2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
Sameek Roychowdhury, MD, PhD, discusses the rationale for targeting FGFR mutations in cholangiocarcinoma and urothelial carcinoma, the potential efficacy of irreversible pan-FGFR inhibitors in these populations, and how continued research in this area may yield more personalized therapeutic approaches for patients with FGFR mutations.
Novel irreversibly binding FGFR inhibitors on the horizon in cholangiocarcinoma and urothelial carcinoma may be key for overcoming resistance to first-generation agents, according to Sameek Roychowdhury, MD, PhD.
One small molecule pan-FGFR inhibitor currently under investigation is KIN-3248. A phase 1 trial (NCT05242822) is ongoing to investigate the agent’s safety, tolerability, pharmacokinetics, and preliminary efficacy in adult patients with intrahepatic cholangiocarcinoma, urothelial carcinoma, and other solid tumors harboring FGFR2 and/or FGFR3 gene alterations.1
“Hopefully, we can understand which mutations and genes can benefit from these novel drugs and bring them forward for those patients,” Roychowdhury said.
In an interview with OncLive®, Roychowdhury discussed the rationale for targeting FGFR mutations in cholangiocarcinoma and urothelial carcinoma, the potential efficacy of irreversible pan-FGFR inhibitors in these populations, and how continued research in this area may yield more personalized therapeutic approaches for patients with FGFR mutations.
Roychowdhury is an assistant professor in the Department of Internal Medicine and the Department of Pharmacology at The Ohio State University and a member of the Translational Therapeutics Program at The Ohio State University Comprehensive Cancer Center–James in Columbus.
Roychowdhury: The gain-of-function activating mutations in FGFR receptors are common in intrahepatic cholangiocarcinoma and urothelial or bladder cancer. In intrahepatic cholangiocarcinoma, we tend to see more of the FGFR2 variety, in anywhere from 10% to 15% of that subset. In bladder cancer, we tend to see more FGFR3, in around 15% to 20% of that subset.
In bladder cancer, we tend to see more of point mutations, or single nucleotide changes, whereas in cholangiocarcinoma, we tend to see more fusions or rearrangements of FGFR2. However, for both cancers, we can see other varieties of FGFR receptor type, as well as the type of mutation that could be gain-of-function.
Finding gain-of-function activating FGFR alterations, whether they are fusions or point mutations in cholangiocarcinoma and bladder cancer, helps us identify a subset of patients that is likely to benefit from an FGFR-targeted kinase inhibitor. Patients who don’t have these FGFR activating mutations are probably not going to benefit at all. Only the subset of patients [with these mutations is likely to benefit].
In cholangiocarcinoma, anywhere from 70% to 80% of patients have clinical benefit; whether it’s a partial response [PR] or stable disease [SD], they’re experiencing disease control and clinical benefit. For bladder cancer, we’re seeing a benefit rate of around 40%, with patients having PRs and SD.
When this defined group of patients receives FGFR kinase inhibitors as targeted therapy, we unfortunately see progression of their cancer and acquired resistance. We are beginning to see secondary mutations, more so in the patients with cholangiocarcinoma than in the patients with bladder cancer. That is 1 class of resistance mechanism.
In each instance, for cholangiocarcinoma, or liver cancer, and bladder cancer, the average time to progression is around 6 months. Some patients may progress before [6 months], at around 6 months, or sometimes [after 6 months], having longer duration of benefit. The mutations tend to interfere with how the drug interacts with the kinase part of the receptor.
We’ve also seen other mechanisms, mutations, genes, and pathways that tend to [activate] downstream of the receptor. We often see evidence of MAPK or PI3K activation downstream after FGFR resistance develops.
The first-generation inhibitors of FGFR kinases are irreversibly binding. Some FGFR kinase inhibitors are being studied or will soon be studied in clinical trials. Some of these have covalent binding, meaning once they bind with the receptor, that bond cannot be broken. This potentially allows for greater inhibition of the kinase and target, preventing it from cycling on and off, and could be beneficial as a means of overcoming resistance to the other first-generation drugs. We’ll be excited to see the current and upcoming data for some of the irreversibly binding drugs.
We’re also beginning to see other drugs that have a new design. We are considering the known acquired resistance mutations and finding a way for the new class of drugs to still bind and inhibit the enzyme and receptor activation. Those are potentially opportunities for new therapies and hope for these patients. We’re looking forward to offering these agents to patients.
The unique eligibility criteria for some of the current FGFR inhibitors that are in clinical trials, aside from the usual criteria of the patient’s wellness, are about molecular eligibility. Patients need to have evidence, usually through genetic testing, of an activating gain-of-function FGFR mutation, including gene fusions and point mutations.
We’re now beginning to understand that there are additional gain-of-function mutations, such as the extracellular domain mutations, that are small deletions. Some of these are known because of what we’ve learned from other hereditary disorders involving FGFR, where there’s gain-of-function mutations in the extracellular domain, as well as in the transmembrane domain. The current studies can include all these patients.
As we begin to see second-generation drugs for FGFR, one of our challenges and opportunities will be to sort out which patient needs which drug. We could be in a [position to choose] which therapy might benefit one patient with one resistance mutation, vs another patient who has a different resistance profile.
In gastrointestinal stromal cell tumors, as well as in chronic myeloid leukemia, when secondary mutations arise, we can select the appropriate kinase inhibitor or other second-class drug. Right now, we’re talking about kinase inhibitors. But there are potentially other ways to target FGFR, for example, antibody-mediated approaches. Having more therapies in the pipeline is always beneficial to our patients, and [we need] more choices so we can take care of them in a better way.
We’ve learned that though cholangiocarcinoma and bladder cancer have these gain-of-function FGFR alterations, other cancer types harbor similar FGFR gain-of-function activating mutations. The challenge is, they’re not as common. Working together with the pharmaceutical industry and their efforts for FGFR [is important for determining] how we make sure we reach these patients and identify which patient subsets are going to respond and benefit.
In my experience, we’ve been lucky enough to treat many patients with pancreatic cancer with FGFR fusions. Compared with [patients with] cholangiocarcinoma, they’ve done quite well, with astounding responses that are far more durable than what we see in cholangiocarcinoma. However, [these responses are] quite rare. That’ll be one of our upcoming challenges for the field of FGFR drug development: How do we advance beyond cholangiocarcinoma and urothelial carcinoma? Certainly, we’re bringing new generations of drugs to those 2 diseases, but how do we make sure we can reach all patients who can benefit?
Pay close attention to FGFR receptor alterations. There are some known gain-of-function fusions and mutations, but we’re beginning to see that there are other gain-of-function mutations that are less common. Whenever you see an FGFR mutation, look closely to see if you can get [that patient] matched to a clinical trial that accepts that mutation or type of mutation. Keep an eye out for these patients.