Coverage from the 64th Annual American Society of Hematology Meeting and Exposition, December 10-13, 2022, New Orleans, Louisiana.
Matthew J. Maurer, DSc, statistician at Mayo Clinic in Rochester, Minnesota, and director of the statistics and informatics core of the Lymphoma Epidemiology of Outcomes (LEO) cohort, discusses findings and implications of research presented at the 64th American Society of Hematology Annual Meeting and Exposition. Investigators found that up to a quarter of patients are excluded from frontline clinical trials of diffuse large B-cell lymphoma (DLBCL) based on 5 organ function lab values.1 In an interview with Evidence-Based Oncology™ (EBO), Maurer, who presented results on behalf of lead study author Arushi Khurana, MBBS, of the Division of Hematology at Mayo Clinic, said the team’s work suggests that patients from minority populations are being excluded from clinical trials for reasons that have nothing to do with the mechanism of therapies.
EBO: What 5 lab values were included in the LEO cohort and what differences did you find based on race or ethnicity?
Maurer: The 5 lab values that we looked at related to organ function were hemoglobin, neutrophil count, creatinine clearance, bilirubin, and platelets. We did see very significant differences in some of these lab values by race and ethnicity. For example, Black patients had significantly lower hemoglobin than White patients in the LEO cohort. In addition, Black patients also had significantly higher creatinine levels compared to White and other minority patients in the study. Depending on what criteria that were applied in terms of the cutoffs on the trials, these criteria may have excluded patients differently based on race and ethnicity.1
EBO: What other differences by race and ethnicity did you find in patients with DLBCL?
Maurer: Strikingly, Black patients were much younger than White patients with DLBCL who enrolled in the LEO cohort. Median age for Black patients with DLBCL in our cohort was 51 years compared to 65 in White patients. So despite the fact that Black patients were significantly younger [and] had better kidney function than White patients on trials, Black patients were excluded from trials based on these organ function labs at a much higher rate.
EBO: What prior research helped build the LEO cohort? Is there something suggesting why there was a trend for Black patients who received lymphoma diagnoses at younger ages than the overall patient population?
Maurer: These results are largely reflective of a previous study by Dr Chris Flowers [of The University of Texas MD Anderson Cancer Center], 1 of the LEO co-principal investigators.2,3 He and colleagues looked at this previously in data from the SEER [Surveillance, Epidemiology, and End Results) registry and identified that Black patients with DLBCL tend to be younger than White patients at the time of presentation and also that Black patients tend to be more symptomatic. We saw that on the LEO cohort as well. We also saw some differences in treatment patterns in the LEO cohort, with White patients much more likely to be to go on a frontline trial than Black, Hispanic, or other non-White, non-Hispanic patients.
EBO: How did you enroll patients in the LEO cohort?
Maurer: Participants in the LEO cohort are all patients with newly diagnosed lymphoma, including DLBCL, who are within 6 months of diagnosis. We enrolled patients from 8 cancer centers across the United States. This is an observational cohort, so patients consent to the research study and then we follow them prospectively and collect details on the treatments, their outcomes, and survivorship.
EBO: Why do the lab criteria exist?
Maurer: These criteria exist so that we ensure that patients are healthy enough to enroll on a trial and that they’re healthy enough to receive full-dose standard of care. We don’t want to do harm in the patients that we put on trials, and so we want to ensure that these patients don’t have other comorbidities or characteristics that would potentially make it worse if they got additional agents or additional experimental therapies beyond standard of care. But what we know is that with some of these organ function labs, [values] can be the actual disease; it’s the lymphoma that’s causing some of these values to be outside the ranges that are expected on clinical trials.
EBO: Were some trials you looked at more restrictive than others?
Maurer: It’s interesting that these clinical trials can have a fair amount of variation in how these criteria are set, so that’s why we looked at a number of recent phase 3 clinical trials in frontline DLBCL. For example, for hemoglobin, some studies used a cutoff of 9 grams per deciliter, some used 10, some didn’t have a hemoglobin cutoff at all—it was not part of the exclusion criteria. So I think we need to be really thoughtful about where these values come from and how we build the eligibility criteria on our trials.
As we showed in our study here within the LEO cohort, these choices have a huge impact on the patients that can go on the trials. And we’re seeing quite a bit of inequity in how these are set in terms of the differential impact [they have] across racial and ethnic groups. To build equitable trials moving forward, we need to be thoughtful about how we’re designing studies and what these criteria are.
EBO: What associations between trial eligibility and event-free and overall survival did you find?
Maurer: We showed in our previous analysis that patients who are excluded based on these 5 organ function labs have significantly worse event-free survival. And some of that we might expect because these patients might be sicker; they might not be able to get full-dose therapy. But what we identified also is that these patients are at a much higher risk of dying from progressive lymphoma. And they were not dying at a higher rate for some treatment-related mortality or other causes. So when we looked at this in the LEO cohort, we confirmed these results. In particular, [this occurred] among Black patients who are ineligible; despite being younger, they have the highest rates of lymphoma-related mortality of the subsets that we looked at.
EBO: How did these findings from the LEO cohort analysis either support or build upon those from your epidemiology resource?
Maurer: These results build strongly off results [from a prior] cohort that included patients who were enrolled at the University of Iowa and Mayo Clinic—thus, in the Upper Midwest—with limited racial and ethnic diversity. The LEO cohort is a larger cohort; the racial and ethnic makeup of patients with DLBCL is largely representative of what we see in the United States, based on SEER estimates. And it’s also a larger study. What we see when we look at a larger, more diverse, more recent data set is that these results are confirmed. So not only did we confirm these findings in the larger, more diverse LEO cohort, we also have the sample size and the representative cohort to start to examine how these eligibility criteria impact potentially patients differently by racial and ethnic group.
EBO: Building off changes in eligibility criteria, is there anything else you hope to see from future clinical trials that takes these findings into account?
Maurer: Dr Khurana has another poster at ASH which examines the dose intensity of patients who are ineligible, [which aims] to understand what are the clinical characteristics of patients who might will be able to receive full-dose therapy vs patients who, perhaps due to comorbidities or other issues, cannot receive full-dose therapy.4 We’re trying to better understand which patients would be good candidates for clinical trials based on modifying the criteria that we’re using. We’re [accustomed] to seeing someone in clinic and the clinic saying, “Boy, this patient be a great candidate for the study, [but] their hemoglobin was low and they didn’t meet the eligibility criteria. But otherwise, they would have been a great candidate.”
So how do we change the way we think about these clinical trial designs? And how do we change the way we think about eligibility for studies to be more inclusive so that we’re capturing patients who want to participate in research? When should these types of patients be enrolling on our trials?
1. Khurana A, Mwangi R, Nastoupil LJ, et al. Evaluating the impact of lab-based eligibility criteria by race/ethnicity in frontline clinical trials for diffuse large B-cell lymphoma (DLBCL): a LEO cohort analysis. Presented at: 64th American Society of Hematology Annual Meeting and Exposition; December 10-13, 2022; New Orleans, LA. Abstract 850. https://ash.confex.com/ash/2022/webprogram/Paper169433.html
2. Chen Q, Ayer T, Nastoupil LJ, et al. Population-specific prognostic models are needed to stratify outcomes for African-Americans with diffuse large B-cell lymphoma. Leuk Lymphoma. 2016;57(4):842-851. doi:10.3109/10428194.2015.1083098
3. Blum KA, Keller FG, Castellino S, et al. Incidence and outcomes of lymphoid malignancies in adolescent and young adults (AYA) patients in the United States. Br J Haematol. 2018;183(3):385-399. doi:10.1111/bjh.15532
4. Khurana A, Mwangi R, King RL, et al. Dose intensity and reasons for dose alterations in patients excluded from frontline diffuse large B-cell lymphoma clinical trials based on eligibility criteria: a Mayo Clinic cohort study. Presented at: 64th American Society of Hematology Annual Meeting and Exposition; December 10-13, 2022; New Orleans, LA. Abstract 2966. https://ash.confex.com/ash/2022/webprogram/Paper169575.html
After years of talk and some early steps, the pandemic forced clinical trials to move beyond the walls of academic centers—and the FDA wants to maintain that momentum “to bring the research enterprise to the patient,” said Nicole Gormley, MD, acting director of the Division of Hematology Products at the FDA.
Gormley opened a symposium on decentralized trials held December 9, 2022, ahead of the 64th American Society of Hematology (ASH) Annual Meeting and Exposition in New Orleans, Louisiana. The hope, said Gormley and speakers from across academia and industry, is that making the process easier on patients will attract more of them—and more from minority groups who have been underrepresented.
“It’s hoped that these efforts will translate into greater diversity of the trial participants as the hurdles to trial participation are minimized,” Gormley said.
There’s some belief that decentralized trials may be less expensive to run because patient recruitment will be less time consuming. But some speakers in the session warned that setting up a decentralized trial requires its own infrastructure; the focus should be on getting the right patient population.
What’s emerging, Gormley said, is a “hybrid approach” in which some parts of the trial occur remotely and other elements take place at a central location. Technology can allow remote collection of online consent and patient-reported outcomes and some trials may also use local laboratories or clinic visits.
“It’s important to keep in mind that decentralized clinical trials don’t have to be an all-or-nothing endeavor,” Gormley said. “What’s much more common is a hybrid approach, where some aspects are done at the central location and other aspects are conducted remotely.”
Although interest in decentralized clinical trials soared during the pandemic, it’s not a new idea. Gormley highlighted examples dating to 2003 and noted that the FDA issued guidance in March 2020 because of the public emergency based on existing practices. It was updated in August 2021.1
“The overarching message of this guidance was that it’s important to ensure the safety of trial participants and sponsors were encouraged to consider alternative methods for safety assessments such as phone contact or virtual visits,” she said. “Although this guidance was developed in response to the COVID-19 pandemic, many if not all of the decentralized trial components were already in existence.”
There are some considerations, however. Gormley said it’s essential to ensure that data collected are fit for use, that additional steps are taken to ensure patient privacy, and that care is taken by trial leaders to ensure they have full access to patient records if some parts of the study are conducted remotely.
Also, Gormley said trial leaders and sponsors must recognize that not every patient has the same access to technology, so proper technological supports must be in place.
Academic perspective. Following Gormley’s talk, Kami J. Maddocks, MD, of Ohio State University Hospital, and Michael R. DeBaun, MD, MPH, of Vanderbilt University School of Medicine, led a session on the benefits of decentralized trials in hematology. Grzegorz S. Nowakowski, MD, the enterprise deputy director for clinical research at Mayo Clinic and chair of the ASH subcommittee on clinical trials, said it’s important to understand why academia supports this approach.
Nowakowski said these trials could be less expensive and lead to faster drug development—but he said the real benefit is getting more patients to participate in trials. “For me, it really comes down to the basic principle that the best therapy for the patients is typically participation in a clinical trial,” he said. “To be able to offer our patients best care everywhere—from academic sites to small hematology practices—we really have to move with this idea of decentralization and be able to deliver this locally.”
At present, only 5% to 7% of patients enroll in trials, he said. “So if you flip it around, over 90% of patients are not receiving the best therapy.”
The current model puts up too many barriers—from travel to regulatory requirements. Nowakowski offered the example of mailing samples from “routine clinical labs” that could be processed locally. Many steps currently in place were set up to avoid as much risk as possible, and that’s laudable.
“We want to be perfect or as close to professional as we can,” he said. “But we cannot allow this risk aversion to stop us and basically slow down our clinical infrastructure to the point where patients literally can’t get to the trials.”
After Mayo Clinic researched how far patients were driving to trials, it set up a partnership with the Leukemia & Lymphoma Society to extend trials into the community with the hope of reaching rural patients using the hybrid model that Gormley discussed. Mayo Clinic plans to collect data on patient satisfaction with this system, Nowakowski said.
Industry perspective. Lilli Petruzzelli, MD, PhD, senior vice president of early clinical development, Genentech, said the pharmaceutical industry shares academia’s interest in boosting patient participation in trials.
“A major issue for us is to increase diversity in our trials for better health equity, and to ensure that we’re learning as much as we can about the right drug in the right population, she said. If only 5% to 7% of the patients are taking part in trials, and 75% of trials are industry sponsored, “really, the onus is on us to play a huge role in trying to make this happen.”
“Numbers mean a lot,” she said. “A third [of clinical trials] don’t even list the data on race.”
The idea that moving trials outside academic centers will harm data validity is a myth that springs from a few anecdotes, Petruzzelli said. By contrast, there’s opportunity for all involved “if we can improve health equity, if we can get the right drugs to patients, if we can really make this happen.”
In her role in early drug development, the need is clear. Right now, Petruzzelli said, “there’s an incomplete understanding of drug behavior across diverse populations early—and that really hampers us in advancing drugs as we want to go to a larger study.” This cycle happens repeatedly, and beyond decentralized trials, she flagged exclusion criteria that unnecessarily keep out too many elderly and minority patients. “One of the biggest challenges we have—and we’ve been stagnant on—is protocol, eligibility, and inclusion criteria….It excludes a lot of patients.”
An abstract on this issue—for which Nowakowski was senior author—was highlighted in an ASH news briefing on December 10, 2022. The study examined a cohort of patients with diffuse large B-cell lymphoma to show that lab values, some of which have nothing to do with the trials or the mechanisms of the drugs, would have caused 9% to 26% of patients to be excluded from 3 recent trials—including POLARIX—with exclusions falling more heavily on minority populations2 (SP62).
Clinical trialist perspective. Caterina Minniti, MD, professor of clinical medicine and pediatrics at Albert Einstein College of Medicine, spoke from her experience with a decentralized clinical trial involving patients with sickle cell disease (SCD). Minniti described SCD as the “canary in the coal mine of heath care” because one can judge the rest of the system by how well these patients are treated. “Even though in the United States it is considered a rare disease, because it affects about 100,000 patients, most of these patients are African or African American and 10% are Hispanic. And often, my patients reside in socioeconomically distressed communities.”
When the pandemic hit, even before telehealth could be reimbursed, she said, “we just had to do it.” And a remarkable thing happened: The number of patients who were no-shows to appointments plummeted, and so did the number of visits to the emergency department. That suggests that getting to the appointment is the challenge.
Lack of trust is a widely acknowledged problem in enrolling minority patients, but Minniti said lack of awareness is a problem, too. “There is this a big hurdle that is peculiar to the sickle cell community, which is the majority of adults with sickle cell disease do not even see a hematologist or even a caretaker that is familiar with sickle cell disease,” she said. Thus, the patient’s primary doctor may not know about clinical trials that are available.
“There is no research without care. If we don’t have patients engaged in care, we cannot enroll them in a clinical trial, we cannot run a clinical trial,” she said, citing CDC data from the state of Georgia that show that only 21% of adults with SCD have seen a provider in the past 3 years with expertise in the disease.3
Minniti highlighted a decentralized, web-based clinical trial sponsored by Pfizer to measure pain, comparing disease-modifying therapies and patients without therapies. Patient recruitment materials used QR codes that patients could highlight, go to the site, see the eligibility criteria, and take the first step to register. Once they completed these steps, a researcher would call the patients back to confirm eligibility. Online processes are also using Instagram and other tools to recruit patients, she said.
As useful as these tools are, Minniti said, there are pros and cons. “I’ve learned that it takes a lot of time to run a decentralized clinical trial,” largely because of the challenge of gaining access to medical records. Some patients don’t even have a primary care physician, so that must be addressed. This may mean the trial must be designed differently, Minniti said.
“I reiterate that even though it’s a decentralized clinical trial, it does not mean it’s a simple way of doing a clinical trial,” she said.
1. Advancing oncology clinical trials. FDA. Updated July 27, 2022. Accessed December 11, 2022. https://www.fda.gov/about-fda/oncology-center-excellence/advancing-oncology-decentralized-trials
2. Khurana A, Mwangi R, Mastoupil LJ, et al. Evaluating the impact of lab-based eligibility criteria by race/ethnicity in frontline clinical trials for diffuse large B-cell lymphoma (DLBCL): a LEO cohort analysis. Presented at: 64th American Society of Hematology Annual Meeting and Exposition; December 10-13, 2022; New Orleans, LA. Abstract 850. https://ash.confex.com/ash/2022/webprogram/Paper169433.html
3. Snyder AB, Lakshmanan S, Hulihan MM, et al Surveillance for sickle cell disease—Sickle Cell Data Collection program, two states, 2004-2018. MMWR Surveill Summ. 2022;71(9):1-18. doi:10.15585/mmwr.ss7109a1
Mary Cushman, MD, professor of medicine at the University of Vermont, expands on the findings of a study on social determinants of health and disparities in hospitalization, treatment, and mortality in patients with pulmonary embolism (PE).1 These findings were presented at the 64th American Society of Hematology Meeting and Exposition, held December 10-13, 2022, in New Orleans, Louisiana. This interview is edited lightly for clarity and length.
EBO: What were some of the main findings of your study on social determinants of health and pulmonary embolism (PE) treatment and mortality?
Cushman: We studied data from the Nationwide Inpatient Sample, which is a random sample of 20% of all hospitalizations in the United States, between 2016 and 2018. We ascertained all pulmonary embolism hospitalizations, whether a person was admitted for pulmonary embolism or PE, or had PE during the hospitalization, and analyzed patterns of care and outcomes of people based on social determinants of health, specifically, race and ethnicity, insurance type, and income level. We found that there were differences in the presentation of PE, the treatments that were administered, and the fatality rates.
Amongst all people admitted for PE, those admitted with what we call high-risk PE or PE that’s more serious were more likely to be Black individuals and Asian/Pacific Islanders, relative to White people, with no differences between other racial groups and White people. We also showed that the actual overall hospitalization rate differed dramatically by racial group. The rate per 10,000 of the [overall] population for admission for PE was about 15. Amongst Black people, it was 20 per 10,000. Amongst Asian and Pacific Islander people, it was 3 per 10,000. White people had a rate of 13 per 10,000.
We found that about 5.5% of all patients received what we call advanced therapies or more aggressive therapies for their PE, and amongst those with high-risk PE, that percentage was 19%. And there were differences by social determinants in the administration of these advanced therapies. Black people and Asian Pacific Islanders were less likely to receive advanced therapies and people who had their primary health insurance as Medicare or Medicaid were also less likely to receive these more aggressive treatments compared to people with private insurance. We didn’t see any difference by income in the use of those treatments.
For in-hospital mortality amongst all patients, Asian, Hispanic, and people of other racial groups had a greater in-hospital mortality. But if you look just at the people with the most serious PE, which we call high-risk PE, the mortality rates in all the non-White groups were greater than [those in] White people. They ranged between 10% and 50% greater in terms of the mortality in those groups. That’s really important because the mortality rate in that group with high-risk PE was 50%, so 1 in 2 people with that type of PE actually died in their hospitalization.
In sum, non-White people were more likely to experience high-risk PE, the advanced therapies were less often used in Black people and Asian and Pacific Islanders, and Black, Asian, and Hispanic people were more likely to present with high-risk PE and die in the hospital than White people.
EBO: Did your study indicate anything about the relationship between social determinants with treatment and course of PE based on race?
Cushman: We know, historically and persisting for a long time, that Black people have twice the death rates from pulmonary embolism compared to other groups in our country, with other groups all kind of similar to each other. We also know that Black people have a greater incidence of pulmonary embolism. Our findings extend that by showing that they’re less likely to receive more aggressive treatments for their PE, and they’re more likely to die in hospital once they have PE. So the question is, why is this?
Our findings were independent of the other social determinants that we had data for, like neighborhood income level, insurance status, what type of hospital they were at—whether it was a rural or urban hospital, or larger or smaller hospital, or a teaching hospital or not. But we didn’t have a lot of granular data to help further explain it. For example, we don’t know the education level of people. These are administrative data and that’s not data that’s recorded in hospital records, so we didn’t know education level, for example, or quality of education. There are so many more questions that need to be addressed to understand these patterns that we’re seeing. All we’re seeing are patterns. We can’t say why they’re there. We can say, yes, they seem to be independent of these other factors that we adjusted for our analysis, but that’s not a perfect way to try to address those other factors as causes.
Obviously, it’s not the color of a person’s skin or anything about their ancestry that is causing these disparities. My hypothesis is that structural racism is at the root of these differences that we’re seeing. And structural racism is imposed by historical policies of the past that segregated people, for example, which led Black people to have lesser quality education and to live in places where they didn’t have advantages compared to other groups, and leads to interpersonal racism today.
So those things really need to be studied to see if they explain these differences. For example, if a Black person might be less well educated or less in touch with resources, they might not know when they’re having symptoms that are serious. If they’re having shortness of breath or chest pain, symptoms of PE, they may not know to get checked. They may not trust the system; there’s a lot of reasons for this. If they don’t trust the system, they may not seek care as quickly as someone who trusts the system. If we as health care providers aren’t acknowledging that and watching for it to make sure our treatments are being given equitably, then that’s on us.
Farmakis I, Cushman M, Valerio L, et al. Social determinants of health and pulmonary embolism treatment and mortality: the Nationwide Inpatient Sample. Presented at: 64th American Society of Hematology Annual Meeting and Exposition; December 10-13, 2022; New Orleans, LA. Abstract 140. https://ash.confex.com/ash/2022/webprogram/Paper166550.html