Cardiac magnetic resonance-derived atrioventricular coupling index for predicting left ventricular reverse remodeling in dilated cardiomyopathy.

AIM: To investigate the value of the cardiac magnetic resonance (CMR)-derived left atrioventricular coupling index (LACI) in predicting left ventricular reverse remodeling (LVRR) in patients with dilated cardiomyopathy (DCM) receiving standardized medical therapy.

MATERIALS AND

METHODS

This retrospective study enrolled 64 hospitalized DCM patients managed with guideline-directed therapy. Within a one-year follow-up, patients were stratified into LVRR and non-LVRR groups using echocardiographic criteria. Three multivariable models were constructed from predictors identified in univariable analysis (p<0.2), each uniquely integrating either the LACI, left atrial strain, or left ventricular strain. The Delong test was used to compare model performance.

RESULTS

In our study, the LVRR group exhibited significantly lower LACI, serum creatinine, uric acid, and all cardiac volumetric parameters (left ventricular and atrial volumes) (all p<0.05). Conversely, the LVRR group showed higher left ventricular and atrial ejection fractions, all left ventricular strains, and all left atrial strains (all p<0.05). LACI was an independent predictor of LVRR (odds ratio = 0.585; 95% confidence interval [CI]: 0.391-0.874; p=0.009). The combined model incorporating LACI demonstrated the best predictive performance, with an area under the curve (AUC) of 0.825, outperforming the models incorporating left ventricular strain (AUC = 0.805) or left atrial strain (AUC = 0.791).

CONCLUSION

For predicting LVRR in DCM, LACI provides value comparable to or superior to than strain parameters when incorporated into models based on clinical and conventional CMR parameters.