Evaluation of Pathologic Complete Response as a Surrogate for Long-Term Survival Outcomes in Triple-Negative Breast Cancer

Abstract

Background: Pathologic complete response (pCR) is a common efficacy endpoint in neoadjuvant therapy trials for triple-negative breast cancer (TNBC). Previous studies have shown that pCR is strongly associated with improved long-term survival outcomes, including event-free survival (EFS) and overall survival (OS). However, the trial-level associations between treatment effect on pCR and long-term survival outcomes are not well established. This study sought to evaluate these associations by incorporating more recent clinical trials in TNBC. Methods: A literature review identified published randomized controlled trials (RCTs) of neoadjuvant therapy for TNBC that reported results for both pCR and EFS/OS. Meta-regression models were performed to evaluate the association of treatment effect on pCR and EFS/OS. Sensitivity analyses were conducted to assess the impact of divergent study designs. Results: Ten comparisons from 8 RCTs (N=2,478 patients) were identified from the literature review. The log (odds ratio) of pCR was a significant predictor of the log (hazard ratio) of EFS (P=.003), with a coefficient of determination of 0.68 (95% CI, 0.41–0.95). There was a weaker association between pCR and OS (P=.18), with a coefficient of determination of 0.24 (95% CI, 0.01–0.77). Consistent results were found in the exploratory analysis and sensitivity analyses. Conclusions: This is the first study that has shown a trial-level association between pCR and survival outcomes in TNBC. By incorporating the most up-to-date RCTs, this study showed a significant trial-level association between pCR and EFS. A positive association between pCR and OS was also recorded.