April 2, 2024 | You might be forgiven for calling the tone of a panel on diagnostics held at last week’s Precision Medicine Tri-Con “optimistic”. But you would be corrected: it isn’t optimism, it’s math.
Peter Bach has had years of experience as a pulmonary physician, Senior Adviser at the US Centers for Medicare & Medicaid Services, and consultant at GRAIL and Foundation Medicine. In his current role, he’s CMO at Delphi Diagnostics that is developing a lung cancer screening blood test.
Screening diagnostics is primed for success, he argued. “That path is actually incredibly clear,” he said. “We think we have a wicked cool technology, [but] none of that matters. The data matter. We have to generate the data in the intended use population. It needs to be convincing to external experts… It’s a longer road to reimbursement, but it’s in the law that you get reimbursement if the task force endorses your technology or your approach… I feel like it’s on us! We have a technology risk. We have execution risk. We have clinical trial execution risk. It’s critical that we get our story right, our metrics right, are clear about our endpoints. All those things are hard. But they’re in our control. It’s not optimism; it’s cold hard facts.”
Bach’s viewpoint was generally shared by the rest of the panelists, but they were not without suggestions for where there could be improvements, and dives into the particulars of other types of diagnostic development beyond screening tests for precision medicine.
There is an education gap, said Jeffrey Venstrom, CMO at GRAIL, and it’s a problem the entire diagnostics community should work on. Busy primary care clinicians are struggling to keep up with A1C and cardiovascular patients, Venstrom said. Liquid biopsies and other diagnostic screening tests, especially ones that have an AI component, feel like a black box, he said. We must “work together to fill those educational gaps and implementation barriers that are going to be immense for getting the profit we need as a company to stay viable, and impact that we all aspire to have on patients,” he said. He also advocated for simplification of the process, “creating that easy-button referral opportunity for a report that says, ‘There’s a signal of cancer detected in the blood’ is a challenge we’re all living with.”
Venstrom challenged panel moderator Edward Abrahams, President of the Personalized Medicine Coalition, that this might be a task for a neutral organization such as his. Now is the time work on these implementation challenges, Venstrom argued, even as test development and research are ongoing, so that the system will be ready for the effective tests to come.
Christopher Conn, director of diagnostics strategy at Amgen, agreed with Venstrom’s proposed order of priority. “The science quickly outpaces what’s adopted in the community setting where the majority of patients are being treated,” he said. “Anything we can do to bridge that gap in awareness can really help facilitate our decision-making in moving forward.”
A big pharma might be able to take a brute force approach to adoption, said Paul Beresford, Vice President and General Manager, CDx, at Agilent Technologies, referencing Pfizer’s partnership with Exact Sciences. “Pfizer put 3,000 people on their sales force, and drove testing into the market,” he said. But it’s not the model everyone can emulate. “We’ve got to think up some efficiencies here to get these tests into the market and drive the commercial equation.”
Justin Odegaard, VP of clinical development and product at Guardant Health, pointed out that the assurance of reimbursement doesn’t carry through to all diagnostics. In advanced cancer and early cancer diagnostics, in particular, there are no guarantees of reimbursement because there’s instability and lack of clarity in the legal, regulatory, and reimbursement frameworks.
“Exact Sciences is a great success story; hopefully we can recapitulate that,” Odegaard said, but, “how do we make that sort of impact without the predictability that those sorts of legal frameworks gave—there’s legal frameworks, there’s regulatory frameworks, there’s reimbursement frameworks… At the size of these investments, this is the only way we can bring drugs to market because of the predictability of those types of frameworks.”
Before we glorify Exact Sciences too much, Abrahams pointed out that the company is still not profitable. He challenged the panel to identify ways to make the path to profitability less arduous for successful science.
“As I understand it, part of the key to profitability is a low-cost platform—especially for screening,” Bach said. “Any time you have screening, you’re talking about millions of people… You can’t have a profit model that depends on a price point people will balk at. You need a low-cost structure so that you can be unit-level profitable.”
At GRAIL, Venstrom said, they have focused on optimizing operations. “It really is the less sexy side of innovation, but is absolutely critical for us in the diagnostics industry for becoming profitable.” GRAIL has a 100,000 square foot lab in the Research Triangle area of North Carolina outfitted with automation inspired by high-throughput drug screening and the automotive industry. Automation is a key component of being able to do this fast, cheap, and at scale, he said.
Engaging the Medical Record
When Abrahams opened the floor for questions, the first one directed the panel back to the very issue both Venstrom and Amgen’s Conn brought up: educating and engaging care providers. How can we get these sorts of tests—both screening diagnostics and early stage diagnostics—the last mile: into the electronic medical record where physicians can engage with them?
Guardant’s Odegaard agreed that it’s a particular challenge. “None of these tests are worth anything if the information doesn’t get to the point-of-care and it doesn’t get there on time. One of the things we’ve seen through the tissue testing experience, if your results arrive late, it’s worthless. One of the things we’ve seen through the EHR experience, if your [test result] PDF goes into the media tab that nobody ever looks at, it’s worthless. So one of the things we’ve been investing in… is really getting an integration not just into the EHRs, which is important, but into the EHR as structured laboratory data.” But this is no easy feat Odegaard, warned. “They do sort of consider their EHR their own fiefs… These are extraordinarily different systems that will not play by the same rules… These are not bad things; they’re just the shape of reality right now.”
Odegaard, too, challenged Abrahams and the Personalized Medicine Coalition to take on the issue. A common set of standards for representing and displaying data in electronic medical records would be truly disruptive.