Ventricular volume ratio as a novel imaging biomarker for disease prediction in 26,516 UK Biobank participants

Abstract

Abstract Background/Introduction Right and left ventricular volumes are established imaging biomarkers for disease discrimination and outcome prediction. Limited research in select disease cohorts suggests additional clinical value from consideration of disproportionate (or imbalanced) dilatation of one ventricle over the other. The prognostic utility of this "ventricular volume ratio" (VVR) has not been tested in large, unselected cohorts. Purpose To examine associations of right-to-left and left-to-right VVR with cardiovascular phenotypes, prevalent disease, and incident cardiac outcomes in the UK Biobank. Methods The analysis includes 26,516 participants with cardiac magnetic resonance (CMR) available. We calculated right-to-left ventricular volume ratio (RV/LV), and the reciprocal left-to-right ventricular ratio (LV/RV), and identified participants with asymmetry above the 95th percentile of healthy cases. Multivariable linear and logistic regression models were used to calculate associations between ventricular volume asymmetry and sample characteristics, other cardiac imaging features, existing disease and incident disease. Incident disease associations were measured using Cox proportional hazards models adjusted by age, sex, systolic blood pressure, body mass index, smoking, Townsend deprivation score, hypertension, diabetes, hyperlipidaemia and other key imaging metrics (LV mass indexed to body surface area, LV ejection fraction and RV ejection fraction). The average follow-up time was 4.6 years (SD 1.3 years). Results Ventricular volume asymmetry in either direction was associated with increased age, lower RVEF, poorer LV global function index, poorer global longitudinal strain and greater maximum wall thickness. While RV-dominant volume asymmetry was associated with increased risk of all-cause mortality (HR= 1.56, 95% CI= [1.08, 2.26], p=0.019), LV-dominant volume asymmetry was found to associate with a much wider array of factors in the UK Biobank population. LV dominance was significantly associated with smoking, hypertension, diabetes, hyperlipidaemia, existing respiratory disease and existing heart disease. In fully-adjusted models, LV dominance was associated with increased risk for ischaemic heart disease (1.46 [1.12, 1.91], p=0.005), non-ischaemic cardiomyopathies (3.62, [1.93, 6.79], p < 0.001) and a two-fold increased risk for heart failure (2.30, [1.63, 3.26], p < 0.001). Conclusion(s) LV/RV is easily calculated from existing imaging metrics, and could potentially have important added value as a biomarker in predicting heart disease, incremental to well-established risk factors and imaging metrics.