Anastomotic leak after restorative proctosigmoidectomy for cancer: what are the chances of a permanent ostomy?

OBJECTIVE

The objective of this study is to identify rates and factors associated with permanent diversion following restorative proctosigmoidectomy for rectal cancer when complicated by an anastomotic leak.

DESIGN

This study is a retrospective review.

SETTING

The setting of this study is a tertiary referral hospital

PATIENTS

Patients involved in this study were those who underwent restorative rectal cancer surgery from 1997 through 2008 identified from an institutional cancer database.

INTERVENTIONS

No interventions were performed in this study.

MAIN OUTCOME MEASURES

Factors associated with time to ostomy closure and rates of permanent diversion following anastomotic leaks

RESULTS

One hundred and thirty patients (mean age 59.7 years) were identified, 111 (85%) of whom had stoma diversion at the index surgery. Asymptomatic occult radiological leaks occurred in 52 patients (40%). Seventy-eight patients (60%) underwent ultimate ostomy closure at a median time of 6.3 months after the index surgery, which was not significantly affected by previous radiotherapy (p = 0.08). Twelve patients (9%) underwent anastomotic reconstruction. Pathologic stage II or greater (OR 4.42; 1.95-10.04), symptomatic presentation (OR 4.13; 1.86-9.19) and anastomotic disruption >5 mm (OR 4.42; 2.01-9.74) were independently associated with permanent diversion. Among all survivors, 33% did not have their ostomy reversed. Stoma diversion constructed after leak detection in 19 patients (15%) did not affect permanent stoma rate (OR 0.86; 0.31-2.34) or time to stoma closure (p = 0.29).

CONCLUSIONS

The majority of anastomotic leaks after restorative rectal cancer surgery can be salvaged without anastomotic reconstruction. However, in one third of the patients, an anastomotic leak results in a permanent stoma.