Management of Acute Cholecystitis in Cancer Patients -a Comparative Effectiveness Approach

Thejus Thayyil Jayakrishnan, MD, Ryan T Groeschl, MD, Ben George, MD, James P Thomas, MD, PhD, Sam G Pappas, MD, T Clark Gamblin, MD, MS, Kiran K Turaga, MD, MPH. Medical College of Wisconsin, Loyola University Medical Center.

Background
Host factors and therapy characteristics predispose cancer patients to a high risk of acute cholecystitis. Management of cholecystitis is often difficult given complex decision making involving the underlying cancer, possible interruption of treatment and surgical fitness of the patient.
Methods
A management pathway was developed for cholecystitis in cancer patients that incorporated patient-specific survival and risks of recurrence. Estimates were obtained from a multi-stage systematic review. A decision tree with a lifetime horizon was constructed to compare conventional strategies (conservative treatment (CT), percutaneous cholecystostomy (PC) and definitive cholecystectomy (DC)) with the new pathway (NP). The decision tree was optimized for highest estimated survival. Sensitivity analyses were performed.
Results
In low surgical risk patients with cancer-specific survival of 12 months, the NP yielded estimated survivals of 11.9 months vs. 11.8months (CT) vs. 11.8months (PC) vs. 11.9months for DC arm. For high risk patients, the estimated survival was 11.6 (NP), 9.9 (DC), 11.4(PC) and 11 (CT) months respectively. The decision to perform a definitive cholecystectomy at 6 weeks after a PC was optimum in patients expected to survive 24months (23.2months from NP) or with a shorter expected survival but a high recurrence risk (20%) as shown in Table 1. Model estimates were robust in sensitivity analyses.
Conclusions
Incorporation of the surgical risk and the risk of recurrent cholecystitis, while balancing the patient-specific survival and the impact of anti-neoplastic therapy in the management of cholecystitis yields improved survival. This work provides measures to evaluate surgical judgment and can augment the physician patient decision making.

« Return to SAGES 2014 abstract archive