Background: Ileosigmoid fistula (ISF) in Crohn's disease (CD) is a challenging clinical condition. The current role of diagnostic modalities and specific surgical management options for ISF in CD are not well characterized.
Methods: Patients from a prospectively collected CD database who underwent surgery for ISF during 2002-2007 were included. Demographics, disease extent, diagnostic modalities, operative approach, perioperative outcomes, concurrent medication use, and smoking status were retrospectively reviewed. Overall length of hospital stay included any postoperative readmission and stoma reversal. Converted cases were considered as laparoscopic procedures in an intent-to-treat analysis.
Results: A total of 61 patients underwent operative management for ISF (37 females, median age 37(range 18-78) years, 19(31%) laparoscopic). All patients had ileocolic resection. Management of the sigmoid colon included primary repair 14(23%), segmental resection 45(74%), or subtotal colectomy 2(3%). Additional CD findings were identified in 25(41%) patients, including ileovesical fistula 8(13%), enterocutaneous fistula 8(13%), and small bowel disease requiring resection 8(13%), strictureplasty 5(8%), or both 2(3%). Sensitivities of colonoscopy, CT scan and fluoroscopic contrast studies for ISF were 40%(21/53), 47%(21/45) and 54%(14/26), respectively. The combination of all diagnostic studies resulted in a preoperative diagnosis of ISF in 35(57%) patients. Protective stoma was used in 33(54%) patients and was more frequent if additional small bowel disease required surgery (77% vs. 44%, p=0.03), for open vs. laparoscopic surgery (64% vs. 32%, p=0.02), when a phlegmon or abscess were present (85% vs. 46%, p=0.007), and was associated with the use of intraoperative ureteral stents (33% vs. 4%, p=0.003). Sigmoid resection was more common in laparoscopic vs. open approach (95% vs. 69%, p=0.02). There were no deaths. Overall morbidity was 35% and leak rate 8%. Neither was affected by stoma diversion, use of laparoscopic technique, or treatment of the sigmoid colon with resection vs. primary closure. Overall length of hospital stay was non-significantly shorter with laparoscopic compared to open surgery (median 6 vs. 9 days, p=0.25).
Conclusions: ISF in CD remains an often incidental surgical finding. Sigmoid resection and primary sigmoid repair have comparable morbidity if appropriately individualized. Laparoscopic treatment is acceptable in select cases and may allow reduction in diverting stoma rates and overall length of hospital stay with similar morbidity.