SNs. The approach to ECE was not specified. Characteristics of patients with and without ECE were compared with Fisher’s exact test and the Wilcoxon rank sum test. Results: From 8/10-11/12, 2157 invasive breast cancer patients had BCS; 381 had LN metastasis, 287 met Z11 selection criteria, and ALND was avoided in 242 (84%). ECE was present in 111 (39%), of whom 23% had ≥3 +SNs (vs 2% without ECE; p<0.0001) and 35% had ALND (vs 3% without ECE; p<0.0001). The presence of ECE was associated with tumor size (1.9cm vs 1.6; p=.01) but not with age, grade, or receptor status. The degree of ECE was associated with age, grade, number of +SNs, and performance of ALND (Table). In 45 cases, ALND was advised for ≥3 +SNs (n=29) or <3 +SNs with ECE (n=16). 39 patients had ALND and 34 of these had ECE. Additional +LNs were seen in 5/9 patients with ≤2mm ECE and 20/25 with >2mm ECE; median of 1 additional +LN in each group. Seven or more additional +LNs were seen in 6 patients with >2mm ECE; 1 patient with ≤2mm ECE had 6 additional +LNs, the remainder had ≤3. Conclusions: The presence of ECE was associated with ≥3 +SNs and the need for ALND. Only a minority of patients with ≤2mm of ECE had ≥3 +SNs, and nodal disease at ALND in this group was limited, suggesting that ≤2mm ECE may not be an indication for ALND. [Table: see text]