Prospective evaluation of the combined value of physical examination and biomarker variables in screening for preclinical dilated cardiomyopathy in Doberman Pinschers

Screening to assess likelihood of preclinical dilated cardiomyopathy (PC-DCM) prior to advanced diagnostic tests in Doberman Pinschers (DP) is desirable. To investigate the combined value of physical examination (PE), N-terminal pro B-type natriuretic peptide (NTproBNP) and cardiac troponin I (cTnI) for identifying PC-DCM in DP. All dogs underwent: PE, echocardiogram, 3-min ECG and cardiac biomarker measurement. Asymptomatic DP (414) were classified based on 3-min ECG and echocardiogram as: No-DCM/MMVD or myxomatous mitral valve disease (MMVD), PC-DCM based on echocardiogram (PC-DCM-Echo), PC-DCM based on arrhythmias with a normal echocardiogram (PC-DCM-ECG), equivocal DCM (EQ-DCM), and MMVD. Receiver operator characteristic curves and prediction models were derived. Heart murmurs and arrhythmias were rare and gallop sounds were absent in No-DCM/MMVD DP. Dogs ≥ four years old and males had higher probabilities of PC-DCM-Echo. Prediction models incorporating PE variables with NTproBNP had an area under the curve (AUC) of 0.940 for distinguishing between PC-DCM-Echo and all other groups, which was similar to the AUC for NTproBNP (0.939) or cTnI (0.932) alone. Discrimination between No-DCM/MMVD and all other groups was similar for NTproBNP (0.781) and cTnI (0.742) as individual tests, however, models combining PE variables and NTproBNP increased the AUC to 0.812. An NTproBNP cut-off of ≥548 pmol/L, was 100% sensitive and 77.3% specific for detecting PC-DCM-Echo. Both NTproBNP and cTnI had good utility as sole tests to discriminate PC-DCM-Echo DP from all others. Models differentiating No-DCM/MMVD DP from all other DP were improved by using PE and NTproBNP together.