Landscape of targetable alterations discovered by next generation sequencing demonstrates the role a community hospital can play in precision-guided oncology: Experience from Lenox Hill Hospital.

e19279 Background: Next generation sequencing (NGS) has become standard of care in aiding diagnosis and treatment of advanced solid cancers, and in conducting clinical trials at large centers. It is imperative that this is expanded to other hospitals that care for the majority of cancer patients. We analyzed NGS reports from our hospital to assess the number of patients who can benefit from approved or investigational targeted therapies. Methods: We analyzed NGS data for 511 solid tumor samples sequenced between January 1, 2018 and December 31, 2019. NGS was performed by GenPath Dx (77%), Caris Life Sciences (16%) or Foundation Medicine (7%). Results: The majority of samples represented advanced stage malignancies and 21% were stage I or II. They originated from the following primaries: 195 lung (37%), 150 gastrointestinal (29%), 64 primary brain (12%), 20 gynecological (4%), 14 skin (3%), 13 head and neck (3%) and 11 sarcomas (1%). 72 samples had mutations in EGFR (14%), 170 in TP53 (33%), 124 in KRAS (24%), 28 in BRAF (6%), and 3 in RET (0.6%). 34/107 (32%) early stage samples harbored an actionable mutation (20 EGFR mutations, 1 MET exon 14 skipping, 4 KRAS G12C, 7 BRAF V600E, 1 FGFR3 amplification and 1 CD74/NRG1 fusion). In contrast, only 58/317 (18%) advanced stage samples had a targetable mutation (p-value = 0.02, χ² test). 76% of EGFR-mutated samples were lung adenocarcinomas, and 19% were primary brain tumors; 54% of these are targetable by FDA-approved EGFR inhibitors. KRAS mutations were found in gastrointestinal (54%), lung (37%) and pancreatic (5%) malignancies. 17 patients had a KRAS G12C mutation and, therefore, could benefit from one of the KRAS G12C inhibitors in early clinical trials. Four samples harbored crizotinib-sensitive mutations (2 MET amplifications and 2 MET exon 14 skipping mutations). Regarding gene fusions, one glioblastoma sample had a PTPRZ-MET fusion and one lung adenocarcinoma sample harbored a CD74-NRG1 fusion. Conclusions: We identified 92/511 samples (18%) with clinically actionable mutations; distributed in 32% early stage and 18% advanced stage disease, indicating that actionable mutations are present at an increased frequency in early stage solid malignancies in our data set and trials to investigate targeted therapy in such settings should be considered. Furthermore, we show that a community-based hospital can be a site for future clinical trials of small molecule inhibitors and bring precision-guided medicine to additional patients.