Skip to main content
Premium Trial:

Request an Annual Quote

OSUCCC Researchers Refining Liquid Biopsy Use for Detecting FGFR Fusions, Guiding Cancer Therapy


NEW YORK – Researchers at the Ohio State University Comprehensive Cancer Center-James are exploring how to best use liquid biopsies to detect FGFR alterations, including hard-to-detect fusions, and use the technology to inform treatment decisions for cancer patients.

A team led by Sameek Roychowdhury, an oncologist and member of the Translational Therapeutics Program at OSUCCC, recently received two grants to advance the application of liquid biopsy testing for patients with FGFR-positive cholangiocarcinoma.

In one study, funded by an NIH grant, researchers are validating a liquid biopsy test that can gauge FGFR alterations from circulating tumor DNA (ctDNA), particularly gene fusions and rearrangements that might be missed by commercial DNA liquid biopsy assays available today. In another telemedicine-enabled study, supported by a grant from the nonprofit Gateway for Cancer Research, Roychowdhury's team is enrolling cholangiocarcinoma patients remotely, collecting their serial blood samples, and using cell-free DNA testing to track changes in the levels and types of FGFR alterations in their blood during treatment. The data may help establish when patients should be tested to monitor for progression and guide therapy selection.

The turnaround time and cost advantages with liquid biopsy testing have garnered oncologists' interest. Roychowdhury cited research from Japan looking at test turnaround times in two precision oncology studies and found that it took a median of 11 days to return results to patients who had ctDNA testing and 33 days for those who had tissue-based genetic testing. The faster turnaround times with ctDNA analysis allowed around 10 percent of patients to get on a study drug based on an identified biomarker, while only 4 percent of those receiving tissue-based testing matched to a therapy.

"Patients and doctors are okay to wait 11 days and try and get you on a trial," Roychowdhury said, "but people get antsy when their cancer is growing, and you get results 33 days later."

There are also potential cost advantages, he added, since collecting a blood sample is much simpler than resecting a piece of tumor tissue, which requires a patient see a radiologist and perhaps be sedated for a biopsy procedure. With blood testing, "I can do it often," Roychowdhury added. "I can't subject patients to biopsies regularly to monitor their cancer."

While there is enormous interest in the field around liquid biopsy, oncologists like Roychowdhury also have outstanding questions around how to best apply this relatively new technology in everyday practice. For example, "We don't know how sensitive we are in detecting FGFR alterations, including the fusions or rearrangements, which are a little harder to detect," Roychowdhury said. His team is hoping to answer some of these questions in the process of developing a ctDNA test optimized to detect FGFR fusions and exploring the best way to interpret test results in the context of treatments patients are receiving.

The US Food and Drug Administration to date has approved three FGFR inhibitors, erdafitinib (Janssen's Balversa), pemigatinib (Incyte's Pemazyre), and infigratinib (QED Therapeutics' Truseltiq). Erdafitinib treats advanced urothelial cancer patients if they harbor specific FGFR3 or FGFR2 mutations and have progressed on platinum chemotherapy. Pemigatinib and infigratinib are for second-line treatment of advanced cholangiocarcinoma patients with an FGFR2 fusion or other rearrangements. FGFR2/3 mutations are more common in bladder cancer, occurring in around two-thirds of patients, and easier to detect than FGFR2 fusions, which are estimated to occur in around 14 percent of intrahepatic cholangiocarcinoma patients.

In approving pemigatinib and infigratinib, the FDA also approved Foundation Medicine's next-generation sequencing FoundationOne CDx for detecting FGFR2 fusions or rearrangements in DNA extracted from patients' tumor tissue. According to the test's technical specifications, the NGS panel can detect fusions in FGFR1/2/3 genes through hybrid capture of specific intronic regions.

However, comparisons of DNA versus RNA-based NGS methods have revealed weaknesses of the former in detecting fusions. Published studies in lung cancer, for example, have demonstrated that RNAseq can capture fusions missed by DNA intron sequencing approaches.

As more FGFR inhibitors make their way through clinical investigations and enter the market in coming years, likely for a range of tumor types with FGFR alterations, Roychowdhury predicted that better liquid biopsy methods will be needed to detect and monitor these biomarkers, particularly FGFR fusions and rearrangements, since currently available liquid biopsies assessing DNA may exclude capture of highly repetitive intronic regions of genes where fusions may occur.

"A lot of the existing NGS approaches shy away from repetitive regions of the genome … [because if] you try to sequence them, you end up using excessive sequencing bandwidth, which is an issue for gene panels covering [between] 100 and 150 genes," he explained. "The problem is, if you have a repetitive region where a gene fusion happens … then you may miss that."

OSU researchers' interest in the second study is to learn more about how to interpret liquid biopsy tests. That trial, funded by Gateway for Cancer Research and launched earlier this month, will involve telemedicine-enabled enrollment of between 100 and 150 cholangiocarcinoma patients around the country and track the tumor mutational landscape in their blood using cell-free DNA testing. OSUCCC will partner with community practices in Ohio and other states to consent and enroll patients remotely using telemedicine and coordinate the collection of blood samples, which will be sent back to the academic cancer center for testing.

This study will "help us learn how to interpret liquid biopsy for FGFR [alterations] and cholangiocarcinoma, but also give us a stepping stool to think about how we turn this into a useful tool in a clinical trial setting and for clinical practice," Roychowdhury said.

He detailed the experience of a cholangiocarcinoma patient who had an FGFR fusion based on liquid biopsy and tissue testing. He was doing well on gemcitabine and cisplatin, but since he wasn't keen on receiving chemotherapy, in May 2019, his doctors at OSU switched him to the FGFR inhibitor infigratinib. His tumors quickly calcified and got smaller, and he kept responding to the targeted drug for more than a year. Unfortunately, in July 2020, his cancer returned, at which point his doctors decided to put him back on gemcitabine-cisplatin, "and we were pleasantly surprised that chemotherapy was working," Roychowdhury recalled.

Throughout this patient's treatment, Roychowdhury's team had been collecting blood samples to track how levels of tumor DNA changed. They found that the levels of the FGFR fusion went down with chemotherapy and became undetectable with targeted treatment and stayed that way for 18 months. Around July 2020, however, the levels of the FGFR fusion in blood tests shot back up and testing also detected five new resistance FGFR2 mutations.

Blood-based testing detected these resistance mutations driving different cancer cell clonal populations several months before imaging scans showed tumor progression. Ultimately, doctors decided to put the patient back on gemcitabine-cisplatin, and "what's exciting here, is that chemotherapy deleted the [resistant] clones" and the FGFR fusion levels also dropped, Roychowdhury said. This patient responded to chemotherapy for another year, but ultimately died not because he progressed on treatment, but due to cancer-related liver failure.

Still, this one patient's experience has given researchers an idea for another trial exploring the efficacy of cyclical therapy with chemo and an FGFR inhibitor. "We've seen that chemotherapy can be effective against subclone mutations that are resistant to the FGFR inhibitor," Roychowdhury reflected. His group is now leveraging that insight into an idea for a trial where patients receive alternating treatment with chemo and an FGFR inhibitor to see if the strategy forestalls resistance and limits cumulative side effects.

"We're brainstorming right now, and maybe we'll get one of our industry partners to help us fund it," he said.

In the Gateway-supported trial, Roychowdhury's group is also hoping to explore "liquid biopsy kinetics," which can help doctors optimize the best timeframe to test patients during treatment. The team wants to "determine how often and when to assess blood [for progression]," Roychowdhury said.

In the case of another FGFR2 fusion-positive cholangiocarcinoma patient at OSUCCC who is on an FGFR inhibitor, according to imaging scans, the patient's tumor is presently shrinking in response to treatment. But the schedule for this particular drug is three weeks of treatment and one week off. Weekly cfDNA testing of blood samples showed fusion levels dropping during the weeks he's on treatment but rising again during the fourth gap week.

If doctors decided to test this patient's blood to monitor response on that fourth week, they might get the false impression that he is progressing on therapy. "Whether it's a chemotherapy, a kinase inhibitor, or immunotherapy, we need to understand the kinetics of what happens in the blood, because if you look at the wrong time, you're going to misinterpret [the test data]," Roychowdhury said, noting that the particular blood kinetics observed in this patient may or may not be unique to FGFR inhibitors. "We need to partner with patients to study the kinetics of liquid biopsy and our telemedicine study supported by Gateway will help make this possible."

For Roychowdhury, the types of research underway at OSUCCC are critical to improving understanding of how to use liquid biopsy to advance precision oncology. "I'm not sure I'm ready to replace tumor biopsies yet. That's too soon," Roychowdhury said. "But [liquid biopsy testing] is giving me some insights."