A case of colovesical fistula caused by the eroded rectal stent.

To cite: Barbaryan A, Jabri H, Attakamvelly S, et al. BMJ Case Rep Published online: [please include Day Month Year] doi:10.1136/bcr-2013009044 DESCRIPTION A 59-year-old man with squamous cell carcinoma of the anus diagnosed and treated with wide local excision followed by chemotherapy and pelvic radiation in 2008 presented 4 years later with constipation. Colonoscopy revealed rectosigmoid stricture for which the patient underwent balloon dilation and colonic stent placement. Biopsy results were negative for tumour recurrence. After stent placement, the patient had been admitted several times with rectal pain with the most recent one being 1–2 months before. That time, colonoscopy did not show any stent malfunctioning, and biopsy was negative for malignancy and diverticulitis. This time the patient presented complaining of dysuria, frequency and fecaluria accompanied by abdominal pain for several weeks. Urine analysis showed pyuria, bacteria and positive nitrite. A CT scan of the abdomen and pelvis showed urinary bladder filled with contrast and gas as well as small defect within the superior urinary bladder wall which extended to the inferior wall of the distal sigmoid colon (figure 1). Cystoscopy demonstrated a large connection between the colon and the bladder from the eroded rectal stent (figure 2). The patient subsequently underwent laparoscopic colostomy placement with plans for further repair of the fistula at a later date. The majority of colorectal stents are placed to relieve obstruction caused by malignancy; meanwhile, their role in benign colorectal pathology (strictures, fistulas and diverticular diseases) is less well investigated. Self-expandable metallic stents (SEMSs) were first described in the early 1990s. They have been mainly used as a definite palliative treatment and as a bridge to single-stage surgical approach. The overall technical and clinical success rate approaches above 88%. Meanwhile, technical and clinical failure rates reached 8% and 5%, respectively. Complication rates are less compared with palliative surgery and include stent migration (11.81%), reobstruction (7.34%) and perforation (3.76%). Migration is more common with covered stents as well as after laser pretreatment, chemotherapy and benign aetiology. At the same time, covered stents are associated with less reocclusion because of tumour ingrowth. Perforation rates are higher in those studies that practiced balloon predilation 10% compared with 2% in non-dilation group. Other risk factors for perforation include excessive manipulation with the guidewire and inexperienced operators. SEMSs are considered more cost effective compared with surgical decompression. Our extensive search of the literature found only one case of colovesical fistula caused by the eroded rectal stent. Another contributing factor