128P Long-term outcomes of neoadjuvant immunotherapy plus chemotherapy in early-stage triple-negative breast cancer: An extracted individual patient data and trial-level meta-analysis

Abstract

Neoadjuvant therapy combining immunotherapy (IO) and chemotherapy (CT) is the standard of care in early-stage triple-negative breast cancer (ESTNBC), following the results of the KEYNOTE-522. However, overall survival (OS) gain is uncertain so far. This study aims to evaluate long-term outcomes of neoadjuvant IO plus CT for ESTNBC. We conducted a systematic review using PubMed, Embase, and Cochrane databases until February 13th, 2023. Randomized phase II or III trials evaluating neoadjuvant IO plus CT versus CT alone for ESTNBC were selected. A meta-analysis of extracted individual patient (pt) data (eIPD) was performed to evaluate event-free survival (EFS) and OS outcomes. Individual pt data was reconstructed from reported Kaplan-Meier plots through WebPlotDigitizer and the R package IPDfromKM and further combined. In addition, we conducted a trial-level meta-analysis using fixed and random effects models. Comparisons were made with Cox regression. The literature search resulted in 107 items, and 4 trials with EFS or OS data available were included (Keynote 522, IMpassion 031, I-SPY 2 and GeparNuevo). 1456 patients were included in EFS analysis (900 in IO arm versus 556 in control arm) and 1348 patients in OS analysis (872 versus 476 patients, respectively). The addition of neoadjuvant IO to CT was associated with better EFS and OS in eIPD analysis (Table). Median OS and EFS were not achieved. The trial-level meta-analysis yielded similar results (fixed effects model: HR 0.66, 95%CI 0.48-0.90; random effects model: 0.57, 95% CI 0.33-1.01). Table: 128PeIPD meta-analysisIO armControl armHR95% CIP4-year EFS84%73%0.600.47 - 0.76< 0.0014-year OS90%84%0.670.49 - 0.930.015 Open table in a new tab This meta-analysis suggests that the addition of IO to neoadjuvant CT in ESTNBC provides a significant improvement in EFS (absolut gain of 11% in 4 years) and a possible OS benefit (significant in the eIPD and in the trial-level fixed effects model meta-analysis). Longer follow-up with mature OS data are awaited to confirm these results.