Mapping the hallmarks of lung adenocarcinoma with massively parallel sequencing

PMID: 22980975 · DOI: 10.1016/j.cell.2012.08.029 · Journal: Cell (2012)

TL;DR

This study performed whole-exome sequencing (WES) of 183 and whole-genome sequencing (WGS) of 24 lung adenocarcinoma tumor/normal pairs, identifying 25 significantly mutated genes. Beyond confirming known drivers (TP53, KRAS, EGFR, STK11, KEAP1, NF1, BRAF), the study nominated novel recurrently mutated genes U2AF1, RBM10, and ARID1A, discovered a novel oncogenic in-frame EGFR C-terminal deletion, and proposed “Epigenetic and RNA deregulation” as a potential new cancer hallmark.

Cohort & data

• 183 LUAD tumor/normal pairs from the luad_broad cohort.
• 159 cases with WES, 24 with WGS (23 had both).
• Median exome fold coverage: 92x (range 51-201); median genome coverage: 69x tumor, 36x normal.
• Clinical composition: 27 never smokers, 17 light smokers, 118 heavy smokers, 21 unknown; stages I-IV.
• Complementary SNP array analysis for somatic copy number alterations.

Key findings

• Mean exonic somatic mutation rate of 12.0 mutations/Mb (median 8.1/Mb; range 0.04-117.4/Mb).
• Smokers had significantly higher mutation rates than never smokers (median 9.8/Mb vs 1.7/Mb; P = 3.0 x 10^-9, Wilcoxon rank sum).
• InVEx/MutSig analysis identified 25 significantly mutated genes (q < 0.25).
• U2AF1 p.S34F hotspot mutation found in 3% of cases (P = 2.0 x 10^-6) — first report in an epithelial tumor; associated with reduced progression-free survival (P = 0.00011, log-rank).
• RBM10 mutated in 7% of cases with recurrent truncating mutations (4% nonsense/frameshift/splice).
• ARID1A mutated in 8% of cases with significant accumulation of truncating mutations (P = 0.027, LOF InVEx).
• Five mutation spectrum clusters identified correlating with smoking history: Cluster 1 enriched in never/light smokers (P = 1.4 x 10^-9); Cluster 4 enriched in advanced stage (P = 0.0063).
• EGFR mutation anti-correlated with KRAS mutation (P = 3.3 x 10^-4) and correlated with never/light smoker status (P = 2.0 x 10^-6).
• Novel in-frame EGFR C-terminal deletion (exons 25-26, vIVb variant) identified by WGS; functionally validated as oncogenic in NIH-3T3 and Ba/F3 models; sensitive to erlotinib.
• SIK2 in-frame kinase domain duplication identified in a tumor without other known oncogene mutations.
• Frequent somatic copy number gains: TERT (42%), MYC (31%), EGFR (22%), NKX2-1 (18%).
• Only 47% of tumors harbored a mutation in a known proliferative signaling driver; 15% had no hallmark alteration identified.

Genes & alterations

Gene	Alteration	Frequency	Significance
TP53	Inactivating mutations	50%	Known tumor suppressor; TP53 mutations associated with reduced PFS (P = 0.0014)
KRAS	Activating mutations	27%	Known oncogene; anti-correlated with EGFR
EGFR	Activating mutations + novel exon 25-26 deletion	17%	Novel C-terminal deletion oncogenic and erlotinib-sensitive
STK11	Inactivating mutations/deletions	15%	Enriched in spectrum cluster 2 (P = 0.0026)
KEAP1	Inactivating mutations	12%	Anti-correlated with never/light smoking
NF1	Inactivating mutations	11%	Co-occurred with U2AF1 mutations (P = 0.0011)
BRAF	Activating mutations	8%	Known oncogene
ARID1A	Truncating mutations	8%	Novel; SWI/SNF complex member
RBM10	Truncating mutations	7%	Novel RNA-binding protein
SMARCA4	Inactivating mutations	Significant	Epigenetic regulator; could not be mapped to any classic hallmark
U2AF1	p.S34F hotspot	3%	Novel in epithelial tumors; associated with poor PFS
SMAD4	Inactivating mutations	3%	Known tumor suppressor
PIK3CA	Activating mutations	Significant	Known oncogene
SETD2	Mutations	Significant	Epigenetic regulator
BRD3	Truncating mutations	Significant	LOF InVEx significant

Clinical implications

• U2AF1 p.S34F mutations are associated with significantly reduced progression-free survival (P = 0.00011) and represent a potential prognostic biomarker in LUAD.
• The novel EGFR exon 25-26 deletion (vIVb) is oncogenic and sensitive to erlotinib, suggesting patients with this alteration may benefit from EGFR tyrosine kinase inhibitor therapy.
• Mutation spectrum analysis can impute smoking status from sequencing data, potentially useful for patient stratification when clinical history is unavailable.
• Approximately 45% of LUAD tumors lack a known proliferative signaling driver mutation, highlighting the need for additional genomic characterization.

Limitations & open questions

• WGS was limited to 24 cases, likely underpowered to detect rare fusions (ALK, RET, ROS1 rearrangements not identified).
• Rarely mutated oncogenes (AKT1, ERBB2, NRAS, HRAS with 3 or fewer events) could not be statistically nominated despite known relevance.
• Functional validation was performed only for the EGFR exon 25-26 deletion; other novel candidates (U2AF1, RBM10, SIK2) lack functional data in this study.
• The “11th hallmark” of epigenetic/RNA deregulation is proposed but not experimentally validated as a distinct oncogenic mechanism.
• The cohort was chemotherapy-naive surgical specimens, limiting generalizability to advanced/treated disease.

Citations from this paper used in the wiki

• “Identical c.101C>T, p.S34F mutations were seen in 4 of 5 U2AF1 mutant cases […] this study is the first report of U2AF1 mutations in an epithelial tumor.”
• “RBM10 was frequently mutated (12/183 cases; 7%) and subject to recurrent nonsense, frameshift, or splice site mutations, present in 7 of 12 mutated cases (4% of overall cohort).”
• “ARID1A, encoding a key protein in the SWI/SNF chromatin-remodeling complex, was mutated in 8% of cases.”
• “We observed significantly higher exonic mutation rates in tumors from smokers (median: 9.8/MB; mean: 12.9/MB) compared to never smokers (median: 1.7/MB; mean = 2.9/MB, P = 3.0 x 10^-9).”
• “Cells stably expressing this transgene [EGFR exon 25-26 deletion] demonstrated colony formation in soft agar […] interleukin-3 independent proliferation that was blocked by treatment with […] erlotinib.”

This page was processed by paper-compiler on 2026-05-06.