Country-Weighed Estimate of Prostate Cancer Hypofractionated Radiotherapy Toxicity Risk

Purpose/Objective(s) The crude incidence rate of prostate cancer worldwide varies over 100-fold, from 0.9 (Kuwait) to 109.0 (Martinique) cases per 100000 inhabitants (CONCORD-3). These differences cannot be explained by the screening programs of the respective countries. We hypothesize instead that the variability reflects different genetic-environmental causes, which might putatively modulate the individual response to different fractionation schedules. We performed a meta-analysis of randomized trials to evaluate the relative risk (RR) of gastrointestinal (GI) and genitourinary (GU) grade 2+ toxicity, with and without weighing by countries incidence. Materials/Methods Published trials were identified through PubMed, Google Scholar, Scopus, and Web of Science. Selection criteria were: trials that compared short course hypofractionated radiation (HF) with fraction dose > 2 Gy versus conventional radiation with 1.8–2 Gy/fraction (CF), reporting GI and GU toxicities grade 2+, with a toxicity follow-up ≥ 3 years. Trials with brachytherapy, proton- or neutron-therapy, and dose-intensification trials were excluded. RR was computed from the area under the curves of toxicities over time by digitizing the graphs if available, else by using the published raw rates considering early and late periods. The RR variances were estimated from the number of events and number of patients randomized. Country-weighing assigned a weight proportional to prostate cancer incidence in the country where the trial originated. Two-sided statistical test used the pooled RR variances. RR is reported relative to CF as reference. Results Seven trials matched the selection. The trials originated from 6 countries with incidence from 37.9 to 70.2 per 100000. The incidence did not correlate significantly with the GI’RR and the GU’RR, correlation coefficients were 0.16 (P=0.74) and 0.51 (P=0.24), respectively. The RRs displayed large variability between countries, with GI’RR range from 0.70 (Canada) to 1.56 (US), and GU’RR range from 0.98 (Italy) to 1.60 (Australia). The unweighted pooled GI’RR was 1.10 (95% confidence interval 1.03–1.18), indicating a 10% proportional increase of GI toxicity risk, P=0.004. The unweighted pooled GU’RR was 1.20 (1.13–1.26), indicating a 20% proportional increase of GU risk, P<0.001. The weighed GI’RR and weighed GU’RR were 1.11 and 1.21 respectively. These did not differ significantly from the unweighted RRs, P= 0.97 and 0.92, respectively. Conclusion The study fails to demonstrate a relationship between the incidence of prostate cancer and the risk of toxicity. Instead it finds a significantly increased risk of gastrointestinal and genitourinary toxicity associated with hypofractionation.