Effect of BMI and BMI and Charlson Comorbidity Index on mortality in pancreatic cancer: A single-institution study.

Abstract

e16473 Background: Pancreatic cancer is an extremely aggressive disease known for its poor outcomes. There are limited prognostic factors identified to date. While elevated Body Mass Index (BMI) has been associated with increased risk of pancreatic cancer incidence in observational cohorts, its impact on mortality among diagnosed patients remains unclear. Our objective was to study impact of pre-treatment BMI and Charlson Comorbidity Index (CCI) on mortality in pancreatic cancer patients undergoing any cancer-directed treatment at our institute. Methods: Patients age 18+ diagnosed with pancreatic adenocarcinoma (ICD10- C25) from 3/1/2021 to 1/1/2024, and who received any pancreatic cancer treatment, were included. Demographic data, tumor characteristics, and treatment details were collected and managed in REDCap. BMI and CCI values at the time of diagnosis were captured. Patients who did not receive treatment were excluded. To determine the impact of BMI and CCI, we fitted a multivariable Bayesian cox proportional hazard regression model. BMI was fitted with splines with three knots to allow for non-linear effects on mortality. We specified a prior distribution that was normal, centered on 0, and where 95% of the distribution was between 0.2-5.0 hazard ratio (Normal (0, log(5)/1.96). Results: A total of 150 patients were included in the study. Sex was evenly split (M = 47% and F = 53%), most of the population was white (73%) and non-Hispanic (77%), and patients had a median age of 70 (62-77 IQR). The median BMI was 26.5 (23-31 IQR), tumor size was 35 mm (25-45 IQR), and most patients were stage 1 (35%). We did not find a relationship in the HR of mortality with BMI, when a BMI of 25 was used as a reference point. When exploring the non-linearity of BMI by clinical stage, there was no relationship either. CCI did not differ in HR when comparing values of 1, 2, 3+ to 0. Conclusions: We were unable to identify the relationship between pre-treatment BMI and CCI on mortality. Subgroup analysis of individual stages demonstrated similar results. Prospective studies with larger sample sizes are warranted to confirm and further elucidate these associations, including the role of metabolic factors. Analysis by type of treatment alone (surgery vs. chemo and radiation) may be useful in identifying a relationship.