Probability of late rectal morbidity in 125I prostate brachytherapy

Abstract

Purpose: Rectal toxicity is a concern in prostate brachytherapy because it is difficult to avoid delivering a dose equal to, or greater than, the prescription dose to the anterior surface of the rectum. The purpose of this study was to define the probability that a patient will experience Grade 2 (bleeding/ulceration) late rectal morbidity after 125I prostate brachytherapy according to the rectal dosimetry. Methods and Materials: Ninety-eight consecutive patients who received monotherapy 125I prostate implants for treatment of Stage T1-T2, favorable-risk adenocarcinoma of the prostate were evaluated for Radiation Therapy Oncology Group Grade 2 late rectal morbidity. All reported incidences of late morbidity were retrospectively confirmed by colonoscopy. All patients had at least 15 months follow-up after implantation. The median follow-up was 32 months (range 15–54). The rectal dosimetry was based on a CT scan obtained at 3–9 weeks after implantation. The rectum was contoured on each CT image between the base and apex of the prostate. A dose-surface histogram was compiled for each implant, and the relative surface area that received a dose ≥100, ≥150, ≥200, ≥300, ≥400, and ≥500 Gy was recorded. The probability of developing late rectal toxicity was calculated by logistic regression analysis as a function of dose and the percentage of the rectal surface that received that dose. Results: Of the 98 patients, 10 developed Grade 2 late rectal morbidity. The percentage of the rectal surface that received 100, 150, 200, and 300 Gy was significantly greater ( p ≤0.02) for patients who experienced late rectal morbidity. The probability of late rectal morbidity increased with both the dose and the percentage of the rectal surface that received that dose. The probability was ≤1% when 20%, 7%, and 0% of the rectal surface received 100, 150, and 200 Gy, respectively. The probability increased to ≤5% when 31%, 19%, and 9% of the rectal surface received these doses. The probability of late rectal morbidity can also be expressed in terms of the maximal rectal dose. The probability of late morbidity was 0.4%, 1.2%, and 4.7% when the maximal rectal dose was 150, 200, and 300 Gy, respectively. Conclusion: The percentage of the rectal surface that receives a dose ≥100 Gy is predictive of Grade 2 (bleeding/ulceration) late rectal morbidity after 125I prostate brachytherapy. The probability of late morbidity depends on both the dose and the percentage of the rectal surface that received that dose. Our results indicate that the rectum can tolerate doses of 100, 150, and 200 Gy to approximately 30%, 20%, and 10% of the rectal surface with a ≤5% risk of late morbidity. Our results also indicate that the practical guideline for limiting the incidence of late morbidity to 1%, 3%, or 5% is to keep the maximal rectal dose to <200, 250, and 300 Gy, respectively.