Follow-up
The New York Heart Association (NYHA) class, percentage of systolic anterior motion (SAM), and LVOTO were significantly improved after the surgery. There were no differences in the incidences of LVOTO and SAM among the four groups. Detailed information is presented in Table 3. In addition, coronary artery computed tomography was performed in 49 patients 1 year after the surgery (15 in the CABG with LIMA, 19 in the CABG with SVG group, and 15 in the myocardial unroofing group). The graft remained patent in only 2 (13.3%) patients in the LIMA group and 16 (84.2%) patients in the SVG group. However, no MB was detected in the myocardial unroofing group.
The mean follow-up period was 34.2±17.2 months. A total of 31 events occurred: 24 patients died (20 patients without MB, 1 patient in the CABG group, 1 patient in the unroofing group, and 2 patients in the untreated group) and 7 patients (2 patients in the CABG group, 1 patient in the unroofing group, and 4 patients in the untreated group) had a non-fatal MI. Of the aforementioned deaths, 4 were due to cancer, 10 due to sudden cardiac death, 8 due to heart failure, 1 due to MI, and 1 due to infective endocarditis. The 3-year cumulative event-free survival of all-cause death (p=0.89) and cardiovascular death (p=0.63) were similar among the four groups (Figure 3 A and B). However, compared with the other three groups, the 3-year cumulative event-free survival of nonfatal MI (p<0.001) and combined endpoints of all 3 events (including all-cause death, cardiovascular death, and nonfatal MI) (p=0.02) were significantly lower in the no-treatment MB group (Figure 3 C and D).
Univariable and multivariable Cox regression analysis were performed to investigate the predictors of combined endpoints in the follow-up period (Table 4). Cox regression analysis indicated that non-treated MB was a significant independent predictor of combined endpoints (HR: 4.06, 95% CI: 1.60–10.32, p<0.001). In addition, the NYHA class III or IV (HR: 4.72, 95% CI: 1.12–19.92, p=0.04) and postoperative hospital stays (HR: 1.07, 95% CI: 1.03–1.11, p=0.001) were also independent predictors of combined endpoints of all three events.