The proportion of racial and ethnic minority patients in NCI-funded clinical trials has nearly doubled over two decades—from 14% in 1999 to 25% in 2019, according to data from NCI’s National Clinical Trials Network and the NCI Community Oncology Research Program.
“Individuals of diverse populations need access, and these two networks do provide access for patients to embark upon a journey into clinical trials or improved outcomes for themselves as well as others,” Worta McCaskill-Stevens, chief of the Community Oncology and Prevention Trials Research Group in NCI’s Division of Cancer Prevention, said in a virtual meeting of the Board of Scientific Advisors and National Cancer Advisory Board June 15. “And finally, diverse enrollment is a major aim of many national research policies.”
In absolute numbers, participation by minority patients rose from 13,784 in the beginning of the studied period to, most recently, 21,125.
In a related study conducted by SWOG Cancer Research Network, a member of NCTN, researchers found that Black patients are better represented in taxpayer-funded clinical trials testing new cancer treatments, compared to trials run by pharmaceutical companies.
SWOG analyzed the data from a total of 358 trials—85 industry trials and 273 SWOG trials—that enrolled 93,825 patients being treated for 15 different cancer types. Enrollments spanned the years 2003-2018. In those 15 cancers, the rate of Black enrollment in industry trials was 3%, compared to 9% in SWOG trials. Blacks account for 12% of the U.S. cancer population.
Noteworthy improvements in representation of minority groups in NCI’s NCTN and NCORP trials from 1999 to 2019 include:
The percentage of African Americans patients enrolled increased from 8% to 11%,
Hispanic or Latino representation have increased from 4% to 10%, and
Accrual of minority patients to phase III trials increased from 14% to 27%.
“We too rarely, I think, see presentations like you’ve just presented, Worta. So, that was fantastic. While we see needs for improvement, we have so much work to do, and we want to engage in that work and support you,” Cheryl Willman, director and CEO of the University of New Mexico Comprehensive Cancer Center, said at the joint meeting.
“I think one of the barriers we see from a minority-serving institution has been the issue of eligibility criteria on so many clinical trials that limit eligibility due to comorbidities, and challenges that we too often unfortunately see in minority and underserved populations,” Willman said. “I know the NCAB has tried to do a lot of work around that, but how deeply we’ve actually impacted changing eligibility criteria, for us, is a huge issue.
“And second is developing trials in areas where we have higher incidences, in certain populations. And I know when you visited us, we look at tribal communities who have some of the highest rates of liver and kidney cancer in the United States, and yet, the trial menus in those cancers are more limited.”
Race is a sociological—as opposed to a biological—construct. It does not reflect the actual genetic admixture of many African Americans, said Otis Brawley, the Bloomberg Distinguished Professor of Oncology and Epidemiology at Johns Hopkins University.
“I think that this data speaks to access and fairness. And the question is: Do Blacks, do Hispanics, do Native Americans have access, and fair access to clinical trials?” Brawley said at the meeting. “And I speak to that because I’ve written about this 27 years ago, even, on the NIH Revitalization Act of 1993, the congressional mandate actually calls for subset analysis amongst the races as if Black is a biological category, different from white, different from Native American.
“I think the politicians got that totally wrong. I’d like to point out that the OMB Directive 15 that defines race in the United States, actually says that they are socioeconomic categories. And if you look at 23AndMe or Ancestry.com, most African Americans in this country have at least 10% Caucasian admixture, that actually makes sense that the category Black, is socioeconomic and not biologic, for example.”
Studies designed to characterize groups according to risk may be more useful for targeting specific populations that are at high risk of certain cancers, Brawley said.
“In order to get the complete answer to the question of some disparities, we need to start thinking about certain very focused studies that look at very particular groups of individuals who do have an increased risk of a particular disease, because of—I will use the category—area of geographic origin, as opposed to race,” Brawley said.
“For example, there are some Native American populations—and Dr. Willman may be able to better define the population than I—who have increased risk of gallbladder cancer; the African, sub-Saharan African, increased risk of prostate cancer. Again, I’m getting away from race and getting close to the area of geographic origin and asking the question, should we do some very defined studies on particular problems that are really significant problems?
“I’ll just end by saying, the work that Dr. McCaskill-Stevens presented is very wonderful in my mind—11% of the American population is Black and 11% of people who are on NCI clinical trials are Black. So, from a national perspective, we’re very representative of the population, but I’m thinking about these very specific targeted questions on specific areas of geographic origin.”
NIH and cancer center directors need to be cognizant of preserving hospital funds for programs that may not be related to cancer—for instance, budgets for clinical trials in general, Brawley said.
“If I can just add one more thing, as a former cancer center director in a poor hospital, there is the tendency or the pressure on the cancer center director, especially in academic hospitals, to divert resources away from programs that can actually help the population and to build a program to put people on clinical trials. And that can actually hurt that hospital,” Brawley said.
“We have to be careful that the NIH in general is not encouraging cancer center directors to take hospital resources that should be going into other things, maybe even other non-cancer things, that would help their community, in order to build a cancer research environment,” Brawley said. “That just means that the federal government needs to pay more for this accrual from hospitals, especially hospitals that take care of a lot of poor folks, you have a lot of Medicaid coming in.”
McCaskill-Stevens’s remarks at the virtual NCAB-BSA meeting follow:
I’m going to present today the accrual of minorities in the NCTN and the NCORP clinical trials in 20 year view.
The scientific evidence gained from clinical trials provides the most important and reliable information for the effective care and management of cancer patients, as well as those individuals at risk of developing cancer. It is for these reasons that the trials must provide evidence that’s both valid and generalizable to all populations. And hence, the frequent interest and frequent questions as to who really participates in NCI clinical trials.
Individuals of diverse populations need access, and these two networks do provide access for patients to embark upon a journey into clinical trials or improved outcomes for themselves as well as others. And finally, diverse enrollment is a major aim of many national research policies.
Today’s discussion will include accrual from 1999 to 2019 as reported using the Office of Management and Budget categories. Minority accrual in newer-generation trials are included. It allows us to take a look at the trials that have been completed, the current trials and how we should look at under-representation for future trials.
We use the Cancer Therapy Evaluation Program Enterprise System for abstracting the accrual. We looked at the cooperative groups from 1999, hence, we have the reorganization into the NCTN, and, similarly, for the Community Clinical Oncology Program and the minority-based CCOPS and its reorganization and transition to the NCORP. We looked at three-year intervals for accrual numbers and percentages and by age intervals for subsets of minorities, specifically African Americans and Hispanics.
Minority enrollment NCTN, NCORP trials
So, we first looked at all phases of trials. As you can see, in the circled number, over 355,000 individuals enrolled in clinical trials over this time period. Of those, 122,000 met the criteria for the racial and ethnic minorities. We started in 1999, somewhat arbitrarily, but it’s six years after the NIH Revitalization Act, which mandated the inclusion of women and minorities in clinical trials. This is important, because institutions will have had time to think about how they were going to collect these data, but more importantly, to allow intervention so the patients could self-report the accuracy of this information.
You can see that the percentage of minority enrollment over time is 19%. You can see that there have been increases and there are actually variations that vary depending upon the availability of trials. Included over these years are the large adjuvant trials in breast and bowel, in particular. There were two large prevention trials for breast and prostate that affect some of the variations that you may see. But we had begun, about this period of time in the mid-2000s, to inform our investigators that we were going to utilize new technology and molecularly characterized trials, that would be smaller, and there would be less of the adjuvant trials.
As an example, TAILORx began in 2006 for it’s accrual. So, you can see that as we get closer to the last two of the three-year intervals, that better reflects the transition to the new generation of trials.
Accrual of African Americans
In this slide, we focused, as a subset, on minorities and African Americans. Over this time period, over 57,000 African Americans were enrolled into clinical trials. You can see that they pretty much mirror the enrollment overall for clinical trials and the total minority accrual is about 9%. But as you look through the years, we’d go up an 8% to the last three-year interval of 11%.
We looked at age. We’re very interested in what was happening in the pediatric population, as well as the increasing interest in adolescents and young adults and, hence, selecting that age interval between the ages of 15 and 39.
We can note that for African Americans, the majority of the patients enrolled were in the age group of 40 to 69. We were also very concerned about the under-representation of minorities and the elderly patients, paying close attention to the chronic comorbidities that we know are significantly seen in these populations.
If we look at the younger populations, we can see that if you combine the younger children and the adolescents and young adults, the percentage is about 19%. African Americans do not have the highest distribution of younger populations within that group.
Accrual of Hispanics, Latinos
This again, all phases, we’ve focused our attention on enrollment of Hispanics, or Latinos. You can see that beginning from 1999, as we traverse over the years to the last three intervals, we’ve pretty much doubled the enrollment, such that over the period of time, over 43,000 individuals were enrolled into clinical trials.
Of note, when you look at the younger population, we can see that the Hispanic population, first, is the youngest of the racial ethnic populations throughout the country with about a third or 18% between the ages of one to 18, and another third or 15 million between 18 and 33. So, in contrast to what we saw in the African Americans, that the percentage, if you combine the younger children with the adolescents and young adults, you have a 42% representation as to 19% seen in African Americans. So, this population is much younger.
If one were to ask the question, “Well, what’s the percentage of cancers in this group?”, it’s pretty much driven by leukemia. And the Hispanic population amongst the young has the highest percent of leukemia among the younger population. So, we know that we looked at 2018 and it was reported and estimated that about 2,700 patients would be diagnosed with cancer in the zero to 14 group. So, depending upon the availability of leukemia trials, we can say that we’re closely matching that age range.
We then looked at phase III trials, which represents about 87% of the total phases. We can see that over 500,000 patients were enrolled. And I wanted to direct your attention to the minority enrollment, because if you just focus on phase III trials, actually, in the last three years of enrollment, it was up to 27%. Not shown here are the phase II trials—we did not include them because we did not observe significant differences in the trends of enrollment of racial and ethnic minorities.
Accrual of other minority groups
We then looked at the other four categories in the OMB criterion, and the overall Asian enrollment is about 3%. So, it’s a little less than what we’re seeing throughout the U.S. Certainly, for American Indians, as well as in the Alaska natives, as well as the native Hawaiians and other specific islanders, there are less than 1%. You can see what has happened over the years, we haven’t seen significant trends, but I know that there are significant efforts to engage these populations by various initiatives.
We then asked the question, what do we see in terms of enrollment from the community setting and the academic setting? So, what you see here, and we looked, we just started at the inception of NCORP in 2014 and looked at the similar time point for the NCTN groups. You see the total of 23,000 enrollments for the NCORP and 90,000 from the NCTN—the overall contribution being 26% on the right hand side, and the minority enrollment contribution 27%. Please note that the NCORP program has a specific component dedicated to minority underserved populations, to which the enrollment for racial and ethnic minorities is 55%. Next slide.
Additionally in 2014, we included the science of cancer care delivery research. The goal of this was to improve clinical outcomes and patient well-being by intervening on patients, clinicians and organizations, factors that influenced the delivery of care, which gave us an important insight as to other areas it might influence the enrollment of underserved populations, specifically in the organizations. You can see that over this period of time, 6,900 patients were enrolled; 26% of them were racial and ethnic minorities. And over 60% of that came from the minority underserved NCORPs.
Enrollment in immuno-oncology trials
This was a picture taken from an article published in the New York Times in 2016. About the time when there was tremendous beginning of excitement about immunotherapy. And essentially, the article states that the participants were overwhelmingly white. And that researchers were looking at ways to enhance minorities enrolled into those trials.
We took a snapshot of two slides that had enough targeted therapy that we thought we could present to you. The first trial is one that utilizes the immunotherapy nivolumab in the perioperative setting, where patients who were going to undergo nephrectomy for renal carcinoma. What you see here is the overall minority accrual is 20%.
We thought of this in the context of the participation in the clinical trials that give the drug approval. For each of the categories that are shown here, at this particular time, the minority representation for all groups exceeds that of the representation and at the time of FDA approval, and with the exception of Asian populations because the Asian population actually have the highest number among racial minorities in the FDA approval clinical trials.
The second trial that we wanted to bring to your attention was a trial that was using pembrolizumab in the adjuvant setting with triple-negative breast cancer, in which patients will be randomized to the drug if they had residual disease after receiving neoadjuvant therapy. And importantly, here, you see the minority enrollment is 28%.
And if one looks at the populations in which we have the greatest incidence of triple-negative breast cancer, Blacks and Hispanics, we can see that one, they exceeded the enrollment into the FDA approval trials. But in addition, this is about equal to the African American population within the U.S. And importantly, it is consistent with the incidence of triple-negative breast cancer in the African American population, similarly for Hispanics.
Moving forward
The impetus for having these data presented to you was the question regarding disaggregation within race and ethnicity, which is a very important one. Currently, we have begun to [collect] NCORP information on African Americans as to whether they are African American, whether they come from the Caribbean or whether they’re from the sub-Sahara in Africa. We are going to begin to look at this in the Hispanic population, which is incredibly important as we see increased numbers throughout the country and estimated increases over time.
We have focused our attention on language, but we are also very interested in the country of origin, not only for the nuances of language and culture, but we want to have ways that inform strategies for enrolling. For example, we need to know whether that individual comes from a country that has had a registry, has had access to clinical trials or the existence of clinical trials, so that we can better understand ways to overcome barriers from those areas.
We have implemented a clinical child log for pediatrics, as well as adults, with a goal to expand the demographic collection of information, including education, socioeconomic factor, barriers, comorbidities that will help us at the NCI, at the site level, but also importantly in trial design as we think about that information.
We had a consensus when we began to enroll patients into the tissue acquisition studies, and this is important for minority populations, because most of these studies are focusing on advanced disease and then providing information to the treatment trials.
We have not, in the past, provided accruals for these. We are currently doing that at this time, moving forward, which will allow us to monitor the participation of these populations that are very, very important as we see that this is the direction which we’re going to be going with our clinical research.
Quality of life studies expertise is within the NCORP, as well as the review. We have a number of quality of life studies that are embedded in treatment trials. And the quality of life of the populations for whom there’s significant obesity, and other chronic diseases, and how they interact with the new drugs that are coming on board, is very important. And so accrual also is being monitored and given to the sites for participation.
And for those important tools that are included in the quality of life accruals, we have found that many of the tools are not translated, in populations, to languages that will be inclusive. We have Moonshot funding, to look and validate those tools for various languages, but that’s led by Diane St Germain in the NCORP group.
And finally, we have the expertise and interest across the research bases and the NCTN groups, to look at other groups for under-representation, including the elderly, the adolescents and young adults, sexual and gender minorities, as well as rural populations, which was identified as an actual population eligible to participate in NCORP because of rurality.
