Serum proteomic patterns as a predictor of severity in acute pancreatitis.

ABSTRACT

BACKGROUND: Approximately 20% of patients with acute pancreatitis (AP) develop a severe and potentially life-threatening course. Serum proteomic pattern analysis for disease diagnosis is a promising novel and rapidly expanding field based on the hypothesis that serum patterns of low molecular mass biomarkers can specifically reflect an underlying organ-specific pathologic state. AIM: To evaluate the potential differences in proteomic profiles between patients with mild and severe AP. METHODS: Sera from 21 patients with mild AP and 7 patients with severe AP were analyzed using surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF MS). Samples were profiled in duplicate on IMAC3 ProteinChips arrays. RESULTS: Of 79 spectral peak clusters (classifiers) detected, 18 had significantly different signal intensities between mild AP (MAP) and severe AP (SAP) sera (p < 0.01; Mann-Whitney U test, averaging for technical replicates) and were considered as potential classifiers in classification and regression tree (CART) analysis. The CART analysis returned simple classification trees consisting of one primary splitter, at 11,720 Da. Training data performance delivered nearly 100% sensitivity and 81% specificity for discrimination of SAP. The next top performing classifier was indicated at 4,283 Da m/z peak. CONCLUSIONS: These initial data suggest that serum proteomic profiles contain features that discriminate MAP and SAP. Larger sample sizes will be required for the development and validation of more specific predictive algorithms.