166 Prediction of left ventricular reverse remodelling using factors from baseline echocardiograms

Introduction Dilated cardiomyopathies are heart muscle diseases characterised by the enlargement of the ventricles. This can lead to a diagnosis of heart failure with reduced ejection fraction and further cardiovascular complications including sudden death. We therefore embarked on a prospective cohort study, to identify which factors from baseline echocardiograms are associated with left ventricular reverse remodelling (LVRR), and thereby improved function. Methods We recruited 259/350 patients with a left ventricular ejection fraction < 40% on echocardiography who subsequently underwent cardiovascular magnetic resonance (CMR). Patients were excluded if they had claustrophobia during CMR assessment. T-test and Chi-squared test using IBM SPSS were used to compare clinical, medicinal, echocardiographic and CMR variables between the group of patients that had reverse remodelled left ventricles and those that failed to. Results Of the total 259 patients with left ventricular dilatation diagnosed on echocardiography, 30% (n=77) had recorded reverse remodelled left ventricles on following CMR assessment. There was no observed statistical significance due to clinical factors such as age, male gender, diabetes, hypertension, hypercholesterolaemia and stroke. However, atrial fibrillation (AF) showed 55% (42) vs 29% (52), p < 0.001. Patients with mitral regurgitation on echocardiogram showed statistical, significance with 45% (82) vs 31% (24), p = 0.04, where the left ventricle remained dilated (table 1). Conclusion Left ventricular function was recovered in 30% of patients within this cohort. Predictions of these factors are significant for the clinical approach to management and treatment of patients with ventricular dilatation as it would enable clinicians to identify patients who would benefit from early CMR and therefore improve prognosis. Future studies can use logistic regression with a larger sample size to better quantify which factors can predict LVRR. Conflict of Interest N/A