Comprehensive annular and subvalvular repair of chronic ischemic mitral regurgitation improves long-term results with the least ventricular remodeling.

Abstract

Undersized ring annuloplasty for ischemic mitral regurgitation (MR) is associated with variable results and >30% MR recurrence. We tested whether subvalvular repair by severing second-order mitral chordae can improve annuloplasty by reducing papillary muscle tethering. Posterolateral myocardial infarction known to produce chronic remodeling and MR was created in 28 sheep. At 3 months, sheep were randomized to sham surgery versus isolated undersized annuloplasty versus isolated bileaflet chordal cutting versus the combined therapy (n=7 each). At baseline, chronic myocardial infarction (3 months), and euthanasia (6.6 months), we measured left ventricular (LV) volumes and ejection fraction, wall motion score index, MR regurgitation fraction and vena contracta, mitral annulus area, and posterior leaflet restriction angle (posterior leaflet to mitral annulus area) by 2-dimensional and 3-dimensional echocardiography. All groups were comparable at baseline and chronic myocardial infarction, with mild to moderate MR (MR vena contracta, 4.6±0.1 mm; MR regurgitation fraction, 24.2±2.9%) and mitral annulus dilatation (P<0.01). At euthanasia, MR progressed to moderate to severe in controls but decreased to trace with ring plus chordal cutting versus trace to mild with chordal cutting alone versus mild to moderate with ring alone (MR vena contracta, 5.9±1.1 mm in controls, 0.5±0.08 with both, 1.0±0.3 with chordal cutting alone, 2.0±0.4 with ring alone; P<0.01). In addition, LV end-systolic volume increased by 108% in controls versus 28% with ring plus chordal cutting, less than with each intervention alone (P<0.01). In multivariate analysis, LV end-systolic volume and mitral annulus area most strongly predicted MR (r(2)=0.82, P<0.01). Comprehensive annular and subvalvular repair improves long-term reduction of both chronic ischemic MR and LV remodeling without decreasing global or segmental LV function at follow-up.