EP2022 - Complex Iatrogenic Colovesical and Rectovesical Fistula Repair Following Hartmann’s Reversal: A Case Report

Purpose/Background: Colovesicular fistulas (CVFs) are abnormal communications between colon and bladder and make up the majority of enterovesical fistulas. Usually due to diverticular disease, they also occur in the setting of malignancy, history of pelvic radiation, inflammatory bowel disease, trauma, and iatrogenic injury. A rectovesicular fistula (RVF) is less common and involves direct communication between the rectum and bladder. Due to their deeper pelvic location they can present an even greater technical challenge. Simultaneous CVF and RVF have not been described in literature and present a novel scenario in complex fistula repair.



We present the case of a 54-year-old male who had urgent sigmoid colectomy and end colostomy for diverticulitis, then Hartmann's reversal six weeks later. He suffered unrecognized iatrogenic colovesicular and retrovesicular fistulas at the Hartmann's reversal. Shortly after, he manifested abdominal pain, fecaluria, and pneumaturia, then large volume urine per rectum. After developing symptoms, he relocated to the United States and underwent a staged repair at University of Iowa.

Methods/Interventions: Initial diagnostic approach included cross-sectional imaging and flexible sigmoidoscopy. On flexible endoscopy, a circular anastomosis with scattered staples from mid rectum (7 cm) to bladder near trigone was visualized and intubated. With intubation of bladder via rectum, an additional circular stapled anastomosis from bladder dome to colon was visualized and intubated. The patient underwent fecal diversion via laparoscopic transverse colostomy. Several months later, he underwent an open repair of both fistulas, jointly performed by a Colorectal Surgeon and a Urologist specializing in reconstruction. The procedure commenced with abdominal exploration, takedown of colovesicular fistula and creation of stapled colorectal anastomosis. Then, a large cystotomy was performed to repair both CVF as well as RVF, with the proctotomy repaired via anterior approach. A peritoneal flap was placed between proctotomy and cystotomy. Bladder repair was accompanied by placement of ureteral stents. TAMIS was also employed to reinforce proctotomy repair.

Results/Outcomes: The patient was admitted post-operatively and discharged on POD 10 with indwelling foley catheter and ureteral stents. Cystogram performed 2 weeks post-operatively demonstrated no leak and foley was removed. Ureteral stents were removed 4 weeks post-operatively. Colorectal anastomosis and proctotomy repair were studied with rectal-contrasted CT which demonstrated no leak or recurrent fistula. He was treated for a urinary tract infection 6 weeks post-operatively. He is scheduled for ostomy takedown approximately 5 months post-operatively.

Conclusion/Discussion: This novel case of tandem CVF and RVF repair following iatrogenic injury illustrates the importance of sound technique when anastomosing bowel and demonstrates one option for managing complex bladder-bowel fistulae. A staged approach starting with fecal and occasionally urinary diversion allows nutritional optimization and time for infection and inflammation to resolve. This notable case also highlights the importance of a multidisciplinary approach to operative management. Through joint efforts of colorectal and urologic teams, this patient's complex fistulas are repaired.

Cross-sectional Imaging of Rectovesicular and Colovesicular Fistula
Sagittal view of rectovesicular fistula (left) with noted staples at fistula site and direct communication from mid-rectum to bladder. (Right) Coronal view of more proximal colovesicular fistula with staples noted to the left lateral aspect of the dome of the bladder connecting to the distal descending colon.
Proposed Mechanism of Injury
Pictured is the proposed mechanism of injury to explain the iatrogenic creation of both a colovesicular fistula and rectovesicular fistula. The End-to-End Anastomosis (EEA) stapler within the rectal stump is​thought to have deployed the stapler shaft anteriorly through the rectal stump, including the caudal and cephalad portion of the decompressed bladder at the time of the Hartmann's reversal. This proposed mechanism would still allow the shaft of the EEA to mate with the anvil within the distal descending colon with inclusion of the bladder within the anastomosis.