ABSTRACT: BACKGROUND:In a colorectal cancer xenograft model, we investigated the therapeutic effect of metformin on tumor hypoxia with [18F]flortanidazole ([18F]HX4) small-animal positron emission tomography (?PET). We also assessed the additive effect of metformin on long-term radiotherapy outcome and we studied the potential of [18F]HX4 as a predictive and/or prognostic biomarker within this setup. METHODS:Colo205-bearing mice (n = 40) underwent a baseline [18F]HX4 hypoxia ?PET/computed tomography (CT) scan. The next day, mice received 100 mg/kg metformin or saline intravenously (n = 20/group) and [18F]HX4 was administered intravenously 30 min later, whereupon a second ?PET/CT scan was performed to assess changes in tumor hypoxia. Two days later, mice were further divided into four therapy groups (n = 10/group): control (1), metformin (2), radiotherapy (3), and metformin + radiotherapy, i.e., combination (4). Then, they received a second dose of metformin (groups 2 and 4) or saline (groups 1 and 3), followed by a single radiotherapy dose of 15 Gy (groups 3 and 4) or sham irradiation (groups 1 and 2) 30 min later. Tumor growth was followed three times a week by caliper measurements to assess the therapeutic outcome. RESULTS:[18F]HX4 uptake decreased in metformin-treated tumors with a mean intratumoral reduction in [18F]HX4 tumor-to-background ratio (TBR) from 2.53 ± 0.30 to 2.28 ± 0.26 (p = 0.04), as opposed to saline treatment (2.56 ± 0.39 to 3.08 ± 0.39; p = 0.2). The median tumor doubling time (TDT) was 6, 8, 41, and 43 days in the control, metformin, radiotherapy and combination group, respectively (log-rank p < 0.0001), but no metformin-specific therapy effects could be detected. Baseline [18F]HX4 TBR was a negative prognostic biomarker for TDT (hazard ratio, 2.39; p = 0.02). CONCLUSIONS:Metformin decreased [18F]HX4 uptake of Colo205-tumors, but had no additive effect on radiotherapy efficacy. Nevertheless, [18F]HX4 holds promise as a prognostic imaging biomarker.
