Aim: To assess the predictors of the onset of impotence 1 year after radiotherapy for prostate cancer.
Patients and methods: In a multi-centric prospective study, the International Index of Erectile Function (IIEF) questionnaire-based potency of 91 hormone-naïve and potent patients (IIEF1-5 > 11 before radiotherapy) was assessed. At the time of this analysis, information on potency 1 year after treatment was available for 62 of 91 patients (42 treated with hypofractionation: 2.35-2.65 Gy/fr, 70-74.2 Gy; 20 with conventional fractionation: 74-78 Gy). Prospectively collected individual information and Dmax/Dmean to the penile bulb were available; the corresponding 2 Gy-equivalent values (EQD2_max/EQD2_mean) were also considered. Predictors of 1‑year impotency were assessed through uni- and multi-variable backward logistic regression: The best cut-off values discriminating between potent and impotent patients were assessed by ROC analyses. The discriminative power of the models and goodness-of-fit were measured by AUC analysis and the Hosmer-Lemeshow (H&L) test.
Results: At 1‑year follow-up, 26 of 62 patients (42 %) became impotent. The only predictive variables were baseline IIEF1-5 values (best cut-off baseline IIEF1-5 ≥ 19), Dmax ≥ 68.5 Gy and EQD2_max ≥ 74.2 Gy. The risk of 1‑year impotence may be predicted by a two-variable model including baseline IIEF1-5 (OR: 0.80, p = 0.003) and EQD2_max ≥ 74.2 Gy (OR: 4.1, p = 0.022). The AUC of the model was 0.77 (95% CI: 0.64-0.87, p = 0.0007, H&L: p = 0.62). The 1‑year risk of impotency after high-dose radiotherapy in potent men depends on the EQD2_max to the penile bulb and on baseline IIEF1-5 values.
Conclusion: A significant reduction in the risk may be expected mainly when sparing the bulb in patients with no/mild baseline impotency (IIEF1-5 > 17).
Keywords: Dose-response; Erectile dysfunction; Predictive models; Prostate cancer; Radiotherapy.