Extreme hypofractionation for high-risk prostate cancer: Dosimetric correlations with rectal bleeding

We explored the association of dosimetric parameters with late rectal bleeding among high-risk prostate cancer patients treated with hypofractionated simultaneous in-field boost (H-SIB) to prostate with nodal treatment.

Rectal toxicity results and dose-volume histogram (DVH) information from patients treated on FASTR and SATURN were combined. Patients in both trials received long-term androgen deprivation and H-SIB with prescription dose 40 Gy to the prostate and proximal seminal vesicles and 25 Gy to the lymph nodes delivered over 5 weekly fractions using image guidance with cone beam computed tomography. Mean rectal DVH values at 5-Gy intervals and mean DVH curves were compared between patients with rectal bleeding (B) versus no bleeding (NB).

There were 12 B and 33 NB patients in the pooled group. Rectal bleeding was more frequent and of higher grade among FASTR patients (8/15, 5 grade 2 or higher) than among SATURN patients (4/30, all grade 1). For any bleeding (grade ≥1), individual dose-volume points in the 20 to 40 Gy range were significantly different (2-sided P < .05) between the B and NB groups, with the 40 Gy point being the most significant (B: V40 = 1.53%, standard deviation (SD), 1.32; NB: V40 = 0.69%, SD, 1.46; P = .006). For grade ≥2 bleeding, the V20 Gy was most significant (B: 68.4%, SD, 4.76; NB: 40.45%, SD, 13.9; P < .001).

The higher relative dose volumes to the rectum (V20-V40) were most strongly associated with clinically significant bleeding in this analysis and are consistent with findings of series that used H-SIB to treat prostate only. Differences in the prostate target volumes and planning margins likely account for the differences in the rates and grades of rectal bleeding observed between trials.