Evaluation of a fecal pancreatic elastase-1 enzyme-linked immunosorbent assay: assessment versus an established assay and implication in classifying pancreatic function

Disagreement continues regarding the two commercially available fecal pancreatic elastase-1 (PE-1) ELISAs and their respective capabilities to assess pancreatic function. Our objectives were to validate the newer PE-1 ELISA and evaluate the test against the previously established assay, to investigate the PE-1 correlation with fecal fat, and examine the PE-1 result distribution of clinical specimens. Methods: The BioServ Diagnostics PE-1 ELISA was validated and performance characteristics compared to the previously validated ScheBo(R) Biotech PE-1 ELISA. A split sample study was accomplished using Deming regression and Bland-Altman plot analysis. Data mining was implemented to evaluate PE-1 and fecal fat correlation, and to explore PE-1 result distribution. Results: Regression analysis shows limited quantitative agreement; slope = 0.9640, intercept = 10.787, R2 = 0.633. The means were 228.8 and 226.2 μg PE-1/g stool for the BioServ and ScheBo assays respectively. Bland-Altman analysis indicated 91% of paired values within two standard deviations of their means (100% within 2.2 standard deviations). There was good qualitative agreement between assays with 91% of cases having equivalent pancreatic function classification. The remaining 9% varied by only one classification level with no bias towards either test evident. The distribution of typical clinical specimens for infants through young adults is dichotomous, with few subjects classified with moderate pancreatic insufficiency. There is minimal agreement between PE-1 and fecal fat. Conclusions: The BioServ Diagnostics PE-1 ELISA is an acceptable alternative to the ScheBo Biotech PE-1 ELISA. We also recommend that PE-1 replace fecal fat analysis for the evaluation of pancreatic function.