Comprehensive Cancer Network guideline 3 acknowledges the lack of representation of T1a and T1b tumors in prior randomized trials and, thus, recommends the consideration of chemotherapy for tumors 1 cm or smaller at the discretion of clinicians. In the absence of randomized trial data on the role of chemotherapy for small tumors, the cutoff size at which chemotherapy should be omitted remains unclear. Using a national-level hospital registry, we conducted an observational cohort study to address this knowledge gap.
MethodsThe US National Cancer Database was queried for female patients with HR-positive, ERBB2-positive, pT1a-bN0 breast cancer diagnosed between 2010 and 2015 who received hormone therapy with or without chemotherapy. Ethical approval was waived by the Roswell Park Comprehensive Cancer Center institutional review board, because the National Cancer Database is a deidentified data set. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.The Kaplan-Meier method and Cox multivariable analysis were performed to analyze overall survival. Propensity score matching was based on the nearest neighbor method in a 1:1 ratio without a replacement. The standardized difference between variables was less than 0.1, indicating appropriate matching. 4 All P values were 2-sided, with P < .05 considered statistically significant. R statistical software version 3.6.1 (R Project for Statistical Computing) was used for all analyses. Data analysis was performed from November 2019 to January 2020. Additional details are shown in the eAppendix in the Supplement.
ResultsA total of 10 065 patients (median [interquartile range] age, 59 [51-67] years) were identified, including 5346 patients who received chemotherapy and 4719 patients who did not (ERBB2-directed therapy was coded distinctly from chemotherapy during 2013 to 2015, and only 15% of such patients underwent either chemotherapy or ERBB2-directed therapy alone; data not shown) (Table ). The median (interquartile range) follow-up was 41.8 (24.3-62.6) months. On multivariable analysis, multiagent chemotherapy was associated with improved overall survival (hazards ratio [HR], 0.69; 95% CI, 0.52-0.90; P = .006), and tumor size as a continuous variable was associated with worse mortality (for every 1-mm increase, HR, 1.07; 95% CI, 1.03-1.12; P = .002). There was a statistically significant interaction between multiagent chemotherapy and tumor size (P for interaction = .02).Cox multivariable analysis was repeated with each tumor size cutoff ranging from 2 mm to 9 mm, and an 8-mm cutoff was statistically significant (P for interaction = .01), with a large effect size and narrow 95% CI on subgroup analysis. Multiagent chemotherapy was not associated with improved