The Food and Drug Administration on Wednesday approved lirafugratinib, sold as Lyrfigtu, for adults with a genetically defined form of advanced bile duct cancer after prior treatment. The decision creates another targeted option for patients whose unresectable, locally advanced or metastatic cholangiocarcinoma carries an FGFR2 gene fusion or rearrangement—a molecular feature that must be found through tumor testing before the drug can be considered.

The approval rests on 116 patients in the REFOCUS study who had not previously received an FGFR inhibitor. According to the FDA, 46% had a measurable tumor response, with a 95% confidence interval of 36% to 55%. Among patients who responded, the median response lasted 11.8 months. Those results are clinically meaningful in a rare, difficult-to-treat cancer, but the evidence has important limits: REFOCUS was open-label and single-arm, so it did not compare lirafugratinib directly with another drug or establish that it prolongs survival.

A narrow approval built around a tumor biomarker

Cholangiocarcinoma begins in the bile ducts, the network of tubes that carries bile from the liver and gallbladder toward the small intestine. The Cancer Institute describes it as a rare cancer with intrahepatic and extrahepatic forms and no routine screening test for people without symptoms. That often leaves treatment decisions to be made after disease is locally advanced or has spread.

Lirafugratinib is not approved for every patient with bile duct cancer. Its label is confined to adults whose tumors harbor an FGFR2 fusion or other rearrangement and whose disease has already been treated. FGFR2 encodes a receptor involved in cell growth and division; altered signaling can help sustain cancer. The National Library explains that FGFR2 normally transmits growth-factor signals across the cell membrane, while abnormal activity can promote cancer-related functions including proliferation and blood-vessel formation.

The drug is an oral, highly selective inhibitor designed to block FGFR2. Preclinical and early clinical work published in Cancer Discovery showed why selectivity matters: earlier FGFR medicines inhibit several related receptors, which can produce off-target effects, while lirafugratinib was engineered to bind FGFR2 more specifically. Selectivity does not make the medicine free of toxicity, and it does not mean every biomarker-positive tumor will respond. It does, however, explain the rationale for matching the medicine to the alteration.

What the REFOCUS data show

REFOCUS, registered as NCT04526106, is a multicenter phase 1/2 program evaluating lirafugratinib in advanced cancers with FGFR2 alterations. For the FDA efficacy population, the agency reviewed 116 patients with unresectable or metastatic cholangiocarcinoma who had received chemotherapy or chemoimmunotherapy but had never taken an FGFR inhibitor. They received 70 milligrams by mouth once daily until their disease progressed or side effects became unacceptable.

An independent review committee assessed scans using standard RECIST criteria. The 46% objective response rate means nearly half the study population had tumors shrink by the protocol-defined amount; it does not mean 46% were cured. The 11.8-month median duration of response means half of responding patients maintained a response longer and half for less time. An updated JCO abstract presented the same trial as an open-label phase 1/2 study, reinforcing that the headline numbers come from a nonrandomized cohort rather than a controlled comparison.

That design is common in rare, biomarker-defined cancers where recruiting a large randomized population can be difficult. It can demonstrate that tumors shrink and that responses persist, especially when an independent committee reviews imaging. It cannot by itself separate the drug’s effect from patient selection, determine whether patients live longer than they would with another available therapy, or provide a reliable head-to-head ranking against other FGFR inhibitors. Longer follow-up and comparative evidence will be needed to resolve those questions.

Existing evidence also shows why resistance remains part of the story. A 2026 laboratory study examining clinical resistance found that secondary FGFR2 mutations can emerge during treatment, although lirafugratinib retained activity against several mutations associated with resistance to broader FGFR inhibitors. That work helps explain the drug’s design but is not proof that it prevents resistance in patients. Cancer can evolve through multiple pathways, and clinical benefit may differ from what molecular experiments predict.

Safety monitoring is central, not optional

The prescribing information carries warnings for eye toxicity, high blood phosphate levels with soft-tissue mineralization, and harm to a fetus. The manufacturer’s detailed safety summary, which reflects the approved labeling, says retinal pigment epithelial detachment occurred in 31% of 385 patients exposed across studies, including grade 3 events in 1.8%. Dry eye occurred in 38%, blurred vision in 18%, and corneal toxicity or keratitis in 11%.

The label therefore calls for a comprehensive eye examination, including optical coherence tomography of the macula, before treatment, every two months during the first 13 months and every four months thereafter. New visual symptoms require urgent evaluation. Depending on severity, clinicians may interrupt treatment, reduce the dose or stop it.

High phosphate levels occurred in 21% of exposed patients, with a median onset of 15 days. Because sustained hyperphosphatemia can contribute to mineral deposits in soft tissue and blood vessels, serum phosphate must be monitored throughout therapy and managed as appropriate. Serious adverse reactions occurred in 32% of patients in the manufacturer’s safety population; infection, pneumonia, fatigue and hemorrhage were the serious events reported in at least 2%. One fatal hemorrhage was reported. These figures describe the combined safety database, not just the 116-patient efficacy cohort.

Testing now has a larger practical consequence

The approval increases the value of performing sufficiently broad molecular testing early enough that results are available when first-line treatment stops working. Targeted therapy depends on finding the relevant biomarker; the Cancer Society notes that biomarker testing is often needed to determine whether a cancer may respond to a targeted medicine. In practice, that can require adequate tumor tissue, validated sequencing and coordinated review by pathology and oncology teams.

Lirafugratinib also enters a field with existing FGFR-directed options, including pemigatinib and futibatinib for previously treated FGFR2-positive intrahepatic cholangiocarcinoma. In a phase 2 study published in the medical journal, futibatinib produced durable responses in a similarly selected population. Cross-trial comparisons are unreliable because eligibility rules, patient characteristics, follow-up and assessment methods differ. The new approval should therefore be viewed as an added choice, not evidence that one FGFR inhibitor is universally superior.

Access will depend on more than the FDA decision. Laboratories must detect and accurately report qualifying fusions or rearrangements; oncology teams must interpret the result in the context of prior therapy, organ function, drug interactions and the patient’s goals; and payers must establish coverage. The need for serial eye examinations and laboratory monitoring also adds logistical demands, especially for people traveling to specialty cancer centers.

A meaningful option with unanswered questions

For patients who match the label, lirafugratinib offers an oral treatment backed by a substantial response rate and responses lasting a median of nearly a year. The FDA also granted priority review, breakthrough therapy and orphan-drug designations, reflecting the seriousness and rarity of the condition and the agency’s interest in speeding review of promising therapies.

But the approval does not settle the larger clinical questions. REFOCUS did not randomize patients, did not establish an overall-survival advantage and excluded people previously treated with another FGFR inhibitor from the pivotal efficacy group. Evidence is still needed on sequencing after other targeted medicines, long-term safety, resistance patterns, quality of life and which patients benefit most.

The immediate system implication is clearer: molecular characterization is no longer an academic add-on in advanced cholangiocarcinoma. It can determine whether a patient is eligible for a treatment that would otherwise never enter the discussion. The FDA’s decision expands that treatment menu, while the study design and safety profile argue for disciplined interpretation, careful monitoring and candid conversations about both promise and uncertainty.