The inclusion criteria for this study were RCTs comparing routine revascularization therapy, in combination with current standard OMT, versus initial OMT alone in patients with CCS defined by coronary angiography or a positive functional study consisting of exercise or pharmacologic stress testing. To be eligible for analysis, the patients in the trials should have received a stent, regardless of whether it was a bare-metal stent (BMS) or drug-eluting stent (DES). OMT was defined as a medical regimen consisting of at least an antiplatelet, antianginal, antihypertensive and lipid-lowering therapy. Additionally, only trials where more than half of the patients in the medical therapy group received a statin were included. Trials sponsored primarily by a pharmaceutical company were considered eligible. Trials that compared intervention to other strategies, such as lifestyle changes, were excluded, and only those directly comparing intervention to medical therapy in CCS were included.

Outcomes of interest were death mainly cardiovascular; myocardial infarction (MI) including procedural and nonprocedural MI; heart failure; and angina relief. The events reported at the longest follow-up time point were abstracted from each of the trials. There were some differences in the definitions of acute myocardial infarction and procedure-related myocardial infarction, as noted in the Supplementary Appendix.

Databases including PUBMED/Medline, EMBASE, and CINAHL were systematically searched. These databases were searched between 2000 and 2024 and yielded 51 results. The search strategy combined the two concepts of intervention, revascularization and OMT, investigating the impact on outcomes of interest.

Studies were selected and the PRISMA flow diagram sequence (Figure 1) was followed for the inclusion of the RCTs that were to be evaluated. A summary of the literature review, including the reasons for exclusion, is shown in Figure 1. The literature search identified 51 articles or abstracts using the MeSH terms Stable Ischemic Heart Disease, Revascularization, Optimal Medical Therapy, Cardiovascular Disease in women and using only English-language articles as a constraint for the research. The four studies included were COURAGE,26 BARI-2D,27 FAME-228 and ISCHEMIA13 and the study characteristics are described in Table 1.

Figure 1.

PRISMA flow diagram sequence.

Two reviewers (M.C.C. and I.H.) independently assessed the titles and abstracts of the identified citations to select eligible studies. The full texts of the selected citations were also screened independently by these two reviewers. Any disagreements were resolved through discussion and consensus and after discussion with a third reviewer (B.T.). The final data from the accepted articles were extracted by both reviewers, who cross-checked each other's work to ensure accuracy and resolve inconsistencies.

Study quality was assessed based on the design including randomization method, success of enrolment, data management, blinding of participants and personnel, follow-up and reporting of outcomes, funding, publications that included both the data and editorial and critical reviews of the study in question.

Following the COCHRANE tool, the risk of bias was assessed. These studies differ due to their chronological separation, reflecting advancements in strategies and standard-of-care at the time. Revascularization methods varied: COURAGE and FAME-2 used only PCI, while BARI-2D and ISCHEMIA used either PCI or CABG. Stents used were BMS in COURAGE, DES and BMS in 35% and 56% of PCI in BARI-2D, and DES in over 97% of PCI subjects in FAME-2 and ISCHEMIA.

In BARI-2D and ISCHEMIA, there was no standardized CABG strategy, but ISCHEMIA specifically encouraged the use of a left internal mammary graft (LIMA). CABG strategies varied, including on-pump and off-pump CABG, hybrid revascularization combining CABG and PCI, and total arterial revascularization using arterial grafts instead of saphenous vein grafts.

Except for ISCHEMIA, participants were enrolled and randomized after coronary angiography, introducing enrollment bias. Clinician bias, assuming revascularization is superior to OMT, may have led to the exclusion of participants. Some patients may have been considered unsafe in the OMT arm alone without evidence that revascularization is superior. ISCHEMIA randomized patients before angiography, which was performed only for those in the revascularization arm. Patients in the OMT arm did not undergo routine angiography; non-invasive testing was used to demonstrate ischemia. To avoid including patients with left-main (LM) disease in the OMT arm, all patients underwent a protocol-mandated CCTA to ensure LM patients were not mistakenly included.

The risk of bias pertaining to sequence generation was low in all the four trials. The risk of bias related to allocation concealment was high in all the trials because both the investigator and the patient knew in which arm of the trial the patient was entered.

Outcomes were assessed by independent board members in each trial, and the risk of bias was low in all four. Data reporting was complete, with very low attrition and exclusions, all of which were reported. There was no evidence of selective outcome reporting in any of the trials.

Patients enrolled in the four trials had similar age ranges and the presence of comorbidities such as hypertension and diabetes, except for the BARI-2D trial, which specifically enrolled patients with diabetes. They also had comparable left ventricular ejection fraction (LVEF) and history of stroke. The prevalence of smokers was also similar across the trials.

The mean percentage of women enrolled in these studies was 22%. The studies showed that, for the most part, the women were older, had better LVEF, and less extensive CAD compared to men. For example, in the COURAGE trial, 14% of women versus 24% of men had triple-vessel disease.

In the ISCHEMIA trial, which randomized participants before cardiac catheterization to an initial invasive or conservative management strategy, women had less severe ischemia on imaging than men. Fewer women underwent coronary computed tomographic angiography (CCTA) than men (67.4% versus 70.0%; p=0.02), and women were less likely to undergo revascularization (73.4% of invasive-assigned women received revascularization versus 81.2% of invasive-assigned men; p<0.001) because women had less severe CAD on invasive angiography, with more non-obstructive disease (12.3% versus 4.5% with no stenosis ≥50%; p<0.001) and less triple-vessel CAD (29.8% versus 42.7%; p<0.001). In the PCI group, women had fewer coronary stenoses per patient. Even within the subgroup of invasive-assigned patients who underwent CCTA, women still had a higher likelihood of non-obstructive CAD on invasive angiography (10.2% versus 3.9%; p<0.001). Consistently, the most common reason revascularization was not performed was the absence of obstructive CAD, which was the case for 71.7% of women not revascularized versus 53.3% of men. The likelihood of revascularization did not differ by sex in analyses stratified by the number of diseased vessels at coronary angiography. In the invasive strategy group, the first revascularization was more likely to be percutaneous coronary intervention in women compared with men (79.1% versus 72.8%; p=0.008). In those with double or triple-vessel CAD, including severe proximal left anterior descending artery stenosis on CCTA, there was no difference in the rate of revascularization (80.5% in women versus 81.9% in men; p=0.825). There was no difference in the rate of coronary artery bypass grafting by sex (30.3% in women versus 41.4% in men; p=0.223) in those with multivessel disease who underwent revascularization. This contrasted with BARI-2D, where women who underwent CABG received fewer grafts than men, likely due to their less extensive and severe disease.

A large proportion of women showed elevated baseline levels of LDL and HbA1c. In BARI-2D, women were less likely than men to achieve the target goal for LDL-C of 100 mg/dL over the five years of follow-up (adjusted OR 0.62, 99% CI: 0.50–0.77, p<0.0001), a result not observed in the COURAGE trial. The adjusted odds for achieving an HbA1c <7% or blood pressure <130/80 mm Hg over five years of follow-up were not significantly different between sexes in both studies. Mirroring BARI-2D, in ISCHEMIA, despite the trial's rigorous design and focus on lifestyle intervention and intensive use of OMT, women were less likely to achieve targets for blood pressure, LDL-C, and HbA1c. The goal of achieving systolic blood pressure (SBP) <140 mm Hg was met by 73.6% of women compared to 77.9% of men (p=0.003) and the LDL-C goal of <70 mg/dL was attained by 50.2% of women versus 61.3% of men (p<0.001). The utilization of OMT in women was unexpectedly lower than in men throughout the study's follow-up period. Women were less likely to be prescribed high-intensity statin therapy (60.7% versus 64.3%; p=0.025), even among those with obstructive CAD identified during cardiac catheterization.

Revascularization did not demonstrate a statistically significant difference in benefit over OMT alone in terms of the primary composite endpoint, with no significant interaction by sex. In the COURAGE trial, PCI did not show additional benefit over OMT in reducing death and non-fatal MI (HR 0.89; 95% CI 0.77–1.03; p=0.10 for women; HR 1.02; 95% CI 0.96–1.10; p=0.52 for men) over 4.5 years. Similarly, in the BARI-2D trial, revascularization showed no significant benefit (HR 1.11; 99% CI 0.85–1.44).

In the FAME-2 trial, it was hypothesized that fractional flow reserve (FFR)-guided PCI could provide better outcomes compared to medical therapy in patients with stable CAD, which was not seen (HR 0.70; 95% CI 0.48–1.04). There was no significant difference in incidence of death or MI (HR: 0.86; 95% CI: 0.36–2.09) between men and women.29

In ISCHEMIA, no significant difference was found (HR 0.93; 95% CI 0.77–1.13; p=0.47). The rate of periprocedural MI was lower among women (adjusted HR 0.40; 95% CI 0.21–0.76; p=0.005), consistent with lower rates of revascularization among women. When considering only those in the invasive strategy group who underwent revascularization, there was no difference in the four-year rate of procedural MI between women and men.

However, the COURAGE trial showed a significantly reduced need for subsequent revascularization in the PCI plus OMT group compared to OMT alone (p<0.001 for both), which was also observed in the BARI-2D trial. It was noted that women were more likely than men to have CHF. Furthermore, in the COURAGE trial, fewer women in the PCI plus OMT arm were subsequently hospitalized for heart failure compared to those who did not undergo PCI and remained on OMT.

Women exhibited a longer duration of angina, increased angina frequency, and more angina-related physical limitations, indicating poorer angina-related health status. This was reflected by a higher proportion of women in the more unfavorable Canadian Cardiovascular Society functional classes (classes 3 and 4), despite having less extensive CAD.

In these studies, freedom from angina was similar between men and women, regardless of treatment assignment. However, in the COURAGE trial, patients in the OMT arm showed less improvement compared to those in the PCI plus OMT arm (p=0.002). This disparity might explain the lower rate of subsequent revascularizations in the OMT subgroup. In contrast, in the BARI-2D study, women remained more symptomatic than men (adjusted OR 1.51, 99% CI 1.21–1.89, p<0.0001), experiencing more angina and a lower functional status as measured by the Duke Activity Status Index. Women who underwent PCI in FAME-2 were significantly less likely to undergo urgent revascularization compared with women treated medically (HR 0.26; 95% CI 0.1–0.68; p=0.007). Persistence of angina was reported in 18.6% of women in the OMT group and 8.9% of women in the PCI group (p=0.04) at one year, and 10.3% versus 4.5%, respectively, at three-year follow-up (p=0.15).30

Interestingly, in ISCHEMIA women had more emergency room admissions for noncardiac chest pain than men (3.3% versus 1.8%, p=0.004) but no increased admissions for the same reason.