BACKGROUND – Lung cancer is the most common cause of cancer-related death globally. Alterations within FGFR genes in de novo mNSCLC have been described, but the data is still limited. FGFR fusions are thought to be rare, but present as an acquired resistance mechanism for EGFR TKIs. These fusions may be actionable with FGFR inhibitors, which could in turn provide another therapeutic target upon EGFR TKI resistance. However, the overall prevalence and genomic landscape of FGFR alterations in mNSCLC patients—as well as in patients who received EGFR TKI treatment—is largely unknown.
METHODS – We used the Tempus database to retrospectively analyze deidentified records from 6,973 adult patients diagnosed with mNSCLC. Patients who underwent tissue-based next generation sequencing (NGS) via the Tempus xT assay (DNAseq of 595-648 genes, whole exome capture RNA-seq) within 30 days of diagnosis were assessed for prevalence of EGFR and FGFR1-4 alterations, including: pathogenic/likely pathogenic single nucleotide variants (SNVs), deep copy number losses or amplifications >=8, and fusions. An additional 413 patients with mNSCLC who underwent Tempus xT sequencing after being treated with an EGFR-directed TKI (afatinib, dacomitinib, erlotinib, gefitinib, or osimertinib) for at least 6 months were also identified and the prevalence of FGFR alterations was similarly assessed. All analyses were performed in R version 4.2.0
RESULTS – Of the 6,973 mNSCLC patients in our cohort, 942 (14%, 95% CI 13-14%) were found to have an FGFR alteration: 380 (5.4%) had an alteration in FGFR2 only, 270 (3.9%) in FGFR3 only, 178 (2.6%) in FGFR1 only and 17 (0.2%) in FGFR4 only. A further 97 (1.4%) patients had alterations in multiple FGFR genes. Of 413 mNSCLC patients who underwent sequencing after being treated with an EGFR-directed TKI for at least 6 months, 55 (13%, 95% CI 10-17%) were found to have an FGFR alteration. Of these, 23 (5.6%) had an alteration in FGFR2, 15 (3.6%) in FGFR3, 10 (2.4%) in FGFR1, and 1 (0.2%) in FGFR4. Alterations in multiple FGFR genes were observed for 6 (1.5%) patients. Of the 55 patients in the FGFR-altered population, 33 (60%) were treated with osimertinib, 6 (11%) with erlotinib, 2 (3.6%) with afatinib, 1 (1.8%) with gefitinib, and 13 (24%) were treated with more than one of these TKIs.
CONCLUSION – To our knowledge, this is the first detailed descriptive analysis using real world evidence to provide insight into the prevalence of FGFR alterations in mNSCLC as well as the prevalence of FGFR alterations within a population treated with prior EGFR TKIs. In both cohorts, we observed substantial prevalence of FGFR alterations (13-14%) suggesting that this FGFR altered population merits further inquiry, particularly to understand whether targeting FGFR alterations in both scenarios can provide therapeutic benefit.