1243P - Tumour mutation burden (TMB), PD-L1, IFN-γ signaling identify subgroups of patients (pts) who benefit from durvalumab (D, anti-PDL1) or D and tremelimumab (T, anti-CTLA4) treatment in urothelial bladder cancer (UC)

Abstract

BackgroundWe conducted NGS of baseline UC tumors to elucidate molecular predictors of clinical outcomes on D/D+T. MethodsCP1108/NCT01693562 was a nonrandomized phase I/II trial including 201 2L UC pts who were treated with D (10mg/kg Q2W). Study 10/NCT02261220 was a phase I trial including 190 2L UC pts treated with D (20mg/kg Q4W) +T (1mg/kg Q4W). Pts with ≥25% PD-L1 expression in tumor or immune cells were scored as PD-L1+ by IHC. RNA-seq (n=62) and whole exome sequencing (WES; n=37) were used in CP1108. The IFN-γ gene signature (IFNGS) was calculated as described (Higgs B et al., 2018). TMB was calculated as somatic coding mutation count per mega base from WES data with TMB≥top tertile as TMB high. FMI assay (n=62) was conducted in Study 10 with TMB≥top tertile as TMB high. ctDNA dynamics using Guardant360 were calculated using an established method (Raja R, et al., 2018). Wilcoxon, log-rank tests and COX-PH model were used in relevant statistical tests. ResultsAmong ITT pts, 55% and 44.5% pts were PD-L1+ in CP1108 and Study 10 respectively. PD-L1 status correlated with IFNGS (p<0.001), whereas PD-L1 did not correlate with TMB. PD-L1+ correlated with better OS in CP1108 (HR=0.46, p<0.001) but not in Study 10 (HR=0.75, p=0.2). IFNGS significantly correlated with better PFS in CP1108 (HR=0.5, p=0.02). Higher TMB in CP1108 using the median cutoff trended towards better OS (HR=0.51, p=0.18), but no clear trend using the top tertile cutoff due to small sample size. Higher TMB in Study 10 correlated with improved OS (HR=0.34, p=0.01). In Study 10, TMB and PD-L1 double positive pts (n=14, 24%) had best OS and pts low in both had worst (n=19, 32%, HR=0.23, p=0.01). Clearance of ctDNA after treatment correlated with better OS (HR=0.23, p=0.006) and PFS (HR=0.33, p=0.01) in combined UC pts from the 2 studies. ConclusionsBoth PD-L1 and TMB predict for survival benefit in UC pts treated with D/D+T. They identify overlapping but generally distinct patient populations. Double highs perform the best. ctDNA loss following treatment can be predictive for better survival. Clinical trial identificationNCT01693562; NCT02261220. Legal entity responsible for the studyAstraZeneca. FundingAstraZeneca. DisclosureC. Massard: Advisory / Consultancy: Amgen; Advisory / Consultancy: Astellas; Advisory / Consultancy: AstraZeneca; Advisory / Consultancy: Bayer; Advisory / Consultancy: Celgene; Advisory / Consultancy: Genentech; Advisory / Consultancy: Ipsen; Advisory / Consultancy: Novartis; Advisory / Consultancy: Roche; Speaker Bureau / Expert testimony: Sanofi; Advisory / Consultancy: Janssen; Advisory / Consultancy: Lilly; Advisory / Consultancy: Pfizer; Advisory / Consultancy: Orion. H. Si: Full / Part-time employment: AstraZeneca. Q. Zhang: Full / Part-time employment: AstraZeneca. B. Higgs: Shareholder / Stockholder / Stock options, Full / Part-time employment: AstraZeneca. R. Raja: Shareholder / Stockholder / Stock options, Full / Part-time employment: AstraZeneca. S.E. Abdullah: Travel / Accommodation / Expenses, Shareholder / Stockholder / Stock options, Full / Part-time employment: AstraZeneca. A. Gupta: Shareholder / Stockholder / Stock options, Full / Part-time employment: AstraZeneca; Shareholder / Stockholder / Stock options, Licensing / Royalties, Use of immunotherapy in the treatment of cancer-patent: Bristol-Myers Squibb. W. Li: Shareholder / Stockholder / Stock options, Full / Part-time employment: AstraZeneca. M. van der Heijden: Advisory / Consultancy: AstraZeneca; Advisory / Consultancy, Research grant / Funding (institution): BMS; Advisory / Consultancy: MSD; Advisory / Consultancy, Research grant / Funding (institution): Roche/Genentech; Advisory / Consultancy, Research grant / Funding (institution): Astellas; Advisory / Consultancy: Seattle Genetics.