Use of three SNPs in the preparation of a product for predicting survival of patients with non-small cell lung cancer

By detecting the polymorphisms or genotypes of rs469783, rs13040574, and rs36071574 in the human genome as biomarkers, the problem of the inability of existing technologies to effectively predict the survival of non-small cell lung cancer patients has been solved, and the accuracy of survival prediction has been improved, especially in the prediction of ten-year survival.

CN114959020BActive Publication Date: 2024-11-22HENAN CANCER HOSPITAL
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110215224.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-02-25
Publication Date
2024-11-22
Estimated Expiration
2041-02-25

AI Technical Summary

Technical Problem

The lack of effective predictive biomarkers in current technologies to identify non-small cell lung cancer patients who can benefit from immunotherapy results in many patients not benefiting from current immunotherapy.

Method used

Polymorphisms or genotypes of rs469783, rs13040574, and rs36071574 in the human genome were detected. Their relationship with the survival of non-small cell lung cancer patients was assessed by multivariate Cox proportional hazards regression analysis. Through analysis of gene expression levels and variant sites, these SNPs were found to be significantly associated with patient survival and can be used as biomarkers to predict survival.

Benefits of technology

It improves the accuracy of predicting the survival of patients with non-small cell lung cancer, especially significantly improving the prediction of ten-year survival, and provides more precise treatment options.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114959020B_ABST
    Figure CN114959020B_ABST
Patent Text Reader

Abstract

The present invention discovered three functional SNPs (ERAP1rs469783T>C, PSMF1rs13040574C>A and NCF2rs36071574G>A) that were independently and significantly associated with patient survival. The HRs for overall survival of the three SNPs were 0.83, 0.86 and 1.31, respectively. In the functional analysis, the alleles ERAP1rs469783C and PSMF1rs13040574A were associated with high levels of expression of the corresponding genes. The allele NCF2rs36071574A was associated with low expression levels of NCF2. Therefore, these three SNPs in the MHC-I pathway are biomarkers for predicting NSCLC survival, and this function is achieved by regulating the expression of the corresponding genes.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and in particular to use of three SNPs in preparing a product for predicting the survival of patients with non-small cell lung cancer. Background Art

[0002] Lung cancer is one of the most common malignancies in humans and has the highest cancer-related mortality rate worldwide. It is estimated that there will be approximately 228,150 new cases and 142,670 deaths from lung cancer in the United States in 2019. Non-small cell lung cancer (NSCLC) is the most common histological type, accounting for approximately 90% of all lung cancer patients. Current treatment options for advanced NSCLC include chemotherapy, radiotherapy, and targeted therapy, but the first two treatments only modestly improve survival rates, and patients inevitably develop drug resistance. In recent years, the role of the immune system in the occurrence and development of cancer has been widely recognized. Immunotherapy is now established as the "fourth pillar" of cancer, and patients with high PD-L1 expression can be treated with immunotherapy alone. Immunotherapy combined with chemotherapy is now the current standard first-line treatment for patients with metastatic NSCLC. However, many patients cannot benefit from current immunotherapy, so there is an urgent need to identify patient-related predictive biomarkers to maximize the identification of those who can benefit from immunotherapy.

[0003] Major histocompatibility complex class I (MHC-I) proteins are the main regulators of cellular immunity, and they control the function of cytotoxic T lymphocytes (CTLs) through the process of antigen presentation. For example, effective CTL responses depend on the ability of MHC-I proteins to present various peptides. On the other hand, the killing effect of cellular immunity in tumor cells is highly dependent on the distribution of MHC-1 activated CTLs on the surface of cancer cells and dendritic cells.

[0004] Since hypothesis-free genome-wide association studies (GWAS) only target the most significant SNPs / genes with stringent P values ​​after correction for multiple testing, most of the identified most significant SNPs lack functional annotations. To date, few novel functional SNPs associated with the prognosis of lung cancer patients have been discovered.

[0005] To date, few new functional SNPs have been associated with the prognosis of lung cancer patients in GWAS studies of European ancestry populations. Therefore, as a promising hypothesis-driven research method in the post-GWAS era, biological pathway-based methods have been used to reanalyze published GWAS datasets to evaluate the cumulative effects of SNPs between multiple genes in the same biological pathway. Since there are fewer SNPs in candidate genes of important biological pathways, the trouble of multiple testing of clinically irrelevant SNPs with no biological significance can be avoided. This improves the detection ability of SNPs with potential biological functions but small differences in significance. Summary of the invention

[0006] The invention provides a use of a substance for detecting the polymorphism or genotype of rs469783, rs13040574 and / or rs36071574 in a human genome in preparing a product for predicting the survival of patients with non-small cell lung cancer.

[0007] The invention provides use of a substance for detecting the polymorphism or genotype of rs469783, rs13040574 and rs36071574 in a human genome in preparing a product for the prognosis of patients with non-small cell lung cancer.

[0008] The invention provides an application of a substance for detecting the polymorphism or genotype of rs469783, rs13040574 and rs36071574 in a human genome in the preparation of a product of single nucleotide polymorphism associated with patients with non-small cell lung cancer.

[0009] rs469783 is 5:96121524 (GRCh37); rs13040574 is 20:1123070 (GRCh37), and rs36071574 is 1:183545603 (GRCh37).

[0010] The present invention provides any of the following applications:

[0011] B1) Use of polymorphisms or genotypes of rs469783, rs13040574 and rs36071574 in the human genome in a product for predicting survival of patients with non-small cell lung cancer;

[0012] B2) Use of the polymorphisms or genotypes of rs469783, rs13040574 and NCF2 rs36071574 in the human genome in preparing a product for the prognosis of patients with non-small cell lung cancer.

[0013] B3) Use of the polymorphisms or genotypes of rs469783, rs13040574 and rs36071574 in the human genome in the preparation of a product for single nucleotide polymorphisms associated with patients with non-small cell lung cancer.

[0014] Optionally, the material for detecting the polymorphism or genotype of rs469783, rs13040574 and rs36071574 in the human genome is a PCR primer and / or a single base extension primer for amplifying a genomic DNA fragment including rs469783, rs13040574 and rs36071574.

[0015] Optionally, the survival is survival period or disease-specific survival rate.

[0016] Optionally, the non-small cell lung cancer is lung adenocarcinoma or lung squamous cell carcinoma.

[0017] The technical solution of the present invention has the following advantages: the present invention conducted a multivariate Cox proportional hazard regression analysis on two previously published whole genome association studies (PLCO and HLCS), evaluated the relationship between the variation of MHC-I related pathway genes and the survival of patients with non-small cell lung cancer, and then analyzed the relationship between gene expression and variation sites. In the stage of discovering variation, 7811 single nucleotide polymorphisms (SNPs) in 89 genes of 1185 NSCLC patients in the PCLO study were significantly associated with the survival of patients, and in the stage of verifying variation, 24 SNPs were verified in 984 NSCLC patients in the HLCS study. In the multivariate stepwise regression analysis, the present invention found that three functional SNPs (ERAP1rs469783T>C, PSMF1 rs13040574C>A and NCF2rs36071574G>A) were independently and significantly associated with the survival of patients. The hazard ratios (HR) of overall survival (OS) for the three SNPs were 0.83 [95% confidence interval (CI) = 0.77-0.89, P = 8.0 × 10 -7 ], 0.86 (0.80-0.93, P=9.4×10 -5 ) and 1.31 (1.11-1.54, P = 0.001). In functional analysis, alleles ERAP1 rs469783C and PSMF1rs13040574A were associated with high expression levels of the corresponding genes. Allele NCF2 rs36071574A was associated with low expression levels of NCF2. Therefore, these three SNPs in the MHC-I pathway are biomarkers for predicting survival in NSCLC, and they do so by regulating the expression of the corresponding genes. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 is an experimental flow chart of the present invention; abbreviations: PLCO, prostate, lung, colorectal and ovarian cancer screening test; NSCLC, non-small cell lung cancer; HLCS, Harvard lung cancer susceptibility study; ERAP1, endoplasmic reticulum aminopeptidase 1; PSMF1, proteasome inhibitor subunit 1; NCF2, neutrophil cytoplasmic factor 2;

[0020] Figure 2 is the survival prediction of SNPs in ERAP1 and PSMF1 combined with risk genotype and eQTL analysis;

[0021] (a) KM survival curves of OS for risk genotypes in the PLCO dataset;

[0022] (b) KM survival curve of OS of binary group of NUG in PLCO dataset;

[0023] (c) KM survival curve of DSS of risk genotypes in the PLCO dataset;

[0024] (d) Kaplan–Meier survival curve of DSS of NUG binary group in PLCO dataset;

[0025] (e) In 373 Europeans from the 1000 Genomes Project, the ERAP1 rs469783C allele was associated with high ERAP1 mRNA expression;

[0026] (f) In 373 Europeans from the 1000 Genomes Project, the PSMF1 rs13040574A allele was associated with high mRNA expression;

[0027] (g) In normal lung tissue from the GTEx project, the ERAP1 rs469783C allele was associated with high ERAP1 mRNA expression;

[0028] (h) In whole blood from the GTEx project, the ERAP1 rs469783C allele was associated with high ERAP1 mRNA expression;

[0029] Figure 3 Differential mRNA expression analysis and OS analysis of three genes in the TCGA database

[0030] (a) Among 109 paired NSCLC tumor tissues and adjacent normal tissues, higher ERAP1 mRNA expression levels were found in adjacent normal tissues (p = 0.0001);

[0031] (b) Among 109 paired NSCLC tumor tissues and adjacent normal tissues, higher PSMF1 mRNA expression levels were found in tumor tissues (p = 0.0006);

[0032] (c) Among 109 paired NSCLC tumor tissues and adjacent normal tissues, higher NCF2 mRNA expression levels were found in adjacent normal tissues (p < 0.0001);

[0033] (d) Kaplan-Meier survival curve of overall survival (OS) of patients with high and low ERAP1 mRNA expression. Among NSCLC patients, patients with high ERAP1 mRNA expression had a higher survival rate than those with low ERAP1 mRNA expression.

[0034] (e) Kaplan-Meier survival curve of overall survival (OS) in patients with high and low PSMF1 mRNA expression. Among NSCLC patients, patients with high PSMF1 mRNA expression had a higher survival rate than those with low PSMF1 mRNA expression.

[0035] (f) Kaplan-Meier survival curves of overall survival (OS) in patients with high and low NCF2 mRNA expression;

[0036] Figure 4 Prediction of ten-year NSCLC survival of three SNPs by ROC curve in PLCO dataset.

[0037] (a) OS time-dependent AUC estimates based on clinical variables (age, sex, smoking status, histology, tumor grade, tumor stage, chemotherapy, surgery, the first four principal components of Table 6 ) and risk genotypes of the three genes.

[0038] (b) Prediction of 10-year NSCLC OS by ROC curve;

[0039] (c) DSS time-dependent AUC estimation: based on clinical variables (age, sex, smoking status, histology, tumor grade, tumor stage, chemotherapy, surgery, the first four principal components in Table 6 ) and risk genotypes of the three genes;

[0040] (d) Prediction of ten-year NSCLC DSS by ROC curve;

[0041] Fig. 5 Correlation between rs36071574 and rs13040574 genotypes and their corresponding mRNA expression levels

[0042] (a) In 373 Europeans from the 1000 Genomes Project, the correlation between the NCF2 rs36071574A allele and mRNA expression levels was not significant;

[0043] (b) In normal lung tissues from the GTEx project, the NCF2 rs36071574A allele was not found to be associated with the expression level of NCF2 mRNA;

[0044] (c) In whole blood from the GTEx project, the NCF2 rs36071574A allele was not found to be associated with the expression level of NCF2 mRNA;

[0045] (d) In normal lung tissue from the GTEx project, there was no significant correlation between the PSMF1 rs13040574A allele and mRNA expression in normal lung tissue (P = 0.933);

[0046] (e) In whole blood from the GTEx project, PSMF1 rs13040574A was associated with high PSMF1 mRNA expression;

[0047] Figure 6 Correlation between rs12094228 genotype and its corresponding mRNA expression level;

[0048] (a) The NCF2 rs12094228G allele was not associated with NCF2 mRNA expression in 373 Europeans from the 1000 Genomes Project;

[0049] (b) NCF2 rs12094228G allele is associated with lower mRNA expression of NCF2 in normal lung tissues from the GTEx project;

[0050] (c) There is no whole blood data for the NCF2 rs12094228G allele in the GTEx project.

[0051] Figure 7 Differential mRNA expression analysis in TCGA database

[0052] (a) Among 58 paired LUAD tumor tissues and adjacent normal tissues, higher ERAP1 expression levels were found in adjacent normal tissues;

[0053] (b) In 51 paired LUSC tissues, higher ERAP1 expression levels were found in adjacent normal tissues;

[0054] (c) In unpaired tumor and normal tissues in LUAD, it was found that the ERAP1 mRNA expression level in normal tissues was higher than that in tumor tissues;

[0055] (d) In unpaired tumor and normal tissues of LUSC, it was found that the ERAP1 mRNA expression level in normal tissues was higher than that in tumor tissues;

[0056] Figure 8 Differential mRNA expression analysis in TCGA database.

[0057] (a) In 58 pairs of LUAD tumor tissues and adjacent normal tissues, no high expression level of PSMF1 mRNA was found in tumor tissues (relative to normal tissues);

[0058] (b) In 51 paired LUSC tumor tissues and adjacent normal tissues, high expression levels of PSMF1 were found in tumor tissues;

[0059] (c) In unpaired tumor and normal tissues of LUAD, it was not found that the expression level of PSMF1 mRNA in tumor tissues was higher than that in normal tissues;

[0060] (d) In unpaired tumor tissues of LUSC, the expression level of PSMF1 mRNA was higher than that in normal tissues;

[0061] Fig. 9 Differential mRNA expression analysis in TCGA database

[0062] (a) Among 58 pairs of LUAD tumor tissues and adjacent normal tissues, higher expression levels of NCF2 were found in adjacent normal tissues;

[0063] (b) In 51 paired LUSC tumor tissues and adjacent normal tissues, high expression levels of NCF2 were found in adjacent normal tissues.

[0064] (c) NCF2 mRNA expression levels were found to be higher in unpaired normal tissues than in tumor tissues in LUAD;

[0065] (d) NCF2 mRNA expression levels were found to be higher in unpaired normal tissues than in tumor tissues in LUSC. DETAILED DESCRIPTION

[0066] Example 1

[0067] For the specific experimental process, see Figure 1 .

[0068] 1. Experimental methods

[0069] In a two-stage analysis, a GWAS dataset of lung cancer patients of European descent from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial was used as the discovery dataset. PLCO is a randomized controlled study conducted by the National Cancer Institute (NCI) that included 77,500 men and 77,500 women aged 55-74 years, recruited from 10 medical centers in the United States between 1993 and 2011. Patients were randomly assigned to an intervention group that received trial screening or a control group that received standard care, and then followed up for at least 13 years after enrollment. Blood samples and personal information about smoking status and family history were collected at enrollment. Cancer diagnosis, histopathology, tumor stage, and treatment were collected during follow-up (Hocking WG, Hu P, Oken MM, Winslow SD, Kvale PA, Prorok PC, Ragard LR, Commins J, Lynch DA, Andriole GL, Buys SS, Fouad MN, Fuhrman CR, Isaacs C, Yokochi LA, Riley TL, Pinsky PF, Gohagan JK, Berg CD, Team PP. Lung cancer screening in the randomized Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. J Natl Cancer Inst 2010;102:722-31). Among these 150,000 patients, after excluding two individuals with no follow-up information, a total of 1,185 NSCLC patients met the requirements for survival analysis (Oken MM, Marcus PM, Hu P, Beck TM, Hocking W, Kvale PA, Cordes J, Riley TL, Winslow SD, Peace S, Levin DL, Prorok PC, Gohagan JK, Team PP. Baseline chest radiograph for lung cancer detection in the randomized Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. J Natl Cancer Inst 2005; 97: 1832-9).Genomic DNA extracted from whole blood samples of these participants was genotyped using Illumina HumanHap240Sv1.0 and HumanHap550v3.0 (dbGaP accession: phs000093.v2.p2 and phs000336.v1.p1). All participants provided written informed consent for the use of the datasets by the PLCO trial, and the institutional review board at each participating institution approved the use of the datasets.

[0070] Another GWAS dataset of 984 histologically confirmed Caucasian NSCLC patients collected from the Harvard Lung Cancer Susceptibility Study (HLCS) in 1992 was used as a validation dataset. In the HLCS study, whole blood samples and personal information were collected after diagnosis, and DNA was extracted from blood samples using the Auto Pure Large Sample Nucleic Acid Purification System (QIAGEN Company, Venlo, Limburg, The Netherlands), and genotyped using the Illumina Humanhap610-Quad array. The genotyping data were used for estimation with Mach3 software based on sequencing data from the 1,000 Genomes Project [Zhai R, Yu X, Wei Y, Su L, Christiani DC. Smoking and smoking cessation in relation to the development of co-existing non-small cell lung cancer with chronic obstructive pulmonary disease. Int J Cancer 2014; 134: 961-70].

[0071] The use of these two GWAS datasets in this study was approved by the Duke University School of Medicine Internal Review Board (Project No. Pro00054575) and the National Center for Biological Information (NCBI) for access to the NCBI dbGaP database of genotypes and phenotypes (Project #6404). Table 1 lists the characteristics compared between the PLCO trial (n = 1185) and the HLCS study (n = 984).

[0072] Gene and SNP selection

[0073] Genes involved in the MHC-I-related pathway with the keyword "MHC IAND class peptide" were selected through the Molecular Marker Database (http: / / software.broadinstitute.org / gsea / msigdb / index.jsp). After removing 85 duplicate genes, 1 pseudogene and 3 genes on chromosome X, 89 genes remained as candidate genes for further analysis. These candidate genes were imputed with IMPUTE2 and 1,000 Genomes Project data (Phase 3). Afterwards, SNPs in these genes and their ±2kb flanking regions were extracted according to the following criteria: estimated information score ≥0.8, genotyping rate ≥95%, minor allele frequency (MAF) ≥5%, and Hardy-Weinberg equilibrium (HWE) ≥1×10 -5 As a result, a total of 7,811 SNPs (527 genotyped and 7,284 imputed) were selected from the PLCO GWAS dataset (dbGaP accession numbers: phs000093.v2.p2 and phs000336.v1.p1) for further analysis. The HLCS study was used as a validation dataset, and the SNP inclusion criteria for the HLCS genotyping data were the same as those for the PLCO genotyping data.

[0074] Table 1. Comparison of characteristics of the PLCO trial and the HLCS study

[0075]

[0076]

[0077] *Chi-square test was used to compare the characteristics of each clinical variable between the PLCO trial and the Harvard study

[0078] (II) Experimental results

[0079] 1. Association between SNPs in MHC-I-related pathway genes and NSCLC survival

[0080] The workflow of this study is shown in the figure Figure 1As shown. The basic characteristics of the 1,185 NSCLC patients in the PLCO trial and the 984 NSCLC patients in the HLCS study have been described elsewhere (Wang Y, Liu H, Ready NE, Su L, Wei Y, Christiani DC, Wei Q. Genetic variants in ABCG1 are associated with survival of nonsmall-cell lung cancer patients. Int J Cancer 2016; 138: 2592-601) and are also shown in Table 1. In the discovery PLCO genotype dataset, single-locus multivariate Cox regression analysis was performed on the selected 7,811 SNPs. For multiple testing correction, none of the SNPs passed the Bonferroni correction (P>0.05) or the false discovery rate (>0.20). This may be due to the high LD in the SNPs generated by the attribution. In addition, the purpose of this pre-screening in this example is to identify functional candidate SNPs for further analysis. Therefore, the inventors used the BFDP method. After multiple testing correction with BFDP≤0.80, a total of 206 SNPs were identified to be significantly associated with NSCLC OS (P<0.05), of which 24 SNPs remained significant after further validation with the HLCS genotype dataset ( Figure 1 ). Subsequently, a combined analysis of the PLCO and HLCS datasets of these 24 newly discovered SNPs revealed that SNPs in ERAP1 and PSMF1 were associated with better survival, whereas SNPs in NCF2 were associated with poorer survival, with no heterogeneity between the two studies (Table 2).

[0081] Table 2. Association of 24 validated significant SNPs with OS in the discovery and validation datasets from two previously published NSCLC WAS

[0082]

[0083] aAdjustable for age, sex, stage, histology, smoking status, chemotherapy, radiotherapy, surgery,

[0084] obtained in the additive genetic model of PC1, PC2, PC3, and PC4;

[0085] bAdjustable for age, sex, stage, histology, smoking status, chemotherapy, radiotherapy, surgery,

[0086] The additive genetic model of PC1, PC2, and PC3 was obtained;

[0087] cPhet: heterogeneity P value obtained by Cochrane's Q test;

[0088] d Meta-analysis in fixed-effects models;

[0089] e SNPs rs30380, rs469758, rs30378, rs246453, rs246454 and SNP rs30379 were highly linked, with basically consistent results;

[0090] f SNPs rs26510, rs27710, rs27529 and SNP rs30187 were highly linked, with basically consistent results.

[0091] 2. Independent SNPs associated with NSCLC survival in the PLCO dataset

[0092] When 24 validated SNPs were included in the multivariate stepwise Cox regression model of the PLCO dataset (because the HLCS study dataset did not have detailed genotyping data), only three SNPs remained significantly associated with survival. This example then expanded the model by further including 15 other previously reported survival predictive SNPs from the same PLCO dataset, and the three newly discovered SNPs were still significantly associated with survival (Table 3).

[0093] Table 3. Multivariate stepwise Cox regression model analysis of 3 independent SNPs, including adjusted other covariates and previously published SNPs in the PLCO trial GWAS dataset

[0094]

[0095] aStepwise analysis included age, sex, smoking status, tumor stage, histology, chemotherapy, radiotherapy, surgery, PC1, PC2, PC3, PC4, and SNPs;

[0096] bStepwise adjustment using 15 published SNPs. Five SNPs were reported in a previous publication (PMID: 27557513); one SNP was reported in a previous publication (PMID: 29978465); two SNPs were reported in a previous publication (PMID: 30259978); two SNPs were reported in a previous publication (PMID: 26757251); three SNPs were reported in a previous publication (PMID: 30650190); two SNPs were reported in a previous publication (PMID: 30989732).

[0097] In the PLCO dataset fully adjusted for available covariates, patients with either the ERAP1 rs469783C or PSMF1 rs13040574A allele had a reduced risk of death or longer survival (P < 0.05 for OS). trend=0.0003; Ptrend for DSS = 0.0008, Ptrend for OS = 0.006; Ptrend for DSS = 0.014), while patients with NCF2rs36071574A allele had a high risk of death or a shorter survival (Ptrend for OS = 0.018, Ptrend for DSS = 0.012) (Table 4). In particular, compared with the TT genotype, the ERAP1 rs469783C variant genotype had a longer survival (TC: HR for OS = 0.94, 95% CI = 0.81-1.10, P = 0.462 and HR for DDS = 0.95, 95% CI = 0.81-1.12, P = 0.539; CC: OS HR=0.65, 95%CI=0.52-0.80, P<0.0001, HR of DSS=0.64, 95%CI=0.51-0.81, P=0.0002; TC+CC: OS HR=0.86, 95%CI=0.74-0.99, P=0.041, HR of DDS=0.86, 95%CI=0.74-1.01, P=0.063). Compared with CC genotype, PSMF1rs13040574A variant genotype had better long-term survival (CA: OS HR=0.79, 95%CI=0 .66-0.93, P=0.005, and HR for DSS=0.83, 95%CI=0.69-0.99, P=0.038; AA: HR for OS=0.75, 95%CI=0.61-0.92, P=0.007, HR for DDS=0.76, 95%CI=0.61-0.95, P=0.015; CA+AA: HR for OS=0.78, 5%CI=0.66-0.91, P=0.002, HR for DSS=0.81, 95%CI=0.68-0.96, P=0.015) (Table 4). In contrast, the NCF2 rs36071574A variant genotype was associated with shorter survival compared with the GG genotype (GA: HR for OS = 1.26, 95% CI = 1.00-1.60, P = 0.051, HR for DDS = 1.29, 95% CI = 1.01-1.65, P = 0.043; AA: HR for OS = 2.35, 95% CI = 0.87-6.38, P = 0.093, HR for DSS = 2.68, 95% CI = 0.99-7.29, P = 0.053; GA+AA: HR for OS = 1.29, 95% CI = 1.03-1.62, P = 0.030 and HR for DSS = 1.32, 95% CI = 1.04-1.68, P = 0.022) (Table 3).

[0098] Table 4. Association between three independent SNPs and survival of NSCLC patients in the PLCO trial (Table 4-1 for OS and Table 4-2 for DSS)

[0099] 4-1

[0100]

[0101]

[0102] aAdjusted for age, sex, smoking status, histology, tumor stage, chemotherapy, surgery, and primary component.

[0103] bData of those lost to follow-up were excluded.

[0104] c Data from 10 individuals who were lost to follow-up were excluded; d The data of 10 subjects who were lost to follow-up were excluded; e10 subjects who were lost to follow-up were excluded; f Unfavorable genotypes were ERAP1 rs469783TT, PSMF1 rs13040574CC, and NCF2 rs36071574GA+AA.

[0105] Trend test: Trend test.

[0106] 4-2

[0107]

[0108]

[0109] aAdjusted for age, sex, smoking status, histology, tumor stage, chemotherapy, surgery, and primary component.

[0110] bData of those lost to follow-up were excluded.

[0111] c Data from 10 individuals who were lost to follow-up were excluded; d Data from 10 subjects who were lost to follow-up were excluded; e The data of 10 subjects who were lost to follow-up were excluded; f unfavorable genotypes ERAP1 rs469783TT, PSMF1 rs13040574CC, NCF2 rs36071574GA+AA.

[0112] Trend test: Trend test.

[0113] Since the available validation dataset from the HLCS study lacks detailed genotyping data, this example uses the PLCO dataset to evaluate the combined effects of three independent SNPs on NSCLC OS and DSS. For the combined analysis, this implementation swapped some alleles according to their actual direction of action to align the direction of allele action at different loci. First, the risk genotypes (i.e., ERAP1 rs469783TT, PSMF1 rs13040574CC, and NCF2rs136071574GA+AA) were merged into multiple risk genotype (NUGs) cumulative scores. As shown in Table 4, increased NUG scores evaluated in the multivariate analysis of the PLCO dataset were associated with decreased survival (Ptrend<0.0001 and 0.0002 for OS and DSS, respectively).

[0114] Then, to facilitate further stratified analysis, this example uses the dichotomous NUG score to divide all patients into 0-1 scores and 2-3 scores. Compared with the 0-1 score group, the survival rate of the 2-3 score group was significantly reduced (HR for OS = 1.52, 95% CI = 1.24-1.86, P < 0.0001; HR for DSS = 1.51, 95% CI = 1.22-1.87, P < 0.0001) (Table 4). This example further proposes Kaplan-Meier survival curves to describe these associations between risk genotypes and NSCLC OS and DSS ( Figure 2 a-2d).

[0115] 3. ROC curve and time-dependent AUC

[0116] This example further evaluated the predictive value of the three SNPs in the PLCO dataset using the time-dependent AUC and ROC curves at the 60th month (or 5-year survival) (follow-up time ranged from 0.03 to 155.83 months, with a median follow-up time of 19.80 months). Compared with the covariate model including age, sex, smoking status, histology, tumor stage, chemotherapy, radiotherapy, surgery and the first four principal components (Table 5), the SNP combined with covariates (age, sex, smoking status, histology, tumor stage, chemotherapy, radiotherapy, surgery and the first four principal components (Table 5) model time-AUC plot did not improve the model's prediction performance at 60 months (or 5-year survival). However, when the time-related AUC and ROC curves were performed at 120 months (or 10-year survival), the predictive performance of the model was significantly improved. The AUC for OS increased from 84.41% to 86.68% (P = 0.022), and the AUC for DSS increased from 84.88% to 87.08% (P = 0.030) ( Figure 4), suggesting that these newly discovered SNPs are helpful in predicting overall survival of NSCLC patients in the PLCO dataset.

[0117] Table 5. Association of the top 10 principal components with OS in NSCLC in the PLCO trial

[0118]

[0119] *The first 4 variables were used for overall stratification adjustment in the multivariate stepwise Cox model

[0120] 4. eQTL analysis

[0121] In this example, eQTL analysis was performed to explore the correlation between the genotypes of the three newly discovered survival predictive SNPs and their corresponding mRNA expression levels. Using the data from the 1000 Genomes Project of 373 lymphoblastoid cell lines of European descent, eQTL analysis was performed and it was found that both ERAP1 rs469783C and PSMF1 rs13040574A alleles were significantly correlated with increased mRNA expression levels of their genes (P values ​​were P<2.0x10 -16 and P = 0.004; Figure 2e and 2f) (Lappalainen T, Sammeth M, Friedlander MR, t Hoen PA, Monlong J, Rivas MA, Gonzalez-Porta M, Kurbatova N, Griebel T, Ferreira PG, Barann ​​M, Wieland T, Greger L, van Iterson M, Almlof J, Ribeca P, Pulyakhina I, Esser D, Giger T, Tikhonov A, Sultan M, Bertier G, MacArthur DG, Lek M, Lizano E, Buermans HP, Padioleau I, Schwarzmayr T, Karlberg O, Ongen H, Kilpinen H, Beltran S, Gut M, Kahlem K, Amstislavskiy V, Stegle O, Pirinen M, Montgomery SB, Donnelly P, McCarthy MI, Flicek P, Strom TM, Geuvadis C,Lehrach H, Schreiber S, Sudbrak R, Carracedo A, Antonarakis SE, Hasler R, Syvanen AC, van Ommen GJ, Brazma A, Meitinger T, Rosenstiel P, Guigo R, Gut IG, Estivill X, Dermitzakis ET. Transcriptome and genome sequencing uncovers functional variation in humans. Nature 2013; 501: 506-11); however, the correlation between the NCF2 rs36071574A allele and mRNA expression levels was not significant (P = 0.701) ( Figure 5a Then, this example used the data of 369 whole blood samples and 383 normal lung tissues from the GTEx project to perform eQTL and found that the rs469783C allele was still associated with higher mRNA expression levels of ERAP1 in normal lung tissue and whole blood, tissue (P = 6.2x10 -15 ) and whole blood (P = 1.2x10 -19 )See Figure 2g and 2h, (ConsortiumGT.Human genomics.The Genotype-Tissue Expression (GTEx) pilot analysis:multitissue gene regulation in humans.Science 2015;348:648-60); however, there was no significant correlation between the PSMF1rs13040574A allele and mRNA expression in normal lung tissue (P=0.933) and whole blood (P=0.498) ( Figure 5d and 5e); and the relationship between the NCF2 rs36071574A allele and the mRNA expression level in normal lung tissue (P = 0.326) or whole blood (P = 0.671) ( Figure 5b , c) (Consortium GT. Human genomics. The Genotype-Tissue Expression (GTEx) pilot analysis: multitissue gene regulation in humans. Science 2015; 348: 648-60). Next, this example evaluates the eQTL of SNPs with high linkage disequilibrium (LD) (r2>0.80) with NCF2 rs36071574 to determine whether SNPs with high linkage disequilibrium (LD) with rs36071574 may have an effect on the mRNA expression of NCF2. This example found that the NCF2 rs12094228G allele was in high LD with the NCF2 rs36071574A allele (r2=0.93) and was associated with a lower expression level of NCF2 in normal lung tissue (P=0.027) ( Figure 6 b)(Consortium GT.Human genomics.The Genotype-Tissue Expression (GTEx) pilot analysis:multitissue gene regulation in humans.Science 2015;348:648-60)). In addition, there is no whole blood data for the NCF2 rs12094228G allele in the GTEx project, such as Figure 6 (c); The NCF2 rs12094228G allele was not associated with NCF2 mRNA expression in 373 Europeans from the 1000 Genomes Project, e.g. Figure 6 (a).

[0122] Finally, this example uses SNPinfo (Xu Z, Taylor JA. SNPinfo: integrating GWAS and candidate gene information into functional SNP selection for genetic association studies. Nucleic Acids Res 2009; 37: W600-5), RegulomeDB (Boyle AP, Hong EL, Hariharan M, Cheng Y, Schaub MA, Kasowski M, Karczewski KJ, Park J, HitzBC, Weng S, Cherry JM, Snyder M. Annotation of functional variation in personalgenomes using RegulomeDB. Genome Res 2012;22:1790-7) and Haploreg (Ward LD, Kellis M. HaploReg v4: systematic mining of putative causal variants, cell types, regulators and target genes for human complex traits and disease. NucleicAcids Res 2016;44:D877-81) online tool was used to predict the bioinformatics functions of these four SNPs. Although none of the four SNPs had obvious functions based on SNPinfo, they had some bioinformatics functions based on RegulomeDB and Haploreg. For example, ERAP1rs469783T>C had effects on enhancer histone marks, DNAse and motifs; PSMF1rs13040574C>A had effects on motifs; NCF2rs36071574G>A had effects on enhancer histone marks and motifs; NCF2rs12094228T>G had effects on enhancer histone marks.

[0123] 5. Analysis of Differential Expression of Different mRNAs

[0124] This example evaluates the mRNA expression levels of three genes identified by SNPs in 109 pairs of tumor and adjacent normal tissue samples in NSCLC obtained from the TCGA database and unpaired tumor and normal tissue samples in the UALCAN database (http: / / ualcan.path.uab.edu / ). This example also evaluates the association between mRNA expression levels and survival probability on www.kmplot.com. Probes were selected according to the web recommendations (probes for ERAP1, PSMF1, and NCF2 are 209788_s_at, 236012_at, and 209949_at, respectively). Figure 3 a and Figure 7 a and Figure 7 As shown in b, in all samples, the expression of ERAP1 mRNA level in tumor tissues was lower compared with adjacent normal tissues. In the UALCAN (http: / / ualcan.path.uab.edu) database, the mRNA expression level of ERAP1 was also much lower in tumor tissues of LUAD (P=0.059) and LUSC (P=7.8×10-5) ( Figure 7 c and 7d). In addition, higher expression levels of ERAP1 mRNA were associated with better NSCLC survival ( Figure 3 d) Similarly, if Figure 3 As shown in Figures 8a and 8b, compared with adjacent normal tissues, the PSMF1 mRNA expression level in tumor tissues was higher in NSCLC (P = 0.006) and LUSC samples (P < 0.0001), but not in LUAD (P = 0.571). Similar results were also found in the UALCAN (http: / / ualcan.path.uab.edu) database. That is, the mRNA expression level of PSMF1 in tumor tissues was higher in LUSC tissues compared with normal tissues (P = 3.3 × 10 -7 ); however, in LUAD, the mRNA expression level of PSMF1 in tumor tissues was not higher in LUAD tissues compared with normal tissues (P = 0.340) ( Figure 8 c and d). However, higher PSMF1 mRNA expression levels were associated with better NSCLC survival ( Figure 3 e), which may be due to the indirect upregulation of PSMF1 in tumor tissues as a part of the proteasome inhibitor PI31.

[0125] Finally, if Figure 3As shown in Figures 9a and b, compared with adjacent normal tissues, tumor tissues had lower NCF2 mRNA expression levels in the combined LUAD and LUSC samples (P < 0.0001), LUAD (P < 0.0001), and LUSC (P < 0.0001). Similar results were obtained in the UALCAN (http: / / ualcan.path.uab.edu) database. That is, compared with normal tissues, LUAD (P < 1.0 × 10 -12 ) and LUSC (P = 1.6 × 10 -12 ) showed a low mRNA expression level of NCF2 in tumor tissues ( Fig. 9 c and d). In addition, higher expression levels of NCF2 mRNA were also associated with better survival in NSCLC ( Figure 3 f).

[0126] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

Claims

1. Use of a substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in a human genome in the preparation of a product for predicting the survival of patients with non-small cell lung cancer; the substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in a human genome is a PCR primer and / or a single base extension primer for amplifying a genomic DNA fragment including PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A.

2. Use of a substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in the human genome in the preparation of a product for the prognosis of patients with non-small cell lung cancer; the substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in the human genome is a PCR primer and / or a single base extension primer for amplifying a genomic DNA fragment including PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A.

3. Use of a substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in the human genome in the preparation of a product for detecting single nucleotide polymorphisms related to the survival or prognosis of patients with non-small cell lung cancer; the substance for detecting the polymorphism or genotype of PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A in the human genome is a PCR primer and / or a single base extension primer for amplifying a genomic DNA fragment including PSMF1 rs13040574C>A and / or NCF2 rs36071574G>A.

4. The use according to claim 1 or 3, characterized in that: The survival is the survival period or the disease-specific survival rate.