Measurement of canine serum insulin has routinely relied on methods developed to measure human insulin. A canine-specific enzyme-linked immunosorbent assay (ELISA) for measuring insulin has become commercially available, with antibodies, calibrators, buffer, and analytic range optimized for canine samples. Antibodies used in this ELISA recognize both porcine and canine insulin. Unlike human assays, in which results are reported as activity (with units calculated based on activity of the human insulin molecule), the canine ELISA results are reported as weight per volume (ng/L). The objective of this study was to validate the canine ELISA for determination of serum insulin concentration in dogs. Serum samples were collected from healthy dogs and from dogs examined for a variety of illnesses and disorders; most were suspected of having insulinoma. Evaluation steps included determination of intra- and interassay coefficients of variation (CV), recovery after dilution and spiking, assessment of linearity, and biologic response to glucagon. Intra- and interassay CVs were 4.3% to 7.8% and 4.4% to 7.7%, respectively. Mean recovery after dilution was 99% and recovery after spikingwith porcine insulin was 116%. The linearity study supported the manufacturer’s claim about the low end of the analytic range (20 ng/L). The range of values measured in samples from ill dogs was slightly higher than the range in samples from healthy dogs, and the highest insulin concentration measured was well below the upper limit of the analytic range (1500 ng/L). Serum insulin concentrations increased significantly after glucagon injection. Results using the canine ELISA were also compared with results using an ELISA for human insulin, and the canine and human ELISAs correlated well. Stability of serum insulin during storage was also determined. Insulin was stable for 30 days in 6 serum samples stored at -20°C and in most samples for 8 days at 4°C to 8°C. Insulin was stable for < 3 days at room temperature (20°C). The results of this study indicate that the new canine serum insulin ELISA had good precision and accuracy and correlated well with the previously used assay.

Commentary: The aim of this study was to validate a canine-specific insulin assay and to assess correlation of this test with the currently accepted human assay for insulin measurement. The justification for use of a newer canine-specific ELISA is to minimize possible inaccuracies in testing due to variation and cross-reactivity of antibodies and to avoid the intrinsic drawbacks of using the human assay for dogs. This study found that the test was reasonably precise, with small variations within and between assays. Excellent correlation was demonstrated between human and canine insulin assays. Suggested uses for this assay include diagnosis of insulinoma and differentiation between absolute and relative insulin deficiency in diabetic dogs. Future studies are needed to establish an appropriate reference range for dogs. It would also be interesting to validate the canine versus human insulin assays in dogs with known insulinoma or diabetes.
Jennifer Ginn, DVM, Diplomate ACVIM

Validation of a species-optimized enzyme-linked immunosorbent assay for determination of serum concentrations of insulin in dogs. Öberg J, Fall T, Lilliehöök I. VET CLIN PATHOL 40:66-73, 2011.