The genomic landscape of advanced gastrointestinal (GI) neuroendocrine tumors (NET) and carcinomas (NEC).

656 Background: The impact of specific genomic alterations in advanced GI neuroendocrine malignancies on prognosis and treatment decisions has not been fully characterized. In this study, we seek to evaluate next-generation sequencing (NGS) panels and corresponding clinical outcomes in patients with metastatic GI NET/NEC. Methods: Patients with GI-NET/NEC were identified from a database of NGS reports performed at the Cleveland Clinic, and clinical variables were extracted by chart review through an IRB-approved protocol. We provide an analysis of this cohort. Results: Of a total n=45 patients, 53% were male. Median age at diagnosis was 60 (range: 35-90 years) and 82% were non-Hispanic white, while 9% were black. There were n=25 GI-NETs, mostly intermediate or high grade (11 G3, 7 G2), and n=20 GI-NECs. There was a balanced distribution of primary sites, including colorectal (27%), pancreas (24%), small intestine (18%), unknown primary (24%) and a single case of biliary primary. High grade GI-NETs and GI-NECs were over-represented in the colorectal primary group vs other sites. (n=10 of 12; p=0.021). The majority of patients (73%) had at least one somatic variant, including 12 patients (27%) with one or more mutations deemed actionable, but only one patient (with MSH2 mutation) has received a mutation-matched drug (pembrolizumab). Some recurrent alterations appeared to co-segregate with aggressive histology (NEC vs NET), including 12 of 13 TP53 mutations (p<0.05 by Fisher Exact Test), 7 of 8 RB1 (p<0.05), and 5 of 6 APC (p=0.074). The most common mutations per person based on primary site is shown. Microsatellite instability high was rare (n=1) and median tumor mutational burden (TMB) was 3/mb, with TMB lower in NETs vs NECs. A number of patients (35%) underwent germline testing, with 6 of 16 having a pathogenic germline variant. Conclusions: As 42% of the 12 patients with actionable mutations were alive at time of analysis, it is possible that additional patients could be treated with mutation-matched therapies. More input will be needed from providers as to the role played by alternative effective treatments and barriers such as insurance coverage and prior authorization to implementing mutation-matched therapies. Given the overall low TMB of GI-NETs, with several tumors having zero or 1 reported mutation on targeted NGS, whole genome sequencing may prove useful. Finally, certain mutations such as RB1 and TP53 were associated with more aggressive histology (i.e. NEC). Lastly, only 9% of our cohort with NGS were black, a significant under-representation based on epidemiological data, which raises concerns about disparate accessibility and barriers to NGS testing. Further studies are needed to address any possible racial disparity.[Table: see text]