Representatives from leading pharmaceutical firms said at a conference last week that they see minimal, or molecular, residual disease testing as the “next big game changer” in precision medicine development, and hopefully in the clinic, though they continue to grapple with how to build that bridge.
Discussing MRD with technology developers and clinicians at Cambridge Healthtech Institute’s Precision Medicine Tri-Conference in San Diego, representatives from Pfizer and Johnson & Johnson agreed that there is no question that this testing will increasingly have a place in precision medicine for solid tumors. At the same time, they were sober about some of the hurdles in translating technological innovation into clinical impact.
Cost was raised early in the panel discussion. “It’s not cheap to do MRD testing,” said Fernando Cruz-Guilloty, director of oncology at Johnson & Johnson Innovative Medicine. “From my perspective that is a big hurdle.”
It’s a hurdle that clinicians and diagnostics firms also seem to recognize. Although the genesis of MRD technologies takes place largely in academia, the “heavy lifting” in terms of validation and clinical development comes from drug development, said Justin Odegaard, VP of Project Management at liquid biopsy firm Guardant Health.
In that sense, the onus on those developing new biomarker technologies is then to “pick the right question,” so that the community can then pick the right trial design to answer that question, he added, expressing some surprise at how much discussions of clinical utility seem to be lagging excitement around clinical validity.
Areas where drugmakers said they are particularly excited include disease monitoring to identify resistance or lack of response which could guide adaptive shifts in treatment, as well as patient selection in postsurgical or adjuvant treatment settings. But the field is learning that not all MRD technologies may be fit for all purposes.
Howard Scher, chair and head of the biomarker development program at Memorial Sloan Kettering, agreed that asking these clinical utility questions “has gotten lost at the party,” in terms of being part of upfront trial design.
“We haven’t done the job,” he said, and the potential for missteps is not trivial.
Jean-Francois Martini, Pfizer’s executive director and translational oncology lead, raised the example of frustrating results from recent efforts to establish tumor-informed MRD as a predictive tool to guide the use of adjuvant targeted therapy, particularly in breast cancer, something other pharmaceutical firms have also faced, including GlaxoSmithKline with its PARP inhibitor Zejula (niraparib).
“One of the challenges for us has been collaborative trials where we tried to utilize bespoke panels to identify those patients who had disease after either curative intent surgery or after adjuvant treatment and [find] those patients who needed additional treatment,” said Martini.
“[Between] delivering the tumor tissue to the vendor, having it tested, having the panel derived so that the first sample could finally be tested, depending on the indication … we had a lot of patient loss for enrollment because by the time the panel was ready for the blood sample to be tested, the patient actually had already started standard-of-care treatment or had already progressed and had already been shown to be metastatic.”
For a test to be successful in a trial, it must be implementable in the indication for which it is being assessed.
Odegaard noted that the panel’s discussion was “strikingly similar” to conversations taking place a decade ago when tumor genotyping was first being integrated into precision medicine for advanced-stage patients. “We had good technologies that worked, but they couldn’t help if you couldn’t get an answer back in time.”
Guardant, as a liquid biopsy testing firm, is obviously promotional about the speed advantages of blood-based testing, but Odegaard said that liquid biopsy tests like its Guardant360 have not usurped tissue-based tests but have ultimately come to complement them.
If he were to “read the tea leaves,” he’d expect to see the same multi-option toolbox emerging for MRD, he added.
“If we look at the lessons we learned before … in indications where you do have a time pressure, maybe [tissue-based MRD] isn’t the right path. In neoadjuvant situations where all you have is about 17 and a half cells in a smear, there’s no way you’re going to be doing tissue diagnostics,” Odegaard said. “Those are great areas to use your ‘screwdriver’ with plasma-only tests.”
Alternatively, if there is a case where adjuvant treatment starts at eight-weeks post-surgery and there is a good resection specimen, that could be where one “picks up the hammer” of a tumor-informed test.
According to Odegaard, Guardant is betting on a tissue-optional future. “Use it when you have it but not when it doesn’t fit.” The company’s job is to show that its blood-only approach is as accurate or as predictive as tumor-informed tests. But that can be a challenge because interventional trials using two different diagnostics are a “headache and a half” and likely unattractive to pharma.
At risk of “beating a dead” horse, he reiterated the similarity to prior debates over next-generation sequencing. “We’ve come a long way, and it’s not been easy or perfect, but I wonder if a lot of those strategies can be applied,” he said.
The pharma voices on the panel were also sober about the challenges of a complex regulatory framework in Europe and an uncertain one in the US as the Food and Drug Administration moves toward oversight of laboratory-developed tests.
Pfizer’s Martini said that some prior efforts by his company to bring biomarker-based trials using multiple LDTs to fruition failed because the firm couldn’t effectively benchmark the different tests. They had to start over with a single centralized assay. The fear is that the same may happen with MRD.
“There’s a great deal of diversity in the LDT space, and we have even more so with MRD because of the variety of different technologies … [but] I think we’re missing the catalytic role that the regulators can play. I think the FDA actually has a great potential to foster new technologies,” said Guardant’s Odegaard.
“If you have a performance bar, you’re not chasing this sort of diversity as much. You don’t have as high of a risk. And I think that is something that also helps us from a developer’s perspective because we know what to demonstrate, we know who our audience is,” he argued.
Ongoing public-private efforts to harmonize tests include efforts by Friends of Cancer Research, and BLOODPAC, but the FDA has positioned itself mainly as a listener when participating in such discussions.
“We can’t force everybody to use the same platform, but as FDA did with some of the heme MRD tests in early days with standardizations and … reference materials … having a reliable arbiter of analytical quality means we can get to the hard work of matching up assays with clinical indications,” Odegaard said.
“Basically, it gets me and mine out of the risk equation so you guys can tackle the hard work,” Odegaard added, addressing his fellow panelists from the clinical and pharma spheres.
