Impact of access to effective post-progression therapies on survival outcomes in first-line randomized oncology trials.

Abstract

1500 Background: Effective anticancer systemic therapies are expected to improve overall survival (OS) or quality of life in randomized controlled trials (RCTs) for patients with cancer. However, access to effective post-progression therapies in control arms is inconsistent and may confound the observed OS benefits, especially for agents with known survival benefit in later lines. Here, we investigated whether control-arm post-progression access to effective therapies is associated with improved OS. Methods: This cross-sectional analysis included 52 phase II–III RCTs published between 2010 - 2025 in selected high impact journals. Eligible trials evaluated first line targeted therapies (TT), immune checkpoint blockers (ICB), or hormone therapy (HT) in the experimental arm for advanced lung, breast, or prostate cancer, where the investigational agent had proven OS benefit in later lines. Post-progression access to effective therapies was quantified as the proportion of patients in the control arm receiving the investigational agent (or similar class) relative to all pts randomized to control arm (Ratio_Effective/Control), and relative to patients receiving any subsequent treatment (Ratio_Effective/Therapy). Correlations between the ratio of post-progression access and OS HRs were assessed, with prespecified subgroup analyses by treatment class and tumor type. Geographic distribution was evaluated in trials reporting enrollment by region. Results: Across all trials, the mean proportion of patients in the control-arm receiving any post-progression therapy was 66% (SD 14%), of which 42% (SD 22%) received an effective therapy (or, 62.5%, considering relative numbers). Post-progression access differed significantly by tumor type (p < 0.001), with the lowest rate in breast cancer, followed by prostate and lung cancer. The median proportion of pts enrolled outside the United States and Europe (US/EU) was 36%. The overall median HR for OS was 0.78. No correlation was observed between rate of post-progression access to effective therapies and the HR for OS across studies. However, in TT trials, higher post-progression access to effective treatments was moderately associated with a less pronounced OS benefit (r = 0.44, p = 0.056, Ratio_Effective/Control and r = 0.46, p = 0.046, Ratio_Effective/Therapy). No association was observed in RCTs evaluating ICB. The difference in the correlation between the TT and ICB trials was statistically significant (p = 0.032). No correlation was observed between ratio of effective therapies and tumor type, and enrollment outside of US/EU. Conclusions: Post-progression access to effective therapies in control arms influenced OS benefits in TT trials but not ICB trials in this study. These results warrant caution interpreting TT survival benefits and underscore the need for RCTs to report and standardize post-progression therapy access.