Exploratory subgroup analyses of EV-302: a phase III global study to evaluate enfortumab vedotin in combination with pembrolizumab versus chemotherapy in previously untreated locally advanced or metastatic urothelial carcinoma

Abstract

Background: In the phase III EV-302 study (NCT04223856), enfortumab vedotin (EV) plus pembrolizumab (P) demonstrated superior efficacy and safety versus platinum-based chemotherapy in patients with previously untreated locally advanced/metastatic urothelial cancer (la/mUC). We report the efficacy of EV+P in prespecified subgroups, including those defined by cisplatin eligibility status, the presence or absence of liver metastases, and metastatic disease sites.

Methods: Patients with previously untreated la/mUC were randomly assigned 1 : 1 to receive either EV 1.25 mg/kg and pembrolizumab 200 mg, or gemcitabine plus cisplatin or carboplatin, all intravenously. The two primary endpoints were progression-free survival (PFS) and overall survival (OS). Confirmed objective response rate was one of the secondary endpoints.

Results: Overall, 886 patients were randomized: 442 to EV+P and 444 to chemotherapy. Baseline characteristics were balanced across treatment groups. Efficacy and safety data for the intention-to-treat (ITT) population, along with PFS and OS data for cisplatin-eligible and -ineligible patients, were previously published (Powles et al. N Eng J Med, 2024). In this analysis, EV+P showed benefit across prespecified subgroups that was consistent with the ITT population. OS benefit in the EV+P arm versus chemotherapy was seen across all subgroups, including patients with liver metastases (OS 19.1 versus 10.1 months), patients without liver metastases [OS not estimable (NE) versus 17.9 months], patients with visceral metastases (OS 25.6 versus 13.6 months), and in patients with lymph node-only disease (OS NE versus 27.5 months). In addition, confirmed objective response rate and PFS benefit with EV+P versus chemotherapy was seen across all examined subgroups.

Conclusion: Along with previously published safety data, EV+P demonstrated benefit compared with chemotherapy across all prespecified subgroups, consistent with the ITT population and supporting EV+P as the standard of care for first-line treatment of la/mUC.

Keywords: bladder cancer; enfortumab vedotin; metastatic urothelial carcinoma; pembrolizumab; phase III; subgroup analysis.

Figure 1

Confirmed ORR by BICR in patients who are cisplatin eligible and ineligible, with and without liver metastases, and who have visceral metastases and lymph node–only disease. BICR, blinded independent review committee; CI, confidence interval; EV, enfortumab vedotin; P, pembrolizumab.

Figure 2

DOR by BICR. Data are shown for patients who were (A) cisplatin eligible or (B) ineligible, (C) patients with liver metastases, (D) patients without liver metastases, (E) patients with lymph node-only disease, and (F) patients with visceral metastases. BICR, blinded independent central review; CI, confidence interval; DOR, duration of response; EV, enfortumab vedotin; NE, not estimable; P, pembrolizumab; PD, disease progression.

Figure 3

PFS by BICR and OS in subgroups defined by the presence or absence of liver metastases. Data are shown for patients with liver metastases (A, B) and without liver metastases (C, D). BICR, blinded independent central review; CI, confidence interval; EV, enfortumab vedotin; HR, hazard ratio; NE, not estimable; OS, overall survival; P, pembrolizumab; PFS, progression-free survival. ^aCalculated using stratified Cox proportional hazards model; an HR <1 favors the EV+P arm.

Figure 4

PFS by BICR and OS in subgroups defined by sites of metastases. Data are shown for patients with LN-only disease (A, B) and visceral metastases (C, D). BICR, blinded independent central review; CI, confidence interval; EV, enfortumab vedotin; HR, hazard ratio; LN, lymph node; NE, not estimable; OS, overall survival; P, pembrolizumab; PFS, progression-free survival. ^aCalculated using stratified Cox proportional hazards model; an HR <1 favors the EV+P arm.