Non-invasive diagnosis of heart failure with preserved ejection fraction with left atrioventricular coupling index

Background Right-heart catheterisation (RHC) is the gold standard for diagnosing heart failure with preserved ejection fraction (HFpEF), but it is invasive and not easily accessible. The left atrioventricular coupling index (LACI), which is the ratio of left atrial minimal volume to left ventricular end-diastolic volume, may serve as a surrogate marker for left ventricular diastolic dysfunction. Purpose This study compares three different imaging modalities for LACI assessment and evaluates LACI performance in diagnosing HFpEF. Methods We retrospectively recruited 9,130 patients from two academic hospitals in The Netherlands and United Kingdom. All patients underwent diagnostic RHC and cardiac magnetic resonance (CMR), with a subset additionally undergoing cardiac computer tomography (CCT), and/or echocardiography. Three cohorts were established: A (The Netherlands only), B (two-thirds from each hospital), and C (remaining one-third from both hospitals). Left atrial volume and left ventricular volume were measured at end-diastole. Agreement between all three imaging modalities were evaluated with Bland-Altman plots. The predictive performance of LACI for HFpEF was evaluated using ROC-analysis, with specific cut-off values that were determined in a development cohort and validated in an independent cohort. Results Cohort A (n = 114) assessed LACI measurement accuracy across all three imaging modalities. CMR and echocardiography showed strong agreement while CCT systematically overestimated LACI due to limitations in establishing true end-diastole. CMR LACI demonstrated superior diagnostic accuracy for HFpEF compared to echocardiography (AUC = 0.84 vs 0.65) (Figure 1). Cohort B (n = 1088) served as the development cohort for LACI cut-off values, identifying LACI ≥ 36% as the optimal threshold (sensitivity: 64%, specificity: 83%). Cohort C (n = 526) functioned as the validation cohort, testing the cut-off values determined in Cohort B and confirming LACI ≥ 36% as the optimal diagnostic cut-off for HFpEF (sensitivity: 58%, specificity: 82%). Conclusion LACI ≥ 36%, particularly measured via CMR, demonstrates strong predictive potential for HFpEF, providing a novel, non-invasive alternative to RHC.