And, of course, in the case of ESR1, “It would be unexpected to find such mutations at initial diagnosis anyway,” Dr. Bean says.
The Food and Drug Administration Oncologic Drugs Advisory Committee was scheduled to meet April 30 to discuss the new drug application for camizestrant tablets. Per the agency’s website, the proposed indication is in combination with a CDK4/6 inhibitor for treating HR-positive/HER2-negative advanced or metastatic breast cancer, based on the presence of an ESR1 mutation during firstline endocrine-based therapy.
Says Dr. Bean: “The details of what they decide actually matter a lot—whether it’s tied to a companion diagnostic, what indications they say exist for monitoring. All of those details will shape how this goes.”
But however things unfold, he says, good communication between pathology, laboratory medicine, and the ordering physician is important.
For all the questions and concerns, Dr. Lee says that in his crystal ball prediction, “I am relatively to highly confident that this will be the method moving forward,” rather than awaiting clinical imaging confirmation. “By the time you see it, it’s too late. And liquid biopsies offer an early detection and an early intervention. And that is always better,” he says. “There are very few examples where it’s better to treat later, so I’m pretty confident in my prediction.”
Meanwhile, the work continues. He is part of another trial—one he hopes will be equally disruptive—looking at breast, lung, and colorectal genotypes. The work is being funded by a national consortium of “highly mathematical AI groups,” he says. The aim is to explore the feasibility of doing real-time multiomic analysis on tumors for prediction purposes. “Let’s look at all the potential mechanisms of resistance and all the potential kinds of therapies we can apply.”
“We need to take a multiomic approach” as a logical next step, Dr. Lee says. “Just one platform is never going to give you all the answers. One mutation is not how a tumor becomes resistant; it becomes resistant in an ecosystem of other things.”
Dr. Lee can hardly contain his enthusiasm as he looks ahead. Liquid biopsy “is such a great way of visualizing a tumor,” he says. “It’s really impressive. We’re doing epigenomic profiling; we’re doing fragmentomics.” This goes well beyond identifying a point mutation, in other words. By using circulating tumor DNA to measure methylation in the epigenome, “It’s possible to predict gene expression back in the tumor.” His astonishment is clear. “No one would have thought we could do that from blood.”
The artificial intelligence component, Dr. Lee says, is going to be important as the impact of SERENA-6 and other similar trials is felt. “We can collect more and more data on patients now, so it’s imperative to have ways to turn that data into knowledge. I think AI is going to be critical to that.”
“We can carry on research and show mutations forever,” he adds. AI is also a strong candidate for supplying clinicians with decision support. “They’re getting overwhelmed. It’s really hard on clinicians—the pathways for treatment are really complex. I think AI and clinical decision support is going to be essential to move these things forward.”
And for pathologists? “They are at the center of it all. It’s important that they understand the potential tsunami of information that’s coming their way.”
Karen Titus is CAP TODAY contributing editor and co-managing editor.