Obesity paradox in gastrointestinal cancer patients on immunotherapy: A retrospective multi-institutional US cohort study.

Abstract

11016 Background: Obesity is a well-established risk factor for different types of cancers, including gastric cancers. However, emerging research suggests its association with improved survival outcomes in cancer treatment, a phenomenon referred to as "the obesity paradox". Recent literature demonstrates this paradoxical effect in cancer patients receiving immunotherapy. We explored this effect across various types of gastrointestinal (GI) malignancies. Methods: De-identified data from the TriNetX database were used to identify adult patients (age > 18 yrs) with GI malignancies (hepatocellular, colorectal, pancreatic, upper GI (gastric and esophageal) who received immune checkpoint inhibitors (ICIs) between 2013 and 2025. Propensity score matching was performed to balance cohorts by demographics, comorbidities, ECOG and albumin level. The primary outcome was 5-year all-cause mortality, compared between patients with obesity (BMI ≥ 30 kg/m²) and those with normal BMI (18–24.9 kg/m²). The secondary outcome was the incidence of ICI-related adverse events (irAEs). Kaplan–Meier was used for time-to-event outcomes, differences were assessed using the log-rank test, and Cox proportional hazards using the TriNETx platform. Results: After propensity score matching, 4,956 patients were included in each cohort. Mean baseline characteristics were comparable between cohorts. The BMI > 30 cohort had BMI(kg/m2) 35.4 ± 5.4, albumin(g/dl) 3.54 ± 0.61, and mean ECOG 0.68 ± 0.96 while the normal BMI cohort had BMI 22.8 ± 4.5, albumin 3.48 ± 0.62, and mean ECOG 0.67 ± 0.83. Obesity was associated with significantly improved all-cause mortality compared with the normal BMI group (26.8 months vs 13.9 months; hazard ratio [HR], 0.7 [0.67-0.74]). Survival probabilities at 1 year (63.7% vs 52.6%), 3 years (46.2% vs 31.9%), and 5 years (39.1% vs 25.4%) were higher in the BMI ≥ 30 kg/m² cohort compared with the normal BMI cohort. Among GI cancer subtypes, the strongest associations between obesity and improved survival were observed in colorectal (HR 0.67 [0.62-0.72]) and pancreatic cancers (HR 0.66 [0.56-0.76]), with more modest effects seen in hepatocellular (HR 0.78 [0.72-0.84]) and upper GI cancers (HR 0.71 [0.62-0.80]). There were no significant differences between cohorts in the incidence of pneumonitis (HR 1.12 [0.80-1.57]), GI (HR 1.07 [0.98-1.17]), thyroid (HR 1.05 [0.97-1.14]), or cutaneous (HR 1.10 [0.99-1.23]) irAEs. Conclusions: In this analysis of 9,912 patients with GI malignancies treated with immunotherapy, higher BMI was associated with improved survival. Confounding by disease burden was partially addressed through adjustment for ECOG and albumin. These findings may reflect obesity-related immune modulation, as preclinical studies suggest leptin-mediated upregulation of PD-1 expression. However, further prospective studies are needed to validate and define these results