Our study represents the first prospective trial reporting results of the highest dose escalation using doses of 86 Gy at 2 Gy/fraction, for the IMRT treatment of patients with localized intermediate-risk prostate cancer without ADT. Out of 39, 7 patients (18%) reported G2 late GI toxicity, one patient (2.5%) reported G3 late GI toxicity and one patient (2.5%) reported G4 late GI toxicity. In this feasibility study, ≥G2 late GI toxicity was higher than expected from cases treated at our Institute with IMRT at doses of 80 Gy and from the literature [15–18]. However, the observed actuarial ≥ G2 late GI toxicity (21%) was lower to that found in the study RTOG 9406 conducted by Michalski et al.  reporting a rate of ≥ G2 GI complication ranging from 30% to 33% for 24 months at dose level V (78 Gy) but higher than that (4%) reported by Cahlon et al. . The higher observed ≥ G2 late GI toxicity might be due to the lack of specific dose constraints for rectum volume within the PTV and to the fact that also seminals vesicles received the full treatment dose. In fact a statistically significant correlation was observed between dose volume histograms of the volume of rectum enclosed in the PTV and ≥ G2 late GI toxicity. It is worth noting that patients were enrolled in this study before the publication of Quantec , where it is stated that “Reducing the V75 by just 5% from 15% to 10% has a significant impact in the predicted complication probability …” but “the proposed dose–volume constraints might be unachievable … but every effort should be made to be as close as possible to the constraints especially in the high doses”. Nevertheless, methods allowing the reduction of the PTV, such as CBCT and/or markers for IGRT, could further reduce the incidence of rectal toxicity [31, 32], considering that the prostate and the anterior rectal wall, i.e. the area most susceptible to receive an high dose, cannot be seen using EPID images only. In randomized dose-escalation trials employing 3D-CRT the incidence of ≥ G2 late GI toxicity ranged between 17% and 32% [3–7]. This GI toxicity are similar to our results, even if in our trial higher doses were delivered. Moreover, pre-radiotherapy ADT has been reported as a protective factor for GI late toxicity due to the expected reduction of PTV volume .
No patients experienced G4 late GU toxicity and three patients (8%) developed G3 late GU toxicity, two of which were previously treated for urethral stricture. The observed 5-year incidence of ≥ G2 late GU toxicity was 12.8%, which seems comparable to the 5-year actuarial risk (of 16%) reported by Cahlon et al. , and to the 3-year actuarial risk of 19% G2 late GU reported by Fonteyene et al., with doses between 72 Gy and 78 Gy . However, comparisons of patients across study cohorts are difficult and should be interpreted with caution. In particular, the role of hormone therapy in the setting of dose escalation could introduce some bias, thus confounding the analysis, which needs to be evaluated in a randomized trial. The observed five years FFBF of 87%, according to the Phoenix definition, is comparable with the results of 85% reported by Cahlon et al. , using a total dose of 86.4 Gy (1.8 cGy/fraction) in combination with neoadjuvant or concurrent ADT. The true role of androgen deprivation in dose escalation schedules in patients with intermediate prognosis risk is currently unknown, the fact that hormonal therapy was not used in this study did not seem to impact on the outcome, even though, more patients and a longer follow up are needed to clearly state the role of ADT. Cell killing by hormone-therapy could reduce the tumor burden, enhancing local control, and maybe decreasing the rate of distant metastases . Eade et al.  suggested that the use of doses >80 Gy for localized prostate cancer results in better local control and less distant failures when compared to doses <80 Gy, analyzing a cohort of patients free from ADT. In this report, the authors observed a reduced risk of biochemical recurrence of 2.2% at 8 years for the addition of each Gy over 80 Gy and concluded that the plateau on the dose–response curve for prostate cancer lies well above 80 Gy. Also, feasibility studies of single Institutions and some randomized trials of dose escalation showed improved results in the treatment of localized prostate cancer [1–8]; analyzing the effects of increased doses between prognostic categories, the best results are observed in the intermediate risk [3–9, 15, 34–36]. Even though, with a larger number of enrolled patients a multivariate analysis could better clarify the results observed, we believe that the current series demonstrates the advantage in terms of disease control of using ultra-high doses in the treatment of intermediate risk prostate cancer while the incidence of toxicity observed could be lowered by applying stricter requirements on the dose volume constraints at the interface of the rectum with the posterior portion of the prostate gland and introducing a more advanced imaging protocol, i.e. cone beam CT imaging. Moreover, authors are aware that quality of life questionnaires to investigate treatment effects as reported by patients could have added information to the overall rating of treatment results; for this reason, since then, great effort has been made to introduce in our policy also this additional tool of evaluation.