Evaluating surrogates for overall survival in the adjuvant treatment of bladder cancer with chemotherapy.

INTRODUCTION: Surrogates for overall survival (OS) can expedite the development of adjuvant treatments for bladder cancer. We evaluated whether disease-free survival (DFS) or distant metastasis-free survival (DMFS) are valid surrogates for OS in patients with muscle-invasive disease treated with cisplatin-based chemotherapy after radical cystectomy.

METHODS: We analyzed individual patient data from 1075 patients enrolled in 9 randomized controlled trials (RCTs) identified by systematic review. These RCTs compared adjuvant cisplatin-based chemotherapy combined with local treatment versus local treatment alone and excluded neoadjuvant chemotherapy. We measured the patient-level association between DFS/DMFS and OS using Spearman's correlation coefficient (ρ), and the trial-level association between hazard ratios (HRs) using R2. For both measures, values close to 1.00 are required for surrogate validation. We assessed the intent-to-treat (ITT) populations and subgroups defined by lymph node status.

RESULTS: The evaluation of DFS in the ITT population showed ρ = 0.89 (95% confidence interval [CI] 0.87-0.90) and R2 = 0.69 (95% CI, 0.34-1.00). Corresponding measures for DMFS were ρ = 0.91 (95% CI, 0.89-0.92) and R2 = 0.90 (95% CI, 0.74-1.00). Patient-level associations were moderate or strong regardless of the lymph node status. At the trial level, DFS displayed weak association with OS in lymph node-positive patients, but associations were strong for lymph node-negative patients and for DMFS.

CONCLUSION: In the adjuvant treatment of bladder cancer with cisplatin-based chemotherapy, DFS is a moderate to strong surrogate for OS, while DMFS is a strong surrogate for OS.