Prostatourethral-rectal fistula after prostate brachytherapy

BACKGROUND: Brachytherapy (BT) has seen increased utilization as a potentially curative treatment for patients with localized initial or recurrent prostate carcinoma. This modality can be delivered by palladium 103 (Pd(103)) or iodine 125 (I(125)) implant with or without external beam radiotherapy (EBRT). Prostatourethral-rectal fistula (PRF) is a serious complication of this approach, and its incidence, clinical presentation, and risk factors for occurrence have not been documented thoroughly. Thus, the authors sought to determine these factors in a large series of patients who were treated at two institutions. METHODS: Seven hundred sixty-five patients received outpatient BT using a computed tomography (CT)-guided or transrectal ultrasound (TRUS)-guided technique between July 1994 and June 1999 using either Pd(103) or I(125) implants. Of the 754 patients with follow-up, 640 patients received BT monotherapy, 69 patients received BT monotherapy as a boost after EBRT, and 45 patients received BT as salvage therapy after locally recurrent prostate carcinoma that was treated initially with BT (20 patients), EBRT (20 patients), surgery plus EBRT (3 patients), surgery and high dose rate radiotherapy (HDR) (1 patient), or EBRT plus HDR (1 patient). CT dosimetry of the TRUS-guided implants was carried out in all patients 1-7 days postprocedure. Patient follow-up and clinical status were compiled in a data base. RESULTS: Seven PRFs developed in 754 patients (1%) between 9 months and 12 months after treatment. One PRF (0.2%) occurred in patients who were treated with BT monotherapy. PRFs occurred in patients who were treated with combination therapy (2 of 69 patients; 2.9%) and in patients who underwent salvage BT (4 of 45 patients; 8.8%) patients. All six patients who developed fistulas in the context of combination BT/EBRT or salvage BT had biopsy of an anterior rectal lesion overlying the prostate noted on physical examination during routine follow-up. Gastrointestinal endoscopic evaluation alone was not associated with any PRF. Five of the seven PRFs resolved with either surgical repair (3 patients) or conservative management (2 patients). CONCLUSIONS: There is a low incidence of PRF formation after BT monotherapy. Because all patients who developed PRF did so subsequent to prior rectal biopsies, the authors currently are discouraging such practices strongly if the rectal lesion is consistent with radiation-induced effects.