Targetable fusions detected using FMI DNA CGP are associated with improved clinical benefit from matched therapy compared to fusions detected using other testing methods. (A) Distribution of orthogonal testing modalities (N = 5879 tests; left axis, gray bars) and results (right axis, teal dots) from 3959 patients with NonSqNSCLC who were assessed for progression and underwent both FMI DNA CGP and additional fusion testing. Each testing modality is only counted once per patient. However, a single patient could be counted toward multiple modalities if a patient underwent multiple types of orthogonal testing such that the sum of all bars may exceed 100%. Ns above bars indicate the number of patients who underwent each type of testing. The fusion-positive percentage was calculated based on the total number of tests performed in each category (N = 7299 tests across all modalities), even if multiples of the same testing modality were performed for a single patient. (B) Forest plot depicting the adjusted rwPFS HR of receiving a matched TKI in the 1L for NonsqNSCLC subcohorts defined by concordance/discordance between FMI DNA CGP and orthogonal fusion testing results. (C) Among patients who received 1L-matched TKI, patients who were negative for ALK, NTRK, RET, and ROS1 fusions on FMI DNA CGP had shorter rwPFS than patients who were positive by FMI DNA CGP. Unadjusted Kaplan-Meier plot is shown. All analyses are indexed to the start of 1L therapy. In addition to univariable Cox model HRs, adjusted HRs are presented for a multivariable Cox model that includes established prognostic variables (Supplementary Fig. S6). A small number of cases were excluded from multivariable analyses due to missingness of covariables. Abbreviations: FMI DNA CGP, Foundation Medicine Tissue DNA comprehensive genomic profiling; HR, hazard ratio; TKI, tyrosine kinase inhibitor; rwPFS, real-world progression-free survival; 1L, first advanced line of Tx.