Objectives: Two enzyme-linked immunosorbent assay methodologies are used to detect pancreatic insufficiency: monoclonal and polyclonal. We sought to compare these assays in patients with cystic fibrosis and to correlate these with the coefficient of fat absorption (CFA).
Study design: As part of a larger study, subjects had stool elastase measured by both methods while taking exogenous enzymes. Subjects subsequently stopped enzymes and had a fecal fat balance study performed; the CFA was then calculated.
Results: One hundred twenty-four subjects participated in this substudy. The median values for the monoclonal and polyclonal assays were 0.3 and 22.75 microg/g, respectively. The correlation coefficient between the 2 tests was 0.86 (P < 0.0001). The correlation coefficients between CFA and the monoclonal and polyclonal assays were 0.24 (P = 0.006) and 0.10 (P = 0.27), respectively. If the criterion standard definition of pancreatic insufficiency was set at a CFA < 90% and the test definition of pancreatic insufficiency was set at <100 microg/g, then the monoclonal and polyclonal assay positive predictive values were 97.6% (120 of 123) and 97.4% (111 of 114), respectively.
Conclusions: The positive predictive value of both monoclonal and polyclonal fecal elastase in patients with cystic fibrosis is extremely good; however, correlation of either test with CFA was poor. The median value for the polyclonal elastase assay is higher than for the monoclonal assay, which could potentially lead to lower sensitivity of the polyclonal assay at lower cutpoints for the monoclonal assay is used.