This story has been updated with additional comments from 23andMe.
NEW YORK – A new pharmacogenomics collaborative community called Standardizing Laboratory Practices in Pharmacogenomics, or STRIPE, aims to bring diverse stakeholders together to find the best way to ensure patient health and safety.
The US Food and Drug Administration, which had previously said that collaboration among stakeholders would be the preferred way to shape the agency's approach to PGx, joined STRIPE in late August, and the community now seems poised to be the primary channel for stakeholders to work with FDA to help standardize the field. However, it is unclear if all stakeholders will come to the table or what the ultimate outcome will be.
Pharmacogenomics strives to incorporate individual patient genetic information into the decision-making process about which pharmaceuticals would be most effective and which might be useless, or dangerous.
The stakeholders in the PGx space are diverse. They include genetics researchers, consortiums and working groups synthesizing and harmonizing the knowledge base, academic and commercial labs offering genetic tests to help guide physician prescribing, as well as clinicians, patient advocates, genetic counselors, pharmacists, software and analytics providers, insurers, and of course, regulators at FDA.
Some of the problems faced by the PGx community are perhaps semantic. For example, most PGx tests are considered lab-developed tests, but the specific definition of that term and the oversight of LDTs has been in flux in recent years.
Some of the problems in the community have apparently grown out of a lack of consensus and standardization, particularly around what is an acceptable level of evidence to support a drug-gene interaction and what constitutes a complete report for a doctor or patient.
And some problems seem to stem from historically inconsistent and confusing FDA regulatory actions in the PGx space. In response FDA proposed joining a collaborative community of pharmacogenomics stakeholders, as previously reported, to help chart a path forward.
A collaborative community is a specific approach to incorporating the collective needs of stakeholders in the medical device space and attempting to find a path forward on a range of issues, according to FDA.
"Collaborative communities can work together to address challenges and opportunities, including minimizing duplication of activities and generating harmonized solutions," an FDA spokesperson explained in an email.
The agency's Center for Devices and Radiological Health had made participating in collaborative communities a strategic priority in 2018. These communities are supposed to be ongoing forums for public and private sector participants to solve problems to "in an environment of trust and openness," according to an agency statement. The role of CDRH is to be present and listen, but also to "foster a community spirit," in part by ensuring that the communities have broad and fair representation.
The FDA spokesperson said that CDRH's decision to participate as a member of any given collaborative community is influenced by the public health impact of the community, its alignment with the regulatory mission, and current priorities and resources.
"CDRH makes participation decisions on a case-by-case basis, and may consider whether the community has a governance structure, a convener, a plan to measure success and a mechanism for sustained engagement to help determine whether CDRH will participate," the spokesperson said.
Jeanette McCarthy, the founder of Precision Medicine Advisors, spearheaded a summit of PGx stakeholders in February. According to a white paper from the summit, efforts to form a pharmacogenomics collaborative community that would meet the FDA's requirements were underway at that time, and were facing challenges.
At least two independent groups began to coalesce. One included members of the Clinical Pharmacogenetics Implementation Consortium (CPIC) and the Precision Medicine Coalition, and one was led by the American Society of Pharmacovigilance, or ASP, who called its community STRIPE. As the communities were forming, labs joined up with both groups, McCarthy said.
Troy Moore, chief scientific officer at Huntsville, Alabama-based genetic testing lab Kailos Genetics, said that he had been a part of the grassroots CPIC-involved group, but that the group was very diverse, so it was challenging to quickly pull together the required materials to put in front of FDA.
"The FDA made it clear they only had the capacity to work with one group," Moore said. "They reviewed each group's charter and membership to make a decision, [and] at the end of the day STRIPE was the only one to submit" its information to the agency.
The formation of any community at all is a net win for the PGx space, in Moore's opinion, no matter who is spearheading it. However, "Keep in mind the FDA is really an observer or listener," he said. "Their participation will not drive the direction," he also said, noting that many in the PGx space are generally curious to see how the effort will shape up.
ASP and STRIPE
Benjamin Brown cofounded the American Society of Pharmacovigilance about three years ago and serves as the organization's executive director.
In an interview, Brown said that ASP is particularly passionate about adverse drug reactions. More than 200,000 deaths are caused by adverse drug events each year in the US, and they are the fourth leading cause of death, he said. The majority of these deaths are preventable, but there had previously been no dedicated advocacy group.
Pharmacogenetic testing is a critical tool to prevent adverse drug events, said Sara Rogers, a pharmacist specializing in oncology who serves as director of clinical affairs at ASP.
Rogers and Brown both participate in a genomics literacy working group at the National Institutes of Health called Inter-Society Coordinating Committee for Practitioner Education in Genomics, or ISCC-PEG. They also co-chair an organization called the Pharmacogenomics Access and Reimbursement Coalition, or PARC, focused on bringing together health economists and payors.
"Through the confluence of all these organizations that we participate in, it became very clear to us there was a paramount opportunity for us to act as a connective tissue," Brown explained.
Historically, "There have been a lot of great initiatives or organizations that interact with the pharmacogenomics community that are working on one project or another," Brown said, "But never before has there been one sole initiative that brought all of them together."
Rogers said that ASP responded to the call for a collaborative community in part due to intensified regulatory actions for labs offering pharmacogenomics testing last year. "There wasn't necessarily specific guidance around what companies could do to come under compliance," she said.
The immediate priority for the STRIPE community is to develop a standardized approach to evidence evaluation for pharmacogenetics associations, Rogers said.
"We are all working from the same body of evidence, but the question is whether we are reviewing that evidence differently," she said.
One of the benefits of a community framework, she said, is that it may make it more likely that stakeholders who participate will adopt the solutions, especially given FDA is a part of the process.
Stakeholders can get involved with STRIPE through free general membership by registering on the community website. Further, they can choose to join any of the nine proposed pathways for participation, which are focused on different shared challenges and barriers to good outcomes, Brown noted.
Examples of these subcommittees include evidentiary standards, consensus development, technology information systems, and patient access, advocacy, and equity. There is also a steering committee with seats for representatives from professional organizations and advocacy groups.
Overall, Brown said that through developing a shared lens to evaluation evidence, uniformity will help increase the diffusion and adoption of PGx technology.
Bernard Esquivel, chief medical officer at OneOme, said his company has been navigating the changing regulatory landscape of the past year and a half and is maintaining an open dialogue with the FDA and customers to find solutions that support safe and effective PGx testing.
OneOme has been part of the STRIPE community from its inception, Esquivel said.
"I am heavily involved in several aspects, as is OneOme's laboratory director," he said. "As an organization, we believe this type of industry-wide collaboration is important to the future of PGx testing."
Esquivel said he believes FDA's decision to join STRIPE is "both confirmation of the good work the community is already doing and a signal of even more progress to come."
Pharmacogenomics has reached an inflection point, Esquivel said. "Clinical validity, clinical utility, and cost-effectiveness have been continuously growing around PGx," he said, and now the field has reached the moment of implementation.
"These are times when stakeholders from across the healthcare ecosystem need to align on best practices that will bring the benefits of PGx testing to patient care, and even make PGx the standard of care for medication management," Esquivel said.
The path forward
While FDA's spokesperson emphasized that "collaborative communities can work together to address challenges and opportunities, including minimizing duplication of activities and generating harmonized solutions," whether the STRIPE collaborative community specifically will be embraced by all PGx stakeholders and include diverse representation remains to be seen.
Representatives from CPIC, the Association of Molecular Pathologists PGx working group, and the Pharmacogene Variation Consortium (PharmVar) declined to comment for this story.
McCarthy said she thinks that some in the PGx space may be concerned that ASP's scope for STRIPE might be rather broad for the needs of the community.
"You've got this huge structure with all these different goals, when really what they need is a surgical strike to the regulatory issues," she said.
There is of course common ground among all the PGx stakeholders. At the same time, "There are good tests out there," McCarthy said. "These tests can help prevent severe adverse drug reactions and can help choose the right dose of a drug to be more effective."
The tensions seem to derive from questions about who is the ultimate arbiter, or whose approach to standardizing the field should have final authority.
Ruben Bonilla-Guerrero, vice president of medical affairs and medical director at Admera Health, suggested that all the stakeholders are important to be involved, particularly the academic and professional organizations.
"Everybody who is impacted, if they are genetically and ethically responsible, should join" the collaborative community, Bonilla-Guerrero said. Furthermore, when groups distance themselves, perhaps out of a desire to assert final authority, "it creates animosity and distrust among the PGx community as a whole," he said.
Distrust and skepticism may have also been cultivated by previous interactions between stakeholders and FDA.
Historic high water marks and low points in the history of FDA regulation of PGx include Halloween of 2018, when the agency issued an alert to physicians and patients about using unapproved pharmacogenetics tests shortly after granting consumer genomics firm 23andMe marketing authorization to sell several tests directly to consumers. Other industry players found the timing confusing.
In addition, FDA warned PGx testing service provider Inova Genomics Laboratory in April 2019 the relationship between CYP2C19 genotype and drug response to certain SSRIs is not established. This was followed by enforcement overtures to a number of other labs last summer.
FDA has defended its various enforcement actions as necessary for patient safety, but McCarthy said uneven enforcement has been a huge problem. "There were labs that were told to remove drug information, and others who never even got contacted by the FDA and are still offering their tests," she noted.
In February 2020 the agency published a table of pharmacogenetics associations it deemed actionable. The table somewhat overlaps with a broadly accepted one curated by CPIC and PharmGKB, which was revised in June with "Very Important Pharmacogenes," or VIPs, pulled out and sorted into two tiers based on the strength of the evidence supporting gene-drug interactions.
Bonilla-Guerrero said the conundrum is that FDA's table is not complete. It is based primarily on information in drug labeling, he said, and the labels are not up to date. Thus, "You may align 100 percent [with the FDA table], and limit your test to just that, and by doing so, you run into gene coverage problems."
To solve this, Bonilla-Guerrero said FDA needs to talk to professional organizations to reach an agreement on the evidence. "This can be solved overnight," he said. "The information is already there, you just need the proper parties to talk and actually come to a consensus, and for the regulatory organization to produce guidelines."
When all is said and done, Bonilla-Guerrero said he is "cautiously hopeful" that the collaborative community will help sort out all of these issues.
Others, too, are optimistic. Moore at Kailos said his company welcomes FDA's involvement and guidance because "In order to operate, you need to know what the rules and boundaries are." With the formation of STRIPE, "Now there is a channel for dialog," he said.
Esquivel, meanwhile, said he is "confident that the unified purpose of advancing PGx will enable this community to come together and achieve much more than we could separately."
One dark cloud over these efforts is a possibility that they may be moot because the very nature of FDA's ability to regulate PGx assays, which for the most part fall into the category of lab-developed tests, has been called into question by a US Health and Human Services statement last month. HHS declared essentially that LDTs could only be regulated through an onerous, time-consuming, process called notice-and-comment rulemaking.
The impact of the HHS statement is hard to predict. FDA has so far declined to comment and instead referred all questions about the statement to HHS.
Jeff Gibbs, a director at the law firm Hyman, Phelps & McNamara, who recently filed a citizen petition against what he and others see as a problematic regulation of pharmacogenomics by the FDA, insisted that companies "have to take it seriously right now, because it represents current HHS policy." The statement will be important legally, he said, and won't vanish if there is a change in the federal government in November.
Meanwhile, the commercial entities interviewed for this story do not plan to make any changes until the dust settles. Specifically, Kailos' Moore said that he will wait and see whether a new administration might roll back the HHS statements. At Admera, Bonilla-Guerrero said the company will not alter its course, either.
"We all have our own conversations with FDA, and we act on the result of those conversations," he said. "I'm the gatekeeper making sure everything that is reported has the highest clinical level of evidence ... FDA or not, HHS or not, we report only the highest level of evidence, period."
More "inconsistent" regulation
Both McCarthy and Bonilla-Guerrero cited a recently cleared test from 23andMe as an example of the type of inconsistency in the regulatory process that they hope will be resolved in the future, potentially through the collaborative community.
Last month FDA granted 510(k) clearance to 23andMe for a test to provide customers with interpretive drug information about whether their CYP2C19 variation may influence their ability to respond to citalopram, an SSRI sold under the brand name Celexa by AbbVie and clopidogrel, a heart medication also known as Plavix and originally sold by Bristol-Myers Squibb.
The interpretative drug information is offered directly to consumers. McCarthy said by her own analysis FDA appears to have evaluated whether patients could read their test reports, and whether the tests were analytically valid. In terms of clinical validity, the agency appears to have looked at whether the few variants tested are associated with a particular outcome.
"What they didn't consider is how predictive those tests were, and whether they were complete or not," she said.
McCarthy has previously suggested in a blog post that 23andMe's PGx tests are not complete because they don't test for all the variants associated with particular drug response. Not testing the whole gamut could lead to "false negative" results, she said, whereby a clinician believes a particular drug is safe or efficacious for a particular patient when, in fact, it may not be due to another genetic variant that was not assessed.
"The fact that [23andMe's tests] are FDA approved and they are suboptimal tests is really concerning," McCarthy said. "That sends the wrong message, especially to consumers."
Bonilla-Guerrero concurred, characterizing the recent 23andMe clearance as essentially "playing the label card," because the company only submitted a very narrow claim.
"Some of the labels are quite old, and they only recognize specific alleles for certain genes that were known at the time" the drug was approved, he said. "Now, there are a lot more alleles with the same functional and clinical impact on these medications that are not included."
Bonilla-Guerrero also said FDA approval leads people to believe a test is superior, but that there is a "huge gap" in the 23andMe tests that could potentially lead to false negative results. "I don't know how you could approve something that, if you know what you are doing, you know it is incomplete," he said.
Kathy Hibbs, 23andMe's chief legal and regulatory officer, commented in an email that one of the primary factors the FDA evaluated in the firm's latest pharmacogenetic clearance was the level of evidence that supports the drug response claims.
23andMe asserted that its CYP2C19 test covers all the highest level, Tier 1 variants recommended by the Association for Molecular Pathology, and noted that the clinical significance of Tier 2 variants is uncertain at this time.
"Consistent with our overall approach to providing health information to our customers, we will only report genetic drug response associations with the highest level of clinical and scientific evidence," Hibbs said.
Pharmacogenetic associations with lower levels of evidence, or newly emerging evidence, will be reevaluated over time for consideration, Hibbs said, adding, "23andMe will continue to lead with science and pioneer increased access to genetic information, including pharmacogenetic information, driven by the highest scientific standards."
23andMe is not currently a member of STRIPE, but the firm said that may change in the future.
The regulatory status of 23andMe's tests could potentially impact other labs, as Moore pointed out, since they could claim their tests are substantially equivalent. Taking that a step further, if a lab is able to obtain some form of regulatory clearance, perhaps the issues it has with LDT regulation become moot, as does the drive to participate in a collaborative community to get the ear of FDA and shape the regulatory process.