Abstract
Background/aim
Bladder-preserving therapies are increasingly important for patients with bacillus Calmette-Guérin (BCG)-unresponsive non-muscle-invasive bladder cancer with carcinoma in situ (CIS) who are unfit for or decline radical cystectomy. Because sustained complete response (CR) is critical for bladder preservation, we compared duration of response (DoR) among CR responders treated with TAR-200 or nadofaragene firadenovec using reconstructed individual patient data from published Kaplan-Meier curves.Patients and methods
Published DoR Kaplan-Meier curves for CR responders were digitized from SunRISe-1 cohort 2 evaluating TAR-200 monotherapy and from the phase III NCT02773849 CIS cohort evaluating nadofaragene firadenovec. Reconstructed individual patient data (rIPD) were reconstructed using IPDfromKM package. Restricted mean survival time (RMST) for DoR was estimated at 6, 12, 18, and 24 months. A Weibull mixture cure model was fitted as an exploratory analysis to estimate the durable CR fraction.Results
RMST differences for DoR between TAR-200 and nadofaragene firadenovec were not statistically significant through 24 months. RMST differences were -0.29 months at 6 months, -0.04 months at 12 months, +0.60 months at 18 months, and +1.34 months at 24 months, with confidence intervals crossing zero at all timepoints. In the exploratory Weibull mixture cure model, the estimated durable CR fraction was larger for TAR-200 than for nadofaragene firadenovec (0.521 vs. 0.256; difference +0.265).Conclusion
In this exploratory rIPD-based cross-trial analysis, CR durability among responders did not differ significantly between TAR-200 and nadofaragene firadenovec through 24 months. Later RMST point estimates and model-based durable CR fraction estimates were numerically higher for TAR-200, but these findings require cautious interpretation because of cross-trial comparability limitations, responder-conditioned selection, and limited follow-up.