Invasive therapy versus conservative therapy for patients with stable coronary artery disease: An updated meta‐analysis

Abstract Background Heart disease remains the leading cause of death in the United States. Although there are clear indications for revascularization in patients with acute coronary syndromes, there is debate regarding the benefits of revascularization in stable ischemic heart disease. We sought to perform a comprehensive meta‐analysis to assess the role of revascularization compared to conservative medical therapy alone in patients with stable ischemic heart disease. Hypothesis There is no significant difference in all‐cause mortality or cardiovascular mortality between invasive and medical arms. Methods We performed a systematic literature search from January 2000 to June 2020. Our literature search yielded seven randomized controlled trials. We analyzed a total of 12 013 patients (6109 in revascularization arm and 5904 in conservative medical therapy arm). Primary outcome was all‐cause mortality. Secondary outcomes included major adverse cardiac events (MACE) (death, myocardial infarction [MI], or stroke), cardiovascular mortality, MI, and stroke. Additional subgroup analysis for all‐cause mortality was performed comparing percutaneous coronary intervention (PCI) with bare metal stent versus conservative therapy; and PCI with drug eluting stent versus conservative therapy. Results There was no statistically significant difference in primary outcome of all‐cause mortality between either arm (odds ratio [OR] = 0.95; 95% CI [confidence interval], 0.83 to 1.08; p = .84). There were statistically significant lower rates of MACE (death, MI or stroke) in the revascularization arm when compared to conservative arm. Conclusions Our analysis did not show any survival advantage of an initial invasive strategy over conservative medical therapy in patients with stable coronary artery disease (CAD).

ESC/EACTS guidelines on myocardial revascularization. 4 On the contrary, benefits of revascularization in stable ischemic heart disease have remained a topic of debate and controversy. Several large clinical trials [5][6][7][8] have not shown any clear difference in mortality between revascularization and conservative therapy arms. New evidence from long term outcomes of FAME-2 9 trial has reported lower incidence of MI with PCI. 9 With publication of the results of the ISCHEMIA 10 trial and the 5 years follow-up results of FAME-2 9 trial, we sought to provide in this meta-analysis a comprehensive and updated assessment of the role of coronary revascularization coupled with medical therapy compared to conservative medical therapy alone (conservative therapy from here-on) in patients with stable ischemic heart disease.

| METHODS
We performed a systematic literature search on PubMed, Cochrane, Embase and Ovid MEDLINE database from January 2000 to June 2020. Search terms included (stable coronary artery disease OR stable angina OR angina) AND (medical therapy OR conservative management OR conservative strategy) AND (PCI OR expanded PCI OR revascularization OR CABG OR surgery). Inclusion criteria were randomized controlled trials (RCTs), trials in humans and English F I G U R E 1 PRISMA flow diagram depicting summary of study selection process language. Non-randomized trials, including observational studies, were excluded. Trials investigating treatment strategies other than directly comparing intervention to medical therapy were excluded. This led to exclusion of three major trials namely: ORBITA 11 trial where the outcomes analyzed were angina relief, rather than hard end points, and follow up period was very short; DEFER trial 12 which looked at outcomes of safety in deferring intervention in patients where fractional flow reserve (FFR) value was >0.75 (no conservative therapy alone arm) and 3 the trial by Hambrecht et al 13 which compared PCI with a 12 month exercise training program (no conservative therapy alone arm). There has been significant evolution in medical therapy for stable CAD. Pivotal trials looking at inhibitors of HMG-CoA reductase (statins) from 1990's 14,15 showed a significant improvement in mortality in patients with coronary artery disease. This has led to statins becoming the cornerstone of medical therapy in CAD. On the other hand, trials from 1990 to 2000's which looked at benefit of revascularization to medical therapy; had balloon angioplasty as the primary means of intervention. PCI with stents (preferably drug eluting over bare metal) is now standard practice for revascularization. Based on these significant paradigm shifts in therapies in both arms (revascularization and conservative), we excluded trials before 2000 as they did not reflect the current standard of care. This meta-analysis was T A B L E 1 Baseline characteristics of the included trials

| RESULTS
Our literature search yielded seven RCTs. We analyzed a total of 12 013 patients (6109 in revascularization arm and 5904 in conservative medical therapy arm). The mean age was 65 years in both arms.
The longest follow up available was 10 years. The baseline characteristics of the included patients in each trial are given in Table 1.
Outcomes of interest from included trials are given in (Table S1).

| Primary pooled analysis
The results of this analysis show no statistically significant difference  (Figure 3).

| DISCUSSION
This meta-analysis of over 12 000 patients shows no difference in primary outcome of all-cause mortality between revascularization with medical therapy and conservative therapy for patients with symptomatic but stable CAD. However, secondary outcomes analyses showed that revascularization is associated with reduced incidence of MACE (death, MI, or stroke) which is driven by a nearly 14% reduction in MI and cardiovascular death when compared to conservative therapy alone.
Current ACC/AHA/AATS/PCNA/SCAI/STS guidelines on stable CAD support coronary revascularization if the patients continue to have ischemic symptoms on optimal medical therapy or in whom revascularization may alter prognosis, such as in patients with reduced EF, where revascularization has a mortality benefit over conservative therapy. 18 Therefore, current guidelines established PCI or CABG as more of a second line therapy in patients with stable CAD. 19,20 Although CABG and PCI are acceptable means of coronary revascularization, the trials included in our meta-analysis have PCI as the predominant means of revascularization, except for BARI-2D, 21  Although the field of coronary interventions has come a long way in PCI techniques and equipment, multiple prior meta-analyses have either shown limited [30][31][32][33][34] or no difference in outcomes between PCI and conservative therapy. 35 Cumulatively, however, the increased peri-procedural MIs in the invasive group negated the effect of late onset MI's and there was no significant difference between both arms at 3 years.
Our analysis updates the current available literature and includes the 5 years follow-up data from the FAME 2 9  points. Large peri-procedural MI have been shown to be independently associated with increased mortality, however the clinical significance of smaller biomarkers elevations is unclear. 45,46 Periprocedural MI appears to be more of a marker of baseline patient risk, atherosclerosis burden, and procedural complexity (calcifications, tortuosity). 47 The differences in the definitions and our inability to separate the procedural and spontaneous MIs in many of the analyzed trials can contribute to a decreased apparent clinical benefit of revascularization.

| CONCLUSION
The current meta-analysis of seven contemporary RCTs did not show any survival advantage of an initial invasive strategy over conservative medical therapy in patients with stable CAD. There was a significantly lower risk of MACE in the revascularization group, predominantly driven by reduced spontaneous M. Shared clinical decision making based on patient's symptoms should be pursued to determine further management options.

| LIMITATIONS
The results of our meta-analysis need to be interpreted with some inherent limitations. We did not have access to patient level data. Patients with stable ischemic heart disease and low left ventricular ejection fraction or left main disease were excluded from all trials and hence the results should not be applied in those disease states. Although, the aim of our meta-analysis was to assess the benefit of revascularization for stable CAD, the results predominantly reflect use of PCI, given small number of patients who underwent CABG. We were unable to stratify the results of revascularization based on CABG versus PCI. Choice of CABG over PCI may reflect higher burden of disease and ischemia and expectedly outcomes like stroke and periprocedural MI could be different. Additionally, the type of troponin assay used and definitions of periprocedural MI have been varied in different trials and has been mostly linked to biomarker elevation in the absence of addressable symptoms.

DATA AVAILABILITY STATEMENT
The authors declare that all data supporting the findings of this study are available within the article (and its supplementary information files).