4485 - Primary efficacy results from B-F1RST, a prospective Phase II trial evaluating blood-based tumour mutational burden (bTMB) as a predictive biomarker for atezolizumab (atezo) in 1L non-small cell lung cancer (NSCLC)

Background

Blood and tissue TMB have shown promise in predicting benefit from PD-L1/PD-1 inhibitors. High bTMB enriched for a PFS benefit in patients (pts) treated with atezo monotherapy in retrospective analyses from the randomized, 2L NSCLC Ph III OAK and Ph II POPLAR studies. We report primary results from the fully enrolled prospective, Ph II, B-F1RST study evaluating a novel bTMB assay as a predictive biomarker for atezo in 1L NSCLC.

Methods

Of 152 ITT pts, 119 comprised the biomarker-evaluable population (BEP) that had adequate circulating tumor (ct)DNA (maximum somatic allele frequency [MSAF] ≥ 1%); 29 had inadequate ctDNA (non-BEP [MSAF <1%]). A prespecified bTMB cutoff (high ≥ 16; low < 16) was used to evaluate clinical efficacy. Coprimary endpoints were ORR and PFS. Statistical tests were 2-sided at a 0.1 level.

Results

With minimum follow-up of ≥6 mo, confirmed ITT ORR was 14.5% (22/152). ORR in BEP was 10.1% (12/119) and non-BEP was 34.5% (10/29). In bTMB high vs low pts, ORR was 28.6% (8/28) vs 4.4% (4/91) and mPFS was 4.6 vs 3.7 mo; HR=0.66 (90% CI, 0.42–1.02; P=0.12). mPFS at exploratory bTMB cutoffs are presented in the Table. bTMB high vs low mOS was NE vs 13.1 mo; HR=0.77 (90% CI, 0.41–1.43; P=0.48). 13% of pts had treatment-related serious AEs and 20% had treatment-related Gr 3/4 AEs. 15% of pts had AEs leading to discontinuation.

Conclusions

The B-F1RST primary analysis is the first full prospective dataset evaluating clinical utility of bTMB as a predictive biomarker for 1L pts receiving atezo monotherapy. Consistent with interim data, pts at the prespecified bTMB ≥16 cut-off had numerical benefit for PFS, ORR and OS. Pts will be followed for ≥18 mo per protocol. Additional exploratory analyses on the bTMB biomarker will be reported.

Clinical trial identification

NCT02848651

Editorial Acknowledgement

Christopher Lum, PhD, Health Interactions