Fatty Pancreas and the Association Between Pancreatic Manipulation Burden and Post-ERCP Pancreatitis.

BACKGROUND

Fatty pancreas has been associated with pancreatitis after ERCP, but its interaction with procedure-related pancreatic injury is unknown.

AIMS: We evaluated whether fatty pancreas modified the association between cumulative pancreatic manipulation and postprocedural pancreatitis.

METHODS

This multicenter retrospective cohort included 679 patients with native papillae undergoing biliary endoscopic retrograde cholangiopancreatography. Fatty pancreas was defined as a pancreas-to-spleen attenuation ratio < 0.7 on preprocedural noncontrast computed tomography. A pancreatic manipulation burden score (range, 0 - 4) assigned one point each for difficult biliary cannulation, double-guidewire technique, pancreatic duct opacification, and transpancreatic sphincterotomy. Multivariable regression assessed main effects and interaction; marginal standardization estimated adjusted risks.

RESULTS

Fatty pancreas was present in 183 patients (27.0%), and 67 (9.9%) developed pancreatitis. Fatty pancreas (adjusted odds ratio, 2.17; 95% confidence interval, 1.27 - 3.73) and each one-point score increase (adjusted odds ratio, 1.53; 95% confidence interval, 1.26 - 1.86) were associated with pancreatitis. In the continuous interaction model, the adjusted per-point score odds ratio was 1.28 (95% confidence interval, 0.98 - 1.68) without fatty pancreas and 2.08 (95% confidence interval, 1.52 - 2.84) with fatty pancreas (interaction odds ratio, 1.62; 95% confidence interval, 1.07 - 2.46; P = 0.023). The categorical analysis yielded directionally concordant interaction estimates but was not statistically significant (P = 0.067). Sensitivity analyses were directionally consistent.

CONCLUSIONS

Computed tomography-defined fatty pancreas was associated with post-ERCP pancreatitis, and the interaction analyses suggest that the association between manipulation burden and post-ERCP pancreatitis may differ according to fatty pancreas status. Larger prospective studies are needed to validate these findings.