Disproportionate left ventricular remodelling in nonsyndromic mitral valve prolapse with significant mitral regurgitation: a CMR pilot study

Background Disproportionate left ventricular remodelling (DLVR) refers to excessive LV dilatation that cannot be explained by mitral regurgitation (MR) severity alone. While DLVR has been described in mitral valve prolapse (MVP) with mild-to-moderate MR, its significance in patients with severe MR remains unclear [1-3]. Purpose To characterise the structural phenotype of DLVR using cardiac magnetic resonance (CMR) in patients with preserved LV ejection fraction (LVEF) and isolated significant MR (grade 3+/4+) due to MVP who were referred for mitral valve repair (MVR). Methods In this prospective pilot study, 66 patients (20% female) undergoing preoperative CMR were analysed. LV volumes were derived from short-axis cine stacks and indexed to body surface area (BSA). MR regurgitant volume (RVol) was calculated as the difference between total LV stroke volume and aortic forward stroke volume. DLVR was defined as LV dilatation exceeding age- and sex-adjusted upper limits after correction for RVol: (LVEDV − RVol)/BSA − ULN(LVEDVi) > 0. Results DLVR was present in 19 patients (29%). MR grade distribution did not differ between groups (p = 0.586). Compared with patients exhibiting proportional remodelling (n = 47), those with DLVR showed significantly greater LV dilatation across all linear and volumetric indices: LVDDi (36±4 vs 31±3 mm/m²), LVESDi (23±4 vs 18±3 mm/m²), LVEDVi (156±30 vs 115±14 mL/m²), and LVESVi (63±18 vs 41±8 mL/m²; all p < 0.001). Eccentric hypertrophy was more prevalent in the DLVR group (95% vs 68%, p = 0.027). Biventricular and biatrial involvement was evident, with higher RVEDVi (94±17 vs 76±16 mL/m², p < 0.001) and larger atrial dimensions (all p ≤ 0.016). Guideline-based surgical thresholds were more frequently met in the DLVR group (LVESD≥40 mm: 58% vs 15%; LVESDi≥20 mm/m²: 84% vs 21%; both p < 0.001), consistent with a higher-risk phenotype. RVol was higher in DLVR (89±30 vs 70±18 mL, p = 0.002) and correlated strongly with LVEDV (r = 0.80, p < 0.001). However, regurgitant fraction (RF) (52±9% vs 48±10%, p = 0.212) was comparable between groups. Multiple linear regression analysis showed that RVol and male sex were significant independent predictors of LVEDVi, whereas age, MVP subtype and EROA (PISA, TTE) were not. Despite preserved LVEF, patients with DLVR had lower LVEF (60±5% vs 64±4%, p < 0.001). This finding, together with comparable systolic blood pressure and larger end-systolic dimensions, suggests reduced LV end-systolic elastance and possible intrinsic myocardial dysfunction. Conclusions Nearly one-third of patients with nonsyndromic MVP and significant MR exhibit DLVR. Although RF appropriately classifies MR severity, it does not adequately capture the extent of LV remodelling. RVol may better reflect the true haemodynamic burden. Whether DLVR represents an intrinsic cardiomyopathy or a time-dependent progression of MV disease warrants further investigation.