Predictive biomarkers of benefit from chemotherapy in invasive lobular carcinoma of the breast.

Abstract

e12552 Background: Treatment guidelines for breast cancer do not differ by histology. This study explores whether systemic therapy with chemotherapy followed by endocrine therapy (CET) confers an overall survival (OS) benefit vs. endocrine therapy alone (ET) in hormone receptor-positive and HER2-negative (HR+/HER2-) invasive lobular carcinoma (ILC) of the breast, and investigates various potential predictive biomarkers. Methods: Data on female patients with pathologic stage I-III, HR+/HER2- ILC diagnosed between 2010-2017 were extracted from the National Cancer Database (NCDB). A ten-percent year-stratified sample was randomly selected for this study. Potential differences in OS between ET and CET were explored using multivariate Cox Proportional-Hazards with considerations of socio-demographics, histologic subtype (pure lobular vs. mixed ductal-lobular), stage, nodal status, and OncotypeDX score (ODX). Results: N = 17,653 cases of HR+/HER2-, stage I-III ILC were randomly selected. This analysis focused on the 86.5% patients who received ET or CET (n = 15,271). 50.6% had AJCC stage I disease at diagnosis (of which 4.7% were node-positive, or ‘N+’), 35.7% had stage II (of which 45.5% were N+), and 13.6% had stage III (of which 97.8% were N+). 3.2% of cases had high-risk scoring per ODX (≥ 26), 33.2% had low-risk, while for 63.6% ODX data was unknown. CET administration was significantly associated with younger age (47.5% in age < 55, 34.0% in 55-64, 26.1% in 65-71 and 9.7% in ≥72), advanced staging (10.6% stage I, 38.6% stage II and 80.2% in stage III used CET), ODX (13.3% low ODX, 66.6% high ODX and 36.6% unknown ODX) and insurance (37.8% in private and 21.5% in non-private). OS was similar between ET and CET (adjusted 5-year cum hazard: 4.3% vs 4.5%, p = 0.57) in cohort overall. A benefit of CET was not observed in stratified analyses by age, histologic subtype, and ODX (high vs. low vs. unknown). However, stage-stratified analysis revealed an OS benefit with CET in stages II-III ILC. In nodal status-stratified analyses: CET had worse OS vs. ET with N- ILC (adjusted 5-year OS: 95.1% vs 96.4%, p = 0.02), but better OS in N+ (91.6% vs 88.9%, p = 0.008) disease. In further stratified models, the CET benefit was also observed in stage II with N+ (96.1% vs 92.0%, p = 0.02), unknown ODX with N+ (89.6% vs 85.4%, p = 0.001), pure lobular with N+ (95.8% vs 92.7%, p = 0.02) and mixed ductal-lobular with N+ (96.6% vs 91.2%, p = 0.0007), without differences in their N- counterparts. Although benefit of CET was observed in N+, the use of CET in N+ disease varied by overall AJCC staging: 33.1% vs. 57.1% vs. 80.8% in stages I, II and III respectively (p < 0.001). Conclusions: Benefit of CET over ET was observed in stages II and III ILC, particularly in N+ disease. No significant difference was observed in high ODX group. While treatment approach is associated with overall staging, it appears that nodal positivity specifically is most predictive of a benefit of CET over ET.