Abstract 6080: The heterogeneity and evolution of lung neuroendocrine tumors within TRACERx

Abstract Introduction: Primary pulmonary neuroendocrine tumors (NETs) contribute to 20-25% of all lung cancer diagnoses and lie on a spectrum of malignant behavior, ranging from low grade classical carcinoid tumors to high grade, aggressive small cell carcinomas. The evolutionary relationships between NETs of different grades remain poorly understood, here we seek to use the TRACERx study to explore the genomic evolution of these lung cancers. Methods: Tissue and blood samples were taken from a cohort of 20 patients with pulmonary NETs in a sub-study within TRACERx. Multi-region sampling (84 total regions) and whole exome sequencing was performed on fresh frozen tissue from each tumor. Histological classification of each tumor region was confirmed and scoring of tumor infiltrating lymphocytes (TILs) was performed. The data were then analyzed using the established TRACERx bioinformatics pipeline. Results: Phylogenetic evolutionary histories were generated, illustrating the genomic divergence in each tumor and identifying points at which specific known drivers were acquired during tumor evolution. Lower grade classical and atypical carcinoids were characterized by a low mutational and copy number burden with a high fraction of subclonal mutations, with 5/12 cases having no apparent somatic mutational driver mutation. In contrast higher grade large cell neuroendocrine carcinomas and small cell carcinomas had higher mutational and copy number burden, as well as higher rates of genome instability and whole genome doubling. We identified no evolutionary link between subtypes such as commonly shared clonal drivers, suggesting that atypical or classical carcinoids later evolve into higher grade carcinomas. Dn/ds was used to investigate for evidence of selection in the carcinoid tumors; while there was no evidence for selection in known cancer driver genes associated with NETs or lung cancer, there was a high global dn/ds value suggesting the existence of unknown drivers for this cancer type. Furthermore, there was evidence of parallel evolution in one atypical carcinoid case with two distinct subclonal MEN1 frameshift mutations co-occurring with loss of the other allele. One small cell carcinoma patient was identified as having two genetically unrelated tumors. A germline predisposition for this tumor was suspected since we identified a variant in BRCA2 (rs11571833) that was maintained under loss of heterozygosity in one tumor with a corresponding somatic BRCA2 mutation in the other, suggesting two independent double hit events. Conclusions: Genomic differences are identifiable between different members of the pulmonary NET family, with high grade tumors showing increased whole genome instability indices and phylogenetic divergence than low grade tumors. This work has shed light on the genomic landscape of pulmonary NETs and enabled the identification of distinct evolutionary trajectories in the different subtypes. Citation Format: Robert Bentham, David A. Moore, Lorelle Brownlee, Raju Veeriah, Lizi Manzano, Fiona Blackhall, Colin R. Lindsay, TRACERx Consortium, Nicholas McGranahan, Charles Swanton, Mariam Jamal-Hanjani. The heterogeneity and evolution of lung neuroendocrine tumors within TRACERx [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 6080.