Predicting early mitral valve disease in Cavalier King Charles

Abstract

INTRODUCTION: Cavalier King Charles Spaniels (CKCS) are predisposed to developing myxomatous mitral valve disease (MMVD). Dogs with stage B2 MMVD benefit from medication. OBJECTIVES: To develop (1) breed-specific cut-offs for individual screening tests and (2) predictive models utilizing physical examination (PE), ECG, radiograph, and blood-based biomarker variables in combination for identification of echocardiographic stage B2 MMVD in preclinical CKCS. ANIMALS: Adult, preclinical CKCS not receiving cardiac medications (N = 226). MATERIALS AND METHODS: Prospective, cross-sectional study. Enrolled CKCS underwent PE, ECG, radiography, Doppler blood pressure measurement, echocardiography, and biomarker testing. Dogs were grouped by MMVD stage using echocardiography only. The discriminatory ability of individual tests to identify stage B2 was assessed, and prediction models were developed using variables derived from four 'tests' (PE, ECG, radiography, and biomarkers). RESULTS: N-terminal pro-B-type natriuretic peptide (NT-proBNP) and radiographic vertebral heart size (VHS) had the best discriminatory ability of individual diagnostic tests to differentiate stage A/B1 CKCS from stage B2, with an area under the curve (AUC) of 0.855 and 0.843, respectively. An NT-proBNP ≥1138 pmol/L or a VHS ≥11.5 had high specificity for predicting stage B2 (90.1% and 90.6%, respectively). Prediction models incorporating variables from multiple tests had better discriminatory ability than single tests. The four-test prediction model had an AUC of 0.971. Three and two-test models had AUCs ranging between 0.925-0.959 and 0.895-0.949, respectively. CONCLUSIONS: Both NT-proBNP and VHS have good utility for predicting echocardiographic stage B2 MMVD in CKCS as individual tests. Prediction models incorporating multiple test variables have superior discriminatory ability.