Professor College of Veterinary Medicine, Seoul National Univsersity Gwanak-gu, Seoul-t'ukpyolsi, Republic of Korea
Abstract: Background - Recognizing the need for a more effective malignant tumor screening method in dogs, we tried to devise algorithm by comparing a combination of several biomarkers.
Objectives - We investigated the possibility of TK1 autoantibody as a tumor biomarker and devised a reliable algorithm after comparative analysis of several biomarker candidates.
Animals - Random serum samples of 1,702 dogs, healthy or diagnosed with certain diseases or tumors were collected from local and referral animal medical centers in Korea.
Methods - Using LFIA methods, the concentrations of thymidine kinase 1 (TK1) autoantibody were measured in serum samples of dogs classified into healthy control, nontumor disease group, and patients with benign and malignant tumors. A multiple logistic regression model was designed combining all biomarker candidates.
Results - TK1 autoantibody was statistically higher in malignant tumor patients than in the control group (Welch t test, P < .0001), and also showed significant differences in carcinoma, hematopoietic tumor, and sarcoma (post hoc Games–Howell test, P < .0001). As a result of the combined analysis of TK1 autoantibody, TK1 protein, CRP, age and breed, AUC was 0.966, which was similar to the result of replacing TK1 autoantibody with ECPKA autoantibody (AUC, 0.974).
Conclusions and clinical importance - TK1 autoantibody proved useful as one of the biomarkers for malignant tumor screening in dogs, and it was confirmed that an algorithm combining several biomarker candidates had high sensitivity and specificity.
