Evaluation of 1-Year vs Shorter Durations of Adjuvant Trastuzumab Among Patients With Early Breast Cancer: An Individual Participant Data and Trial-Level Meta-analysis

IMPORTANCE: The optimum duration of adjuvant trastuzumab among patients with early breast cancer is uncertain but of great therapeutic relevance. OBJECTIVE: To compare shorter durations with 1 year of adjuvant trastuzumab for patients with early breast cancer. DATA SOURCES: PubMed, Embase, Cochrane Central Register of Clinical Trials, and conference proceedings were searched from January 1, 2005, to June 30, 2019, for relevant randomized clinical trials (RCTs). STUDY SELECTION: To be eligible, the trial had to be randomized, compare a shorter duration with 1 year of trastuzumab as adjuvant treatment, and include patients with early breast cancer. DATA EXTRACTION AND SYNTHESIS: Individual patient data for disease-free survival (DFS) and overall survival (OS) were extracted from published survival curves of included RCTs; DFS and OS curves for each trial and the combined population were reconstructed. The DFS and OS hazard ratios (HRs) were estimated from the reconstructed survival curves as well as published estimates. The HR for DFS was used to test noninferiority using the median noninferiority margin of eligible RCTs. This study followed Preferred Reporting Items for Systematic Reviews and Meta-analyses reporting guideline. MAIN OUTCOMES AND MEASURES: The primary outcome was DFS; OS and cardiac toxic effects were secondary outcomes. RESULTS: There were 6 eligible RCTs with a median DFS noninferiority margin of 1.3 (range, 1.15-1.53), 5 of which had extractable individual patient data for 11 376 patients, 1659 DFS events, and 871 deaths. For shorter duration vs 1 year of trastuzumab, the 5-year DFS was 85.42% (95% CI, 84.41%-86.38%) vs 87.12% (95% CI, 86.15%-88.02%) (HR, 1.14; 95% CI, 1.03-1.25, 1-sided P for noninferiority = .004), and OS was 92.39% (95% CI, 91.61%-93.10%) vs 93.46% (95% CI, 92.73%-94.13%) (HR, 1.17; 95% CI, 1.02-1.33). Using trial-level published estimates from 6 RCTs, including 11 603 patients, 1760 DFS events, and 930 deaths, the HR for DFS was 1.15 (95% CI, 1.04-1.26; 1-sided P for noninferiority = .002) and, for OS, 1.17 (95% CI, 1.03-1.33). There was significantly less risk of congestive heart failure with shorter-duration trastuzumab (relative risk, 0.53; 95% CI, 0.38-0.74). CONCLUSIONS AND RELEVANCE: In this study, a shorter duration of adjuvant trastuzumab was noninferior to its 1-year administration and resulted in lower rates of cardiac toxic effects. These results suggest that a shorter duration may be the preferred option for patients with low-risk disease or a predisposition to cardiac toxic effects.