By Deborah Borfitz
September 8, 2020 | Oncologists often make decisions about diagnostic tests with very little information to go on, says Paul Gerrard, M.D., with McDermott+Consulting. The dearth of high-quality evidence has made it difficult for both doctors and institutions to advance personalized medicine without a roadmap for overcoming the challenges, which the Personalized Medicine Coalition (PMC) has been developing, adds Daryl Pritchard, the nonprofit’s senior vice president of science policy and a geneticist by training.
Both men spoke at the recent Next Generation Dx Summit on the topic of how oncologists choose and evaluate such tests. Gerrard served as moderator of the session.
The PMC, comprised of 250 members across all stakeholder types, has a healthcare working group devoted to identifying the barriers and go-forward strategies for personalized medicine, says Pritchard. Forty hospital systems, plus physician and clinical decision support groups, capped off a series of collaborative meetings with a 2017 forum where they laid out the laundry list of challenges.
The tasks, about 100 in all, fall in five broad categories: education and awareness (around novel technologies), patient empowerment (engagement, privacy, and data protection), evidence development and value recognition (“a key area to drive access to personalized medicine technologies”), infrastructure and information management, healthcare delivery approaches and processes (“the pinnacle of where all this is leading”).
The level of evidence necessary to determine clinical utility is “the money question,” Pritchard continues. The inclusion of a diagnostic test for a molecularly targeted therapy in professional guidelines suggests clinical utility has been established, but it takes “many steps to get there.”
For a product to be approved or cleared by the U.S. Food and Drug Administration (FDA) or administered in a lab that has certification under standards of the Clinical Laboratory Improvement Amendments requires that the analytical and clinical validity of the test be identified. Determining clinical utility, the benefit to patients and the healthcare system, is the “hardest part,” says Pritchard.
“Many people believe that to establish utility you need a lot of real-world evidence [RWE] [generated by] practice-based use of a diagnostic test,” Pritchard adds. “That data would be meaningful and useful for patients, payers and [clinical] decision-makers.” But who is responsible for developing the protocols for using RWE, and for gathering that data and putting it in a research context, is an open question.
The quality of evidence generated by these so-called “pragmatic trials” can vary widely, Pritchard says, so standards will be needed to guide the study process. ISPOR (International Society for Pharmacoeconomics and Outcomes Research) is attempting to do just that, he adds. “The future for pragmatic trials is bright, but [owing to the quality problem] has not caught on as quickly as hoped.”
“It’s a new concept and many technologies have only been around for a few years, so the downstream outcomes and effects have not been established yet,” Pritchard notes. “It’s a chicken-and-egg problem and who does the research to solve for that?”
The problem has become clearer over the past few years, thanks to the efforts of data groups like Flatiron and the MITRE Corporation as well as guidance on RWE issued by the FDA, says Pritchard. The focus has been on using information technology to help understand RWE. “How best to do that is the question.”
Pritchard’s advice is for diagnostics companies to involve oncologists as advisors when designing RWE-based studies to ensure the results are meaningful to them. Payers might also then have more faith in the findings.
Greater transparency by payers about how they assess healthcare technologies might help but it’s not a panacea, Pritchard says. The Institute for Clinical and Economic Review (ICER) and other value framework groups have come up against a lot of criticism for being open about what goes into their assessments, given the variety of perspectives on the matter.
Patient Perspectives
Payers, of course, need to be convinced a technology is clinically useful for patients while also worrying about the bottom line. And oncologists aren’t particularly interested in learning about new technologies until they’re available to patients by virtue of insurance coverage, notes Gerrard, who previously worked for a Medicare Administrative Contractor making coverage and payment determinations for new molecular medicine technologies. His observation was that decision-making, “seemed to fall back on the payer rather than the healthcare community.”
Pritchard agrees, noting that Blue Cross Blue Shield and Harvard Pilgrim Health Care have been using evidence to help drive coverage policies around personalized medicine. PMC has found that genomic testing for lung cancer is only mildly cost effective, he adds, because only 65% of patients with actionable mutations were subsequently treated with a targeted therapy rather than standard chemotherapy. When a six-payer panel were shown the vastly improved cost effectiveness if all patients received the best therapy for them, the consensus was to cover testing so the full value of the technology could be realized. “They didn’t want to be a barrier.”
From the perspective of many hospital systems, an important variable in any care strategy is meaningful engagement of patients, continues Pritchard, to include the privacy and security of genomic data and appropriate consenting. Multiple provider organizations now have genetic counselors or social workers to help patients decide whether to order certain tests and make sense of the data that testing generates.
For payers, including the patient perspective in the technology assessment process is starting to become a top priority, says Pritchard. The goal is better outcomes and the challenge is accommodating individual patients who variably want to take charge of their healthcare or follow their doctor’s lead. “ICER has faced a lot of criticism for including the patient perspective of value [in its assessments].”
The good news is that the conversation about personalized medicine technologies is no longer about why they make sense, says Pritchard. “Over the last three to four years we’ve turned a corner. Implementation is now at the forefront and that’s a positive sign, although it comes with a new set of challenges.”