Percutaneous coronary intervention versus medical therapy for stable coronary artery disease: Meta-regression analysis of mortality and morbidity against stent use

Abstract

In a meta-analysis of randomized trials enrolling patients with stable coronary artery disease (CAD), Wijeysundera et al. [1] found that percutaneous coronary intervention (PCI) was associated with greater freedom from angina compared with medical therapy. Meanwhile, a meta-analysis by Jeremias et al. [2] of randomized trials demonstrated that revascularization by PCI in conjunction with medical therapy in patients with non-acute CAD was associated with significantly improved survival compared with medical therapy alone. Most trials included in the meta-analysis, however, compared medical therapy with balloon angioplasty without stenting. In a metaanalysis by Trikalinos et al. [3] summarizing head-to-head (direct) comparisons (only 4 randomized trials) in patients with non-acute CAD, succeeding advancements in PCI with stenting (bare metal) compared with medical therapy did not produce detectable improvements in deaths or myocardial infarction (MI). Also in their network meta-analyses integrating direct and indirect evidence, the point estimates were similar to the direct comparisons. To assess the impact of stenting in the comparison of PCI versus medical therapy in patients with stable CAD, we performed a meta-regression analysis using data of 9497 patients from randomized trials identified in the most recent and comprehensive (published in 2010 with literature search through June 2009) systematic review by Wijeysundera et al. [1]. The analyzed studies were 19 randomized trials: 15 trials [4–18] included in the meta-analysis by Wijeysundera et al. [1]; and 4 trials [19–22] excluded in it [1] because of not reporting on angina. We included 10-year results of the Medicine, Angioplasty, or Surgery Study (MASS II) [14] instead of 5-year outcomes. For each study, data regarding all-cause/cardiac death and any/non-fatal MI in both the PCI andmedical therapy groupswere used to generate odds ratio (ORs) and 95% confidence intervals (CIs). Study-specific estimates were combined using Mantel–Haenzel methods in both fixedand random-effects models. We developed a mixed-effects (unrestricted maximum likelihood) meta-regression model by plotting the natural logarithm of the OR for each outcome against the proportion of patients undergoing stenting in the PCI group. All analyses were conducted using Review Manager version 5.0 (Nordic Cochrane Centre, Copenhagen, Denmark) and Comprehensive Meta-Analysis version 2 (Biostat, Englewood, NJ). Pooled analysis (5470 patients) demonstrated a statistically significant 42% reduction in cardiac death with PCI relative to medical therapy in the fixed-effects model (OR, 0.58; 95% CI, 0.44 to 0.77; p=0.0001; p for heterogeneity=0.24; Fig. 1A). In the meta-regression analysis, we observed a statistically significant positive relationship between the proportion of stenting and the risk of cardiac death (p=0.02; Fig. 1B).When data from 5 stent-predominant trials (proportionof patientsundergoingstenting≥50% [N70% in all the5 trials])were pooled (3074 patients), PCI was not associated with a reduction in cardiac death relative to medical therapy (fixed-effects OR, 0.74; 95% CI, 0.52 to 1.06; p=0.10; p for heterogeneity=0.78; Fig. 1A). Meanwhile, pooled analysis demonstrated no statistically significant differences in all-cause death (fixed-effects OR, 0.89; 95% CI, 0.78 to 1.03; p=0.11; p for heterogeneity=0.30), anyMI (fixed-effects OR,1.18; 95%CI, 0.97 to 1.44; p=0.10; p for heterogeneity=0.33), and non-fatal MI (randomeffects OR, 0.99; 95% CI, 0.66 to 1.49; p=0.98; p for heterogeneity=0.0004).In the meta-regression analyses, we observed no statistically significant relationship between the proportion of stenting and the risk of all-cause death (p=0.16), any MI (p=0.91), or non-fatal MI (p=0.16). The results of our analysis suggest that PCImay reduce cardiac death by 37% over medical therapy in patients with stable CAD. The previous meta-analysis by Jeremias et al. [2] of 21 randomized trials also demonstrated that, compared with medical therapy alone, coronary artery revascularization by PCI for non-acute CADwas associated with a significant reduction in mortality (OR, 0.82; 95% CI, 0.68 to 0.99). However, 17 of the 21 trials compared medical therapy with balloon angioplasty without stenting, and in the remaining only 4 trials bare metal stents were implanted in 72% to 100% of cases. For only 4 randomized direct comparisons of PCI with bare metal stents versus medical therapy in another previous meta-analysis, Trikalinos et al. [3] recorded no significant differences for death (risk ratio [RR], 0.96; 95% International Journal of Cardiology 150 (2011) 90–120