Despite known racial and ethnic differences in patient response to certain drugs, the incorporation of those factors into the construction and execution of global clinical trials remains as complex and controversial as ever.
In the U.S., where about 22 percent of people consider their race to be something other than Caucasian, fewer than 5 percent of participants in trials registered with ClinicalTrials.gov are non-Caucasian, University of California, Davis, researchers recently found. In cancer, one of the most heavily researched fields in medicine, the picture is even worse, with fewer than 2 percent of clinical studies focusing on non-Caucasian ethnic or racial groups, the researchers found.
"The proportion of minorities in clinical research remains very low and is not representative of the U.S. population with cancer," said Moon Chen, an expert in cancer health disparities and an associate director at the UC Davis Comprehensive Cancer Center who led the research, published in Cancer earlier this year.
That failure to reach minorities matters, because minority populations have an outsize claim on the projected number of cases of invasive cancers in the U.S. according to a 2008 analysis by University of Texas researchers. As American demographics change, the researchers estimated a 99 percent increase in cancer incidence among minorities by 2030, especially among Hispanics, Asian/Pacific Islanders and multiracial people, compared to a 31 percent increase among Caucasians during the same period.
Three states (California, Hawaii and Texas), two territories (Puerto Rico and Guam) and the District of Columbia are populated primarily by minorities, Moon pointed out. Even outside those places, Hispanics and Asian Americans account for an ever-increasing portion of the U.S. population.
"In order for us to benefit from clinical trials, we need more people of different phenotypes," he said.
Numerous programs have sought to increase minority participation in trials, but their impact has been mixed and their history is not as long as one might imagine. It wasn't until passage of the 2012 FDA Safety and Innovation Act (FDASIA) that Congress mandated that the FDA issue a public report on the extent to which demographic subgroups including sex, age, race and ethnicity, are included in new drug applications (NDAs) submitted to the FDA.
Simply moving to include more women in clinical trials was a big step at the NIH a couple decades ago, said Moon during a workshop the agency held in July. "Considering the demographic changes in America . . . we want to now have inclusion of everybody," he said.
The good news? As of 2011, the latest year the FDA reviewed in its August 2013 report on the subject, every one of the 31 NDAs approved by its Center for Drug Evaluation and Research (CDER) reported trial composition by race. Within the Center for Biologics Evaluation and Research (CBER), the picture was similar, with all five applications approved in 2011 providing racial and/or ethnicity data. Furthermore, for all of the reviewed applications, race composition of the trials was made publicly available. Race subset safety analysis data were publicly available for about half and, for nearly 80 percent, race subset analyses for efficacy and clinical pharmacology were available, too.
Even though the racial composition in some trials was consistent with the disease's racial prevalence — such as in melanoma and head lice, which primarily affect Caucasians — trials in other indications failed to make that match. In trials testing type 2 diabetes therapies, for instance, African American representation was less than 5 percent, even though African Americans make up about 13 percent of the U.S. population and have a higher prevalence of type 2 diabetes in general.
But do race and ethnicity really count for anything more than a crude stand-in for a patient's genetic ancestry, a profile that's significantly more evaluable in light of the ever-falling cost of genetic testing? Measured by efforts to improve the inclusion of minorities in clinical trials today, the consensus appears to be yes.
"We know that some subpopulations may for whatever reason have a shorter course of progressing to more severe disease – may have other co-morbidities such as concomitant hypertension in diabetes in larger proportions," Jonca Bull, director of the FDA's Office of Minority Health, told BioWorld Today. "We see these data as informing the policy that anchors our oversight in protecting and promoting public health. We're mindful of who the population of patients are who are likely to receive a particular therapy. Keeping that population science in mind, as we look at trial design . . . the fundamental question is, 'Who are the patients?'"
Achieving racial diversity in trials matters, the argument goes, because without including minorities, important differences in drug efficacy and safety among different racial groups can be missed. "If you look at the data for the recent hepatitis C approvals, inclusion of demographic subgroups was part of their approach because there were differences in terms of genotypes. Genotype 1 is more prevalent and harder to treat in certain demographic subgroups, notably in blacks and Hispanics," said Bull. "So taking that kind of data into account from what we know from population science has certainly informed our understanding of where new therapies are bringing about important improvements."
Given the scope of the problem, it's going to take multiple approaches to achieve more regular inclusion of minorities in clinical trials. In light of that, a number of efforts have evolved and expanded over time.
At the FDA, the Office of Minority Health, or OMH, was created in 2010 to advise the agency on ways to reduce health disparities among racial and ethnic subgroups. Developed under a mandate of the Affordable Care Act, the OMH seeks to alleviate differences that can impact how frequently a disease affects a group and how often the disease causes death. Chief among its priorities is addressing disparities that show up in diabetes, obesity, HIV/AIDS, tobacco use and aging.
Giving the effort even more urgency, as part of section 907 of FDASIA, Congress directed the FDA to take a closer look at the inclusion and analysis of demographic subgroups in applications for drugs, biologics and devices – including by sex, race and ethnicity, and age – and report on its findings.
In its review, the FDA found that its statutes, regulations and policies generally give product sponsors a solid framework for providing data in their applications on the inclusion and analysis of demographic subgroups but that there was still a lot of room for improvement, "Section 907 gave us a vehicle for considering improvements," FDA Commissioner Margaret Hamburg wrote in August, prefacing the FDA Action Plan to Enhance the Collection Availability of Demographic Subgroup Data.
"It's a guiding principle of the FDA to meet the health needs of patients across the demographic spectrum," Hamburg wrote, pointing out that diversity in clinical trials is "one of the core tenets of rigorous biomedical research, as well as a guiding principle of the FDA's goal to meet the health needs of patients across the demographic spectrum."
FIRST, UNDERSTAND THE PROBLEMS
To achieve that goal, the agency has prioritized actions in three major categories: quality, participation and transparency. The FDA's quality-targeted actions seek to improve the completeness and quality of demographic subgroup data collection, reporting and analysis. While FDA statutory and regulatory requirements, guidance documents, policies and procedures generally inform sponsors and agency staff about how demographic data should be collected and analyzed, the agency said, there is still room "to further enhance the quality and consistency of these efforts," it acknowledged in August.
Because the list of barriers to clinical trial participation by underrepresented subgroups in the U.S. is so complex, the agency's planned short-term actions revolve mostly around building a better understanding of the issues. To expand on insights about socioeconomic and research design issues uncovered by a survey the FDA's OMH funded with the University of Maryland's Center of Excellence in Regulatory Science and Innovation last year, it will convene a meeting of experts with the Institute of Medicine in 2015 to better understand contemporary barriers to participation by minorities in clinical trials.
Furthermore, the FDA said it plans to work with industry to help ensure appropriate use of exclusion criteria in clinical trial protocols, which it said can at times "unintentionally exclude certain demographic groups, thereby limiting useful information about product performance in diverse populations."
Efforts to make demographic subgroup data more available and transparent, on the other hand, will be more about making sure data that are already available in the agency's vast store of trial information makes it into the hands of patients and other interested groups. Stakeholders have told the agency on multiple occasions that clinical data about demographic subgroups aren't easy to access or understand when making treatment decisions. "Because much of this information already exists but is buried within medical product reviews on FDA's website, we will be taking significant steps to extract the data and make it readily available to the public," the agency wrote in the action plan.
Within one to three years, both CDER and CBER plan to post demographic information from pivotal studies for newly approved drugs and biologics, among other actions to better present data in ways that are sensitive to the needs of underrepresented populations.
To track it all, the FDA is establishing an internal steering committee to oversee and monitor implementation of the action plan. A public workshop will be held sometime before January 2016 to discuss progress.
"I think that for where we are now, these are probably more sociologic constructs than ones that we want to ground in well-characterized science," said Bull. "But the groupings do provide what I like to consider a coarse tool in looking at population science."
FIGURING OUT THE RIGHT APPROACH
Not every effort on the path to greater inclusion will be a success. One notable effort to align therapeutic effect with patient identity came along in the 2000s. Long before the OMH existed, Nitromed Inc. sought to leverage a link between greater incidence of hypertension, one of the major risk factors for congestive heart failure, and African Americans.
Hailed by then-FDA associate director of medical policy Bob Temple as "a striking example of how a treatment can benefit some patients even if it does not help all patients," the agency approved its first – and to date its only – race-targeted medicine, the heart failure drug Bidil (isosorbide dinitrate and hydralazine hydrochloride). Most scientists have remained skeptical about putting energy into incorporating the factor into their assessments, and Bidil's commercial track record didn't encourage drugmakers to emulate its approach either. Nitromed quit selling or promoting it 2008 due to poor commercial performance. (See BioWorld Today, Jan. 17, 2008.)
The medical community at large took a dim view of the venture. "In sum, Bidil is on the scene today mostly in answer to the question of how a company could generate a profit and much less in answer to the question of which drug would best help which group of patients and why," wrote Howard Brody and Linda Hunt in the November 2006 issue of the Annals of Family Medicine. "Family physicians might be leery of offering too much support to this so-called way of 'advancing' medical knowledge," they concluded.
Yet, population-level insights can still have clear value. As recently an February, global researchers running a prospective, multinational, epidemiological study of epidermal growth factor receptor (EGFR) mutations in patients from Asia with newly diagnosed advanced lung adenocarcinoma found high EGFR mutation frequency in tumors from those patients. Their work led them to conclude that mutation testing should be considered for all patients with stage IIIb/IV adenocarcinoma, since such mutations can impact the efficacy of EGFR tyrosine kinase inhibitors such as Iressa (gefitinib, Astrazeenca plc) and (Tarceva (erlotinib, Astellas Pharma Inc. and Roche AG) compared with chemotherapy.
No matter whether the collection of subgroup data on race is merely advisable or mandated, it isn't always relevant, some in the biopharma industry say.
Were anyone to seem a natural to harbor concerns about the shortfalls of racial diversity in clinical trials, it would be Menghis Bairu, chairman and CEO of Serenus Biotherapeutics Inc. His young and growing company is dedicated to in-licensing, registering and commercializing therapies approved in the U.S., Europe and Japan to address unmet medical needs in sub-Saharan Africa. (See BioWorld Today, June 18, 2014.)
But Bairu is quick to point out that Serenus intends to serve all Africans with equal diligence, regardless of race. But the important thing, he said, is not so much race as the genetic makeup of each individual.
"In the era of personalized medicine, there won't be a drug for the white race, and one for the black race," said Bairu. "There will be a drug for each individual depending on their peculiar genetic makeup or code, and the subtle nuanced differences from the societal norm."
Furthermore, Bairu pointed out, most clinical trial programs are multicenter, often spanning the globe, and even when they are U.S.-centered they invariably have African and/or African American representation, making it a given that there will be representation from other regions and races.
In March, Cubist Pharmaceuticals Inc. came under criticism during a meeting of the FDA's Anti-infective Drugs Advisory Committee in which Anna Mazzucco, scientific adviser to the National Center For Health Research, expressed concern that while Hispanic and African American patients are more likely to contract methicillin-resistant Staphylococcus aureus, 85 percent of the patients in all the phase III Sivextro (tedizolid phosphate) trials were Caucasian.
With such a low percentage of minority enrollment, Mazzucco said, "any safety or efficacy problems specific to minority patients, potentially because of different metabolism of the drug, would be invisible because of the small number of patients that were combined into the analysis with a much larger sample of white patients."
In response, Cubist told BioWorld Today, Sivextro studies were open to minority participation, subject to their meeting the medical and age criteria for the trial.
"To my knowledge, serious bacterial infections do not manifest differently across race or ethnicity," said Liz Kalina, Cubist's director of product communications. "We are always working to identify patients who meet inclusion and exclusion criteria for our trials. Efficient recruitment is vital to clinical study."
BIG DATA TO THE RESCUE?
In the end, even a whole-hearted openness to minority participation in trials on the part of investigators is unlikely to be enough to repair deeper issues that stymie progress on inclusiveness.
"As interest in population-based medicine delivery grows, funding to study effectiveness of vaccines and other treatments in ethnic minority populations is hard to come by," Danielle Lee, a biologist and science communicator, wrote as part of a July 2014 article for Ebony. Furthermore, she wrote, "the science community has to proactively assuage lingering concerns of vulnerable populations."
But building trust takes time. Damage done in the African American community by episodes such as the Tuskegee syphilis experiment and refreshed in popular culture by films such as the 2005 film adaptation of John le Carré's "The Constant Gardener" sow fear that's difficult to dispel, even in light of sincere efforts to improve outreach. That's the reason postmarketing trials and the advent of electronic medical records may offer more immediate hope for progress.
"I suspect that the opportunities that electronic health records provide are going to give us a lot more information than we've had historically about who patients are and how products are performing," said OMH's Bull, "giving us ways to look at data in ways that help us make the best decision and the strongest and most complete and informative knowledgebase so that we're optimizing for populations where a drug is most effective as well as minimizing use in populations where there may be safety concerns."
But even that seemingly promising angle may run into trouble if, as Pfizer inc. noted in a recent white paper, FDA restrictions prevent manufacturers from freely sharing data, "including meta-analyses that evaluate data from a number of controlled studies, observational studies that evaluate the 'real world' use of a product over time, and subpopulation data from clinical trials, including information about the effects of medicine by race and gender."
Improvements to both the analysis of population-based data and the creation of those data to start, seem likely to arrive slowly. Unlike the mandates and economic pressures reshaping the face of health care spending and delivery in America and abroad, changes to clinical research appear to be driven less by fiscal considerations and more by a desire to do right by people who have, in many cases and for many reasons, gotten the short end of the stick in modern medical care.
Editor's note: BioWorld's Clinical Trials series continues Thursday with a focus on age.