RESEARCH COMMUNICATIONS OF THE 28th ECVIM-CA CONGRESS

The aims of this study were the validation of human assays for serum fructosamine (SF) and glycated hemoglobin (HbA1c) in dogs and the comparison of the ability of the HbA1c and SF in reflecting the glycemic control in dogs with diabetes mellitus. SF (nitrotetrazole blue method, Fructosamine 17350H, Sentinel Diagnostics) and blood HbA1c (HbA1c, OSR6192 Beckman‐Coulter) were analyzed on an automated chemistry analyzer (Beckman‐Coulter AU 480). Linearity, precision and accuracy were determined; a reference interval for HbA1c and SF was established from 40 healthy dogs, using the percentile method. Performances of HbA1c and SF in assessing the glycemic control were evaluated; correlation between the two variables and a clinical score was studied in 200 re‐evaluations of 47 diabetic dogs treated with insulin q12h. The clinical score used to classify diabetic dogs in good (GGC), moderate (MGC) or poor (PGC) glycemic control was set on the basis of stability of body weight, presence of poliuria/polydipsia, median glucose of the blood glucose curves (BGCs), blood glucose nadir and overall evaluation of BGC. The average intra‐ and interassay coefficient of variation (CV) for HbA1c were 1.5% and 10.9%, respectively; the average intra‐ and interassay CV for SF were 4.1% and 2.5%, respectively. Excellent accuracy was obtained for both assays (r>0.99). The reference interval for HbA1c was 1.6‐4.5% and for SF was 222‐382 μmol/L. In diabetic dogs, HbA1c and SF were significantly correlated (r=0.48) and they were also correlated with the clinical score (r=‐0.33; r=‐0.39, respectively). ROC curves analysis of SF and HbA1c to distinguish GGC from MGC/PGC dogs showed an AUC of 0.69 and 0.66, respectively. Values of HbA1c<5.5% and SF<400 μmol/L had Sp=79%, Se=41%, and Sp=71%, Se=61%, in discriminating dogs with GGC from MGC/PGC dogs, respectively. ROC curves analysis of SF and HbA1c to distinguish PGC from GGC/MGC dogs showed an AUC of 0.75 and 0.69, respectively. Values of HbA1c>6.8% and SF>500 μmol/L had Sp=79%, Se=52%, and Sp=85%, Se=45%, respectively, in discriminating dogs with PGC from GGC/MGC dogs. Using the clinical score as the reference method, SF and HbA1c identified correctly the glycemic control (GGC/MGC/PGC) in 50% and 44% of cases, respectively. In conclusion, SF and HbA1c have similar performances in classifying the glicemic control. Neither of the two parameters should be used as the sole indicator of glycemic control, and must always be evaluated in association with history, physical findings and BGC results.