9P Final results from TAIL: Updated long-term safety and efficacy of atezolizumab (atezo) in a diverse population of patients (pts) with previously treated advanced NSCLC

Abstract

In the pivotal OAK study (NCT02008227), atezo (anti–PD-L1) significantly improved OS vs docetaxel in previously treated NSCLC. The phase III/IV TAIL study (NCT03285763) evaluated the safety and efficacy of atezo in pts with previously treated advanced NSCLC, including those generally excluded from key clinical trials. At the primary analysis, the co-primary endpoints, treatment-related (TR) serious adverse events (SAEs) and TR immune-related adverse events (irAEs), occurred in 7.8% and 8.3% of pts, respectively. No new safety signals were seen. Here we report final safety and efficacy data from TAIL. Pts with advanced NSCLC who progressed after 1-2 lines of chemotherapy were enrolled. Pts received 1200 mg of atezo IV once every 3 weeks. Coprimary endpoints were TR SAEs and TR irAEs. Key secondary endpoints included OS, PFS, ORR and DOR. Safety and efficacy in key pt subgroups were also assessed. Of 619 enrolled pts, 615 received atezo. Median age was 64 y, 60% of pts were male and 90% had ECOG PS 0/1. At data cutoff (26 June 2021), median follow-up was 36.1 mo (range, 0.0-42.3) and 473 pts (77%) had died. TR SAEs and TR irAEs (95% CI) occurred in 8.0% (6, 10) and 9.4% (7, 12) of pts, respectively. The most common Grade ≥3 TR SAEs were pneumonitis (1.0%; including 1 Grade 5 event), pericarditis and colitis (both 0.5%). Safety data for the subgroups were similar to those for the overall population; despite a moderately higher incidence of TR SAEs in the ECOG PS 2 subgroup (13.1%; n=8/61) vs the overall population, the incidence of irAEs was similar (8.2%). Efficacy data are shown in the table. With a median follow-up of ≈3 y, no new or unexpected safety signals were seen. These updated data confirm a positive risk-benefit ratio for atezo in the prior-treated NSCLC setting and may prove useful for informing tx decisions in pts generally excluded from pivotal NSCLC trials.Table: 9PPopulationnORR, %mDOR, momOS, mo36-mo OS rate, %All patients61511.516.611.219.6OAK-like42414.414.714.425.4Age ≥ 75 y7613.218.711.817.5Histological subtypeaSquamous15313.79.512.514.1Non-squamous46110.818.410.420.2ECOG PS 2613.38.93.53.6CNS metastases905.6NE5.17.5Autoimmune disease306.7NE10.16.9Prior anti–PD-1 therapy405.011.65.8NERenal impairment7913.912.711.915.8Active/chronic hepatitis B/C1414.320.714.713.3PD-L1 expressionb (OAK-like)Positive (≥1%)14715.016.615.528.4Negative (<1%)1119.08.611.719.1NE, not evaluable. a One patient was missing data. b Expression on tumour cells using any assay and in biomarker-evaluable patients (n=381). Open table in a new tab