TUMOR LOCATION DOES NOT STRATIFY SURVIVAL IN PATIENTS UNDERGOING RESECTION FOR GALLBLADDER CANCER
Dana A. Dominguez*1,2, John Aversa1, Laurence Diggs1, Brendan Hagerty1, Jeremy L. Davis1, Jonathan M. Hernandez1, Andrew M. Blakely1
1Surgical Oncology Program, National Cancer Institute, NIH, Bethesda, MD; 2University of California, San Francisco - East Bay, Oakland, CA
Background: Tumor location on the gallbladder (peritoneal vs. hepatic side) was included in the recently published eighth edition of the American Joint Committee on Cancer (AJCC) Staging Manual. This was largely based on a multi-institutional study that suggested prognostic significance for stage II disease. However, the prognostic impact of tumor location in gallbladder adenocarcinoma has not been validated in a larger dataset.
Methods: Patients with resected gallbladder adenocarcinoma (Tis-4, any N, any M) and adequate staging and tumor location data were selected from the National Cancer Database (2005-2016). Univariable and multivariable analyses were performed to evaluate factors associated with tumor location. Factors associated with overall survival (OS) were determined with univariate and multivariate Cox proportional hazards modeling. OS was examined by disease stage using the Kaplan-Meier method after excluding Nx patients. Likelihood-ratio test was performed to compare multivariate Cox regression including T, N, and M stage with and without tumor location.
Results: Overall, 2294 patients were available for analysis, of which 1802 (78.6%) had tumors involving the hepatic side of the gallbladder. Patients with hepatic-sided tumors were more likely to have higher T-stage (referent: pTis; pT3: OR 5.97, 95% CI 2.42-14.71; pT4: OR 35.3, 95% CI 6.7-186.6), liver resection (OR 2.21, 95% CI 1.73, 2.83), or positive margin(s) (OR 1.82, 95% CI 1.38, 2.39). Factors independently associated with decreased OS included increasing age, ≥2 comorbidities, lymphovascular invasion, positive margin, increasing T-stage, stage N1-2 or Nx, and M1 disease. Improved OS was associated with liver resection and chemotherapy administration. Hepatic-sided tumor location was associated with decreased OS on univariate (HR 1.21, 95% CI 1.06-1.36) but not multivariate (HR 0.96, 95% CI 0.84-1.10) analysis (Table 1). Among patients with Stage II disease (n=417), OS was similar between hepatic- (n=296, 71.0%) and peritoneal-sided tumors (n=121, 29.0%; log-rank p=0.92) (Figure 1). Likelihood-ratio test did not show improvement in the OS model after addition of tumor location (P(X2)>0.144 = 0.998).
Conclusion: In this large national cohort analysis, tumor location was not independently associated with OS in gallbladder adenocarcinoma. Subset analysis of stage II patients failed to demonstrate a survival difference when stratifying by location. The prognostic value of adding tumor location to the AJCC staging system remains unclear.
Back to 2020 Abstracts