| Summary: | Different lipase assays have variable reported diagnostic accuracies for acute pancreatitis (AP) in dogs. The aims of this retrospective study were to evaluate optimal cutoffs for 1,2-o-dilauryl-rac-glycero-3-glutaric acid-(6′-methylresorufin) ester (DGGR)-lipase to predict diagnostic cutoffs of canine pancreas-specific lipase (cPL; IDEXX). DGGR-lipase activity and cPL from the same blood draw in 301 dogs with a variety of diseases were compared using Spearman’s rank correlation, Cohen’s kappa agreement, and receiver operating characteristic (ROC) curves. Activity of DGGR-lipase (10–15,616 U/L) and cPL concentrations (8.1–≥2000 µg/L) were highly correlated (r<sub>s</sub> = 0.91). Areas under the ROC curves (AUCs) to predict cPL >200 and ≥400 µg/L with DGGR-lipase were 0.97 and 0.99, with optimal cutoffs of 143 U/L (sensitivity (Se) 91.7%; specificity (Sp) 95.3%) and 205 U/L (Se 97.5%; Sp 96.4%), and Cohen’s kappa agreements of 0.87 and 0.92, respectively. AUCs for a clinical diagnosis of AP, assigned to 87/301 dogs, with DGGR-lipase (0.75) and cPL (0.76) did not differ significantly (<i>p</i> = 0.48); optimal cutoffs were 161 U/L for DGGR (Se 67%; Sp 81%) and 235 µg/L for cPL (Se 68%; Sp 84%). To conclude, DGGR-lipase is a highly accurate predictor of cPL with a comparable performance when used to diagnose AP in dogs.
|
|---|