Background: The FDA’s recent work on biomarker development and qualification

• A biomarker, according to the FDA, is “a defined characteristic that is measured as an indicator of normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions.” Some examples of biomarkers include blood values of a given chemical or protein, imaging values such as tumor growth, or pathology results returned from a biopsy – but biomarkers span a wide range of types of measurement.
• There are several different analytical validation, clinical validation and qualification evaluations required to establish a biomarker. The assessment of a biomarker requires three components: materials for measurement (for example, equipment), an assay, and a method to interpret the measurement. Additionally, the evidentiary requirements depend on the type of biomarkers and context of use (COU).
• Currently, biomarkers may be validated during the drug approval process in the investigational new drug application phase, through FDA’s biomarker qualification process under the 21^st Century Cures Act, or through scientific community consensus.
• The FDA maintains a Biomarkers Qualification Program, by which “biomarkers shown to be useful indicators across different development programs may be designated by the FDA as qualified biomarkers.” In conjunction with the program, the agency also has a searchable database of drug development tools, searchable by keyword, program, stage in the qualification process, FDA’s determination on the tool, and other details such as the disease, therapeutic area, and organization associated with the drug development tool.
• Though much of the FDA’s biomarker information and resources are nominally housed within the Center for Drug Evaluation and Research (CDER), some resources (such as the drug development tool database) have shared oversight with the Center for Biologics Evaluation and Research (CBER), and many resources are also useful for CBER-regulated products. A landing page that outlines biomarker and qualification basics has many useful links for FDA resources.
• Together with the National Institutes of Health, the FDA maintains a glossary of resources called BEST (Biomarkers, EndpointS, and other Tools), which “aims to capture distinctions between biomarkers and clinical assessments and to describe their distinct roles in biomedical research, clinical practice, medical product development, and in the regulation of products by FDA.” The BEST resource, a living document, is an important tool in postmarket surveillance activities, but is also a key resource during the development of therapeutics.
• Work on biomarker development is occurring in various parts of the FDA, including in the CDER’s Office of Orphan Products Development (OOPD), which administers grant programs to “stimulate the development of promising products for rare diseases and conditions” and oversees the orphan drug designation program, according to the FDA’s 2022 Regulatory Science report. Some of OOPD’s work includes investigating and developing novel biomarkers, to help further review and regulation of applications for therapies to treat rare diseases. [ Read AgencyIQ’s analysis of the FDA’s 2022 Regulatory Science Report here.]
• Within CBER, work is underway to qualify drug development tools for cell and gene therapies. Drug development tools, which can include biomarkers, are used to evaluate or create clinical endpoints, demonstrating that a method, material or measure is meaningful and can be relied upon during the drug development process. CBER’s Office of Tissues and Advanced Therapies (OTAT) recently put out a contract opportunity to assist with developing a qualification process for drug development tools for gene therapy. OTAT said it envisions a contractor helping to develop “a stem cell-based method for detecting unintended genetic modifications following genome editing procedures, as well as in vitro and in vivo assays for understanding biological consequences of defined genetic alterations,” according to the contract announcement. [ Read AgencyIQ’s analysis of the RFQ here.]
• FDA researchers have also been partnering with academia and industry through efforts coordinated by the Friends of Cancer Research (FOCR) to understand how a blood test that measures fragments of cancer DNA can be used in oncology research, and for clinical and regulatory purposes. Using this circulating tumor DNA (ctDNA) as a biomarker that is a validated early endpoint in clinical trials is a goal of the FOCR work. Use of ctDNA could also allow both more frequent and less burdensome assessments, as compared with direct sampling of tumor tissue for many cancers, and could be used in conjunction with imaging to monitor early response to therapy. Earlier this year, the FDA issued a draft guidance on ctDNA’s use for research and regulatory purposes. [ Read AgencyIQ’s analysis of comments on the draft guidance here.]
• One group, the ctDNA Monitor Treatment Response ( ctMoniTR), is working to establish its use as an important regulatory endpoint to enhance drug development. ctMoniTR is a partnership between the advocacy group Friends of Cancer Research (FOCR) and 25 organizations representing industry, government and academia. The goal of the partnership is to “assess the ability of a rapid and easy-to-use blood test to monitor treatment response.”
• ctDNA is already “being used as a selection marker for clinical trials, mainly in the metastatic setting,” said Julia Beaver, chief of medical oncology for FDA’s Oncology Center of Excellence, in a July FOCR meeting focused on ctDNA’s qualification as an early endpoint. And although current evidence – “which is mainly showing prognostic benefit” – does not yet provide sufficient support for ctDNA’s regulatory use, “we are really excited to be involved in all of the efforts toward this end,” as work continues to generate more evidence to support ctDNA’s use as a regulatory endpoint. [ Read AgencyIQ’s analysis of the FOCR meeting here.]
• The FDA’s pediatric Oncologic Drugs Advisory Committee met in May to discuss the issue of surrogate endpoints in a rare and deadly pediatric cancer. Drug development for pediatric neuroblastoma patients with high-risk disease could be smoothed with adoption of end-of-induction response, an intermediate outcome during the course of treatment, according to researchers and clinicians who appeared before the committee. Using an intermediate pharmacodynamic biomarker can enrich the trial population and weed out ineffective therapies or doses before proceeding further in the drug development process, according to Lisa McShane, chief of the biometric research program and associate director of the Division of Cancer Treatment and Diagnosis at the National Cancer Institute. [ Read AgencyIQ’s analysis of the pedsODAC meeting here.]
• The FDA held a two-day workshop in March to address the complicated question of multi-component biomarkers, discussing pain points and areas for exploration, particularly in the fields of medical imaging and artificial intelligence/machine learning. Multi-component biomarkers are “biomarkers [that] include features based on two or more measurements, potentially including clinical characteristics such as patient demographics that may be used independently and/or in combination through an algorithm as defined characteristics indicating normal biological processes, pathogenic processes, or responses to an exposure or intervention, including therapeutic interventions and environmental exposures,” according to Dan Krainak, Biomarker Working Group Co-Chair at the FDA. A multi-component biomarker is critical to measure a characteristic that is not adequately captured by a single measurement, and the workshop focused on some of the unique regulatory challenges in qualifying biomarkers of this complexity [ Read AgencyIQ’s analysis of the multi-component biomarkers workshop here.]

This week, CBER’s Scientific Symposium included a focus on internal and external biomarker work

• CBER’s annual scientific symposium convenes agency staffers and external researchers for a 3-day recap of research relevant to the agency’s biologics work. This year, the agenda for the symposium included sessions on advanced manufacturing, computational science, gene therapy, immune responses to vaccination, emerging and re-emerging diseases, as well as a session on biomarker discovery.
• The biomarker session was kicked off by Lisa Butterfield, of the University of California San Francisco. Butterfield also serves as the current chair of the FDA’s Cellular, Tissue and Gene Therapies Advisory Committee.
• Butterfield, whose research area is immuno-oncology, cited the need to standardize measures of immune response and reviewed four basic models under which biomarker analysis can be performed. Research scientists in a lab associated with a clinical trial can do the analysis, she said. Advantages to this approach are that the lab is likely to be highly innovative and flexible, and the work is relatively less expensive. However, broad expertise and standardization may be lacking, and personnel turnover is high. An academic core lab can also perform biomarker analysis, with the advantages of high levels of expertise, standardization, and competence in validation. However, there may be less innovation and more limited technologies in these settings. When contract research organizations do analysis, experience and standardization are high, as may be cost, and the burden of transporting samples to a distant location. Finally, when industry performs the analysis, costs are generally higher, and technology may be limited.
• “All of these work; all of these have pluses and minuses,” said Butterfield. What’s needed is that the analysis be accurate, with close agreement to the true value of the biomarker; precise – “We need to know what a normal range is,” said Butterfield, adding that analyses must also be reproducible. These are standards set by the Clinical Laboratory Improvements Amendments (CLIA) for all tests, and they apply in biomarker development and analysis equally as for other tests.
• Turning to specifics, Butterfield reviewed several research areas where she’s involved in biomarker development, including highlighting other FOCR work.  Here, the organization is coordinating academic, industry and FDA researchers who are working on harmonization of tumor mutational burden (TMB). This biomarker is tumor-agnostic; the presence of more genetic mutations in tumors means better response to checkpoint blockade. But variability methods used to assess TMB can complicate therapeutic decision-making – a fact that prompted the FOCR project.
• FOCR work has now identified empirical variability in estimating TMB in 16 different assays, developing “a publicly available calibration tool to align TMB estimates using different panel assays,” according to Butterfield’s presentation.  Ongoing work is exploring some limitations of the biomarker: TMB levels, whether from the blood or the tumor itself, only show the mutations, not “the antigenicity of the relevant neoantigens involved in tumor/immune response, nor do they provide information on the capacity of tumors to present antigens” – factors likely to impact the success of immune checkpoint therapy, according to a 2022 Nature Medicine article.
• Next, an FDA scientist in a CBER molecular virology lab shared intramural FDA work on micro-RNA (miRNA) as potential biomarker for HIV diagnosis. MiRNAs are non-coding RNAs that play a role in regulating gene expression. Krishnakumar Devadas explained that as use of pre-exposure prophylaxis (PrEP) for HIV is on the rise, both incident and acute cases “may be negative for viral markers in plasma or serum,” meaning that individuals may be infected but have a negative viral load. This situation can also occur with latent HIV infection.
• Investigating miRNAs as a biomarker for HIV infection addresses an unmet need, “to identify host and non-viral biomarkers that could be used in addition to, or instead of, viral markers to accurately identify new infections and latency,” said Devadas.
• MiRNAs have some advantages as potential biomarkers: they are stable at extreme conditions, including extremes of pH, through the freeze-thaw cycle, and despite various specimen treatments, including application of hydrochloric acid. Additionally, they are found in cell-free body fluids, including plasma, urine, amniotic fluid, and milk, as well as serum, which make obtaining samples a relatively non-invasive process.
• After a discovery phase, 17 miRNAs were selected for validation, with six eventually being statistically significantly correlated with immunoassay-verified infection. Eventually, Devadas and his collaborators settled on an early HIV-1 mRNA panel of four miRNAs (miR-223-1, miR-16-2, miR-195-1, and miR-20b-1) that detected HIV-1 with 100% sensitivity and specificity.
• The four-item panel, Devadas said, “had significantly higher sensitivity than any single miRNA to distinguish early HIV-1 infection from healthy controls.” The plasma signature of miRNA may also help distinguish acute, early infection from long-term infection, he added. Work here is ongoing.
• Finally, another FDA scientist shared clues that are unlocking a mystery: why, despite high vaccination rates, pertussis cases have been rising. Tod Merkel, a principal investigator in CBER’s Office of Vaccine Research and Review (OVRR), related that although fatalities are now rare, pertussis cases have risen and remain stubbornly high since the introduction of the acellular pertussis vaccine in 1991.
• Coverage for both infants and adolescents, who each receive acellular pertussis vaccine according to the pediatric immunization schedule set by the Centers for Disease Control and Prevention (CDC), both hover around 95%.  “The increase in pertussis cases is not due to a failure to vaccinate,” Merkel emphasized.
• The first step to solving the mystery was to identify a nonhuman primate model, since rodents and other animal models can’t be infected by Bordetella pertussis, the bacterium that causes pertussis, also known as whooping cough.
• Merkel and his collaborators discovered, by vaccinating the olive baboons chosen for the model, that although baboons vaccinated with either acellular or whole-cell pertussis vaccines are asymptomatic, the acellular-vaccinated baboons were still infected. Further, these infected animals could transmit disease, they discovered by placing pertussis-naïve primates in cages 1:1 with acellular- and whole-cell-vaccinated primates.
• The beginning of the answer was found when the researchers discovered that the adaptive immune response was skewed depending on whether the primate had been vaccinated with one of the two formulations, or was convalescent from a pertussis infection but was unvaccinated. Baboons who received the acellular pertussis vaccination only mounted a T-helper 2 (Th-2) cell response, with little Th-1 and no Th-17 response seen. By contrast, both the convalescent baboons and those that received whole-cell vaccinations had robust Th-1 and Th-17 responses, with minimal Th-2 response.
• “We think this explains why the acellular-vaccinated animals don’t clear infection efficiently,” said Merkel, explaining that it was the Th-17 responses that led to clearing extracellular bacteria, which are responsible for transmitting infection. Asymptomatic infection in either unvaccinated individuals or those who received whole-cell pertussis vaccine probably boosted immunity in the real world, he added.
• Further work in Merkel’s lab has confirmed that it takes repeated exposures before an acellular-vaccinated baboon will remain uninfected with a challenge from B. pertussis.
• The identification of the biomarkers that served as correlates of infection and transmissibility have enabled more work: Merkel and colleagues are now looking at candidate antigens for “next-generation” acellular pertussis vaccines, having identified antigens that provoke robust Th-17 response. OVRR is looking to test the hypothesis, identify correlates of vaccine-induced immunity that confer protection against infection, and identify which correlates will indicate durable vaccine-induced immunity.

What’s next

• FOCR’s work on ctDNA and TMB elucidation and validation is ongoing. Given the well-publicized and ongoing issues with accelerated approvals in oncology, researchers must tread carefully as they hope to validate new surrogate endpoints to speed oncology approvals.
• We would also remind readers that sponsors can use the FDA’s searchable database of drug development tools for exploration, to look for potential collaborators, and to ensure they are not reinventing the wheel.
• Basic science work conducted both intramurally within the FDA and in conjunction with collaborators underpins much of FDA’s regulatory business – a fact highlighted in today’s Science Symposium, which also featured many “lightning talks” highlighting other preclinical and clinical work being done within CBER.
• More from the Science Symposium: a notable amount of the agenda was dedicated to vaccine-related topics, including a keynote by Paul Offit of the University of Pennsylvania. Offit, who serves on the FDA’s Vaccines and Related Biologic Products Advisory Committee (VRBPAC), was one of two dissenters when that committee voted on the (currently available) variant-adjusted Covid-19 boosters. Look for analysis from AgencyIQ on these themes, later this week.

https://fda.agencyiq.com/article/00000183-6346-da48-a3e3-e3e7c5760000