How can people who never smoke get lung cancer?

When it comes to lung cancer, our first reaction is usually: old smoker, right? But in fact, about 10%-20% of lung cancer patients have never smoked, and the specific cause is still unclear
.

Compared with smoking lung cancer patients, how are their genome characteristics different? In September, an article published in the journal Nature Genetics (IF=38.
33) answered this question
.

The study included 232 patients (LCINS) who were diagnosed with NSCLC and had no history of smoking (LCINS), of which 189 adenocarcinomas, 36 squamous cell carcinomas and 7 other types of tumors were not treated.
The authors treated their tumors/normal Tissue paired samples were made DNA WGS
.

The authors of the genomic characteristics of LCINS found that compared with smokers, the median TMB of LCINS was more than 7 times lower (Figure 1).
TMB was significantly related to tumor stage, histology, and age, but not related to tumor purity
.

Figure 1.
The main genomic characteristics of TMBLCINS in LCINS and other cancer types in TCGA are shown in Figure 2.
In the RTK-Ras pathway, EGFR has the highest mutation frequency, followed by KRAS, ALK, MET, ERBB2, ROS1 and RET.
The mutation distribution of the seven genes shows strong mutual exclusion
.

RTK–Ras+ and RTK–Ras− also have significantly different mutation patterns.
The former SNVs/indels, SCNAs, SVs, WGD, and BRCA2 loss of heterozygosity have a higher incidence of mutations, but the tumor/normal telomere length ratio is higher.
Low
.

In addition, TP53 mutation and MDM2 amplification are mutually exclusive, and SVs are enriched in hot spots such as MDM2, TERT, 6p21, MYC, and CDKN2A
.

Figure 2.
The genomic characteristics of LCINS SCNAs subtypes.
Through unsupervised clustering of arm-level SCNAs, the sample is divided into three subtypes (Figure 3.
a).
Subtype 1 (ie Piano).
This naming gives me reason.
It is suspected that the author is a music lover~) accounted for 49.
6% of all tumors, manifested by lack of SCNA, tumor growth is slow; subtype 2 (Mezzo-forte) is rich in chromosome arm level amplification, mainly occurring in 1q, 5p, 7p, 7q and 8q; subtype 3 (Forte) is characterized by enriched whole genome doubling (WGD)
.

Combining the CNV data of smoking LUAD patients, the authors found that they mainly fall into the Forte subtype
.

Figure 3.
a Among these three subtypes, several other genomic characteristics of LCINS also show differences (Figure 3.
b)
.

In comparison, piano has a lower TMB; among the 25 genes with a higher mutation frequency, although 24 are identified as TCGA pan-cancer driver genes, their mutation frequencies are also present in the three subtypes Differences, such as TP53 mutations, are mainly enriched in the forte subtype
.

In addition, the study also identified a new potential driver gene UBA1
.

In the mezzo-forte subtype, half of the tumor samples have EGFR mutations, but very few KRAS mutations; the squamous cell carcinoma samples in the piano subtype are rich in chromatin remodeling-related gene mutations.
In general, the piano subtypes The type has lower SNVs/indels, SCNAs, SVs, WGD loads, and higher T/N:TL and subclonal mutation ratios
.

Figure 3.
Mutation characteristics of bLCINS The author used SigProfiler to identify 14 mutation characteristics previously reported by COSMIC through deconvolution analysis of single base substitution (SBS) mutation characteristics
.

Among them, SBS18, which is related to reactive oxygen species damage, is present in nearly half of the samples, especially forte and mezzo-forte subtypes; SBS8, which is related to nucleotide excision repair defects, is present in 13% of samples, especially squamous cell carcinoma
.

Approximately 58% of the samples have more than 100 SNVs, mainly subclones, containing APOBEC mutation characteristics SBS2 and SBS13, with significant inter-tumor heterogeneity (Figure 4)
.

Tumor samples with significantly enriched APOBEC features also showed TP53 deletion and RTK–Ras+
.

The study also found that in all tumor samples, endogenous mutation characteristics occupy the main body
.

The study did not observe the mutation characteristic SBS4 associated with smoking, even in 62 cases of exposure to second-hand smoke (passive smoking)
.

Figure 4.
The mutation profile of LCINS.
The difference in tumor purity between passive and non-passive smokers is not the reason for the lack of this feature, and the mutation patterns of the two groups are also very similar (Figure 5)
.

Figure 5.
Comparison of mutation profiles of passive and non-passive smokers.
Genome instability.
Overall, 36.
2% of tumor samples have WGD mutations, which are more common in the forte subtype
.

Among the genomic mutations of piano subtypes, the proportion of SCNAs is much lower than that of forte and mezzo-forte
.

Kataegis mutation type (a large number of substitution mutations in very close positions of the genome, and often at the same time as large rearrangements) was detected in 49.
6% of tumors, compared with the other two subtypes, the frequency in the piano subtype Lower
.

Mutations in Kataegis events have APOBEC-related features in both APOBEC3A and APOBEC3B-like tumors.
Kataegis often occurs in the MDM2 locus and tends to co-localize with SV
.

In addition, the study found that TL is negatively correlated with age and has nothing to do with tumor purity
.

It is worth noting that the TL of lung adenocarcinoma patients in LCINS was significantly longer than that of smokers' lung adenocarcinoma patients.
The deletion of 9q, 9p and 22q and HLA LOH were significantly related to TL shortening
.

In the forte subtype, the telomeres of the tumor samples were significantly shorter than the matched normal samples, mezzo-forte showed a difference, and in the piano subtype, the telomeres of the tumor were significantly longer than normal tissues
.

Expanded figure.
The relationship between the T/N TL ratio and somatic mutations.
The evolutionary history of LCINS.
For each subtype in LUAD, the study reconstructed the possible timing of genomic abnormalities including SCNAs, WGD and common driver genes (Figure 6)
.

Among the three subtypes, mutations in the driver genes TP53, RBM10, KRAS and EGFR usually occur at an early stage
.

In the copy number variation of the mezzo-forte subtype, gains and LOH are relatively balanced, while in the forte subtype, most of the LOH mutation events
.

Compared with mezzo-forte, WGD in the forte subtype usually occurs after other key SCNAs
.

Figure 6.
Predicted genomic abnormality timing The author used a validated model to estimate the time of the appearance of the most recent common ancestor (MRCA) (Figure 7.
a).
The tumor samples were grouped according to common driving events to predict the occurrence of these events in the individual Time
.

For example, in tumors carrying ERBB2, CDKN2A or TP53 mutations, and chr22q SCNA, etc.
, MRCA appeared more than ten years before clinical diagnosis; tumors with MDM2 amplification or MET, RBM10, HUWE1, KRAS mutations have a shorter incubation period
.

The authors found that the incubation period of piano was significantly longer than the forte subtype (Figure 7.
b), and the age of MRCA was lower
.

Figure 7.
The impact of reconstructing the evolutionary history of LCINS on survival.
Cases with TP53 mutation or MDM2 amplification have poorer survival rates.
Compared with MDM2 amplification, TP53 mutation has a greater impact
.

In addition, EGFR mutations, CHEK2 LOH, 22q deletion, and 15q deletion are also associated with poor survival (Figure 8)
.

Although the overall survival rate is not significantly related to the status of the RTK-Ras pathway, it is related to the genes in the pathway.
For example, patients with ERBB2 mutations have poor survival, and patients with MET mutations have better overall survival
.

Finally, patients with piano subtypes have the best survival, especially those with squamous cell carcinoma
.

Figure 8.
The relationship between genomic abnormalities and clinical outcomes.
This article focuses on lung cancer patients who have not smoked, identified three subtypes, and analyzed their respective molecular characteristics and evolutionary processes, and discovered the accumulation of endogenous mutations The role in the occurrence of the disease is of great significance to the selection of later treatment options for this group
.

If you have something to say, share today’s article here.
If you have a topic or question you are interested in, you can leave a message to the editor, know everything, help you answer the editor’s writing is not easy friendship copy, please contact the talking editor butler 15510012760 (same number on WeChat) )have a nice day! Shengxinren double 11, in progress transcriptome | methylation | resequencing | single cell | m6A | multiomics cytoscape | limma | WGCNA | water bear legend | linux electrophoresis | PCR | sequencing brief History | Karyotype | NIPT | Basic Experimental Genes | 2019-nCoV | Enrichment Analysis | Joint Analysis | Microenvironmental Plague Pursuit | Summary of Ideas | Scholars | Scientific Research |