Erdafitinib Demonstrates Manageable Toxicity, Early Efficacy in Glioma

Erdafitinib (Balversa) had a predictable toxicity profile and showed early efficacy signals in patients with recurrent or progressive IDH–wild-type glioma with FGFR-TACC (F3T3) gene fusions, according to findings from the safety run-in cohort of the phase 2 ETCTN 10559 trial (NCT05859334), presented at the 2025 Society for Neuro-Oncology Annual Meeting.¹ The agent also generated durable responses in this analysis.

The safety run-in cohort completed therapy in October 2024, and 8 mg daily continuous dosing was identified as the recommended phase 2 dose (RP2D) of erdafitinib. One patient experienced a dose-limiting toxicity (DLT), which was grade 3 central serous retinopathy; this was the only grade 3 treatment-emergent adverse effect (TEAE) reported and the only TEAE that led to erdafitinib discontinuation. Any TEAEs and grade 1 TEAEs were reported in all patients during the DLT period of cycle 1; no TEAEs were grade 4 or higher. Grade 2 TEAEs included dyspepsia (n = 2), hyperphosphatemia (n = 1), and hyponatremia (n = 1). Grade 1 hyperphosphatemia was common (n = 4). No TEAEs led to erdafitinib dose reductions or treatment interruptions. Notably, 1 patient experienced grade 3 cerebral edema that was deemed unrelated to treatment.

Preliminary efficacy data were available for 5 patients. Best overall responses were complete response (CR; n = 1), partial response (PR; n = 2), stable disease (n = 1), and progressive disease (n = 1).

“The safety profile of erdafitinib within gliomas is within [the] expected known safety profile [of the agent] in other tumor types, and we were able to identify durable responses in this population,” lead study author Macarena de la Fuente, MD, stated in the presentation.

de la Fuente is an associate professor of neuro-oncology, the chief of the Neuro-Oncology Division, clinical service leader for the Neuro-Oncology Service Line – Sylvester Comprehensive Cancer Center, chair of the Neuro-Oncology Site Disease Group, director of the Neuro-Oncology Fellowship Program, and leader of the Oncology Clinical Service for Neuro-Oncology at the University of Miami Miller School of Medicine in Florida.

Study Rationale

F3T3 gene fusions are the most prevalent gene fusions in adult glioma and are present in approximately 3% to 6% of adult patients with IDH–wild-type glioma. F3T3 fusions are truncal alterations that emerge during gliomagenesis and independently predict favorable outcomes in gliomas; these fusions are also retained in recurrent glioblastoma. Data from in vitro and in vivo studies have shown that F3T3 fusions have strong oncogenic activity and sensitivity to F3T3 inhibitors.

Erdafitinib, a potent, oral pan-FGFR tyrosine kinase inhibitor (TKI), was FDA approved in 2024 for the salvage treatment of patients with locally advanced or metastatic urothelial carcinoma harboring FGFR alterations, as determined by an FDA-approved test, whose disease has progressed during or after treatment with 1 or more prior lines of systemic therapy.² Responses to erdafitinib in patients with glioma have been reported in basket trials, but prior to the ETCTN 10559 study, no clinical trials focused solely on examining the activity of the agent in gliomas harboring F3T3 fusions.¹

ETCTN 10559 Trial Design

This multicenter, single-arm trial enrolled patients with recurrent or progressive F3T3 fusion–positive glioma. The trial was designed to have 2 safety lead-in cohorts. Cohort 1 (n = 6) evaluated the agent at a continuous dose of 8 mg daily. If at least 2 DLT events were observed, patients would be enrolled in the safety lead-in cohort 2 (n = 6) and receive erdafitinib at a continuous daily dose of 6 mg; if 1 or no DLT events were observed in either cohort, an additional 21 patients would be enrolled in the dose-expansion cohort. If at least 2 DLT events were observed at the 6-mg dose level, the trial would be stopped.

The currently reported cohort determined the safety, tolerability, and RP2D of erdafitinib at the 8-mg dose in this population.

Patients had a mean age of 63.3 years (standard deviation, 6.9) and a median age of 64 years (range, 52-72). Three patients were male, and 3 were female. Most patients were White (n = 5) and not Hispanic or Latino (n = 4). All patients had grade 4 glioblastoma and had received prior radiation and temozolomide.

Efficacy of Erdafitinib

de la Fuente highlighted results from 2 case studies from the safety run-in cohort. One patient was a 60-year-old woman with recurrent World Health Organization (WHO) grade 4 F3T3 fusion–positive glioblastoma. This patient underwent biopsy, then received radiotherapy plus temozolomide, followed by 6 cycles of adjuvant temozolomide. At 15 months from her initial diagnosis, this patient presented with clinical and radiographic disease. She enrolled in ETCTN 10559 in April 2024 and completed over 19 cycles of therapy. Treatment is ongoing in this patient, who achieved a PR at cycle 1 and a CR at cycle 13.

The second case study involved a 68-year-old man with IDH–wild-type, F3T3 fusion–positive, WHO grade 4 glioblastoma that was MGMT promoter unmethylated. This patient underwent subtotal resection of a right parieto-temporal mass, then received 3 weeks of hypofractionated radiotherapy with concurrent temozolomide, followed by 4 cycles of adjuvant temozolomide. He enrolled in ETCTN 10559 in October 2024 and achieved a PR at cycle 8.

Next Steps

The dose-expansion cohort of this phase 2 trial began enrollment in February 2025. In total, 90% of the planned 21 patients have been enrolled, and 12 clinical trial sites have been activated.

REFERENCES

1. de la Fuente M, Bhatia A, McFaline-Figueroa JR, et al. A phase 2 study of erdafitinib in patients with recurrent or progressive IDH-wild type glioma with FGFR-TACC gene fusions: safety run-in cohort results. Neuro-Oncology. 2025;27(suppl 5):v140-v141. doi:10.1093/neuonc/noaf201.0557

2. FDA approves erdafitinib for locally advanced or metastatic urothelial carcinoma. FDA. January 19, 2024. Accessed November 22, 2025. https://tinyurl.com/2y5c65za