Abstract
PURPOSE: As hypoxia can mediate resistance to immunotherapy, we investigated the safety, tolerability, and efficacy of combining evofosfamide, a prodrug that alleviates hypoxia, with ipilimumab, an immune checkpoint inhibitor, in immunologically "cold" cancers, which are intrinsically insensitive to immunotherapy, as well as in "hot/warm" metastatic cancers that are, atypical of such cancers, resistant to immunotherapy. PATIENTS AND METHODS: In a phase I, 3+3 dose-escalation trial (NCT03098160), evofosfamide (400-640 mg/m(2)) and ipilimumab (3 mg/kg) were administered in four 3-week cycles. The former was administered on days 1 and 8 of cycles 1-2, while the latter was administered on day 8 of cycles 1-4. Response was assessed using immune-related RECIST and retreatment was allowed, if deemed beneficial, after completion of cycle 4 or at progression. RESULTS: Twenty-two patients were enrolled, of whom 21 were evaluable, encompassing castration-resistant prostate cancer (n = 11), pancreatic cancer (n = 7), immunotherapy-resistant melanoma (n = 2), and human papillomavirus-negative head and neck cancer (n = 1). Drug-related hematologic toxicities, rash, fever, nausea, vomiting, and elevation of liver enzymes were observed in > 10% of patients. The most common drug-related grade 3 adverse event was alanine aminotransferase elevation (33.3%). Two patients discontinued ipilimumab and 4 required evofosfamide deescalation due to toxicity. Of 18 patients with measurable disease at baseline, 3 (16.7%) achieved partial response and 12 (66.7%) achieved stable disease. The best responses were observed at 560 mg/m(2) evofosfamide. Preexisting immune gene signatures predicted response to therapy, while hypermetabolic tumors predicted progression. Responders also showed improved peripheral T-cell proliferation and increased intratumoral T-cell infiltration into hypoxia. CONCLUSIONS: No new or unexpected safety signals were observed from combining evofosfamide and ipilimumab, and evidence of therapeutic activity was noted.