Association between event-free survival and overall survival in early-stage triple-negative breast cancer.

ABSTRACT

Aim: This study evaluated event-free survival (EFS) as a surrogate outcome for overall survival (OS) in neoadjuvant therapy for early-stage triple-negative breast cancer (eTNBC). Methods: Meta-regression analyses based on a targeted literature review were used to evaluate the individual- and trial-level associations between EFS and OS. Results: In the individual-level analyses, 3-year EFS was a significant predictor of 5-year OS (p < 0.01; coefficient of determinations [R2] 0.82 [95% CI 0.68-0.91]). Additionally, there was a statistically significant association between the treatment effect on EFS and OS at the trial level (p < 0.001; R2 0.64 [95% CI 0.45-0.82]). Conclusion: This study demonstrates significant associations between EFS and OS and suggests that EFS is a valid surrogate for OS following neoadjuvant therapy for eTNBC.

What is this article about? Studies of cancer therapies typically use patient survival to understand whether a treatment is helpful, such as overall survival (time from treatment to death) and event-free survival (time from treatment until the cancer progresses). Only using overall survival can slow clinical trials and the ability to assess whether new treatments may be useful. This study examined whether event-free survival was a good surrogate outcome for overall survival in studies of neoadjuvant therapy for early stage, triple-negative breast cancer (eTNBC). Neoadjuvant therapy is used to shrink a tumor before the definitive surgery, and TNBC is a type of breast cancer lacking three common hormone receptors that treatments target. To accomplish this, we first searched for published clinical trials and observational studies that reported overall and event-free survival and extracted their data. Then we tested the association between the two survival outcomes to determine if event-free survival could be used to accurately predict overall survival. Using data from randomized clinical trials, we also tested whether a treatment's effect on event-free survival could predict its effect on overall survival. What did this study find? We found that event-free survival at three years could predict overall survival at 5 years, and that there was a meaningful relationship between a treatment's effect on event-free and overall survival for eTNBC following neoadjuvant treatment. What do the results of the study mean? The results suggest that event-free survival is an accurate and useful surrogate for overall survival following neoadjuvant treatment of eTNBC.