Background
Phase 3, randomised trial; N=542 (269 experimental, 273 control); 129 sites, 21 countries. Previously untreated, locally advanced unresectable or metastatic nonsquamous NSCLC. Tiragolumab = anti-TIGIT mAb; combined with atezolizumab (anti-PD-L1) + chemo vs an ACTIVE comparator of pembrolizumab + chemo.
Results
Interventions and follow up: Arm A: Tiragolumab 600 mg + atezolizumab 1200 mg + pemetrexed 500 mg/m2 + carboplatin AUC5 or cisplatin 75 mg/m2, q3w
Arm B: Placebo + pembrolizumab 200 mg + same chemotherapy, q3w
Primary endpoint: Investigator-assessed PFS and OS
mFollow up: 11.8 mo
Results: PFS: 8.3 vs 9.9 mo; HR 1.27 (95% CI 1.02–1.57); P=.99 — not met, favoured control
OS: 18.9 vs 23.1 mo; HR 1.33 (95% CI 1.02–1.73); P=.98 — not met, favoured control
Arm B: Placebo + pembrolizumab 200 mg + same chemotherapy, q3w
Primary endpoint: Investigator-assessed PFS and OS
mFollow up: 11.8 mo
Results: PFS: 8.3 vs 9.9 mo; HR 1.27 (95% CI 1.02–1.57); P=.99 — not met, favoured control
OS: 18.9 vs 23.1 mo; HR 1.33 (95% CI 1.02–1.73); P=.98 — not met, favoured control
Adverse events
Grade 3–4: 61.4% vs 60.7%
Grade 5: 10.1% vs 5.9%
Grade 5: 10.1% vs 5.9%
Conclusions
Tiragolumab + atezolizumab + chemotherapy was not superior — and trended inferior on both PFS and OS — to pembrolizumab + chemotherapy in first-line nonsquamous NSCLC. The trial was terminated.
Key Limitations
Active-comparator design cannot separate failure of TIGIT blockade from possible inferiority of the atezolizumab backbone vs pembrolizumab. Short median follow-up (11.8 mo) at termination. Numerically higher grade 5 AE rate in the experimental arm (10.1% vs 5.9%).
Clinical Context
Third phase 3 miss for tiragolumab (after SKYSCRAPER-01 in PD-L1-high NSCLC and SKYSCRAPER-02 in ES-SCLC); Roche terminated the study and program in this setting. Pembrolizumab + pemetrexed/platinum (KEYNOTE-189) remains the 1L standard for driver-negative nonsquamous NSCLC. The anti-TIGIT class in NSCLC is in serious doubt.