32nd Annual Meeting and Pre-Conference Programs of the Society for Immunotherapy of Cancer (SITC 2017): Part Two

Abstract

Background Hypoxia is prevalent in head and neck squamous cell carcinoma (HNSCC), where it limits radiotherapy outcomes and may create an immune-suppressive microenvironment. The hypoxia-activated prodrug (HAP) evofosfamide (TH-302) targets hypoxia by undergoing oxygensensitive reductive activation (Figure 1). Evofosfamide is being clinically evaluated in combination with ipilimumab in solid tumours, including HNSCC (NCT03098160). The present study investigated the efficacy and sensitivity determinants of evofosfamide in HNSCC. Methods Case reports were retrieved from an unpublished phase 2 evofosfamide monotherapy expansion cohort in solid tumours (480 or 575 mg.m2 on days 1, 8 and 15 of 28-day cycles). A collection of 22 human papillomavirus-negative HNSCC cell lines derived from lesions of varying TNM stages was assessed for sensitivity to 3 HAPs - evofosfamide, PR-104A and SN30000 - in addition to cisplatin, 5-fluorouracil and the active metabolite of evofosfamide, bromo-iso-phosphoramide mustard (Br-IPM). Reductive activation of evofosfamide in cultured cells was measured by LC-MS/MS. The lines were molecularly characterised by RNAseq and whole-exome sequencing and their genomic and transcriptomic features compared to public domain HNSCC clinical samples. Molecular predictors of evofosfamide sensitivity were investigated using hierarchical clustering, differential expression and correlation analyses. The antitumour activity of evofosfamide (50 mg.kg-1 qdx5 for 2-3 cycles with/without 10 Gy radiation on day 5 of cycle 1) was evaluated in two HNSCC xenografts and two HNSCC PDX models derived inhouse. Evofosfamide was evaluated in combination with CTLA-4 blockade in the murine SCC-VII model. Hypoxic fraction was assessed using pimonidazole. Results Of 5 metastatic or locally-advanced HNSCC patients who received evofosfamide after failing standard-of-care, two showed partial responses and three showed stable disease. Evofosfamide was highly selective for hypoxic HNSCC cells and more potent and selective than PR-104A or SN30000. Cell line sensitivity to evofosfamide was correlated with Br- IPM and cisplatin but not with PR-104A, SN30000 or 5-FU, indicating distinct sensitivity determinants. Evofosfamide sensitivity was associated with the expression of genes relating to proliferation. Accordingly, a proliferation metagene identified subtypes within the cell lines that were differentially sensitive to evofosfamide. Xenografts chosen on the basis of putative predictive biomarkers (tumour hypoxia, proliferation subtype) showed the expected patterns of response. Two PDX models were also highly responsive to evofosfamide. SCC-VII was refractory to evofosfamide monotherapy but showed increased growth delay when evofosfamide was combined with CTLA-4 inhibition. Conclusions This study provides a rationale for the clinical evaluation of evofosfamide with immunotherapy and/or radiotherapy in genetically defined subsets of HNSCC.