Immediate prediction of post-ERCP pancreatitis: a novel approach using serum trypsin.

Abstract

Background: As trypsin activation occurs early in the pathogenesis of pancreatitis, evaluating serum trypsin (TRY) may allow immediate prediction of post-ERCP pancreatitis (PEP). This study aims to evaluate the diagnostic performance of TRY measurement for predicting PEP.

Methods: This multicenter prospective observational study was conducted at 12 tertiary centers in Japan. Serum TRY, amylase (AMY), pancreatic-type AMY (P-AMY), and lipase (LIP) were measured before ERCP, immediately post-procedure (0 h), and 2 h post-procedure. Predictive performances of pancreatic enzymes for PEP were evaluated using receiver operating characteristic (ROC) analysis, Youden's index, and DeLong's test.

Results: PEP was identified in 48 of the 463 patients analyzed (10.3%). The optimal 0-h TRY cut-off was 2043 ng/mL, yielding an area under the ROC curve (AUC) of 0.83 (68.7% sensitivity, 84.6% specificity), significantly higher than those of 0-h AMY (AUC 0.74, p = 0.002), 0-h P-AMY (0.78, p = 0.007), or 0-h LIP (0.80, p = 0.048). At 2-h, TRY showed an AUC of 0.89, superior to 2-h AMY (0.84) and comparable to 2-h P-AMY (0.89) and 2-h LIP (0.90). When cutoffs were set as 2 × , 2.5 × and 3 × of the upper limit of normal, TRY consistently showed higher sensitivity than AMY or P-AMY. TRY showed the earliest and steepest post-procedural increase among all enzymes in patients with PEP.

Conclusions: TRY measured immediately after ERCP provides significantly better early prediction of PEP than conventional enzymes. Rapid measurement of TRY may enable immediate risk stratification, potentially improving clinical outcomes.