Analysis of the Temporal Sequence of Selected Long-term Toxicities for Patients Treated with Brachytherapy or Brachytherapy Plus External Beam Radiation for Prostate Adenocarcinoma

Abstract

Differential timing of specific long-term toxicities for patients treated with brachytherapy are reported for cervical cancer. We assessed the temporal sequence of selected long-term side effects including dysuria, rectal bleeding, hematuria, and erectile dysfunction for patients treated with low dose rate brachytherapy for prostate cancer. Patients definitively treated for prostate adenocarcinoma at Mt. Sinai Medical Center from 1990-2007 who received either a full I-125 or Pd-103 implant or a partial Pd-103 implant followed by external beam radiation were selected from our database. Patients with less than 2 years of follow-up and salvage implants were excluded. Two thousand forty-seven patients met inclusion criteria. Median follow-up was 5.8 years (range 2-17.8). Patients were selected for a documented report of dysuria of any grade (N = 483), grade 2 rectal bleeding excluding hemorrhoids (N = 92), hematuria of any grade excluding bleeding immediately following the implant (N = 168), and erectile dysfunction as defined as a drop in the Mt. Sinai Erectile Function score from a value of 2 or 3 to 1 or 0 (N = 368). For erectile dysfunction, patients who received any hormonal therapy were excluded (N = 920). Values are reported as the time from implant until the first incident of the toxicity. Kaplan-Meier survival curves were generated to determine the time for the toxicity to occur in 25%, 50%, and 75% of patients who experienced that particular toxicity. Cox-Regression was done for multivariate analysis. With regard to dysuria, the time elapsed for the onset of symptoms in 25%, 50%, and 75% of patient who experienced dysuria was 0.5, 1.1, and 2.3 years, respectively. Time to first onset of symptoms for 25%, 50%, and 75% of patients who experienced grade 2 rectal bleeding was 1.3, 2.1, and 3.1 years, respectively. For hematuria, time to first onset for 25%, 50%, and 75% of patients who experienced the symptom was 2.0, 4.1, and 8.3 years, respectively. Finally, time to the first onset of erectile dysfunction for patients who did not receive hormonal therapy was 0.6, 3.2, and 8.2 years for 25%, 50%, and 75% of the patients who experienced these symptoms. Multivariate analysis for all toxicities found no association in timing of the symptoms with regard to BED, Age, or Risk Group of the patient. Unique patterns of time to first onset of selected toxicities were observed in patients who received a radioactive implant for prostate adenocarcinoma. While dysuria and rectal bleeding occur rapidly and early, hematuria and erectile dysfunction tend to be slightly delayed in first onset and more variable in the overall distribution of first events. Patients might be educated accordingly.