Risk factors for anastomotic leak after recto-sigmoid resection for ovarian cancer

Abstract

Objectives.

Anastomotic leak after recto-sigmoid (RS) resection for ovarian cancer (OC) is a life-threatening complication. Selection of patients for protective diverting stomas has been based on observations from the colorectal literature. Our objective was to identify primary risk factors for anastomotic leak in OC patients undergoing RS resection to better determine who would most benefit from protective diversion.

Methods.

All patients with OC or primary peritoneal cancer who underwent a debulking procedure with RS resection between January 1999 and December 2004 were included. Retrospective chart review including review of operative notes, pathology reports, and medical records including follow-up visits was done. Cases with inadequate postoperative follow-up, primary end colostomies, or diverting stomas were excluded.

Results.

177 patients form our study cohort. There were a total of 12/177 anastomotic leaks (6.8%). The mean time to diagnosis of anastomotic leak was 19 days (range 4–32). The leak rate for primary debulking operations was 8.7% (10/115), whereas the leak rate in secondary debulking procedures was 3.2% (2/62) (NS, P = 0.22). In univariate analysis, only perioperative serum albumin was significantly associated with an increased risk of anastomotic leak (mean 3.4 g/dL vs. 2.4 g/dL, P = 0.002). Based on serum albumin, the leak rate was 6/29 (21%) for levels <3.0 g/dL and 2/58 (3.4%) for patients with albumin greater than or equal to 3.0 g/dL (OR 7.3, 95% CI 1.37–38.87).

Conclusions.

Low serum albumin is associated with an increased risk of anastomotic leak after RS resection for OC. Patients with a low albumin level may benefit from a protective diverting colostomy/ileostomy.

Introduction

Ovarian cancer is the fourth leading cause of cancer death in women in the United States [1]. The most important predictor of survival in advanced stage disease is residual disease (RD) after primary debulking surgery [2], [3]. A recent paper by Aletti et al. demonstrated that patients with advanced ovarian cancer who were optimally debulked had an improved 5-year survival compared with patients who have RD greater than 1 cm [2]. Accordingly, radical surgical techniques are often employed to achieve optimal debulking. As ovarian cancer commonly presents with confluent tumor in the cul de sac, recto-sigmoid resection (RSR), either alone or en bloc with hysterectomy and bilateral salpingooophorectomy, is often necessary to achieve complete clearance of pelvic disease.

Anastomotic leak after RSR for ovarian cancer is a devastating and life-threatening complication. The leak rate reported in the colorectal literature varies from 2.8 to 23% [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], whereas it ranges from 0.8 to 3.2% in the gynecologic oncology literature [14], [15], [16], [17]. Recent publications indicate the mortality rate after an anastomotic leak ranges from 7.3 to 16% [5], [6], [18]. Multiple preoperative and intraoperative factors have been implicated as increasing the risk of an anastomotic leak, including long operating times, previous pelvic irradiation, and short distance of the anastomosis to the anal verge [11], [13], [19], [20]. Traditionally, selection of patients for protective diverting stomas has been based on assumptions taken from the colorectal literature. However, this extrapolation to ovarian cancer patients may not be appropriate. Most RSR performed for advanced ovarian cancer occur within the context of a long operative time, with extensive abdominal dissection and resection of tumor in patients with diffuse carcinomatosis. Additionally, the National Cancer Institute (NCI) recently recommended that all optimally debulked stage III ovarian cancer patients be offered intraperitoneal chemotherapy [21]. As such, the risk factors for anastomotic leak may be different and the consequences of such a complication may be more severe. While there are several papers in the ovarian cancer literature addressing the safety and efficacy of RSR during cytoreductive surgery, few have provided information that would allow perioperative risk assessment for anastomotic leak. Obermair et al. demonstrated that a serum albumin level less than 3.0 g/dL was associated with increased post-surgical morbidity, including leak, but they were unable to characterize the influence of other potential risk factors for leak such as operative time and blood loss/transfusions [17].

In the present study, we explored the risk factors for anastomotic leak after RSR during debulking surgery in a cohort of consecutively treated patients with primary or recurrent ovarian cancer. Our underlying hypothesis was that patients undergoing secondary debulking or having a low serum albumin would be at increased risk for anastomotic leak. The main objective was to identify the primary risk factors for anastomotic leak in ovarian cancer patients undergoing a RSR as part of a debulking procedure. Our secondary aim was to identify those patients who would benefit from a protective diverting colostomy/ileostomy at the time of RSR with reanastomosis.