Session Item

The phase 3 multicenter randomized controlled FLAME trial (NCT01168479) compared effectiveness and toxicity  of external beam radiotherapy (EBRT) with and without a simultaneous integrated boost up to 95Gy to the macroscopic tumor in intermediate and high risk prostate cancer patients. Earlier, we showed that cumulative toxicity rates up to two years were comparable for the standard and dose-escalated arm. The aim of the present study was to investigate the association between radiation dose to the rectum and gastro-intestinal (GI) toxicity (≥ grade 2) in the FLAME study cohort.

All 571 patients with intermediate and high risk prostate cancer of the FLAME trial were included in this study. The dose-volume effect of the absolute and relative rectal dose volume histogram (DVH) parameters (D2cc and D50%) in relation to the cumulative GI toxicity ≥ grade 2 during the first 4 years of follow-up was assessed. A mixed model analysis for repeated measurements was used, including the total study cohort irrespective of randomization. We performed both crude analyses and analyses adjusted for age, cardiovascular disease (CVD), diabetes mellitus (DM), T-stage, baseline toxicity ≥ grade 1, institute and hormonal therapy.  We calculated cumulative incidences for overall GI toxicity ≥ grade 2 and toxicity per GI subdomain. Finally, we assessed the crude association between D2cc and D50% on the GI toxicity endpoints rectal pain, proctitis, fecal incontinence, diarrhea and rectal bleeding. Adjusted odds ratios (OR) for the subdomains were not assessed because of the low prevalence of toxicity per subdomain.

Four years following treatment, the cumulative incidence for GI toxicity ≥ grade 2 was 22.4%. The cumulative incidences for acute and late GI toxicity ≥ grade 2 were 13.1% and 11.6%, respectively. A dose-toxicity effect for both D2cc and D50% was observed with un- and adjusted ORs of 1.05 and 1.09 for cumulative GI toxicity, respectively (Table 1), indicating higher odds of GI toxicity ≥ grade 2 per increase in rectal dose with steps of 1Gy. For the subdomains, an association with proctitis and fecal incontinence for D2cc and D50% was found, with the strongest association for the rectal D50%. The effect estimates for the diarrhea and rectal bleeding domains are not presented since the model did not converge because of very low prevalence ≥ grade 2 toxicity at some time points.

Table 1. Mixed model for repeated measurements GI toxicity versus DVHs in patients with GI toxicity ≥ grade 2 versus patients with < grade 2 toxicity

Rectal dose-volume effects on GI toxicity were observed for D2cc and D50% DVH parameters. The range in rectal dose in the FLAME study was limited due to strict rectal dose constraints that were prioritized over focal boost dose. Nevertheless, even in the small range of dose-variation, a significant dose-toxicity effect between rectal DVH parameters and GI toxicity was observed. Therefore, further increasing the rectal dose is not desirable.