Background: In the IMvigor130 primary analysis in patients (pts) with untreated locally advanced or mUC, a trend toward favorable OS with atezo (anti-PD-L1) monotherapy (Arm B) vs placebo + platinum/gemcitabine (plt/gem; Arm C) was seen in pts with PD-L1-expressing immune cells on ≥5% of the tumor area (IC2/3 per VENTANA SP142 IHC assay; Galsky Lancet 2020). We report clinical outcomes in Arm B vs C from the second interim OS analysis, including efficacy in cisplatin-ineligible pts.

Methods: Pts were randomized 1:1:1 to Arm A (atezo + plt/gem [not reported here]), B or C. Co-primary endpoint OS was analyzed via hierarchical statistical testing. Since OS in Arm A vs C was not statistically significant, no formal comparison of Arm B vs C was performed in intention-to-treat (ITT) and IC2/3 pts. Objective response rate (ORR) and duration of response (DOR), both per RECIST 1.1 (secondary endpoints), and safety are reported; additional exploratory subgroup analyses of OS by PD-L1 status in ITT and cisplatin-ineligible pts were also done.

Results: Efficacy from the updated OS analysis (June 14, 2020 cutoff; median follow-up, 13.3 mo) is shown in the Table. Exploratory subgroup analyses suggested that OS for IC2/3 pts may be higher in Arm B vs C. In ITT pts, ORR was higher in Arm C, but median DOR was longer in Arm B. In the safety population, Grade (Gr) 3/4 treatment-related adverse events (TRAEs) occurred in 16% and 80% of Arm B and C pts, respectively, and Gr 5 TRAEs in 1% of both arms. Gr 3/4 AEs of special interest (AESIs) occurred in 10% and 4%, respectively, and Gr 5 AESIs in <1% of both arms.

Conclusions: These exploratory efficacy data from an updated OS analysis of IMvigor130 support the benefit of atezo monotherapy as first-line treatment for PD-L1 IC2/3 cisplatin-ineligible mUC. Atezo monotherapy showed a better tolerability profile than chemotherapy. With this longer follow-up, no new safety concerns arose.

IMvigor130 efficacy (second interim OS analysis)

Efficacy analysis populationArm B: atezoArm C: placebo + chemotherapya
ITT   
OS events/pts (%) 241/360 (66.9) 246/359 (68.5) 
Median OS (95% CI), mo 15.2 (13.1, 17.7) 13.1 (11.7, 15.1) 
OS HR (95% CI)b 0.99 (0.83, 1.19) 
Responders/ptsc 84/359 157/356 
ORR (95% CI), % 23.4 (19.1, 28.1) 44.1% (38.9, 49.4) 
Median DOR (95% CI), mod 29.6 (15.9, NE) 8.1 (6.3, 8.5) 
PD-L1 IC2/3 population   
OS events/pts (%) 43/88 (48.9) 52/85 (61.2) 
Median OS (95% CI), mo 27.5 (17.7, NE) 16.7 (10.0, 26.1) 
OS HR (95% CI)b 0.67 (0.45, 1.02) 
PD-L1 IC0/1 populatione   
OS events/pts (%) 198/272 (72.8) 194/274 (70.8) 
Median OS (95% CI), mo 13.5 (11.1, 16.3) 12.8 (11.3, 15.0) 
OS HR (95% CI)f 1.05 (0.86, 1.28) 
Cisplatin-ineligible pts with PD-L1 IC2/3   
OS events/pts (%) 28/50 (56) 30/43 (70) 
Median OS (95% CI), mo 18.6 (14.0, NE) 10.0 (7.4, 18.1) 
OS HR (95% CI)f 0.60 (0.36, 1.01) 
Efficacy analysis populationArm B: atezoArm C: placebo + chemotherapya
ITT   
OS events/pts (%) 241/360 (66.9) 246/359 (68.5) 
Median OS (95% CI), mo 15.2 (13.1, 17.7) 13.1 (11.7, 15.1) 
OS HR (95% CI)b 0.99 (0.83, 1.19) 
Responders/ptsc 84/359 157/356 
ORR (95% CI), % 23.4 (19.1, 28.1) 44.1% (38.9, 49.4) 
Median DOR (95% CI), mod 29.6 (15.9, NE) 8.1 (6.3, 8.5) 
PD-L1 IC2/3 population   
OS events/pts (%) 43/88 (48.9) 52/85 (61.2) 
Median OS (95% CI), mo 27.5 (17.7, NE) 16.7 (10.0, 26.1) 
OS HR (95% CI)b 0.67 (0.45, 1.02) 
PD-L1 IC0/1 populatione   
OS events/pts (%) 198/272 (72.8) 194/274 (70.8) 
Median OS (95% CI), mo 13.5 (11.1, 16.3) 12.8 (11.3, 15.0) 
OS HR (95% CI)f 1.05 (0.86, 1.28) 
Cisplatin-ineligible pts with PD-L1 IC2/3   
OS events/pts (%) 28/50 (56) 30/43 (70) 
Median OS (95% CI), mo 18.6 (14.0, NE) 10.0 (7.4, 18.1) 
OS HR (95% CI)f 0.60 (0.36, 1.01) 

IC, tumor-infiltrating immune cells; NE, not estimable.

aConcurrent control arm, which includes pts randomized in Stage 2 when Arm B was added to the study design.

bStratified per randomization stratification factors: PD-L1 status (ITT only), Bajorin risk factor score/liver metastases, investigator's choice of platinum (cisplatin or carboplatin) before randomization.

cPatients with measurable disease at baseline.

dIn pts who had a confirmed response.

eIC0/1 = PD-L1 IC <5%.

fUnstratified.

Citation Format: Ian D. Davis, Matthew D. Galsky, Xavier Garcia del Muro, Se Hoon Park, Ugo De Giorgi, Boris Alekseev, Marina Mencinger, Kouji Izumi, Javier Puente, Jian-Ri Li, José Ángel Arranz, Aristotelis Bamias, Enrique Grande, Eiji Kikuchi, Almut Mecke, Sanjeev Mariathasan, Xiaodong Shen, Hannah (Xinhui) Huang, Maria De Santis. Updated overall survival (OS) analysis of atezolizumab (atezo) monotherapy vs chemotherapy in untreated locally advanced or metastatic urothelial carcinoma (mUC) in IMvigor130 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr CT040.