Are We Any WISER Yet? Progress and Contemporary Need for Smart Trials to include Women in Coronary Artery Disease Trials

Abstract

Despite calls to ensure proportionate representation of both sexes in biomedical research, women continue to be underrepresented in Cardiovascular Disease (CVD) clinical trials. A comprehensive analysis of seven large suspected Ischemic Heart Disease/Coronary Artery Disease (HD/CAD) clinical trials (PROMISE, ISCHEMIA, CIAO-ISCHEMIA, ORBITA, FAME, FAME 2 and COURAGE trial) provides understanding of contributions to barriers to enrollment of women and leads to strategies to address these barriers. Specifically, in the seven trials, enrollment of women did not exceed 27%, while numerous barriers are evident. Proposed strategies to improve womeńs inclusion in clinical trials, include adding reproductive stage/estrogen status, attention to study design inclusion/exclusion criteria using female thresholds, consideration of diagnostic and intervention study design to be inclusive, increasing women and minorities in leadership positions, including sex as a biological variable (SABV) in study design and statistical analysis, and addressing social and race/ethnicity barriers. Dedicated action to actualizing these steps are needed at this time to developing diagnostic and therapeutic strategies resulting in better care and improved outcomes for CVD in women.

Introduction

Cardiovascular disease (CVD) is the leading cause of death in American women, counting for up to 50% of mortality in disease populations.^{1, 2} Although large clinical studies conducted exclusively in women, such as the Women’s Health Initiative,^3–5 Women’s Health Study,^6–8 and the Womeńs Ischemia Syndrome Evaluation (WISE) studies^9–11 have contributed to representation of female-specific biomedical research, women continue to be underrepresented in enrolled cohorts of CVD clinical trials,^12–15 as well as scientific leadership in CVD investigation.^{12, 16, 17}

Women with CVD represent 45% of the target population, compared to 30% of enrollment in CVD trials.¹⁸ Overall, in CVD trials, the lowest enrollment of women is in heart failure (24%) and the highest in pulmonary arterial hypertension (77%).¹⁵ In suspected ischemic heart disease (IHD)/coronary artery disease (CAD) trials, enrollment of women has been slow, ranging from 24% to 28%.¹⁵ Current guideline-directed CAD therapies are based on data that predominantly includes male patients. The reasons why CAD trials populations are not female-representative are unclear, but sex-based differences in clinical presentation of IHD often exclude women from trials for not meeting inclusion criteria. Compared to men, IHD in women is identified less often,^19–21 judged to be non-cardiac more often,^{22, 23} diagnosed at a more advanced age,²⁴ and is treated less aggressively.^{21, 25} Furthermore, normal coronaries or non-obstructive coronary artery disease (CAD) is more prevalent in women with suspected IHD (up to 65%) compared to men (up to 32%).^{26, 27}

Improving equality for women and men in CAD trials would improve outcomes for both.^{28, 29} The objectives of this review are to perform a comprehensive analysis of seven large clinical trials in CAD and barriers reducing female enrollment, and to outline strategies to improve womeńs trial enrollment to reflect the real-world patient population, incorporating sex-stratified analysis and collection of sex-specific data.

Sex Distribution of Suspected IHD/CAD Clinical Trials and Contributing Factors

We performed an extensive literature search on articles that addressed challenges in women enrollment in CVD trials focusing in IHD/CAD trials, and we analyzed the variables in seven large clinical trials in IHD/CAD potentially contributing to enrollment of women. We did not perform a systematic review due to paucity of data to conduct an extensive review. Table 1 summarizes data from these seven IHD/CAD trials (PROMISE,³⁰ ISCHEMIA,³¹ CIAO-ISCHEMIA,³² ORBITA,³³ FAME,³⁴ FAME 2,³⁵ and COURAGE³⁶).

Table 1:

Summary of 7 CVD trials focused on suspected IHD/CAD, according to inclusion/exclusion criteria that contributed to enrollment of women

TRIAL	PROMISE	ISCHEMIA	CIAO-ISCHEMIA	ORBITA	FAME	FAME 2	COURAGE
First Author
Senior Author
No. of patients	10,006	5,179	208	200	1,005	1,220	2,287
Mean age (years)	60	64	63	66	64	63	61
Women Inclusion	52.7%	23%	66%	27%	26%	23%	15%
Primary intent of the trial intervention	Diagnostic	Therapeutic (procedure)	Diagnostic	Therapeutic (procedure)	Therapeutic (procedure)	Therapeutic (procedure)	Therapeutic (procedure)
Type of intervention	CTA vs. Functional testing.	Invasive strategy of MT, angiography, and revascularization when feasible vs. Initial conservative strategy MT alone, with angiography reserved for failure of MT.	SAQ Angina Frequency score and stress echocardiography over one year in ISCHEMIA trial screen failures with INOCA.	PCI vs. Placebo procedure for angina relief	Angiography-guided PCI vs. FFR-guided PCI.	FFR-guided PCI + best available MT (PCI group) vs. Best available MT alone (medical-therapy group).	PCI and optimal MT (PCI group) vs. Optimal MT alone (medical-therapy group).
Inclusion Criteria	Symptomatic outpatients without diagnosed CAD Age > 54 years (men) or > 64 years (women). Or age 45–54 years (men) or 50–64 years (women) with at least 1 cardiac RF (diabetes, PAD, CVD, current or past tobacco use, HTN, or dyslipidemia).	- Obstructive CAD (Stable CAD, after clinically indicated stress testing showed moderate or severe reversible ischemia on imaging tests or severe ischemia on exercise tests without imaging).	- Non-obstructive CAD (Patients with moderate or severe ischemia on stress echocardiography and no obstructive CAD by CTA).	- Obstructive CAD (18–85 years with angina or equivalent symptoms and at least one angiographically significant lesion ≥70% in a single vessel that was clinically appropriate for PCI).	- Obstructive CAD (Multivessel CAD, defined as stenoses of at least 50% of the vessel diameter in at least 2–3 of the major epicardial coronary arteries, and if PCI was indicated)	- Obstructive CAD Referred to PCI for: Stable Angina. Stabilized Angina class IV. Atypical or no chest pain with documented ischemia. AND: angiographic 1,2 or 3 vessel disease.	- Obstructive CAD Stableobstructive CAD CAD and class IV angina stabilized medically. Stenosis at least 70% in 1 proximal epicardial coronary artery and evidence of myocardial ischemia with either exercise or pharmacologic vasodilator stress test. 1 coronary stenosis of 80% and classic angina without provocative testing.
Exclusion Criteria	Unstable hemodynamic status or arrhythmias suspected to be caused by ACS. History of CAD within previous 12 months Clinically significant congenital, valvular, or cardiomyopathic heart disease	Non-obstructive CAD GFR < 30 ml/min. Recent ACS. Unprotected LM stenosis at least 50%. LVEF < 35%. NYHA class III or IV Unacceptable angina despite MT at maximum acceptable doses.		Stenosis ≥ 50% in nontarget vessel. ACS. Previous CABG surgery. LM stem CAD. Contraindic. to DES. CTO. Severe valvular disease. Severe LVEF. Mod. to severe PHTN.	Significant LM CAD. Previous CABG surgery. Cardiogenic shock. Extremely tortuous or calcified coronary arteries. Life expectancy of less than 2 years. Contraindic. to DES. Patients who were pregnant.	Prior CABG. LVEF <30%. LM disease.	Persistent class IV angina. Markedly (+) stress test. Refractory HF or cardiogenic shock. LVEF < 30%. Revascularization previous 6 months. Coronary anatomy not suitable for PCI.

(Abb.: PROMISE, “Prospective Multicenter Imaging Study for Evaluation of Chest Pain” (ClinicalTrials.gov number, NCT01174550); ISCHEMIA, “International Study of Comparative Health Effectiveness With Medical and Invasive Approaches” (ClinicalTrials.gov number, NCT01471522); CIAO-ISCHEMIA, “Changes in Ischemia and Angina Over 1 Year Among ISCHEMIA Trial Screen Failures With no Obstructive CAD on Coronary CT Angiography” (ClinicalTrials.gov number, NCT02347215); ORBITA, “Objective Randomised Blinded Investigation With Optimal Medical Therapy of Angioplasty in Stable Angina” (ClinicalTrials.gov number, NCT02062593); FAME, “Fractional Flow Reserve Versus Angiography for Multivessel Evaluation” (ClinicalTrials.gov number, NCT00267774); FAME 2, “Fractional Flow Reserve (FFR) Guided Percutaneous Coronary Intervention (PCI) Plus Optimal Medical Treatment (OMT) Verses OMT” (ClinicalTrials.gov number, NCT01132495); COURAGE, “Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation” (ClinicalTrials.gov number, NCT00007657; CTA=coronary computed tomography angiogram; CAD=coronary artery disease; RF=risk factor; PAS=peripheral artery disease; CVD=cerebrovascular disease; HTN=hypertension; ACS=acute coronary syndrome; MT=medical therapy; GFR=glomerular filtration rate; LM=left main; LVEF=left ventricular ejection fraction; SAQ=Seattle Angina Questionnaire; INOCA=ischemia with non obstructive coronary arteries; PCI=percutaneous coronary intervention; CABG=coronary artery bypass graft; DES=drug eluting stent; CTO=chronic total occlusion; PTHN=pulmonary hypertension; FFR=fractional flow reserve; HF=heart failure)

Except for the PROMISE and CIAO-ISCHEMIA trials which focused on suspected IHD rather than obstructive CAD as an inclusion criteria, enrollment of women did not exceed 27%. (Figure 1).

Figure 1:

Gender Distribution of Coronary Artery Disease Trials

When the primary intent of the trial intervention was therapeutic, including procedures, there was lower enrollment of women. By contrast, in the two trials where the primary intent of the intervention was diagnostic and not including an invasive procedure, women were well represented. Although contemporary ACC/AHA guidelines recommend invasive management of acute coronary syndrome, irrespective of gender, studies have demonstrated that women are less likely to undergo an invasive procedure such as coronary angiography or coronary intervention.^37–39 Thus, choice of intervention in randomized clinical trials may contribute to female participant under-representation.

Symptoms such as angina, multivessel CAD or angiographically significant coronary lesions were the inclusion criteria in most of these trials. Thus, lower enrollment of women in these trials may be explained by lower number of women referred for screening, as they are less likely to meet the inclusion criteria and may be less willing to participate in a study that involves an invasive procedure.

Failure to meet an angina inclusion criteria may be attributed to sex differences in clinical presentation, as women are more often judged as non-cardiac clinical presentation including the lack of chest pain as the main symptom of CAD.⁴⁰ In CIAO-ISCHEMIA, where the enrollment was two-thirds women, the inclusion criteria was ischemia in the setting of non-obstructive CAD, a condition that is more reported in women compared to men.²⁶

Notably, the first author of the primary trial publication was a woman in 3/7 trials (PROMISE, CIAO-ISCHEMIA, ORBITA), of which 2/3 (PROMISE, CIAO-ISCHEMIA) had high female enrollment. The senior author was a woman only in one trial (CIAO-ISCHEMIA). Further, we note that study of INOCA is predominantly been led by female investigators.

As we discuss later, the under-representation of women in leadership positions could be another factor contributing to lower women enrollment in clinical trials.

Barriers to IHD/CAD Trial Enrollment of Women

Reproductive stages/Role of estrogen

Historically, women of childbearing potential were banned from participating in clinical research, a policy set forth in 1977 by the Food and Drug Administration (FDA) due to the potential for medication causing serious birth defect.⁴¹ It was not until 1993 when the NIH Revitalization Act passed into law this policy was rescinded. However, a review of clinical trial publications from 2000 through 2020 identified that women of childbearing potential remain underrepresented in many disease categories and have a negative association with enrollment in clinical research.⁴²

Another barrier to enrolling women in clinical research is the complex role of endogenous and exogenous estrogen over a woman’s lifespan. The historical misconception was that fluctuating hormones would not only confound trial results but also be more costly due to the added measurements.⁴³ Over the past several decades the role of hormones and cardiovascular health has been extensively studied in both randomized controlled studies such as the Women’s Health Initiative (WHI),⁴⁴ and observational cohort studies such as the Study of Women Across the Nation (SWAN),⁴⁵ the WISE studies,^9–11 and the Nurses’ Health Studies (NHS).⁴⁶ These studies all have contributed to the understanding that different reproductive stages have positive or negative implications on cardiovascular health. For example, the early loss of endogenous estrogen in women who enter premature menopause (<40 years), whether surgically or natural, increases future risk of CVD. While women with longer reproductive lifespan have lower rates of CVD events.⁴⁷ Therefore, when enrolling women into clinical trials reproductive stages should include reproductive, menopausal transition and the age and type of menopause. Several menopause status algorithms have been published such as those used hormones and bleeding patterns in the WISE cohort.⁴⁸

Study design: Inclusion/Exclusion Criteria

While there is an increased awareness of the impact of CAD in women, there remain disparities between men and women in CAD diagnostic assessment. Sex-related differences in clinical presentation and biomarkers may explain under-diagnosis of IHD and low screening and/or enrollment of women in trials (Figure 2). Cardiac biomarkers (troponin and brain natriuretic peptide) and clinical presentation play an important role in diagnosis and risk assessment of CAD. Use of inappropriate diagnostic thresholds in troponin and atypical presentation of myocardial infarction (MI) in women likely contribute to under-diagnosis.

Figure 2:

Sex Distribution in Clinical Presentation and Biomarkers in Coronary Artery Disease

(Abb.: INOCA=ischemia with non-obstructive coronary arteries; LV=left ventricular; BNP=brain natriuretic peptide; MI: myocardial infarction)

Clinical presentation.

There are some clinical characteristics in presentation that explain why in some studies enrollment of women is lower, such as the fact that women have higher prevalence of ischemia with non-obstructive coronary arteries (INOCA) and spontaneous coronary artery dissection (SCAD).^{22, 49–51} Obviously these patients cannot be included in trials where the inclusion criteria is obstructive CAD, such as ISCHEMIA, FAME, and COURAGE. However, some sex-related differences in clinical characteristics and ischemic symptoms may drive under-enrollment of women in CAD trials. For example, women more commonly present with CAD symptoms other than chest pain, such as jaw pain, neck and shoulder pain, fatigue, nausea, vomiting or atypical angina that more often is ascribed to digestive rather than cardiac etiology,^{23, 52} contributing to low enrollment of women in CAD trials focused on angina. Women may also have more comorbidities (given their older age at disease presentation),^{53, 54} such as renal dysfunction or history of a hemorrhagic event, resulting in meeting exclusion criteria. Similarly, since women tend to develop CAD about 10 years later than men,⁵⁵ underrepresentation of women in CAD trials could be explained by lack of success in recruiting older women compared to older men.

Cardiac biomarkers.

Sex differences in cardiac biomarkers of myocardial injury may contribute to lower rate of MI diagnosis and subsequent IHD diagnosis in women with chest pain.⁵⁶ Compared to men, women are less likely to have biomarker evidence of myocyte necrosis, elevated CK-MB or troponin, but more likely to have elevation of brain natriuretic peptide (BNP), a marker used in risk stratification of acute coronary syndromes.⁵⁷ In 2007, the National Academy of Clinical Biochemistry (NACB) and International Federation of Clinical Chemistry (IFCC) committee recommended use of sex-specific reference ranges for some cardiac biomarkers.⁵⁸

Despite the widespread use of cardiac troponin as a marker of MI and that some studies have shown higher circulating troponin levels in males compared to females,^{56, 57, 59, 60} there is no sex-specific reference value reported, with consequent under-diagnosis of MI in women. Thus, women at risk may be missed when using male sex-specific thresholds with standard troponin criteria. High-sensitivity troponin assay using sex specific diagnostic thresholds help closing this gender bias.⁵⁶ High-sensitivity troponins have demonstrated that the 99^th percentile reference limits are nearly two-fold higher in men than in women.^{61, 62} In a prospective cohort study of 1126 patients with suspected ACS (46% women), a high sensitivity troponin assay with sex-specific troponin threshold (women 16 ng/L, men 35 ng/L) doubled the diagnosis of MI in women compared to a contemporary assay with single threshold (50 ng/L).⁵⁶ Thus, trial inclusion criteria that do not use sex specific diagnostic thresholds for MI may underrepresent women, who remain at risk for significant morbidity/mortality despite lower troponin or CK-MB levels.⁶³

Putative mechanisms of sex differences in biomarker levels include that men present with predominant obstructive CAD with atherosclerotic plaque rupture and myocyte necrosis, while women more often have non-obstructive CAD with coronary microvascular dysfunction (CMD) leading to ischemic myocyte injury.⁶⁴ Other potential mechanisms include sex differences in left ventricular mass and remodeling,⁶⁵ troponin release and clearance kinetics, blood pressure, and sympathetic nervous system activity.^{62, 66} When biomarkers are used as inclusion/exclusion criteria, sex-related differences can reduce female trial eligibility. A broader, multi marker approach would be appropriate for CAD studies, and a possible solution to enroll more women in trials.

Study Design: Intervention/Outcome

Pharmacologic interventions.

Enrollment of women in CAD clinical trials have been low (<30%) in both pharmacologic (antithrombotic,^67–69 lipid-lowering,^70–72 anti-inflammatory^73–76) and procedural trials.^{33–36, 77–79} Clinical trials of statins, which are currently recommended for primary and secondary prevention of CAD for both women and men, have historically enrolled lower women, including only 20–27% women.^{70, 71} In meta-analyses, the effect size of prevention with statin therapy was smaller or nonsignificant in women compared with men, leading the American Heart Association (AHA) to state that “evidence-based data supporting a statistically significant reduction of CVD events and all-cause mortality in primary prevention in women are lacking for statins”.⁸⁰ Not unexpectedly, guideline-recommended statin use, persistence and adherence remain lower in women compared to men,^{81, 82} and women are less likely to believe statins are effective or safe compared with men.⁸² A recent systematic review of 60 lipid-lowering randomized clinical trials (N=485,409) conducted between 1990 and 2018 demonstrated an increase in the enrollment of women over time (from 19.5% to 33.6%) but found that a common limiting factor was inclusion of only postmenopausal women or surgically sterile women.⁸³ Indeed, concern for reproductive/fetal safety is a common reason for excluding pregnant and lactating patients in clinical trials.⁸⁴ However, given the rising trend of MI hospitalizations and mortality in young women⁸⁵ and recent reports that CVD is the leading cause of death in pregnant and postpartum women in the United States,^{85, 86} trial criteria that exclude young, pregnant, or postpartum women should be reconsidered. Legislative efforts to improve pregnancy category labelling of drugs resulted in the 2015 “Pregnancy and Lactation Labelling Rule”, which was more descriptive about the risks of using a drug during pregnancy and lactation; however, the challenges associated with unknown drug safety continue to limit the study and the inclusion in clinical trials of this subpopulation. Recently the U.S. Food and Drug Administration (FDA) released a Drug Safety Communication requesting to remove the contraindication against using statins in all pregnant patients, stating that statins may be beneficial in preventing serious events in very high CVD risk pregnant patients.⁸⁷

Clinical trials to specifically study pharmacologic interventions in women should be developed with special considerations for high CVD risk pregnant or lactating women. Although ACE-I/ARB clinical trials should continue to exclude pregnant women, protocols can be designed to enroll women of reproductive age who are willing to avoid pregnancy during the study and undergo a negative urine pregnancy test.^{88, 89}

Procedural interventions.

Because most procedural trials include obstructive CAD as inclusion criteria fewer women are often eligible.^90–94 Indeed, percutaneous coronary intervention and coronary artery bypass graft trial participants have been predominantly male, being 70–85% of the total of participants.^{33–36, 77–79} However, registry data indicate that among patients with stable obstructive CAD, cardiovascular outcomes are not lower among women compared to men.⁹⁵ For example, in an analysis of the CONFIRM registry of patients undergoing coronary computed tomography angiography (CCTA) for suspected CAD, women with non-obstructive left main CAD had a nearly 80% higher risk for adverse cardiovascular events than men, and the relative hazard for 3-vessel CAD was higher for women compared to men.^{96, 97} Thus, efforts to improve representation of women should remain a goal in procedural CAD intervention trials.

Choice of primary outcomes.

Primary outcomes of CAD trials are often composite outcomes of death from cardiovascular causes and myocardial infarction and should include sufficient sex-stratified analysis and reporting.⁹⁸ Patient reported outcomes, such as angina-related symptoms, function, and quality of life are often not selected as primary outcomes, but may help identify important sex-specific long-term prognosis.⁹⁹ For example the recent ISCHEMIA trial found that an initial invasive strategy, as compared to an initial conservative strategy, did not reduce the primary endpoint of all-cause mortality or ischemic cardiovascular events among patients with stable obstructive CAD and moderate or severe ischemia.¹⁰⁰ However, patients randomized to the invasive strategy had greater improvement in angina-related health status (as measured by the Seattle Angina Questionnaire [SAQ]) compared to those randomized to the conservative strategy, and differences were larger for patients with more frequent angina at baseline.¹⁰¹ Sex-differences in improvement in SAQ have not yet been reported in the ISCHEMIA trial, although women had more frequent angina compared to men at baseline.¹⁰² Thus, it is unknown whether the invasive strategy improved angina-related health study in women. Since sex-specific biological and gender-related socio-cultural differences in angina exist, sex-stratified analysis of SAQ should be considered as a pre-specified key outcome in CAD clinical trials.¹⁰³

Choice of primary outcome should be carefully considered in the setting of ischemia with non- obstructive CAD. Indeed, in a secondary analysis of ISCHEMIA, women were more likely to have non-obstructive CAD, less extensive CAD on CCTA, and less severe ischemia on stress imaging.¹⁰² The PROMISE trial similarly demonstrated that women with suspected CAD were less likely to have a positive stress test or obstructive CAD on CCTA and had lower event rates compared to men.^{104, 105} The WISE study demonstrated a high prevalence of CMD in women with chest pain and without obstructive CAD,¹⁰⁶ and CMD in the absence of obstructive CAD can result in abnormal stress tests.¹⁰⁷ Taken together, these studies suggest that higher prevalence of CMD in the setting of non-obstructive CAD in women likely contribute to observed sex differences in test results and event rates. The WISE study also demonstrated that although women without obstructive CAD have better event-free survival than women with obstructive CAD,¹⁰⁸ CMD, persistent chest pain, and evidence of inflammation in women with non-obstructive CAD predicted adverse outcomes.^{26, 109}

In the ORBITA trial, the primary endpoint was the difference in exercise time increment between the groups, determined via serial cardiopulmonary exercise testing.³³ Since elderly women have higher prevalence of physical disability compared to men,¹¹⁰ inclusion of exercise-based procedures could have contributed to lower female enrollment, although sex-differences in participant age was not reported in ORBITA.

Women in Leadership Positions

The proportion of women in clinical trial leadership positions and their contribution as first and senior author in publications has increased over time.¹⁷ However, this up trend has been slow and remains low, as it was demonstrated in different studies (Figure 3). Specifically, Jagsi et al. found that from 1970 to 2004, the proportion of women among U.S. physician authors of original research, published in 6 top tier journals, increased from 5.9% to 29.3% of first authors and from 3.7%t to 19.3% of senior authors.¹¹¹ Ouyang et al. evaluated CVD trials published in 3 prominent journals from 1980 to 2017 and found an increase of 9.5% to 26% for female first author and 5.9% to 17.4% for female senior author.¹¹² Labinaz et al. evaluated preclinical studies between 2006 and 2016 in top CVD journals and found that the proportion of women as first author increased from 32.3% to 46.3%, and for senior author from 11.8% to 26%.¹¹³ According to a recent report, women represent only 10.1% of cardiovascular clinical trial leadership committees.¹⁶

Figure 3:

Women as First and Senior Author in publications over the last decades according to different studies

The most cited factors identified as barriers to the academic advancement of women are the constraints of traditional sex roles, manifestations of sexism in the medical environment and lack of effective mentors.^{113, 114} Carr et al. reported that women who had children published less and had less institutional support than their male colleagues.¹¹⁵ These barriers have been accentuated during the Covid-19 pandemic, likely due to the shift to on-line teaching/curriculum adjustments, and extraprofessional caregiving responsibilities of children and family members which disproportionately affect women.^{116, 117}

Trials with women as first or senior author tended to enroll higher proportions of female participants and stimulate female investigators to take part in cardiovascular clinical research. In fact, a recent analysis of heart failure trials confirmed this finding and found the proportion of women authors per trial was the only independent predictor of enrollment of female participants, possibly due to 1) women authors studying female predominant conditions, 2) higher likelihood of female participants to enroll in trials conducted by women investigators, or 3) study design that is more accommodating to female participants.¹² Thus, one way to increase the participatory rate of women in trials would be to ensure a diverse research team. Womeńs involvement in a research team is associated with higher quality research,¹¹⁸ considering sex as a biological variable (SABV), and with higher rates of reporting sex-specific outcomes.¹¹³

Statistical considerations

SABV is a research strategy to recognize sex as an important variable to consider when designing studies and assessing results. SABV should be incorporated in the study design and statistical analysis of clinical trials. In the study design, the proportion of women enrolled should be viewed relative to a priori expectancies. Sex can be considered as a stratification variable yielding a factorial design, which allows researchers to estimate more representative treatment effects, maximizes power to test the interaction between treatment and sex, and perform sub-group analyses. In the statistical analysis, SABV can be considered at three different levels: (i) adjusting analyses for sex in order to obtain a more accurate estimate of the treatment effect; (ii) testing interaction between sex and treatment when sex differences in the treatment effect are of interest and (iii) performing sub-group (or stratified) analyses by sex yielding treatment effects separated by sex that are not necessarily different when the interaction between sex and treatment is not statistically significant, but can be informative for a future study. Multivariate analysis to identify significant correlates of the primary end-point should include both sex and age. Furthermore, automatic designation of end-points according to patient sex should be performed, as mandated by several journals.¹¹⁹ Some journals do not mandate the reporting of sex considerations but encourage authors to follow the “Sex and Gender Equity in Research (SAGER) guidelines”.¹²⁰

Among the seven listed trials (Table 1), only three of them adjusted their analyses by sex as a covariate even though adjusting for covariates has been shown to improve power in the statistical analysis of clinical trials.^{121, 122} Sub-group analyses were performed in all seven trials and did not find statistically significant interaction effects, and the subgroup results agreed with the main results.

Notably, where differences in the treatment effects for sex are observed such that a p-value indicating the presence of treatment effect for a given sex is statistically significant but not significant for the other sex, claims of sex differences should not be made.¹²³ Researchers often incorrectly conclude there are sex differences solely based on stratified analyses instead of considering those results to plan future studies when the interaction is not statistically significant.¹²⁴

Social Barriers to Study Access/Participations

Although sex may affect subject’s consent to participate in the study and also their completion of a study, literature regarding the determinants is lacking. Some studies show that women are more likely to participate in health research than, while others do not.^125–128 Women are less likely than men to agree to participate in randomized controlled trials, which may be driven by a general discomfort with the randomization process.^{127, 128} Many factors may influence the decision to participate in a clinical trial. We summarize in Table 2 contributors to womeńs decisions to participate in clinical trials. Thus, to ensure eligibility of women in CAD clinical trials, it is essential to increase awareness about the risk of heart disease in women, among women themselves and in health care providers, take into account the interpersonal relationships and social norms of the community where the study is conducted, and recognize the impact of study visit hours, travel time, and scheduling child or older adult care.

Table 2.

Contributors Influencing Womeńs Participation in CVD Clinical Studies and Trials

The Role of Race/Ethnicity

There are additional barriers to racial and ethnic minority groups to clinical trial participation. Racial and ethnicity health inequities exist, including structural inequities, less access to care and less utilization of evidence-based medicines, resulting in minority women have higher mortality/morbidity from IHD.^{129, 130} Racial/ethnic minorities who either are uninsured or underinsured, often seek their medical attention in community health centers and are less likely to receive a specialty referral.¹³¹ Also, women from community health centers were less likely to obtain initial and follow-up consultations than men.¹³² According to the 2019 AHA Survey of Womeńs Cardiovascular Disease Awareness, awareness of the leading cause of death in women – heart disease – declined from 65% in 2009 to 44% in 2019, particularly in non-Hispanic Black and Hispanic Women.¹³³ The role of race and ethnicity in female participants and in leadership positions, likely contributes adversely to under-enrollment of minority groups in CVD clinical trials.

Strategies to Improve Enrollment of Women

Overcoming barriers to enrollment of women success will involve joint work between research team, health care system, community, and governmental authorities. In Table 3 we summarize strategies to improve enrollment of women in CVD studies and trials.

Table 3.

Strategies to improve womeńs enrollment in CVD focused on suspected IHD/CAD Trials

Summary and Conclusions

Despite efforts to ensure proportionate representation of both sexes in biomedical research, women continue to be underrepresented in the CAD clinical trials. Comprehensive analysis of seven large suspected IHD/CAD clinical trials (PROMISE, ISCHEMIA, CIAO-ISCHEMIA, ORBITA, FAME, FAME 2, and COURAGE trial) provide understanding of contributions to barriers to enrollment of women and lead to strategies to address these barriers. Specifically, in the seven trials we analyzed, enrollment of women did not exceed 27%. Numerous barriers are evident including sex differences obstructive CAD prevalence, clinical presentations, biomarkers, CVD outcomes, representation of women in leadership positions, statistical design and power, and social/racial/ethnic representation. Proposed strategies to improve womeńs inclusion in clinical trials, include adding reproductive stage/estrogen status, attention to study design inclusion/exclusion criteria using female thresholds, consideration of diagnostic and intervention study design to be inclusive, increasing women and minorities in leadership positions, including SABV in study design and statistical analysis, and addressing social and race/ethnicity barriers. Dedicated action to actualizing these steps are needed at this time to developing diagnostic and therapeutic strategies resulting in better care and improved outcomes for CVD in women.