Renal function and coronary bypass surgery in patients with ischemic heart failure

Abstract

Objective: Chronic kidney disease is a known risk factor in cardiovascular disease, but its influence on treatment effect of bypass surgery remains unclear. We assessed the influence of chronic kidney disease on 10-year mortality and cardiovascular outcomes in patients with ischemic heart failure treated with medical therapy (medical treatment) with or without coronary artery bypass grafting.

Methods: We calculated the baseline estimated glomerular filtration rate (Chronic Kidney Disease Epidemiology Collaboration formula, chronic kidney disease stages 1-5) from 1209 patients randomized to medical treatment or coronary artery bypass grafting in the Surgical Treatment for IsChemic Heart failure trial and assessed its effect on outcome.

Results: In the overall Surgical Treatment for IsChemic Heart failure cohort, patients with chronic kidney disease stages 3 to 5 were older than those with stages 1 and 2 (66-71 years vs 54-59 years) and had more comorbidities. Multivariable modeling revealed an inverse association between estimated glomerular filtration rate and risk of death, cardiovascular death, or cardiovascular rehospitalization (all P < .001, but not for stroke, P = .697). Baseline characteristics of the 2 treatment arms were equal for each chronic kidney disease stage. There were significant improvements in death or cardiovascular rehospitalization with coronary artery bypass grafting (stage 1: hazard ratio, 0.71; confidence interval, 0.53-0.96, P = .02; stage 2: hazard ratio, 0.71; confidence interval, 0.59-0.84, P < .0001; stage 3: hazard ratio, 0.76; confidence interval, 0.53-0.96, P = .03). These data were inconclusive in stages 4 and 5 for insufficient patient numbers (N = 28). There was no significant interaction of estimated glomerular filtration rate with the treatment effect of coronary artery bypass grafting (P = .25 for death and P = .54 for death or cardiovascular rehospitalization).

Conclusions: Chronic kidney disease is an independent risk factor for mortality in patients with ischemic heart failure with or without coronary artery bypass grafting. However, mild to moderate chronic kidney disease does not appear to influence long-term treatment effects of coronary artery bypass grafting.

Keywords: CABG treatment effect; chronic kidney disease; coronary artery bypass grafting; medical therapy; survival.

Figure 1:

Kaplan-Meier all-cause mortality estimates of the entire STICH population (MED and MED+CABG) separated into CKD stages 1,2 and 3. The figure includes statistical comparisons of the different CKD stages and also shows the number of patients at risk by CKD stage below. Note that the worse renal function, the higher is all-cause mortality. Patients in CKD stage 4/5 were omitted for low patient number (N=28). CKD: Chronic Kidney Disease; CI: Confidence Interval; eGFR: estimated Glomerular Filtration Rate

Figure 2:

Kaplan-Meier all-cause estimates for all-cause mortality by treatment group separated into three panels with CKD stage 1 (A), CKD stage 2 (B) and CKD stage 3 (C). The figure also shows the N-values per group, number of events and 10 year mortality rate. The statistics (p-value and Hazard ratios) are univariate comparisons of CABG vs. MED for each group. MED: Medical Treatment Group; CABG: Coronary Artery Bypass Grafting; eGFR: estimated Glomerular Filtration Rate

Figure 3:

Kaplan-Meier estimates for death or cardiovascular hospitalization by treatment group separated into three panels with CKD stage 1 (A), CKD stage 2 (B) and CKD stage 3 (C). The figure also shows the N-values per group, number of events and 10 year mortality rate. The statistics (p-value and Hazard ratios) are univariate comparisons of CABG vs. MED for each group. The graphs have been truncated once the number of patients at risk were below 10. MED: Medical Treatment Group; CABG: Coronary Artery Bypass Grafting; eGFR: estimated Glomerular Filtration Rate

Figure 4:

Spline curves illustrating the risk all cause death (A), death or cardiovascular hospitalization (B) and cardiovascular mortality (C) at 10 years for MED and CABG+MED patients as a function of the patients’ eGFR. The shaded areas illustrate the confidence intervals. The interaction p-value reports the association of eGFR and treatment interaction with respect to each endpoint of interest. Note that CABG was superior to MED in all patients with an eGFR between 30 and 90 ml/min. The difference becomes larger if the combined endpoint of mortality plus cardiovascular rehospitalization or cardiovascular mortality was used. The lower and the upper end of the curves represent only few patients which is reflected by the broader confidence interval areas. MED: Medical Treatment Group; CABG: Coronary Artery Bypass Grafting; eGFR: estimated Glomerular Filtration Rate

CONSORT Diagram STICH Trial