|
Back to 2015 Annual Meeting Program
Can a Diverting Ileostomy Be Safely Closed in Patients With an Asymptomatic Radiologic Leak After Pelvic Bowel Anastomosis?
Sharon Z. Koh*1, Cindy Kallman2, Karen N. Zaghiyan1, Edward H. Phillips1, Phillip Fleshner1
1Surgery, Cedars Sinai Medical Center, Los Angeles, CA; 2Radiology, Cedars Sinai Medical Center, Los Angeles, CA
Aim:
There is currently no consensus with regards to management of patients who present with asymptomatic leaks after a colorectal, coloanal or ileal pouch-anal anastomosis protected with a diverting ileostomy (DI). It is recommended that closure of DI should only be performed in the presence of a healed anastomosis, generally about 8 weeks later. Delaying closure until anastomotic healing prolongs the inherent morbidity of DI and subjects patients to additional radiation exposure to confirm radiologic healing. We aim to demonstrate safety and efficacy in the restoration of bowel continuity in the presence of a stable radiologic leak (RL) on repeat gastrograffin enema (GGE).
Methods:
A single center retrospective analysis of all patients with a RL after a pelvic bowel anastomosis from January 2004 - October 2014 was performed. Patients with fistulous tracts arising from the RL to other organs were excluded. Clinically stable RL were defined by lack of symptoms of pelvic sepsis (fever, lower abdominal pressure or pain or genitourinary symptoms) and stable size on repeat imaging. DI were closed in patients meeting these criteria.
Results:
14 patients had a median age of 59 (range, 21-88) years and included 9 (64%) males. Index surgical procedures included colorectal (n=3), ileorectal (n=2), coloanal (n=2) and ileal pouch-anal (n=7) anastomoses. None of the patients became symptomatic or had enlargement of the leak during the observation period. 4 (29%) patients had resolution of the RL on follow-up GGE. Comparison of patient groups that had or did not have spontaneous leak resolution demonstrated that only size of the abscess cavity on the initial GGE predicted resolution of the leak. Median area (measured in 2 standard axes) of the leak in those with resolution of the RL (79mm2) was significantly smaller than those that did not resolve (410mm2) (p=0.03). Morphology, lengths and location of the sinus and/or fistulous tracts between the RL and the pelvic anastomoses were not predictive factors for the resolution of the RL. The median time to takedown of the DI from RL recognition was 20 (1-153) days with 9 patients (64%) having closure within 1 month from the last GGE. The delay in closure in one patient was due to a postoperative development of an enterocutaneous fistula from a small bowel anastomosis. None of the patients developed pelvic sepsis requiring takedown of the pelvic anastomosis after reversal of the DI during the median follow up of 3 (range 0.9-30) months. One patient in the group with residual RL required a DI performed due to radiation proctitis 10 months after restoration of bowel continuity.
Conclusion: Closure of the DI in asymptomatic patients with a RL appears to be safe. There may not be a need to repeat GGE to ensure resolution of a stable RL prior to DI reversal, hence minimizing unnecessary delay and additional radiation.
Back to 2015 Annual Meeting Program
|
