Long-term results of surgical ventricular reconstruction and comparison with the Surgical Treatment for Ischemic Heart Failure trial

Objective: The role of surgical ventricular reconstruction (SVR) in patients with ischemic cardiomyopathy is controversial. Observational series and the Surgical Treatment of IsChemic Heart failure (STICH) trial reported contradictory results. SVR is highly dependent on operator experience. The aim of this study is to compare the long-term results of SVR between a high-volume SVR institution and the STICH trial using individual patient data. Methods: Patients undergoing SVR at San Donato Hospital (Milan) were compared with patients undergoing SVR in STICH (as-treated principle) by inverse probability treatment-weighted Cox regression. The primary outcome was all-cause mortality. Results: The San Donato cohort included 725 patients, whereas the STICH cohort included 501. Compared with the STICH-SVR cohort, San Donato patients were older (66.0, lower quartile, upper quartile [Q1, Q3: 58.0, 72.0] vs 61.9 [Q1, Q3: 55.1, 68.8], P < .001) and with lower left ventricular end-systolic volume index at baseline (LVESVI: 77.0 [Q1, Q3: 59.0, 97.0] vs 80.8 [Q1, Q3: 58.5, 106.8], P = .02). Propensity score weighting yielded 2 similar cohorts. At 4-year follow-up, mortality was significantly lower in the San Donato cohort compared with the STICH-SVR cohort (adjusted hazard ratio, 0.71; 95% confidence interval, 0.53-0.95; P = .001). Greater postoperative LVESVI was independently associated with mortality (hazard ratio, 1.02; 95% confidence interval, 1.01-1.03). At 4 to 6 months of follow-up, the mean reduction of LVESVI in the San Donato cohort was 39.6%, versus 10.7% in the STICH-SVR cohort (P < .001). Conclusions: Patients with postinfarction LV remodeling undergoing SVR at a high-volume SVR institution had better long-term results than those reported in the STICH trial, suggesting that a new trial testing the SVR hypothesis may be warranted.