Whole-genome and transcriptome landscape of actionable driver-negative lung adenocarcinoma.

Abstract

8523 Background: Driver-negative lung adenocarcinoma (DN-LUAD) without actionable genomic alterations (AGAs) has a poor prognosis. A deeper understanding of its molecular background, particularly across the entire genome, is crucial for improving risk assessment and therapeutic development. Methods: We performed deep whole-genome sequencing (WGS; tumor depth: 100–120x, normal depth: 30x, paired-end reads: 150 bp) and RNA sequencing using fresh-frozen tissues from four institutions. DN-LUAD was defined as LUAD without known actionable mutations or gene fusions. We analyzed somatic mutations, copy number alterations (CNAs), and structural variations (SVs). Tumors were classified as whole-genome doubling (WGD) if more than half of the autosomal tumor genome showed at least two copies in the major copy numbers of somatic cells. HRDetect score ≥0.7 was used to determine the presence of homologous recombination DNA repair deficiency (HRD) (Davies H, et al. Nat Med. 2017). Gene Set Enrichment Analysis (GSEA) was used for pathway analysis. Results: Among the 745 patients (pts), 517 were classified as having DN-LUAD. WGS identified AGAs undetected by whole-exome sequencing (WES) in 33 (4.4%) pts ( EGFR 13, KRAS 7, BRAF 5, ERBB2 4, MET 4, and HRAS 1). The frequently observed genomic alterations in DN-LUAD are shown in the Table. WGD was observed in 62.1% (n = 321) of DN-LUADs and was significantly associated with higher TMB, CNA, and SV burden. Tumor suppressor gene (TSG) mutations in TP53 , STK11 , and KEAP1 , as well as CDKN2A copy number loss and SVs, were significantly more frequent in WGD pts. HRD was identified in 19 DN-LUAD pts (3.7%), with a significantly higher frequency in the WGD group (WGD vs. without WGD; 4.98% vs. 1.5%, p = 0.016). GSEA showed significant (q < 0.0001) upregulation of cell cycle pathways (E2F targets, G2M checkpoint, and MYC targets) and downregulation of immune pathways (allograft rejection and interferon-gamma response) in DN-LUADs with WGD. Conclusions: This largest-ever WGS study identified AGAs undetectable by WES and uncovered a subgroup of DN-LUAD characterized by increased genomic instability driven by multiple TSG alterations associated with WGD, which was also more likely to exhibit HRD. WGD was associated with the upregulation of cell cycle pathways and the downregulation of immune pathways. These findings highlight the critical role of WGS in elucidating the pathogenesis of DN-LUAD. AllN = 517 WGDn = 321 Without WGDn = 196 p Median TMB (Mutations/Mb) 6.6 10.2 3.0 <0.0001 Median No. of CNAs 101 117 77 <0.0001 Median No. of SVs 185 244 103 <0.0001 TP53 381 (74) 277 (86) 104 (53) <0.0001 SMARCA4 59 (11) 43 (13) 16 (8) 0.094 STK11 56 (11) 42 (13) 14 (7) 0.049 CDKN2A 44 (9) 33 (10) 11 (6) 0.092 KEAP1 40 (8) 34 (11) 6 (3) 0.003 CDKN2A loss 106 (21) 78 (24) 28 (14) 0.009 MET amplification 44 (9) 34 (11) 10 (5) 0.045 CDKN2A 135 (26) 95 (30) 40 (20) 0.028 FHIT 125 (24) 99 (31) 26 (13) <0.0001