Predictors of adverse events after surgical ventricular restoration for advanced ischaemic cardiomyopathy

Abstract

Objective: Post-infarction ventricular remodelling has been graded (I-III) according to the loss of systolic left ventricular silhouette curvature changes. Although surgical ventricular restoration (SVR) has been extended to type III ischaemic cardiomyopathy, the results are less satisfactory. We sought to identify geometric and functional predictors of late outcome after SVR.

Methods: Among 144 patients who underwent SVR since 1998, a subgroup of 31 patients (age: 65.2+/-7.6 years) was analysed. Inclusion criteria were: type III cardiomyopathy, no associated procedure except coronary artery bypass grafting, prior anterior infarction, absent-to-2+ mitral regurgitation, elective operation, follow-up > or =18 months (mean: 44+/-26; longest: 96 months). Probability of events was estimated with the Kaplan-Meier method. A Cox multivariable regression model was constructed selecting eight potential predictors of four adverse events: death, cardiac death, recurrent heart failure (New York Heart Association class III or IV) and left ventricular re-remodelling, defined as a 25% increase of end-systolic volume index after SVR, or an end-systolic volume index > or =50 ml.m(-2).

Results: Early and late mortality were zero and 6% (2/31 patients, one cardiac-related death). NYHA class and all echocardiographic functional variables significantly improved early after SVR. Freedom (+/-standard error (SE)) from heart failure was 97%+/-3%, 93%+/-5%, 77%+/-11% and 64%+/-15%, whereas freedom from left ventricular re-remodelling was 97%+/-3%, 80%+/-8%, 60%+/-12% and 39%+/-15%, respectively, 1, 3, 5 and 7 years after SVR. Multivariable analysis identified baseline mitral regurgitation degree and sphericity index as independent predictors of recurrent heart failure (p=0.025; hazard ratio (HR)=7.80 (95% confidence intervals (CIs): 1.29-47.19)) and left ventricular re-remodelling (p=0.047; HR=2.84 (95% CIs: 1.01-7.95)). Both predictors also correlated with a higher recurrence of end-systolic volume index > or =50 ml.m(-2) at late follow-up.

Conclusions: Despite advanced cardiomyopathy, SVR determines left ventricular volume reduction and improved systolic function. Baseline absent-to-moderate mitral regurgitation and a more spherical left ventricular geometry predict a less favourable clinical and functional outcome, suggesting a possible rationale for wider indications for combined correction of 2+ mitral regurgitation and undersizing of the mitral annulus, particularly in patients with sphericity index > or =0.75.