Background: There are no data to guide the selection of NGS gene panels for clinicians.

Methods: We compared two commercially available NGS gene panels of different size with respect to their clinical utility as quantified by the number of detectable actionable alterations in various solid tumor types based on a systematic review of the literature.

Results: We classified as clinically actionable 86% of the genes on the Ion AmpliSeq Cancer Hotspot PanelÒ covering the hotspots in 50 genes (Panel 50), and 50.3% of the genes in the FoundationOne PanelÒ covering the complete exons of 315 genes (Panel 315). We have identified in total 145 compounds of which 99 were FDA approved, 14 were tested in active phase I/II, 21 in phase II, 1 in phase II/III and 10 in phase III clinical trials. The matched targeted therapies that could be proposed based on the use of the Panel 50 or Panel 315, respectively, were FDA-approved drugs with on- label (42.0% vs 18.9 %) and off-label use (10% vs 6.9%) or experimental (34.0% vs 24.3%) molecules. The number of actionable alterations in various solid tumor types using the Panel 315 was in median 50% higher, compared to the Panel 50 (T-test, p¼ 1.72 x 10-5 ) (Table 7P). This gain was mainly attributed to the higher number of genes related to homologous recombination repair deficiency on the Panel 315 (21 genes vs 1 gene), which can be targeted by PARP inhibitors and the presence of genes asso- ciated with microsatellite instability and response to immune checkpoint inhibitors (8 vs 1 gene).

Conclusions: Our results suggest a substantial gain in actionability with FDA-approved

or experimental drugs using the larger gene panel.