ctDNA monitoring of FGFR3-altered metastatic urothelial cancer treated with dabogratinib (formerly TYRA-300) in the SURF301 trial.

809 Background: Activating FGFR3 mutations and fusions occur in 10-20% of patients with metastatic urothelial carcinoma (mUC). Erdafitinib, a pan-FGFR inhibitor, is an approved targeted therapy for FGFR3 -altered mUC. However, low isoform specificity can lead to side effects, and dose reductions or interruptions occur due to inhibition of FGFR1/2/4. Dabogratinib is a highly selective inhibitor of FGFR3 with potentially less off-target toxicity relative to pan-FGFR inhibitors and has shown preclinical activity against FGFR3 gatekeeper mutations. SURF301 (NCT05544552) is a phase 1/2 study in solid tumors testing oral dabogratinib, with a previously reported 55% (6/11) of patients achieving confirmed partial responses at doses of 90mg QD or above. Here, we report phase 1 results on early circulating tumor DNA (ctDNA) dynamics in FGFR3 -altered patients with mUC enrolled in SURF301. Methods: Blood was collected at baseline, cycle 1 day 15 (C1D15), C2D1 and at each disease assessment visit (every 3 months). Plasma cell-free DNA (cfDNA) was sequenced using a custom targeted panel of 78 mUC genes. Somatic variants were called via an established bioinformatic pipeline, leveraging same-patient serial samples to boost detection sensitivity. The proportion of tumor-derived cfDNA was estimated using orthogonal approaches, and evaluated as an absolute (MTM/mL) and relative measure (ctDNA fraction). On-treatment ctDNA changes from baseline were calculated and statistical significance assessed using Fisher’s exact test and Bayesian posterior estimates derived from Beta–binomial models with Jeffreys priors. Results: Twenty-one mUC patients provided blood at baseline and on-treatment (n=72 samples). Median age was 68 (range 53-84) and 57% had 3 or more prior lines of therapy. Eleven patients received dabogratinib at ≥90mg QD, while 10 received 40-60mg QD. All FGFR3 variants (9 mutations, 7 TACC3 fusions, 1 both) detected in tumor tissue were also identified in baseline ctDNA (when ctDNA was detected; 17/21). On-treatment ctDNA fraction showed significant decreases as early as C1D15 across all dose groups, with 6 patients having reductions to below the level of quantification (ctDNA clearance) including 1 patient treated at 40mg QD. All patients in the higher dose groups had reductions or no baseline ctDNA. Conclusions: We observed frequent on-treatment reductions in plasma ctDNA in FGFR3 -altered mUC patients receiving dabogratinib in the SURF301 trial. Molecular responses, suggestive of anti–tumor effect and adequate target coverage, were seen at all doses, with the deepest molecular responses occurring at doses of ≥90mg QD, in alignment with reported pharmacokinetic and pharmacodynamic evidence of drug activity. Clinical trial information: NCT05544552 .