Impact of volume and location of irradiated rectum wall on rectal blood loss after radiotherapy of prostate cancer.

PURPOSE: To identify dose-volume parameters related to late rectal bleeding after radiotherapy for prostate cancer.

MATERIALS AND

METHODS

Clinical complication data from a randomized trial were collected and linked to the individual dose-volume data. In this trial, patients with prostate cancer were treated with either conventional (with rectangular fields) or three-dimensional conformal radiotherapy to a dose of 66 Gy. Patient complaints, including rectal blood loss, were collected for 199 patients, using questionnaires. Absolute and relative dose-volume histograms (DVHs) of the rectal wall (with and without the anal region) were calculated with and without rectal filling. A proportional hazard regression (PHR) model was applied to estimate the probability of any rectal blood loss within 3 years, as a function of several DVH parameters. In a multivariable analysis, dose-volume parameters were tested together with patient- and treatment-related parameters (age, smoking, diabetes, cardiovascular disease, tumor stage, neo-adjuvant androgen deprivation, conformal vs. conventional and rectal bleeding during treatment).

RESULTS

The estimated incidence of any and moderate/severe rectal bleeding at 3 years was 33% and 8%, respectively. Differences between the conventional and conformal technique were small and not significant. The analysis of relative DVHs of the rectal wall (with and without the anal region), showed significant (p < 0.01) relations between the irradiated volume and the probability of rectal blood loss within 3 years for dose levels between 25 Gy and 60 Gy. This relationship was shown in subgroups defined by dose-volume cutoff points as well as in the PHR model, in which a continuously rising risk was seen with increasing volumes. For absolute DVHs and DVHs of the rectum including filling, less or no significant results were observed. The most significant volume-effect relation (p = 0.002) was found at 60 Gy for the rectum wall excluding the anal region. The probability of rectal bleeding increased from 10% to 63% when the irradiated rectum volume at 60 Gy increased from 25% to 100%. Other factors. including age, smoking, diabetes, cardiovascular disease, tumor stage, neo-adjuvant androgen deprivation, conformal vs. conventional, rectal bleeding during treatment, rectum length. and whole rectum volume. did not have a significant effect in the multivariable analysis. When controlling for the volumes at 60 Gy, the volumes at lower dose levels (25-55 Gy) were no longer significant (p = 0.5).

CONCLUSIONS

For any rectal bleeding within 3 years, an overall incidence of 33% was observed for patients treated to 66 Gy. For this endpoint, a volume-effect relation was found for DVH parameters of the relative rectal wall volume. This relationship appeared to be most significant for the rectum without the anal region and for the higher dose levels (50-60 Gy).