An exploratory analysis indicates that overall survival did not improve with the use of single-agent atezolizumab in untreated locally advanced or metastatic urothelial cancer.
Results from the phase 3 IMvigor130 trial (NCT02807636) showed overall survival was not improved with atezolizumab (Tecentriq) monotherapy vs placebo plus platinum-based chemotherapy and gemcitabine for patients with untreated locally advanced or metastatic urothelial cancer.1
These results which were reported at the 2023 Genitourinary Cancer Symposium also showed a potential increase in benefit with PD-L1 inhibitor in patients with high PD-L1 expression (IC2/3), particularly in those who were cisplatin ineligible. The results showed that the median OS was 15.2 months (95% CI, 13.1-17.7) in the atezolizumab arm (n = 360) vs 13.3 months (95% CI, 11.9-15.6) in the chemotherapy arm (n = 359; HR, 0.98; 95% CI, 0.82-1.16). The 12- and 24-month OS rates with atezolizumab were 57.9% (95% CI, 52.7%-63.1%) and 34.5% (95% CI, 29.4%-39.5%) vs 54.6% (95% CI, 49.2%-59.9%) and 32.3% (95% CI, 27.3%-37.4%) with chemotherapy, respectively.
“Data from this final OS analysis of the IMvigor130 study of first-line atezolizumab in mUC were generally consistent with the findings from the first and second interim OS analyses. Exploratory efficacy data suggest a clinical benefit with atezolizumab monotherapy as first-line treatment for cisplatin-ineligible patients with PD-L1 IC2/3 mUC,” study co-author Aristotelis Bamias, MD, of the National and Kapodistrian University of Athens in Greece, said in a presentation of the data.
The trial enrolled patients with platinum-eligible, locally advanced or mUC who had not received prior systemic therapy in the metastatic setting. An ECOG performance status of 0 to 2 was also required.
Patients were randomly assigned 1:1:1 to atezolizumab plus platinum and gemcitabine (arm A); atezolizumab alone (arm B); or placebo plus platinum and gemcitabine (arm C).
The co-primary end points of the trial were investigator-assessed progression-free survival (PFS) and OS for arm A vs arm C in the intention-to-treat (ITT) population and OS for arm B vs arm C in the ITT and PD-L1 IC2/3 population.
Per the statistical design of the trial, PFS benefit had to be met for arm A vs arm C in the ITT population before OS could be formally tested in arm A vs arm C (ITT), arm B vs arm C (ITT), and arm B vs arm C (PD-L1 IC2/3), respectively. Prior findings from the primary analysis showed a significant improvement in PFS with atezolizumab/chemotherapy vs placebo/chemotherapy.2 However, because the first and second interim OS analyses proved negative, the comparison of atezolizumab and chemotherapy was exploratory only. Findings from the final OS analysis for arms A and C were presented by Enrique Grande, MD, of MD Anderson Cancer Center Madrid in Spain, during the meeting.3
Secondary end points included investigator-assessed objective response rate (ORR), duration of response (DOR), and PFS for arm B vs arm C. Safety and outcomes in the cisplatin-ineligible population were also evaluated.
Forty-nine months had passed from the last patient randomly assigned to data cutoff. At this time, 19% (n = 67) of patients remained on study in the atezolizumab monotherapy arm vs 14% (n = 49) in the chemotherapy arm. Most study discontinuations in the atezolizumab and chemotherapy arms were due to death, at 73% (n = 264) and 74% (n = 265), respectively.
Median survival follow-up was 13.4 months (14.0 months in arm B and 12.0 months in arm C).
Additional survival data showed an advantage with atezolizumab in patients with higher PD-L1 expression. Median OS was 27.5 months (95% CI, 17.7-49.4) with atezolizumab (n = 88) vs 16.7 months (95% CI, 10.0-26.1) with chemotherapy (n = 85) in patients with PD-L1 IC2/3 (HR, 0.70; 95% CI, 0.48-1.03). Patients with PD-L1 IC0/1 had median OS of 13.5 months (95% CI, 11.1-16.3) with atezolizumab (n = 272) vs 12.9 months (95% CI, 11.6-15.0) with chemotherapy (n = 274; HR, 1.03; 95% CI, 0.85-1.24).
Taken further, investigators showed more pronounced benefit with atezolizumab in cisplatin-ineligible patients with PD-L1 IC2/3 expression. Here, median OS was 18.6 months (95% CI, 14.0-49.4) with atezolizumab (n = 50) vs 10.0 months (95% CI, 7.4-18.1) with chemotherapy (n = 43; HR, 0.56; 95% CI, 0.34-0.91). Cisplatin-ineligible patients with PD-L1 IC0/1 expression had median OS of 11.2 months (95% CI, 6.9-14.7) with atezolizumab (n = 140) vs 11.8 months (95% CI, 10.2-14.3) with chemotherapy (n = 140; HR, 1.14; 95% CI, 0.88-1.48).
Response data were also assessed in the ITT and cisplatin-ineligible PD-L1 IC2/3 populations. In the former population, the ORR was 24.2% (95% CI, 19.9%-29.0%) with atezolizumab vs 44.4% (95% CI, 39.2%-49.7%) with chemotherapy. The median DOR was 29.6 months (95% CI, 15.9–not evaluable [NE]) and 8.1 months (95% CI, 6.3-8.5) with placebo plus chemotherapy, and the disease control rates (DCRs) were 38% (95% CI, 33%-43%) and 59% (95% CI, 54%-64%), respectively.
In the cisplatin-ineligible PD-L1 IC2/3 population, the ORR was 40.0% (95% CI, 26.4%-54.8%) with atezolizumab vs 32.6% (95% CI, 19.1%-48.5%) with chemotherapy. The median DOR was NE (95% CI, 7.2-NE) with atezolizumab vs 6.2 months (95% CI, 4.2-10.9) with chemotherapy; DCRs were not available in this population.
Subsequent non-protocol therapy analysis showed that approximately half of patients in the atezolizumab and chemotherapy arms received at least 1 additional treatment, respectively, including chemotherapy (n = 162, 45% vs n = 114, 32%), immunotherapy (n = 13, 4% vs n = 88, 25%), or treatment with the antibody-drug conjugate enfortumab vedotin-ejfv (Padcev; n = 7, 2% vs n = 5, 1%).
Regarding safety, no new signals were reported with additional follow-up. Treatment-related adverse effects (AEs) and AEs leading to discontinuation were less common with atezolizumab vs chemotherapy, at 61% (n = 217) and 9% (n = 31) with atezolizumab vs 96% (n = 372) and 34% (n = 132) with chemotherapy.
Serious AEs and serious treatment-related AEs occurred in 46% (n = 163) and 12% (n = 44) of patients in the atezolizumab arm vs 50% (n = 196) and 26% (n = 101) in the chemotherapy arm. Notably, no new grade 5 treatment-related AEs were recorded in either arm since the first interim OS analysis.
“Overall, the tolerability profile of atezolizumab monotherapy was favorable relative to that of placebo plus chemotherapy,” Bamias said. “The benefit-risk ratio of atezolizumab vs chemotherapy support atezolizumab as first-line treatment for cisplatin-ineligible patients with PD-L1 IC2/3 mUC,” Bamias concluded.
However, in November 2022, atezolizumab was voluntarily withdrawn from the US market for its first-line indication in metastatic urothelial carcinoma.4
Discussant Andrea B. Apolo, MD, of the National Institutes of Health, added, “Chemotherapy plus checkpoint [inhibitor] combinations have different efficacy in a variety of solid tumors. The combination of chemotherapy plus [a] checkpoint inhibitor did not work with atezolizumab and with pembrolizumab [(Keytruda) in mUC]. The chemotherapy used in the chemotherapy plus checkpoint [inhibitor] combinations may make a difference––cisplatin may be a better choice, but additional data from other trials are pending.
“Hierarchical studies limit the statistical testing of subsequent study arms if the first question asks are negative, so this important to consider when designing these large trials. Checkpoint inhibitor monotherapy is no longer an FDA approved standard-of-care option for the first-line treatment of platinum-eligible patients. There are ongoing trials assessing the efficacy of ADCs plus checkpoint inhibitors vs chemotherapy in the metastatic and neoadjuvant settings.”