A Large-Scale Genome-Wide Gene-Gene Interaction Study of Lung Cancer Susceptibility in Europeans With a Trans-Ethnic Validation in Asians

A review of disease risk prediction methods and applications in the omics era

Risk prediction and disease prevention are the innovative care challenges of the 21st century. Apart from freeing the individual from the pain of disease, it will lead to low medical costs for society. Until very recently, risk assessments have ushered in a new era with the emergence of omics technologies, including genomics, transcriptomics, epigenomics, proteomics, and so on, which potentially advance the ability of biomarkers to aid prediction models. While risk prediction has achieved great success, there are still some challenges and limitations. We reviewed the general process of omics-based disease risk model construction and the applications in four typical diseases. Meanwhile, we highlighted the problems in current studies and explored the potential opportunities and challenges for future clinical practice.

Lung Cancer in Ever- and Never-Smokers: Findings from Multi-Population GWAS Studies

BACKGROUND

Clinical, molecular, and genetic epidemiology studies displayed remarkable differences between ever- and never-smoking lung cancer.

METHODS

We conducted a stratified multi-population (European, East Asian, and African descent) association study on 44,823 ever-smokers and 20,074 never-smokers to identify novel variants that were missed in the non-stratified analysis. Functional analysis including expression quantitative trait loci (eQTL) colocalization and DNA damage assays, and annotation studies were conducted to evaluate the functional roles of the variants. We further evaluated the impact of smoking quantity on lung cancer risk for the variants associated with ever-smoking lung cancer.

RESULTS

Five novel independent loci, GABRA4, intergenic region 12q24.33, LRRC4C, LINC01088, and LCNL1 were identified with the association at two or three populations (P < 5 × 10-8). Further functional analysis provided multiple lines of evidence suggesting the variants affect lung cancer risk through excessive DNA damage (GABRA4) or cis-regulation of gene expression (LCNL1). The risk of variants from 12 independent regions, including the well-known CHRNA5, associated with ever-smoking lung cancer was evaluated for never-smokers, light-smokers (packyear ≤ 20), and moderate-to-heavy-smokers (packyear > 20). Different risk patterns were observed for the variants among the different groups by smoking behavior.

CONCLUSIONS

We identified novel variants associated with lung cancer in only ever- or never-smoking groups that were missed by prior main-effect association studies.

IMPACT

Our study highlights the genetic heterogeneity between ever- and never-smoking lung cancer and provides etiologic insights into the complicated genetic architecture of this deadly cancer.

Proteome-wide Mendelian randomization identifies causal plasma proteins in lung cancer

Plasma proteins are promising biomarkers and potential drug targets in lung cancer. To evaluate the causal association between plasma proteins and lung cancer, we performed proteome-wide Mendelian randomization meta-analysis (PW-MR-meta) based on lung cancer genome-wide association studies (GWASs), protein quantitative trait loci (pQTLs) of 4,719 plasma proteins in deCODE and 4,775 in Fenland. Further, causal-protein risk score (CPRS) was developed based on causal proteins and validated in the UK Biobank. 270 plasma proteins were identified using PW-MR meta-analysis, including 39 robust causal proteins (both FDR-q < 0.05) and 78 moderate causal proteins (FDR-q < 0.05 in one and p < 0.05 in another). The CPRS had satisfactory performance in risk stratification for lung cancer (top 10% CPRS:Hazard ratio (HR) (95%CI):4.33(2.65-7.06)). The CPRS [AUC (95%CI): 65.93 (62.91-68.78)] outperformed the traditional polygenic risk score (PRS) [AUC (95%CI): 55.71(52.67-58.59)]. Our findings offer further insight into the genetic architecture of plasma proteins for lung cancer susceptibility.

Genome-wide exploration of genetic interactions for bladder cancer risk

Although GWASs have been conducted to investigate genetic variation of bladder tumorigenesis, little is known about genetic interactions that may influence bladder cancer (BC) risk. By leveraging large-scale participants from UK Biobank, we established a discovery database with 4000 Caucasian participants (2000 cases vs 2000 non-cases), a database with 1648 Caucasian participants (824 cases vs 824 non-cases) and 856 non-Caucasian participants (428 cases vs 428 non-cases) as validation. We then performed a genome-wide SNP-SNP interaction investigation related to BC risk based a machine learning approach (ie, GenEpi). Moreover, we used the selected interactions to build a BC screening model with an integrated interaction-empowered polygenic risk score (iPRS) based on Cox proportional hazard model. With Bonferroni correction, we identified 10 statistically significant pairs of SNPs, which located in 17 chromosomes. Of these, four SNP-SNP interactions were found to be positively associated with BC risk among Caucasian participants (ORs 1.57-2.03), while six SNP-SNP interactions showed negatively associated with BC risk (ORs 0.54-0.65). Only four of the SNP-SNP interactions were consistently identified in non-Caucasian participants located in ST7L-ADSS2, FHIT-CHDH, LARP4B-LHPP and RBFOX3-MPRIP. In addition, the iPRS showed a HR of 1.81 (95% CI: 1.46-2.09) compared the highest tertile to the lowest tertile, with an enhanced AUC (0.91; 95% CI:0.85-0.97) than PRS (AUC: 0.86; 95% CI:0.76-0.95; P-DeLong test  = 2.2 × 10-4 ). In summary, this study identified several important SNP-SNP interactions for BC risk, and developed an iPRS model for BC screening, which may help to identify the people at high-risk state of BC before early manifestation.

Prevalence of BRCA1, BRCA2, and PALB2 genomic alterations among 924 Taiwanese breast cancer assays with tumor-only targeted sequencing: extended data analysis from the VGH-TAYLOR study

BACKGROUND

The homologous recombination (HR) repair pathway for DNA damage, particularly the BRCA1 and BRCA2 genes, has become a target for cancer therapy, with poly ADP-ribose polymerase (PARP) inhibitors showing significant outcomes in treating germline BRCA1/2 (gBRCA1/2) mutated breast cancer. Recent studies suggest that some patients with somatic BRCA1/2 (sBRCA1/2) mutation or mutations in HR-related genes other than BRCA1/2 may benefit from PARP inhibitors as well, particularly those with PALB2 mutations. The current analysis aims to evaluate the prevalence of genetic alterations specific to BRCA1, BRCA2, and PALB2 in a large cohort of Taiwanese breast cancer patients through tumor-targeted sequencing.

METHODS

A total of 924 consecutive assays from 879 Taiwanese breast cancer patients underwent tumor-targeted sequencing (Thermo Fisher Oncomine Comprehensive Assay v3). We evaluated BRCA1, BRCA2, and PALB2 mutational profiles, with variants annotated and curated by the ClinVAR, the Oncomine™ Knowledgebase Reporter, and the OncoKB™. We also conducted reflex germline testing using either whole exome sequencing (WES) or whole genome sequencing (WGS), which is ongoing.

RESULTS

Among the 879 patients analyzed (924 assays), 130 had positive mutations in BRCA1 (3.1%), BRCA2 (8.6%), and PALB2 (5.2%), with a total of 14.8% having genetic alterations. Co-occurrence was noted between BRCA1/BRCA2, BRCA1/PALB2, and BRCA2/PALB2 mutations. In BRCA1-mutated samples, only p.K654fs was observed in three patients, while other variants were observed no more than twice. For BRCA2, p.N372H was the most common (26 patients), followed by p.S2186fs, p.V2466A, and p.X159_splice (5 times each). For PALB2, p.I887fs was the most common mutation (30 patients). This study identified 176 amino acid changes; 60.2% (106) were not documented in either ClinVAR or the Oncomine™ Knowledgebase Reporter. Using the OncoKB™ for annotation, 171 (97.2%) were found to have clinical implications. For the result of reflex germline testing, three variants (BRCA1 c.1969_1970del, BRCA1 c.3629_3630del, BRCA2 c.8755-1G > C) were annotated as Pathogenic/Likely pathogenic (P/LP) variants by ClinVar and as likely loss-of-function or likely oncogenic by OncoKB; while one variant (PALB2 c.448C > T) was not found in ClinVar but was annotated as likely loss-of-function or likely oncogenic by OncoKB.

CONCLUSION

Our study depicted the mutational patterns of BRCA1, BRCA2, and PALB2 in Taiwanese breast cancer patients through tumor-only sequencing. This highlights the growing importance of BRCA1/2 and PALB2 alterations in breast cancer susceptibility risk and the treatment of index patients. We also emphasized the need to meticulously annotate variants in cancer-driver genes as well as actionable mutations across multiple databases.

Bayesian Approaches in Exploring Gene-environment and Gene-gene Interactions: A Comprehensive Review

Rapid advancements in high-throughput biological techniques have facilitated the generation of high-dimensional omics datasets, which have provided a solid foundation for precision medicine and prognosis prediction. Nonetheless, the problem of missing heritability persists. To solve this problem, it is essential to explain the genetic structure of disease incidence risk and prognosis by incorporating interactions. The development of the Bayesian theory has provided new approaches for developing models for interaction identification and estimation. Several Bayesian models have been developed to improve the accuracy of model and identify the main effect, gene-environment (G×E) and gene-gene (G×G) interactions. Studies based on single-nucleotide polymorphisms (SNPs) are significant for the exploration of rare and common variants. Models based on the effect heredity principle and group-based models are relatively flexible and do not require strict constraints when dealing with the hierarchical structure between the main effect and interactions (M-I). These models have a good interpretability of biological mechanisms. Machine learning-based Bayesian approaches are highly competitive in improving prediction accuracy. These models provide insights into the mechanisms underlying the occurrence and progression of complex diseases, identify more reliable biomarkers, and develop higher predictive accuracy. In this paper, we provide a comprehensive review of these Bayesian approaches.

New insights into the biology and development of lung cancer in never smokers-implications for early detection and treatment

Lung cancer is the leading cause of cancer deaths worldwide. Despite never smokers comprising between 10 and 25% of all cases, lung cancer in never smokers (LCNS) is relatively under characterized from an etiological and biological perspective. The application of multi-omics techniques on large patient cohorts has significantly advanced the current understanding of LCNS tumor biology. By synthesizing the findings of multi-omics studies on LCNS from a clinical perspective, we can directly translate knowledge regarding tumor biology into implications for patient care. Primarily focused on never smokers with lung adenocarcinoma, this review details the predominance of driver mutations, particularly in East Asian patients, as well as the frequency and importance of germline variants in LCNS. The mutational patterns present in LCNS tumors are thoroughly explored, highlighting the high abundance of the APOBEC signature. Moreover, this review recognizes the spectrum of immune profiles present in LCNS tumors and posits how it can be translated to treatment selection. The recurring and novel insights from multi-omics studies on LCNS tumor biology have a wide range of clinical implications. Risk factors such as exposure to outdoor air pollution, second hand smoke, and potentially diet have a genomic imprint in LCNS at varying degrees, and although they do not encompass all LCNS cases, they can be leveraged to stratify risk. Germline variants similarly contribute to a notable proportion of LCNS, which warrants detailed documentation of family history of lung cancer among never smokers and demonstrates value in developing testing for pathogenic variants in never smokers for early detection in the future. Molecular driver subtypes and specific co-mutations and mutational signatures have prognostic value in LCNS and can guide treatment selection. LCNS tumors with no known driver alterations tend to be stem-like and genes contributing to this state may serve as potential therapeutic targets. Overall, the comprehensive findings of multi-omics studies exert a wide influence on clinical management and future research directions in the realm of LCNS.

ATHENA: an independently validated autophagy-related epigenetic prognostic prediction model of head and neck squamous cell carcinoma

The majority of these existing prognostic models of head and neck squamous cell carcinoma (HNSCC) have unsatisfactory prediction accuracy since they solely utilize demographic and clinical information. Leveraged by autophagy-related epigenetic biomarkers, we aim to develop a better prognostic prediction model of HNSCC incorporating CpG probes with either main effects or gene-gene interactions. Based on DNA methylation data from three independent cohorts, we applied a 3-D analysis strategy to develop An independently validated auTophagy-related epigenetic prognostic prediction model of HEad and Neck squamous cell carcinomA (ATHENA). Compared to prediction models with only demographic and clinical information, ATHENA has substantially improved discriminative ability, prediction accuracy and more clinical net benefits, and shows robustness in different subpopulations, as well as external populations. Besides, epigenetic score of ATHENA is significantly associated with tumor immune microenvironment, tumor-infiltrating immune cell abundances, immune checkpoints, somatic mutation and immunity-related drugs. Taken together these results, ATHENA has the demonstrated feasibility and utility of predicting HNSCC survival ( http://bigdata.njmu.edu.cn/ATHENA/ ).

Potential Impact of Criteria Modifications on Race and Sex Disparities in Eligibility for Lung Cancer Screening

INTRODUCTION

Low-dose computed tomography (LDCT) screening reduces lung cancer mortality, but current eligibility criteria underestimate risk in women and racial minorities. We evaluated the impact of screening criteria modifications on LDCT eligibility and lung cancer detection.

METHODS

Using data from a Lung Nodule Program, we compared persons eligible for LDCT by the following: U.S. Preventive Services Task Force (USPSTF) 2013 criteria (55-80 y, ≥30 pack-years of smoking, and ≤15 y since cessation); USPSTF2021 criteria (50-80 y, ≥20 pack-years of smoking, and ≤15 y since cessation); quit duration expanded to less than or equal to 25 years (USPSTF2021-QD25); reducing the pack-years of smoking to more than or equal to 10 years (USPSTF2021-PY10); and both (USPSTF2021-QD25-PY10). We compare across groups using the chi-square test or analysis of variance.

RESULTS

The 17,421 individuals analyzed were of 56% female sex, 69% white, 28% black; 13% met USPSTF2013 criteria; 17% USPSTF2021; 18% USPSTF2021-QD25; 19% USPSTF2021-PY10; and 21% USPSTF2021-QD25-PY10. Additional eligible individuals by USPSTF2021 (n = 682) and USPSTF2021-QD25-PY10 (n = 1402) were 27% and 29% black, both significantly higher than USPSTF2013 (17%, p < 0.0001). These additional eligible individuals were 55% (USPSTF2021) and 55% (USPSTF2021-QD25-PY10) of female sex, compared with 48% by USPSTF2013 (p < 0.05). Of 1243 persons (7.1%) with lung cancer, 22% were screening eligible by USPSTF13. USPSTF2021-QD25-PY10 increased the total number of persons with lung cancer by 37%. These additional individuals with lung cancer were of 57% female sex (versus 48% with USPSTF2013, p = 0.0476) and 24% black (versus 20% with USPSTF2013, p = 0.3367).

CONCLUSIONS

Expansion of LDCT screening eligibility criteria to allow longer quit duration and fewer pack-years of exposure enriches the screening-eligible population for women and black persons.

OWL: an optimized and independently validated machine learning prediction model for lung cancer screening based on the UK Biobank, PLCO, and NLST populations

BACKGROUND

A reliable risk prediction model is critically important for identifying individuals with high risk of developing lung cancer as candidates for low-dose chest computed tomography (LDCT) screening. Leveraging a cutting-edge machine learning technique that accommodates a wide list of questionnaire-based predictors, we sought to optimize and validate a lung cancer prediction model.

METHODS

We developed an Optimized early Warning model for Lung cancer risk (OWL) using the XGBoost algorithm with 323,344 participants from the England area in UK Biobank (training set), and independently validated it with 93,227 participants from UKB Scotland and Wales area (validation set 1), as well as 70,605 and 66,231 participants in the Prostate, Lung, Colorectal, and Ovarian cancer screening trial (PLCO) control and intervention subpopulations, respectively (validation sets 2 & 3) and 23,138 and 18,669 participants in the United States National Lung Screening Trial (NLST) control and intervention subpopulations, respectively (validation sets 4 & 5). By comparing with three competitive prediction models, i.e., PLCO modified 2012 (PLCO_m2012), PLCO modified 2014 (PLCO_all2014), and the Liverpool Lung cancer Project risk model version 3 (LLPv3), we assessed the discrimination of OWL by the area under receiver operating characteristic curve (AUC) at the designed time point. We further evaluated the calibration using relative improvement in the ratio of expected to observed lung cancer cases (RI_EO), and illustrated the clinical utility by the decision curve analysis.

FINDINGS

For general population, with validation set 1, OWL (AUC = 0.855, 95% CI: 0.829-0.880) presented a better discriminative capability than PLCO_all2014 (AUC = 0.821, 95% CI: 0.794-0.848) (p < 0.001); with validation sets 2 & 3, AUC of OWL was comparable to PLCO_all2014 (AUC_PLCOall2014-AUC_OWL < 1%). For ever-smokers, OWL outperformed PLCO_m2012 and PLCO_all2014 among ever-smokers in validation set 1 (AUC_OWL = 0.842, 95% CI: 0.814-0.871; AUC_PLCOm2012 = 0.792, 95% CI: 0.760-0.823; AUC_PLCOall2014 = 0.791, 95% CI: 0.760-0.822, all p < 0.001). OWL remained comparable to PLCO_m2012 and PLCO_all2014 in discrimination (AUC difference from -0.014 to 0.008) among the ever-smokers in validation sets 2 to 5. In all the validation sets, OWL outperformed LLPv3 among the general population and the ever-smokers. Of note, OWL showed significantly better calibration than PLCO_m2012, PLCO_all2014 (RI_EO from 43.1% to 92.3%, all p < 0.001), and LLPv3 (RI_EO from 41.4% to 98.7%, all p < 0.001) in most cases. For clinical utility, OWL exhibited significant improvement in average net benefits (NB) over PLCO_all2014 in validation set 1 (NB improvement: 32, p < 0.001); among ever smokers of validation set 1, OWL (average NB = 289) retained significant improvement over PLCO_m2012 (average NB = 213) (p < 0.001). OWL had equivalent NBs with PLCO_m2012 and PLCO_all2014 in PLCO and NLST populations, while outperforming LLPv3 in the three populations.

INTERPRETATION

OWL, with a high degree of predictive accuracy and robustness, is a general framework with scientific justifications and clinical utility that can aid in screening individuals with high risks of lung cancer.

FUNDING

National Natural Science Foundation of China, the US NIH.

A trans-omics assessment of gene-gene interaction in early-stage NSCLC

Epigenome-wide gene-gene (G × G) interactions associated with non-small-cell lung cancer (NSCLC) survival may provide insights into molecular mechanisms and therapeutic targets. Hence, we proposed a three-step analytic strategy to identify significant and robust G × G interactions that are relevant to NSCLC survival. In the first step, among 49 billion pairs of DNA methylation probes, we identified 175 775 G × G interactions with P_Bonferroni  ≤ 0.05 in the discovery phase of epigenomic analysis; among them, 15 534 were confirmed with P ≤ 0.05 in the validation phase. In the second step, we further performed a functional validation for these G × G interactions at the gene expression level by way of a two-phase (discovery and validation) transcriptomic analysis, and confirmed 25 significant G × G interactions enriched in the 6p21.33 and 6p22.1 regions. In the third step, we identified two G × G interactions using the trans-omics analysis, which had significant (P ≤ 0.05) epigenetic cis-regulation of transcription and robust G × G interactions at both the epigenetic and transcriptional levels. These interactions were cg14391855 × cg23937960 (β_interaction  = 0.018, P = 1.87 × 10^-12 ), which mapped to RELA × HLA-G (β_interaction  = 0.218, P = 8.82 × 10^-11 ) and cg08872738 × cg27077312 (β_interaction  = -0.010, P = 1.16 × 10^-11 ), which mapped to TUBA1B × TOMM40 (β_interaction =-0.250, P = 3.83 × 10^-10 ). A trans-omics mediation analysis revealed that 20.3% of epigenetic effects on NSCLC survival were significantly (P = 0.034) mediated through transcriptional expression. These statistically significant trans-omics G × G interactions can also discriminate patients with high risk of mortality. In summary, we identified two G × G interactions at both the epigenetic and transcriptional levels, and our findings may provide potential clues for precision treatment of NSCLC.

A specific immune signature for predicting the prognosis of glioma patients with IDH1-mutation and guiding immune checkpoint blockade therapy

Isocitrate dehydrogenase (IDH1) is frequently mutated in glioma tissues, and this mutation mediates specific tumor-promoting mechanisms in glioma cells. We aimed to identify specific immune biomarkers for IDH1-mutation (IDH1mt) glioma. The Cancer Genome Atlas (TCGA) and Chinese Glioma Genome Atlas (CGGA) were used to obtain RNA sequencing data and clinical characteristics of glioma tissues, while the stromal and immune scores of TCGA glioma tissues were determined using the ESTIMATE algorithm. Differentially expressed genes (DEGs), the protein–protein interaction(PPI) network, and least absolute shrinkage and selection operator (LASSO) and Cox regression analyses were used to select hub genes associated with stroma and immune scores and the prognoses of patients and to construct the risk model. The practicability and specificity of the risk model in both IDH1mt and IDH1-wildtype (wtIDH1) gliomas in TCGA and CGGA were evaluated. Molecular mechanisms, immunological characteristics and benefits of immune checkpoint blockade therapy in glioma tissues with IDH1mt were analyzed using GSEA, immunohistochemical staining, CIBERSORT, and T-cell dysfunction and exclusion (TIDE) analysis. The overall survival rate for IDH1mt-glioma patients with high stroma/immune scores was lower than that for those with low stroma/immune scores. A total of 222 DEGs were identified in IDH1mt glioma tissues with high stroma/immune scores. Among them, 72 genes had interactions in the PPI network, while three genes, HLA-DQA2, HOXA3, and SAA2, were selected as hub genes and used to construct risk models classifying patients into high- and low-risk score groups, followed by LASSO and Cox regression analyses. This risk model showed prognostic value in IDH1mt glioma in both TCGA and CCGA; nevertheless, the model was not suitable for wtIDH1 glioma. The risk model may act as an independent prognostic factor for IDH1mt glioma. IDH1mt glioma tissues from patients with high-risk scores showed more infiltration of M1 and CD8 T cells than those from patients with low-risk scores. Moreover, TIDE analysis showed that immune checkpoint blockade(ICB) therapy was highly beneficial for IDH1mt patients with high-risk scores. The risk model showed specific potential to predict the prognosis of IDH1mt-glioma patients, as well as guide ICB, contributing to the diagnosis and therapy of IDH1mt-glioma patients.

Epigenome-wide gene–age interaction study reveals reversed effects of MORN1 DNA methylation on survival between young and elderly oral squamous cell carcinoma patients

DNA methylation serves as a reversible and prognostic biomarker for oral squamous cell carcinoma (OSCC) patients. It is unclear whether the effect of DNA methylation on OSCC overall survival varies with age. As a result, we performed a two-phase gene–age interaction study of OSCC prognosis on an epigenome-wide scale using the Cox proportional hazards model. We identified one CpG probe, cg11676291_MORN1, whose effect was significantly modified by age (HR_discovery = 1.018, p = 4.07 × 10⁻⁰⁷, FDR-q = 3.67 × 10⁻⁰²; HR_validation = 1.058, p = 8.09 × 10⁻⁰³; HR_combined = 1.019, p = 7.36 × 10⁻¹⁰). Moreover, there was an antagonistic interaction between hypomethylation of cg11676291_MORN1 and age (HR_interaction = 0.284; 95% CI, 0.135–0.597; p = 9.04 × 10⁻⁰⁴). The prognosis of OSCC patients was well discriminated by the prognostic score incorporating cg11676291_MORN1–age interaction (HR_{high vs. low} = 3.66, 95% CI: 2.40–5.60, p = 1.93 × 10⁻⁰⁹). By adding 24 significant gene–age interactions using a looser criterion, we significantly improved the area under the receiver operating characteristic curve (AUC) of the model at 3- and 5-year prognostic prediction (AUC_3-year = 0.80, AUC_5-year = 0.79, C-index = 0.75). Our study identified a significant interaction between cg11676291_MORN1 and age on OSCC survival, providing a potential therapeutic target for OSCC patients.