A non-comparative, randomized, phase II trial of atezolizumab or atezolizumab plus tiragolumab for programmed death-ligand 1-positive recurrent cervical cancer (SKYSCRAPER-04)

Objective: To evaluate tiragolumab (anti-TIGIT) and atezolizumab (anti-PD-L1) as second- or third-line therapy for PD-L1-positive persistent/recurrent cervical cancer. Methods: In the open-label, non-comparative, randomized phase II SKYSCRAPER-04 trial (NCT04300647), patients with PD-L1-positive (SP263 tumor area positivity ≥5%) recurrent/persistent cervical cancer after 1-2 chemotherapy lines (≥1 platinum-based) were randomized 3:1 to atezolizumab 1200 mg with/without tiragolumab 600 mg every 3 weeks until disease progression or unacceptable toxicity. Stratification factors were performance status, prior (chemo)radiotherapy, and disease status. The primary endpoint was independent review committee-assessed confirmed objective response rate per RECIST v1.1 in patients receiving tiragolumab plus atezolizumab. An objective response rate ≥21% (one-sample z-test p≤0.0245) was required for statistical significance versus a historical reference. Results: Protocol-defined independent review committee-assessed objective response rates were 19.0% (95% CI 12.6 to 27.0) in 126 patients receiving tiragolumab plus atezolizumab (p=0.0787 vs historical reference) and 15.6% (95% CI 6.5 to 29.5) in 45 atezolizumab-treated patients. Response rates were higher in PD-L1high (tumor area positivity ≥10%) than PD-L1low (tumor area positivity 5%-9%) subgroups with both regimens. At 8.5 months' median follow-up, independent review committee-assessed progression-free survival was 2.8 months (95% CI 1.7 to 4.1) with tiragolumab plus atezolizumab and 1.9 months (95% CI 1.5 to 3.0) with atezolizumab. In post hoc analyses (10.4 months' median follow-up), median overall survival was 11.1 months (95% CI 9.6 to 14.5) with the combination and 10.6 months (95% CI 6.9 to 13.8) with atezolizumab (crossover permitted). In the combination group, 3% of patients had adverse events requiring treatment discontinuation and 8% had grade ≥3 adverse events of special interest; corresponding values in the single