Use of thymidine kinase 1 autoantibody, thymidine kinase antigen, extracellular protein kinase A autoantibody, and C‐reactive protein for early detection of malignant tumors in dogs

Abstract Background Although the importance of early diagnosis of malignant tumors is a major concern, an easily accessible in‐house method has not been established. Objectives To investigate the most optimal model by combining thymidine kinase 1 (TK1) autoantibody, extracellular protein kinase A (ECPKA) autoantibody, TK1 antigen, C‐reactive protein (CRP), age, breed, and sex. Animals Serum samples from 1702 dogs were collected from local animal hospitals and referral animal medical centers in Korea. Methods TK1 autoantibody, ECPKA autoantibody, and CRP were measured using LFIA methods in serum samples of dog to design a new neoplastic index (NI) for early detection of malignant tumors in dogs. Results AUC of TK1 autoantibody model with TK1 antigen, CRP, age, and breed in multiple logistic regression analysis was 0.966 (TK1 autoantibody, P = .0005; TK1 antigen, P = .0003), and when the cutoff value was 0.417, the specificity was 87.1%, and sensitivity was 96.4%. Regression coefficients were 24.4, 20.5, 24.8, 10.6, respectively in TK1 autoantibody, TK1 antigen, CRP, and age, and the effect by breed (regression coefficient 2.1, 3.0) was the lowest. The same multiple logistic regression analysis on dogs with lymphoma, and AUC of TK1 autoantibody model was 0.981 (all P < .0001, TK1 antigen P = .09), when the cutoff value was 0.352, the specificity was 92.9%, and sensitivity was 93.7%. Conclusions and Clinical Importance The NI including TK1 autoantibody could be useful in the screening test for both lymphoma and other tumors.