E282: Factors influencing surgical management and survival in gallbladder cancer – A SEER-Medicare based study

Introduction: Gallbladder (GB) cancer incidence is rising, yet prognosis remains poor. Literature demonstrates improved outcomes with oncological resection of disease staged T1b or higher, yet many patients do not receive appropriate resection.  This study aims to evaluate factors that may attribute to this discrepancy.

Methods: GB cancer stage T1b or higher patients in SEER-Medicare (2008-2015) were classified as receiving cholecystectomy alone (CCY) or cholecystectomy and liver/biliary resection (complete resection).  Baseline characteristics were obtained. Surgical outcomes and overall survival were compared before and after propensity score matching by stage, sex, age, frailty, race, region, poverty and metro/rural status by Chi-square and Wilcoxon rank-sum tests for categorical and continuous variables, respectively.

Results: 1,129 patients were identified. 830 underwent CCY (58.3% early stage & 41.7% late stage) and 299 had complete resection (54.2% early stage & 45.8% late stage). CCY patients were more often female (73.4% vs 65.6%; p=0.0104), ≥80 years old (48.2% vs 22.4%; p< 0.0001), frail (44.5% vs 27.1%; p< 0.0001), and treated by general surgeons (98.1% vs 84.9%; p< 0.0001) versus surgical oncologists. Overall survival rates at 1-, 3-, and 5-years were better with complete resection. After matching, complete resection had improved overall survival compared to CCY at 1-year (69.2% vs 49.5%; p< 0.0001), 3-years (42.8% vs 24.1%; p< 0.0001), and 5-years (37.5% vs 18.7%; p< 0.0001). Complete resection resulted in increased lymph node yield (LNY) and positive lymph nodes, yet had higher rates of bile leak and readmissions one year after surgery. These outcomes persisted after matching – complete resection compared to CCY had higher LNY (72.9% vs 35.1%; p< 0.0001) and positive lymph nodes (32.8% vs 21.4%; p< 0.0001), but increased bile leak rate (16.4% vs 7.7%; p=0.0011), readmission rate (median (IQR): 1 (1-4) vs 1(0-2); p< 0.0001), and total readmission length of stay (8 (3-22) vs 3 (0-11); p< 0.0001) one year after surgery.

Conclusions: Most GB cancer patients are not receiving appropriate oncological resection. Patients are less likely to receive oncological resection if female, older, frail, or operated on by a general surgeon without surgical oncology involvement. Even when adjusting for these variables, complete resection has improved survival outcomes at multiple endpoints. With further study and awareness, both patients and physicians can make informed decisions, potentially resulting in increased complete resection rates and improved survival for GB cancer patients.