Efficacy of anti-PD1 therapy in PD1-high mRNA tumors across multiple cancer types: results from cohort 1 and cohort 2 of the phase II SOLTI-1904 ACROPOLI trial

Immune checkpoint inhibitors (ICIs) have transformed the treatment of solid tumors, yet responses vary across cancer types, highlighting the need for reliable pan-cancer biomarkers. Prior retrospective studies identified PD1 mRNA as a promising predictor of anti-PD1 response, but prospective validation was lacking. The SOLTI-1904 ACROPOLI trial was a non-randomized, open-label, multicenter phase II study designed to evaluate the efficacy of anti-PD1 monotherapy in patients with advanced solid tumors and high PD1 mRNA expression (cohort 1). An exploratory cohort also enrolled PD1-low tumors from cancer types with known ICI sensitivity (cohort 2). Between April 2021 and March 2022, 1003 patients across 33 cancer types were prescreened; 10.6% were classified as PD1-high. A total of 56 PD1-high and 15 PD1-low cases were enrolled to receive spartalizumab (400 mg intravenously every 4 weeks). The primary endpoint was objective response rate (ORR) in PD1-high tumors. Cohorts 1 and 2 were closed early due to discontinuation of spartalizumab development, before reaching full accrual. In the PD1-cohort, which included heavily pretreated tumors (median two prior lines) across 19 histologies, the ORR was 17.9% (95% confidence interval 7.8% to 27.9%). Among patients with tissue samples collected within 12 months of treatment, the ORR increased to 33.3%. The clinical benefit rate (partial response or stable disease ≥24 weeks) was 30.4%. Responses occurred in cancers typically resistant to ICIs, including pancreatic and microsatellite-stable colorectal cancers. No new safety signals were identified. Programmed death-ligand 1 expression by immunohistochemistry was significantly associated with response, whereas exploratory analyses identified additional potential biomarkers, including B cell gene signatures and tumor proliferation markers. PD1 mRNA may help identify immunogenic tumors across cancer types. However, the trial's early closure and exploratory nature warrant further validation. A composite biomarker strategy integrating immune and tumor-intrinsic features may improve patient selection for ICIs.