Women, Elderly, Minorities Vastly Underrepresented in Clinical Trials Over Decades

Trial sponsors should incentivize recruiting sites to be more representative with their enrollment, says Erin Michos.

Women, Elderly, Minorities Vastly Underrepresented in Clinical Trials Over Decades

Iniquities still run rampant in clinical trial enrollment, according to two new studies looking at the representation of women, older participants, and minorities in both research on lipid-lowering therapies and pivotal trials used to support US Food and Drug Administration approval of cardiometabolic drugs.

The studies underscore the glacial pace of progress in making sure that therapies are actually safe and effective for huge swaths of the population in whom they were never adequately tested.

“Randomized clinical trials serve as the basis for all of our guidelines—they're the best evidence that we have—and so when we take the results of these trials, we want to make sure the results are generalizable to the practice of the patients that we see every single day in practice,” Erin Michos, MD (John Hopkins Ciccarone Center for the Prevention of Cardiovascular Disease, Baltimore, MD), senior author for both papers, told TCTMD. “So, we need to make sure that the drugs are safe and effective in the patients that we treat.”

“This is exactly the same as has occurred over time,” said Nanette Kass Wenger, MD (Emory University, Atlanta, GA), who commented on the studies for TCTMD. “It must have been 20 years ago that I wrote one of the earlier papers saying that women and the elderly are underrepresented in clinical cardiovascular trials, and it has not improved.”

Positive effects from the 2014 Research for All Act, which dictated that women and underrepresented minorities should be more equally included in new US government-funded research, should begin to emerge in the next few years, Wenger said. The FDA has also taken action to ensure more gender equality among clinical trials the agency reviews, issuing guidance in 2014 for sponsors of medical device trials to evaluate sex-specific outcomes.

In the current COVID-19 era, more diagnoses have been observed in men, Wegner noted. “But we see a disproportionate amount of deaths, particularly in the Hispanic and the African American communities. Some of it is related to comorbidities, but much of it seems to be related to the social determinants of health. But as we see all of these disparities, it tells us that if we don't have representation of these populations in our clinical trials, we're not going to get the relative relevant information that I would like to have for a particular patient sitting across the desk from me.”

Women, Older Participants in Cholesterol Trials

For the first study, published last week in JAMA Network Open, Michos along with lead author Safi Khan, MD (West Virginia University, Morgantown), and colleagues looked at 60 randomized clinical trials of lipid-lowering therapies comprised of 485,409 participants conducted between 1990 and 2018. The cohort included both primary (47.8%) and secondary prevention (52.2%) trials; patients with ACS were the most commonly studied population (27.7%), followed by stable CAD (25.0%) and hypercholesterolemia (11.7%). Just over one-third (35.0%) of the trials were conducted in Western Europe, exactly one-third (33.3%) were multiregional, and 21.7% took place in North America.

It must have been 20 years ago that I wrote one of the earlier papers saying that women and the elderly are underrepresented in clinical cardiovascular trials, and it has not improved. Nanette Kass Wenger

Overall, the proportion of women in the trials was 28.5%, increasing from 19.5% in 1990-1994 to 33.6% in 2015-2018 (P = 0.01 for trend). Michos acknowledged that this rise is “pretty modest” and “underwhelming” given the length of time studied.

Women were more likely to be included in primary versus secondary prevention trials (42.2% vs 22.2%; P = 0.02). It was common for studies to either solely include women who were postmenopausal or surgically sterile (28.3%) or exclude pregnant (23.3%) or lactating women (16.6%).

When analyzing the prevalence-to-participation ratio (PPR), the researchers found that women were underrepresented compared to their disease burden in trials looking at diabetes (0.74), heart failure (0.27), stable coronary heart disease (0.48), and ACS (0.51).

Fewer than half of all trials (n = 23) reported the proportion of participants who were at least 65 years old. Overall, this cohort comprised 46.7% of all patients with no increase observed between 1995-1998 (31.6%) and 2015-2018 (46.5%; P = 0.43 for trend).

Only 53.0% and 36.6% of trials included sex- and age-specific results, respectively, neither of which improved over the study period.

Michos said that while she was encouraged by the fact that about 70% of studies between 2015-2018 reported sex-specific outcomes, if women are underrepresented generally, the studies might be underpowered to look for these differences in the first place. “Women are not just smaller men,” she said. “There are a lot of differences related to hormones and drug metabolism, and we just need to make sure that these drugs work the same in women and are effective. And the same thing with older adults, obviously how drugs are metabolized and absorbed in older adults [is important]. Older adults often take other drugs and are at more risk for drug-drug interactions and safety, and so it really is important that we have adequate enrollment of these demographics since these are the patients we see every day in clinic.”

FDA Pivotal Cardiometabolic Trials

In the second study, published online last week ahead of print in the Journal of the American Heart Association, Michos, lead author Muhammad Shahzeb Khan, MD (John H. Stroger, Jr. Hospital of Cook County, Chicago, IL), and colleagues exclusively included 143 pivotal trials from the FDA website that the agency used to approve various cardiometabolic drugs—24 for cardiovascular conditions and 11 for diabetes—between 2008 and 2017. There were a median 5,930 participants per trial supporting each drug.

Women made up 36.4% of participants overall, with no significant increase in prevalence observed over the study period (29.6% to 48.6%; P = 0.29 for trend). Multiregional trials were more likely to enroll women compared with those conducted in Western and Central Europe (39% vs 25%; P < 0.01). Women were most-often enrolled in trials of pulmonary arterial hypertension (77%), hypercholesterolemia (46%) and diabetes (46%).

Trials of coronary heart disease (PPR 0.52), heart failure (0.58), and ACS (0.68) featured proportional underrepresentation of women, whereas those of pulmonary arterial hypertension tended to overrepresent women (1.35).

Women are not just smaller men. Erin Michos

A sex-based analysis of findings was available for all 35 drugs, but only 11 were accompanied by a conclusive statement of safety and efficacy between the sexes. Further, 21 of the 24 novel cardiovascular drugs demonstrated similar effects for men and women.

Race was reported for all but one of the cardiometabolic drug programs. Overall, 81% of the enrolled populations were white, 4% were black, 12% were Asian, and 11% were Hispanic/Latino. The proportion of white participants (P = 0.01) and black participants (P < 0.01) were significantly related to trial location.

Drug approval year was not associated with either the enrollment of women (P = 0.29) or underrepresented minorities (P = 0.45).

These findings are potentially “even more important because once these get FDA approved, these drugs are going to be applied in clinical practice,” Michos said. “And what we really want to know is: is this generalizable to the patients we see every day in practice?”

Incentives, Greater Access

The reasons why these subgroups have long been underrepresented in clinical trials is multifactorial, according to Michos. “Some studies have suggested that women, once they're approached, may be more likely to decline or not participate,” she said. “It may be that women perceive more harm with research and women tend to underestimate their cardiovascular risk, so if they don't think that it applies to them, they may be less likely to participate.”

Michos added that much remains to be done to make clinical trials more accessible to all, whether that’s lengthening the hours during which participants can come in for study visits or even having investigators go to patients instead of vice versa.

Also, “there is some data suggestive for both women and minorities that when the investigators and the trial data collectors are more representative of the patients that we see, they might be more likely to build trust,” she said. “As you increase the diversity among the research team, it may also help increase the diversity of trial participants.”

Subconscious bias also cannot be ignored. “I think most trialists want to enroll women; I don't think anybody's actively saying that they don't want to include women, but sometimes people are just not approached,” Michos said. “There might be some assumptions that they don't want to participate or with older adults there may be some concerns that maybe they're going to have impairments with visual or hearing and won't be able to participate. . . . And for women, rightfully, there might be some reasonable exclusions actually during pregnancy and breastfeeding, but you don't want to not enroll women of any childbearing potential because then you might not enroll women for decades and decades of their life.”

With the threat of these biases, she said she would like to see trial sponsors incentivize recruiting sites to be more inclusive with the participants they enroll. “If you incentivize investigators to make sure they meet certain criteria with number of women or number of minorities, . . . then they go out and find the people,” Michos said.

Until we can see trial results from those initiated after the new equal representation legislation was enacted, Wenger said “the advocacy emphasis that we can have is in the labeling. If elderly advocacy groups see a population and the population studied just doesn't include any elderly individuals, maybe they can say that this is not applicable to an elderly population.” However, this can be a “very political matter,” she added

  • The study published in JAMA Network Open was supported by the Amato Fund for Women’s Cardiovascular Health Research.
  • Khan, Khan, Michos, and Wenger report no relevant conflicts of interest.