Sorafenib plus dacarbazine in solid tumors: A phase i study with dynamic contrast-enhanced ultrasonography and genomic analysis of sequential tumor biopsy samples

Purpose Improved prognostic accuracy for treatment response and a wider understanding of drug effects in humans are crucial for enhancing the utility of sorafenib and other promising targeted therapies. We developed a strategy of global genomic investigation of sequential tumor biopsy samples at baseline and 21 days post treatment, and applied this approach in a phase I study of sorafenib plus dacarbazine in patients with solid tumors. The objective of this study was also to validate functional parameters of DCE-US as surrogate markers to predict earlier response. Experimental design Patients received 21-day cycles of oral sorafenib, 400 mg twice daily and dacarbazine, 1,000 mg/m² in a 1-h intravenous infusion on day 1. Efficacy was assessed using response evaluation criteria in solid tumors. Sequential biopsy samples (baseline and day 21) were obtained from the same tumor. Changes from baseline in global gene expression (GE) measured by genomic microarrays and in tumor vascularity at baseline, D8, D21, D 42 and every 2 cycles using dynamic contrast-enhanced ultrasonography (DCE-US) were analyzed for patients with and without a clinical response to treatment at 3 months. Results Among 23 patients evaluable for treatment efficacy, 17 were eligible for gene expression and DCE-US analyses. One patient achieved a partial response; 14 exhibited stable disease. Ten patients were defined as exhibiting stable disease (SD) and 7, progressive disease (PD) at 3 months. Genomic analyses identified a 237-gene signature that distinguished SD from PD at 3 months. Of note, CDK4 overexpression and PDGFR downregulation were associated with PD. Functional parameters of DCE-US representing the blood volume at baseline, day 8, and day 21 were correlated with disease progression at 3 months. Conclusions This novel approach of sequential investigations in a phase I trial was feasible, detecting early changes in gene expression and tumor vascularity evaluated using DCE-US that may be predictive of clinical outcome.