Extent of Ejection Fraction Improvement After Revascularization Associated with Outcomes Among Patients with Ischemic Left Ventricular Dysfunction

Introduction

The cornerstone of treatment of patients with heart failure (HF) and reduced ejection fraction (EF) continues to be optimal medical therapy, which is associated with significant improvement in survival and quality of life.^1,2 Partial congestive HF patients with reduced EF have an opportunity for recovery of EF to a normal level after a period of appropriate therapy. Patients with recovered EF might have more favorable outcomes, including a lower risk of mortality,^3–7 HF hospitalization^4,6,7 and better quality of life⁸ compared to patients with persistently reduced EF. EF has been reported to be a major predictor of improved survival in patients with HF and reduced EF.^9–11 For example, it has been demonstrated that after a period of appropriated therapy, patients with recovered EF (from EF <35% to EF >40%³ or from EF <45% to EF ≥45%⁵) has a lower risk of long-term mortality compared to patients with unrecovered EF (EF remained ≤40% or <45%). However, the association between EF improvement and survival benefit is not consistent from literatures. One observational study indicated lack of improvement in EF after CABG was not associated with poorer outcome compared with that of patients with improved EF.¹² From the STICH (Surgical Treatment for Ischemic Heart Failure) trial, it was revealed that there was no relationship between changes in EF and subsequent death.¹³ Whether the improvement in EF affects the subsequent risk of mortality needs to be further investigated.

Ischemic etiology is one of important risk factors for lack of EF improvement among patients with HF.^3,6,14 Revascularization including coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) may attenuate the ischemic state and reversing left ventricular (LV) remodeling,^15–18 thus improve the long-term outcomes of patients with LV dysfunction.^19–22 However, revascularization therapy among patients with LV dysfunction is not always effective. The extent and determinants of EF improvement after revascularization have not been well investigated.^16,23–27 The presence of myocardial viability might be one of correlates of EF improvement after coronary revascularization.^23,28,29 However, in different studies, about 12%²⁹ to 64%²³ patients remained EF unimproved after revascularization. We recently reported that diabetes mellitus (DM) associated with greater EF improvement after revascularization among patients with reduced EF.³⁰ The predictive factor associated with change in EF after revascularization needs to be further clarified.

Therefore, this study was performed to investigate 1) the extent of EF improvement following revascularization in patients with coronary artery disease (CAD) and preoperative EF ≤40%; 2) the determinants of absolute change in EF after revascularization; 3) the association between absolute change in EF and clinical outcomes.

Materials and Methods

Patient Selection

This was a real-world cohort study that used data from Beijing Anzhen Hospital. The study was registered in Chinese Clinical Trial Registry (No. ChiCTR2100044378). The study protocol was approved by the hospital’s ethics committee.

CAD patients with reduced EF (≤40%) who underwent CABG or PCI with a drug-eluting stent between January 2005 and December 2014, and with repeated EF measurements during follow-up were enrolled. Patients were excluded if they had concomitant noncoronary surgery, were diagnosed as ST-segment elevation myocardial infarction and had only one record of EF follow-up reassessment within 3 months after revascularization. The final study sample included patients who had EF reassessment by echocardiography 3 months after revascularization. Patients were then categorized according to the absolute change in EF: 1) EF worsened group (absolute decrease in EF >5%); 2) EF unchanged group (absolute change in EF −5% to 5%); 3) EF improved group (absolute increase in EF >5%).³¹

Data Collection and Definitions

Baseline demographic, clinical, laboratory, angiographic parameters, and medical therapy for the study patients were ascertained from Beijing Anzhen Hospital medical records. Baseline EF was captured within 30 days before PCI or CABG. Follow-up EF values were defined as the first EF measurement 3 months³² after revascularization assessed in Beijing Anzhen Hospital. Complete revascularization was defined as successful PCI (residual stenosis of <30%) of all angiographically significant lesions (≥70% diameter stenosis) in 3 coronary arteries and their major branches. A staged procedure within 90 days after discharge was acceptable For CABG, grafting of every primary coronary artery with ≥70% diameter stenosis was accepted as complete revascularization.

Outcome data were obtained from medical records at Beijing Anzhen Hospital and through telephone follow-up. Death was regarded as cardiovascular in origin unless obvious non-cardiovascular causes could be identified. Any death during hospitalization for repeat coronary revascularization was regarded as cardiovascular death. The follow-up time for patients started at the time of the first available EF measurement.^31,33,34

Statistical Analysis

Categorical variables were summarized as frequencies with percentages and continuous variables were expressed as mean ± SD. Baseline characteristics were compared among the EF worsened, EF unchanged and EF improved groups by using Chi-Square test for categorical variables and one-way ANOVA for continuous variables. Multinomial logistic regression was used to identify independent correlates of patients in the three EF categories as defined above: worsened, unchanged or improved EF. Outcome with improved EF was set as the reference category to calculate the relative risk ratios (RRR) of variables to have worsened EF or unchanged EF. Variables of demographics and history, preoperative echocardiography values, angiography and medical therapies as well as clinical chemistry were included in the analysis. Cumulative incidences were estimated by the Kaplan–Meier method and compared by Log rank test. The risks of outcomes were analyzed with a Cox proportional hazards regression model. The proportional hazards assumption was tested for individual covariates and globally on the basis of Schoenfeld residuals. All statistical analyses were based on 2-tailed tests. P<0.05 was considered statistically significant. Statistical analyses were performed with Stata version 14.0 (StataCorp).

Results

Patients Characteristics

Among 1816 initially identified patients, 78 patients who died within 3 months after revascularization, 764 patients were further excluded because EF was not evaluated 3 months after revascularization. Finally, 974 patients who had an initial EF ≤40% and had echocardiography reassessment 3 months after revascularization were enrolled in this study. The differences of the characteristics at baseline between enrolled and excluded patients are indicated in Supplemental Table 1.

The average age at baseline was 64.7±10.9 years (Table 1). Men comprised 83.5% of all subjects. Five hundred and fifty-six (57.1%) received PCI and 418 (42.9%) underwent CABG. After revascularization, 84 (8.6%) had EF worsened, 317 (32.5%) had EF unchanged and 573 (58.8%) had EF improved (Figure 1A). Age at baseline and sex distribution were similar among three groups (Table 1). The EF improved group had a significantly highest prevalence of DM (P=0.008) and lowest prevalence of myocardial infarction (MI) (P=0.001) compared with other two groups. The anatomic severity of coronary artery disease was similar among three groups. There was no significant difference in the proportions undergoing revascularization by PCI or CABG, and the groups had similar percentages of complete revascularization. The EF improved group had a significantly highest proportion of use of clopidogrel or ticagrelor compared with other two groups (P=0.031).

Table 1 Patient Characteristics at Baselinea

Figure 1 Predictive role of EF improvement for survival. (A) Absolute change in EF after revascularization; (B) The predicted probabilities of all-cause death at different values of EF. Abbreviations: EF, ejection fraction; CI, confidence interval.

Echocardiographic Characteristics

In the EF improved group, the preoperative EF was lowest compared with other two groups (P<0.001) (Table 2). Mean (SD) EF improved significantly, from 35.8% (4.7%) to 52.0% (8.6%), in the EF improved group (P<0.001). In the EF unchanged group, the change in EF was small but statistically significant, from 36.8% (3.7%) to 37.6% (4.7%) (P<0.001). In contrast, EF decreased significantly, from 37.3% (2.8%) to 27.5% (4.1%) in the EF worsened group (P<0.001). Both preoperative and postoperative LV end-diastolic diameter (LVEDD) (P<0.001) as well as end-systolic diameter (LVESD) (P<0.001) were smallest in EF improved group compared with other two groups. With the most reduction of LV size, the severity of mitral regurgitation was lowest in EF improved group (P<0.001). The mean duration between the preoperative and follow-up EF measurements in three groups was comparable (worsened group: 5.0±3.6 months; unchanged group: 6.1±2.0 months; improved group: 5.9±2.6 months, P=0.203).

Table 2 Echocardiographic Characteristics a

Predictors of Change in EF

Compared to patients with EF improved, patients who had history of DM had less likely to have worsened EF (relative risk ratios [RRR], 0.48; 95% confidence interval [CI], 0.28–0.83; P = 0.008) or unchanged EF (RRR, 0.74; 95% CI, 0.55–0.99; P = 0.043) (Table 3). Compared to patients with EF improved, patients who had history of MI had higher likely to have worsened EF (RRR, 1.88; 95% CI, 1.18–2.98; P = 0.008) or unchanged EF (RRR, 1.55; 95% CI, 1.17–2.04; P = 0.002). Patients with higher preoperative EF had greater likely of being in the EF worsened group (RRR per 1% increase in EF, 1.10; 95% CI, 1.03–1.17; P = 0.005) or EF unchanged group (RRR per 1% increase in EF, 1.06; 95% CI, 1.02–1.09; P = 0.002). Severity of LV remodeling as indicated by LVESD and LVEDD also significantly associated with change in EF. Neither anatomic severity of coronary vessels (as indicated by multivessel disease and left main disease) nor extent of revascularization (complete vs incomplete) was an independent correlate of change in EF. Compared to patients with EF improved, patients who had clopidogrel or ticagrelor had less likely to have unchanged EF (RRR, 0.67; 95% CI, 0.49–0.91; P = 0.011).

Table 3 Baseline Factors Associated with Worsened EF or Unchanged EF in Comparison with Improved EF

Outcomes

The median follow-up time was 3.5 years, during which 143 patients died and 117 patients were cardiac death. Greater extent of EF improvement after revascularization was significantly associated with lower risk of all-cause death. For each 5-unit increments in EF, the risk of death decreased by 20% (hazard ratio, HR, per 5% increases, 0.80; 95% CI, 0.73–0.86; P<0.001) (Figure 1B). Compared with EF improvement group, patients with EF worsened had significantly higher risk of all-cause death (HR, 3.35; 95% CI, 2.07–5.42; P<0.001) and cardiovascular death (HR, 4.12; 95% CI, 2.45–6.93; P<0.001) (Table 4, Figure 2A and B). Similarly, compared with EF improvement group, patients with EF unchanged had significantly higher risk of all-cause death (HR, 2.05; 95% CI, 1.40–3.01; P<0.001) and cardiovascular death (HR, 2.23; 95% CI, 1.44–3.45; P<0.001). Furthermore, patients with EF worsened had significantly higher risk of all-cause death (HR, 1.63; 95% CI, 1.00–2.66; P=0.048) and cardiovascular death (HR, 1.85; 95% CI, 1.10–3.11; P=0.020) compared with EF unchanged group. Those findings persisted in adjusted model.

Table 4 Risk of Outcomes

Figure 2 The association between EF improvement and survival after revascularization. (A) Kaplan–Meier curves estimating incidence of all-cause death; (B) Kaplan–Meier curves estimating incidence of cardiac death. Abbreviations: EF Worsened, absolute decrease in EF >5%; EF Unchanged, absolute change in EF −5% to 5%; EF Improved, absolute increase in EF >5%. EF, ejection fraction.

In addition, there were 123 (14.6%) patients had repeated revascularization by either PCI or CABG during follow-up. The EF improved group (17.1%) had a significantly highest proportion of repeated revascularization compared with EF unchanged group (12.2%) and EF worsened group (6.9%) (P=0.030).

Discussion

Available data suggest that HF might not always be a progressive disease. Several therapies including guideline-directed medical therapy,¹⁴ cardiac resynchronization therapy³¹ and revascularization^17,18,23 might improve EF for patients with LV dysfunction. Ischemic etiology is one of important risk factors for lack of EF improvement among patients with HF.^3,6,14 However, the extent of EF improvement after revascularization has not been well-established. A study enrolled 47 CAD patients with initial EF <50% who underwent coronary revascularization demonstrated that 42.6% had EF improved ≥10%.¹⁷ In the current study, after revascularization in patients with reduced EF (≤40%), about 10% remained follow-up EF worsened (absolute decrease in EF >5%), about 30% had follow-up EF unchanged (absolute change in EF −5% to 5%) and about 60% had EF improved (absolute increase in EF >5%).

In the current study, DM, no prior MI, lower preoperative EF and less LV enlargement were identified as factors associated with greater EF improvement after revascularization. The mismatch between blood supply and cardiac metabolic requirements in ischemic heart was more severe in diabetic compared with non-diabetic myocardium. Thus, revascularization might have greater effect on LV remodeling in patients with DM.³⁰ Patients with history of MI might have less viable myocardium which resulted in less opportunity to have EF improvement following revascularization. It has been reported that preoperative EF less than 27% was one of three prognostic factors which associated with greater survival benefit from CABG.²⁷ Lower preoperative EF and less LV enlargement might represent a critical cardiac stage that might benefit more from revascularization. In the current study, no guideline-directed medical therapy for HF was found to be predictive of EF improvement. One potential reason was that the effect of medication was attenuated by revascularization. After revascularization, only 200 (20.5%) had a follow-up EF ≤35%. Furthermore, although complete revascularization did not associate with EF improvement, the EF improvement group had highest proportion of repeated revascularization during follow-up. Considering the course of the disease, a possibility can be that EF initially improves after successful revascularization but subsequently worsens, and this feature is suspicious of recurrence of significant coronary stenosis. The association between repeat revascularization and EF improvement needs to be further investigated. In this study that extent of EF improvement was firstly identified, to our knowledge, as a factor associated with clinical outcomes after revascularization among CAD patients with reduced EF. For each 5-U increments in EF, the risk of death decreased by 20%. Compared with EF improvement group, patients with EF worsened or unchanged had significantly higher risk of all-cause death and cardiovascular death. Whether extent of EF improvement has clinical implication for ICD implantation-decision making needs to be further investigation.

All patients in the current study underwent isolated CABG. Moderate to severe mitral regurgitation were not treated simultaneously. However, mitral regurgitation might have a great impact on cardiac function recovery and outcomes.³⁵ This needs to be further investigated.

Since the present study was designed as a single-center and observational study, there is a possibility of patients selection bias. Since institutions and specific methods for measuring EF vary, only EF measurements by echocardiography in Anzhen Hospital were adopted. This restriction improved the accuracy of the EF measurements but increased the number of excluded patients. Of 1948 person-time EF measurements, 948 (97.3%) EF measurements before revascularization and 898 (92.2%) EF measurements during follow-up were by Simpson. It would be better to have myocardial viability test before PCI/CABG, nevertheless, it was sometimes difficult to perform for HF patients. There was little data of myocardial viability in current study cohort. Minimal patients (n = 11) had ICD therapy during the follow-up. This might overestimate the mortality especially for patients with EF worsened according to the current therapeutic strategy.

Conclusion

After revascularization in patients with reduced EF (≤40%), about 10% remained follow-up EF worsened, about 30% had EF unchanged and about 60% had EF improved. Patients with EF improvement were more likely to have DM, have no prior MI, have lower preoperative EF and have less LV enlargement. Changes in EF were inversely associated with the risk of mortality. The extent of EF improvement after revascularization might be a potential factor which defines clinical outcomes.