What question did this study set out to answer?

March 4, 2026

Perioperative enfortumab vedotin plus pembrolizumab versus neoadjuvant pembrolizumab in cisplatin-ineligible patients with muscle-invasive bladder cancer (MIBC): An IPD-based comparative analysis of KEYNOTE-905/EV-303 and PURE-01.

Key Points

To compare the efficacy of perioperative enfortumab vedotin plus pembrolizumab versus neoadjuvant pembrolizumab in treating cisplatin-ineligible muscle-invasive bladder cancer.
Performed an indirect comparison of clinical trials using individual patient data (IPD) from Kaplan–Meier curves.
Endpoints focused on event-free survival (EFS) and overall survival (OS).
Utilized statistical analyses including log-rank test and Cox proportional hazards model.
Conducted analyses using RStudio with patient-level dataset combining KEYNOTE-905/EV-303 and PURE-01.
Pathologic complete response was higher in the enfortumab vedotin plus pembrolizumab group (57.1%) compared to the pembrolizumab group (39.5%).
No statistically significant difference in event-free survival between treatment groups (HR 0.75, p = 0.181).
EFS at 12 months was 77.6% for enfortumab vedotin plus pembrolizumab and 84.5% for pembrolizumab.

Abstract

779 Background: Cisplatin-based chemotherapy is the standard neoadjuvant therapy for patients (pts) with MIBC. However, a significant proportion of pts are ineligible for/refuse chemotherapy, requiring alternative perioperative strategies. PURE-01 tested neoadjuvant pembrolizumab (Pembro) monotherapy before radical cystectomy (RC) in pts with cT2–T4N0M0 MIBC, without an adjuvant period. KEYNOTE-905/EV-303 is a randomized, phase III study testing perioperative enfortumab vedotin (EV) plus Pembro (EVP), followed by RC and adjuvant EVP for 6 cycles, then Pembro alone for additional 6 cycles, versus RC alone. EV-303 included a third arm of Pembro monotherapy, which was stopped in 2022. The contribution of both therapeutic components in the EVP regimen and the neoadjuvant vs adjuvant part remains uncertain. Methods: An indirect comparison of trials was performed using IPD reconstructed from Kaplan–Meier curves. The endpoints were event-free survival (EFS) and overall survival (OS) since initiation of treatment. Individual pt-level data (IPD) were retrieved using a graphical reconstructive algorithm and merged into a single dataset including two treatment groups. EFS was estimated with Kaplan–Meier curves and compared with the log-rank test and the Cox proportional hazards model. Landmark estimates at 12- and 24 months and median values were calculated. Analyses were conducted in RStudio (v.2025.05.1+513). Results: A total of 170 pts from KEYNOTE-905/EV-303 (EVP arm) and 155 from PURE-01 were included. Baseline staging was cT2 in 17.6% of patients in the EVP group and 48.4% in the Pembro group; cT3/T4 in 78.2% and 51.6%, respectively; 4.1% and 0%, respectively, had a clinical N1 stage. Pathologic complete response (pCR) was higher in the EVP group than in the pembrolizumab group (57.1% with 95% CI, 49.3–64.6 and 39.5%, respectively). The IPD-based Cox analysis showed no significant difference in EFS (HR 0.75, 95% CI 0.49–1.14; p = 0.181). Conclusions: An indirect IPD-based comparison between EVP and Pembro did not result in statistically significant differences in EFS. These results emphasize the importance of conducting clinical trials aimed at discriminating the contribution of components that remain uncertain, representing a critical need. Limitations include the retrospective design, indirect comparison, patient, and geographical heterogeneity. Endpoint Time (months) EVP (%) 95% CI Pembro (%) 95% CI EFS 6 90.7 (86.4-95.3) 87.1 (82–92.5) 12 77.6 (71.3–84.4) 84.5 (78.5–90.9) 24 74.1 (67.2–81.6) 71.7 (62.7–82.0) Median NR (37.2-NR) NR (NR-NR)

AIに質問

Bookmark