Abstract 3458: Resistance to selective FGFR inhibitors in FGFR-driven urothelial cancer

Abstract Background Several selective FGFR inhibitors have been developed for the treatment of FGFR-driven urothelial cancer, in particular erdafitinib which received FDA and EMA approval. Molecular mechanisms responsible for resistance to FGFR inhibitors have not been elucidated in urothelial cancer. Understanding these resistance mechanisms is crucial for the development of novel agents and treatment strategies. Methods We identified patients with metastatic urothelial cancer harboring FGFR mutations or rearrangements, treated with selective FGFR inhibitors in the MATCH-R (NCT02517892), MOSCATO (NCT01566019), STING-UNLOCK (NCT04932525), and CTC studies at Gustave Roussy. Post-progression tissue biopsies and circulating tumor DNA (ctDNA) were analyzed with whole exome sequencing and Illumina or FoundationOne assays, respectively. We generated Ba/F3 models containing FGFR3::TACC3 transcript with the FGFR3 kinase domain mutations detected in post-progression samples and exposed them to eight FGFR inhibitors (readouts: cell viability assays and Western blots). We established patient-derived xenografts (PDX) and cell lines to assess resistance mechanisms and perform pharmacological studies. Results Twenty-one patients with post-progression samples (n = 13 both tissue and ctDNA, n = 5 ctDNA only, n = 3 tissue only) were included in our study. The molecular driver was identified as FGFR3 mutations/fusions in 19 cases, whereas one FGFR4 mutation and one FGFR2 rearrangement were found in two patients. Patients progressed to erdafitinib (n = 14), futibatinib (n = 4) or pemigatinib (n = 3). Thirteen, six, and two patients experienced response, stable disease, and progression, respectively. At progression, we detected single mutations in the FGFR kinase domain in six (29%) patients (FGFR3 N540K, V553L, V553M, V555L, V555M, E587Q; FGFR2 L551F) and multiple mutations in one (5%) case (FGFR3 N540K, V555L, L608V). Using Ba/F3 cells, we confirmed that the detected mutations conferred resistance to FGFR inhibitors. FGFR3 V553L, emerging in a patient treated with pemigatinib, was sensitive to erdafitinib and futibatinib. We detected alterations in the mTOR pathway in 11 (52%) patients (n = 5 TSC1/2, n = 5 PIK3CA, n = 1 NF2, n = 1 PTEN), with or without FGFR3 kinase domain mutations. In patient-derived models, erdafitinib combined with pictilisib was synergic in the presence of PIK3CA E545K, while erdafitinib-gefitinib combination overcame bypass resistance mediated by EGFR activation. Conclusions We detected a high frequency of FGFR kinase domain mutations at progression to selective FGFR inhibitors in urothelial cancer. Unlike FGFR2-driven cholangiocarcinoma, polyclonality of FGFR3 kinase domain mutations does not seem to be a hallmark of FGFR-driven urothelial cancer. The identification of off-target events and the preclinical validation sustain the potential clinical applicability of combinatorial treatment strategies. Citation Format: Francesco Facchinetti, Antoine Hollebecque, Floriane Braye, Damien Vasseur, Yoann Pradat, Rastislav Bahleda, Cédric Pobel, Ludovic Bigot, Olivier Déas, Juan-David Florez-Arango, Giorgia Guaitoli, Hayato Mizuta, David Combarel, Lambros Tselikas, Stefan Michiels, Sergey Nikolaev, Jean-Yves Scoazec, Santiago Ponce-Aix, Benjamin Besse, Ken A. Olaussen, Yohann Loriot, Luc Friboulet. Resistance to selective FGFR inhibitors in FGFR-driven urothelial cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 3458.