The importance of introducing diversity earlier in drug development
DDW’s Reece Armstrong speaks to Shantanu Dhamija, VP of Strategy and Innovation at Molecular Devices about challenges of diversity in clinical trials and why induced pluripotent stem cells might be key to representing wider patient groups.
RA: Why is US clinical trial diversity at its lowest point in the last 10 years?
SD: The level of diversity achieved in clinical trials is the result of competing forces. On one hand, there are relatively deep-rooted fears about participating in clinical trials due to atrocities like the Tuskegee study, in which researchers studied the effects of untreated syphilis in African American men without informed consent. This experience has discouraged minority participation in clinical trials.
On the other hand, there are policy and legislative initiatives, as well as advocacy programs, developed to increase diversity in clinical trials. Examples of these are decentralised clinical trial (DCT) approaches that reduce the need for regular clinical trial site visits and advocacy through partnership with Native American tribes and historically Black colleges and universities.
In our current environment of political polarization, efforts to encourage diversity are unable to prevail, and more needs to be done to foster the trust that is necessary for broader participation in clinical trial recruitment.
RA: How does a lack of patient diversity in trials impact how treatments work in certain populations?
SD: Our current clinical development paradigm is rooted in the belief that the efficacy and safety of experimental drugs need to be assessed in the context of a population because individuals can experience the same disease differently.
According to the Food and Drug Administration (FDA), 75% of clinical trial participants for approved molecular entities and therapeutic biologics in 2020 were white, while only 11% were Hispanic, 8% were Black or African American and 6% were Asian1. Meanwhile, according to the 2020 U.S. Census, the overall population is 61.6% white, 18.7% Hispanic, 12.4% Black or African American and 6% Asian.
This mismatch between the population the drug is tested on, and the population it is intended to benefit raises questions about the broad applicability of clinical trial findings, especially as it pertains to off-target effects and drug safety assessment. This undermines public confidence in our healthcare institutions, which in turn can adversely impact the healthcare decisions of people.
RA: Do you think recent draft guidance by the FDA will help improve diversity across trials?
SD: The Food and Drug Omnibus Reform Act (FDORA) signed into law by President Biden in Dec 2022 strengthens the FDA’s 2022 draft guidance on the topic, making action plans for achieving a more diverse clinical trial cohort a requirement—versus a nice-to-have —for all Phase III studies for drugs and other biological products. Furthermore, FDORA establishes a timeline for the implementation of diversity initiatives, for example, requiring the FDA to provide draft guidance on diversity action plans by December 2023 and publish final guidance within nine months of the close of the comments period. I expect this to have a positive impact, but it remains to be seen whether this countermeasure is sufficient.
RA: Whilst recruitment methods are important to increasing diversity, what else can researchers do to ensure that therapies are more representative of patients?
SD: The FDORA also includes provisions that encourage novel clinical trial designs, including retaining certain flexibility measures initiated during the Covid-19 public health emergency. As a part of this, the FDA has until December 2023 to issue or update draft guidance on decentralised clinical trials, which may address the hurdles of geographic access that have prevented the adequate representation of minorities in clinical trials.
In addition to initiatives that specifically focus on improving diverse representation in clinical trials, I believe there is an opportunity to leverage more physiologically relevant in-vitro models such as organoids to incorporate genetic diversity earlier in the drug discovery process.
Researchers such as Dr. Pradipta Ghosh at the HUMANOID Center of Research Excellence at the University of California, San Diego (UCSD) are pioneering a new approach to human clinical trials by introducing Phase ‘0’ studies in which organoids are used to reject ineffective compounds early in the process.
A similar approach can be used, with induced pluripotent stem cells (iPSCs) from an appropriately diverse population and iPSC-derived organoids, for the early identification and mitigation of drug adverse events such as idiosyncratic Drug Induced Liver Injury (DILI).
RA: How can teams use induced pluripotent stem cells (iPSCs) to represent patient groups and improve diversity in the early stages of drug development?
SD: This depends on the question the team wants to answer. One set of questions is related to efficacy and the preclinical identification of the target population most likely to benefit from the investigational drug. Organoids are being used to answer such questions currently, as Merus Pharmaceuticals did (using adult stem-cell derived organoids) to identify the set of KRAS mutations that were inhibited by MCLA-158. The other set of questions is related to the safety of the drugs and off-target effects. iPSCs, which can be obtained by reprogramming skin fibroblasts or peripheral blood mononuclear cells (PBMCs) are self-renewing and can differentiate into all different type of cells (eg. cardiomyocyes, hepatocytes) and organoids (eg. cardiac, liver, gut, brain organoids). Since clinical and genetic assessment can be made in advance of reprogramming, researchers can develop a cell bank (eg. of controls) that represents all patient groups and improved diversity in early stages of drug development.
RA: Are there any challenges or barriers in testing diverse cell mixes?
SD: There are numerous challenges, but none are unsurmountable. Access to iPSCs that represent the appropriate genetic diversity is the first challenge, and one that needs to be navigated very carefully. It is worth mentioning that one of the most widely used cell lines, the HeLa line2, was acquired without consent—as was standard at the time. Though the lack of appropriate consent is less of an issue, the handling of patient data and cellular material needs close monitoring.
The second challenge is related to the relatively low scalability and reproducibility of organoid models and reliance on the experience and expertise of the scientists doing the work. This is a critical hurdle to overcome because consistent results using organoids from a single cell line are foundational in establishing the credibility of a solution that is based on screening drugs against multiple cell lines. Organoid start-ups like Herophilus have shown that automation and machine learning can be leveraged to produce organoids for screening from multiple cell lines at scale.
The third challenge is the lack of complete solutions to perform this kind of research leaving researchers to assemble the different pieces—cell lines, media, automation, data analysis, etc.—of the solution. Addressing this challenge will require collaboration amongst solution providers with complementary offerings, and with academic institutions that have developed novel organoid models.
RA: How exactly can iPSCs help researchers understand how therapies will impact patient populations differently?
SD: Idiosyncratic DILI is a prime example of how iPSCs, and iPSC-derived organoids enable researchers to assess the impact of therapies on patient populations. Idiosyncratic DILI is a rare but potentially serious liver disorder and a major cause of significant liver injury. Causal medications varied by race3, with trimethoprim/sulfamethoxazole being the most common cause among African Americans and amoxicillin/clavulanate being the most common cause in Caucasians.
Idiosyncratic DILI is rare enough that it is impractical to use race as the only criteria for creating an iPSC bank to understand the impact on different populations. However, Takanori Takebe and his colleagues have demonstrated that it is possible to use a patient’s genetic sequence to predict their risk of DILI and confirmed the accuracy of their predictions by treating liver organoids developed using iPSCs obtained from the patient with drugs known to cause DILI. This work is a practical application of the use of iPSCs and iPSC-derived organoids to assess the risk of investigational drugs on different patient populations.
RA: Can testing at an early stage bring about benefits later on in drug development?
SD: Beyond the obvious benefits associated with the early detection of risks associated with a drug development program, in-vitro screening strategies that incorporate population differences may encourage greater diversity in clinical trials. For example, an Asian male might have concerns about whether or not a drug is safe for him. But if someone can say: “We’ve tested this on a human-relevant cell models and results show no-to-low-risk for Asian people,” this individual may be encouraged to enroll in a clinical trial.
This work can also help expand eligibility. Often, for example, clinical researchers do not want to enrol pregnant women because they are unclear about developmental toxicity. Now, because iPSCs basically mimic developmental cues, you can start to assess developmental toxicity in iPSC-driven organoid models.