An immune-related prognostic risk model for colorectal cancer was constructed based on next-generation sequencing

doi:10.21203/rs.3.rs-3595512/v1

Download PDF

Research Article

An immune-related prognostic risk model for colorectal cancer was constructed based on next-generation sequencing

https://doi.org/10.21203/rs.3.rs-3595512/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Objective

Screening the immune-related genes of colorectal cancer based on clinical next generation sequencing and construction of an immune-related prognostic risk model, so as to provide an immediate indicator for evaluating the prognosis and efficacy of immunotherapy of CRC patients.

Methodology:

Univariate and multivariate Cox risk regression analysis was used to analyze the correlation between immune-related genes, clinicopathological features, and prognosis of CRC patients, and a prognostic risk model was established.

Results

A total of 57 mutant genes were detected in 150 CRC patients, of which SNV and Indel had the highest mutation frequency, (92.67% and 47.33%, respectively). The genes with the highest number of mutations were TP53 and APC. In all 150 CRC patients, 12 mutation genes were detected, including 6 drug-treatable genes: KRAS, PIK3CA, PTEN, BRAF, NRAS, and TP53.TP53 and PIK3CA were the genes with the most drug-treatable mutation sites, and KRAS, NRAS, and BRAS had the highest medication grade. The mutation rates of 36 CRC common detection genes in the NGS and TCGA clinical databases were very consistent, and they were enriched in negative regulation of cell proliferation, negative regulation of apoptosis, and kinase pathways such as PI3K-AKT signaling pathway. Through the correlation analysis of immune infiltration, 14 immune-related genes TP53, JAK1, MTHFR, EGFR, KDR, APC, KRAS, TSC2, PMS2, KIT, ALK, BRAF, CTNNB1, and FBXW7 were obtained. The Cox univariate analysis and Cox multivariate analysis showed that late clinical stage (stage), late T stage and ALK gene mutation increased the risk of death and were independent prognostic factors affecting the risk of death. A Cox proportional-hazards model was constructed : h(t,x)/h0(t) = EXP(0.835 clinical stage + 0.878T + 1.094ALK). Among the 112 clinical follow-up patients, the survival time of CRC patients with high-risk scores was significantly shorter than that of CRC patients with low-risk scores. In addition, the risk score of the MSI-L group was significantly higher than that of the MSS group in the 92 CRC patients data sets of the TCGA database.

Conclusion

ALK gene mutation may be related to immune and prognostic survival in CRC patients; the constructed prognostic risk model can be used to evaluate the prognosis and efficacy of immunotherapy in CRC patients.

colorectal cancer

NGS

prognosis

immunity

risk model

According to the 2023 global cancer statistics, the prevalence and mortality of colorectal cancer (CRC) are the third highest in the world, and the incidence of patients in China ranks second [1]. The incidence of colorectal cancer is expected to continue to increase over the next two decades [2]. At present, about 100,000 patients worldwide currently die of CRC each year. This is because about 25% of patients have already metastasized at the time of treatment, and another 50–60% of CRC patients have metastasized during the progression of the disease after treatment, losing the chance of radical surgery [3]. For CRC patients, traditional treatment methods such as surgery, chemotherapyand chemoradiotherapy ignore individual differences. Therefore, the patient's treatment effect and the prognosis is poor.

In recent years, targeted drugs based on molecular markers have significantly prolonged the survival of CRC patients. Molecular targeted drugs are a class of drugs that use the differences in molecular biology between target cells and normal cells (including different characteristics of genes, enzymes, and signal transduction) to target cells, inhibit cell growth and proliferation, and ultimately cause target cell death [4]. Currently, the main categories of molecularly targeted drugs used in clinical research in colorectal cancer patients are as follows : i) The drugs that inhibit epidermal growth factor receptor (EGFR) such as EGFR and human epidermal growth factor receptor-2 (HER2) on the surface of tumor cells are mainly cetuximab and panitumumab. ii) The drugs targeting vascular endothelial growth factor signaling pathways such as vascular endothelial growth factor (VEGF), vascular endothelial growth factor receptor-1/2/3 (VEGFR-1/2/3), and platelet-derived growth factor (PDGF), mainly bevacizumab. iii) The drugs that Inhibit receptor tyrosine kinase targets such as Ret proto-oncogene tyrosine-protein kinase receptor Ret (RET), neurotrophic receptor kinase 1/2/3 (NTRK1/2/3), platelet-derived growth factor receptors (PDGFR), angiopoietin-1 receptor (TIE-2), mast/stem cell growth factor receptor Kit (c-kit), c-ros sarcoma proto-oncogene 1-receptor tyrosine kinase (ROS1), mainly including regorafenib. (iv) The drugs targeting other kinases, such as vrafmurine sarcoma viral oncogene homolog B (BRAF), mitogen-activated protein kinase 1/2 (MEK1/2), phosphatidylinositol-4,5-bisphosphate 3-kinase(PIK3CA), mainly include trimetazidine [5–12].

Targeted therapy makes drug treatment based on target specific gene mutations ; it cannot be used in the absence of mutations. Therefore, it is necessary in clinical practice to screen patients with next generation sequencing (NGS). By understanding the gene mutation information of molecular markers in CRC patients, it is helpful to find an appropriate targeted drug regimen in the clinic. NGS, also known as massively parallel sequencing, has revolutionized genomic research. NGS screens mutant genes at the genetic level for independent individuals and shows the mutation status of CRC patients. Therefore, more and more patients with advanced metastatic CRC are opting for targeted therapy, leading to the widespread acceptance of molecularly targeted drugs in clinical precision therapy. In addition to the targeted drug genes mentioned above, NGS also detects many important tumor-driven genes in the detection of CRC, so NGS can also provide more comprehensive information and support for the development of new targeted drugs [13–14]. However, in clinical practice, mutations in drug targets can currently be detected in only a few CRC patients. In addition, the cost of treatment with targeted drugs is relatively high compared with conventional chemotherapeutic agents, and some targeted drugs are not covered by medical insurance. Therefore, the total number of patients receiving targeted therapy is small, and it cannot be widely used.

In recent years, the immune system has been shown to play an important role in the evolution of tumors. Immunotherapy has gradually become a new way to treat tumors by the enhancing autoimmune response rather than killing tumor cells directly [15–16]. The efficacy of immunotherapy with immune checkpoint inhibitors has been gradually confirmed [17–18]. It mainly includes cytotoxic T lymphocyte-associated antigen-4 (CTLA-4) inhibitor, programmed death receptor-1 (PD-1) inhibitor, and programmed death receptor-1 ligand (PD-L1) inhibitor. PD-1 inhibitors based on nivolumab and pembrolizumab, for example, can effectively block the binding of PD-1 to PD-L1, allowing effector T cells to perform the function of killing cancer cells, and ultimately achieve the purpose of tumor treatment [11]. In 2017, the National Medical Products Administration approved immune checkpoint therapy for CRC patients with mismatch repair deficiency (dMMR) or microsatellite instability-high (MSI-H) [12]. In contrast, for CRC patients with mismatch repair incomplete (pMMR), microsatellite stability (MSS), or microsatellite instability-low (MSI-L), current immunotherapy has little effect [19–20]. Therefore, finding molecular markers for the immune status of CRC patients and predicting the efficacy of immunotherapy is an urgent problem.

In clinical practice, we have found that patients often receive very little information about gene mutations for targeted therapy after NGS detection at high cost, often nothing at all. Therefore, whether these detected tumor driver genes or drug targets are related to the immune status of patients and whether they are related to immunotherapy is the starting point of this study. In this study, the mutations of 57 genes were analyzed in the NGS sequencing results of 150 CRC patients, and 36 common CRC detection genes were screened out. The immune infiltration levels of 36 CRC common detection genes in CRC were analyzed by tumor immune infiltration database TIMER2.0 and single-sample gene set enrichment analysis (ssGSEA), and 14 immune-related genes were obtained. Simultaneously, the clinical data of 590 patients and mutation data of 14 immune-related genes were downloaded from the TCGA database. Univariate and multivariate Cox analyses determined that clinical stage, T classification, and ALK gene were independent prognostic factors, and cox proportional-hazards model was constructed. Follow-up of 112 CRC patients confirmed that the model could be used to assess the prognosis of CRC patients. At the same time, by downloading the clinical data of 92 CRC patients with MSI detection in TCGA database, it is shown that the risk score of CRC patients is positively correlated with MSI-H, suggesting that the prognostic risk model can be used to predict the efficacy of immunotherapy. In this study, bioinformatics analysis tools were used to sort out and analyze the NGS data and TCGA data from 150 CRC patients, and an immune-related prognostic risk model was established. The aim was to find key genes related to immune status, immunotherapy and prognosis of CRC patients, and to provide a valuable reference for clinical immunotherapy research of CRC patients.

Patients and specimens

Included in the study were 150 individuals diagnosed with colorectal cancer who were registered at The Affiliated Hospital of Chengde Medical University (Hebei, China) from January 2018 to December 2022. Research Ethics Review Committee of Chengde Medical University approved the present study (Hebei, China; approval no. 2017003), and all patients provided written informed consent before the start of the study. The inclusion criteria were: i) CRC was a definite clinical diagnosis; ii) The clinical data were recorded accurately and completely, including the patient's age, sex, histological type, clinical stage at diagnosis, T classification, N classification, M classification, the levels of tumor markers carcinoembryonic antigen (CEA), carbohydrate antigen 125 (CA125), carbohydrate antigen 199 (CA199). Clinical staging was based on the American Joint Committee on Cancer (AJCC) Classification (8th edition). and iii) NGS detection was performed. Exclusion criteria: CRC patients who did not receive NGS testing.

Sections of tumor tissue were cut into 5m thick pieces and fixed in 4% formalin at 4 C for 30 minutes, then dehydrated and embedded in paraffin. Ethanol diamine tetraacetic acid anticoagulant tubes (5 ml) were filled with blood samples and gently inverted eight to ten times before being analyzed. All patients were fasting for 8–12 hours before blood was taken, and no previous blood transfusion had taken place within 4 weeks. Blood samples used for NGS did not show any signs of hemolysis, lipid turbidity, or jaundice.

A TCGA dataset containing data from 1,151 CRC patient samples was downloaded from the cBioPortal database (v3.0.2; http://www.cbioportal.org/). For survival analysis, 590 patients from the 1,151 samples in the TCGA dataset who met the inclusion criteria were included in the present study. The inclusion criteria were: i) Patients had a definite clinical diagnosis of CRC; ii) Detailed clinical data were accurately and completely recorded, including the patient's sex, age, histologic type, clinical stage at diagnosis, T classification, N classification, M classification, overall survival status, and time; and iii) The mutations status of 57 genes was described.

Design and general performance of the CRC 57 gene NGS panel

The CRC 57 gene NGS panel was established. The gene panel's inclusion criteria consisted of: i) Genes from CRC treatment drugs approved by the FDA and/or recommended by the National Comprehensive Cancer Network; ii) Genes annotated in databases, such as Clinical Knowledgebase (https://ckb.jax.org/), Oncology Knowledge Base (OncoKB; http://oncokb.org/), cBioPortal, My Cancer Genome database (https://www.mycancergenome.org/) and Clinical Trials (https://clinicaltrials.gov/); and iii) Detailed reports have previously documented the genes that are being studied clinically. The panel included the following 57 genes : RNF43, PTPRD, MYC, WRN, FANCM, TP53, APC, KRAS, PIK3CA, FBXW7, PDCD1LG2, FGFR1, NF1, STK11, TERT, PTEN, EGFR, RB1, CD274, JAK2, PTCH1, SMARCA4, POLE, CDKN2A, ESR1, MET, KIT, CTNNB1, PMS2, TSC1, BRAF, SMAD4, ALK, NRAS, MSH2, ERBB2, VHL, NTRK1, ROS1, MTHFR, MSH6, MAP2K1, XRCC1, TSC2, AKT1, KDR, JAK1, PDGFRA, HRAS, UMPS, GSTP1, G0S, SRC1, NQO1, TP63.

Genetic material (gDNA) was obtained from tissue or blood samples that were embedded in paraffin using the GoldHi Plasmid Mini Kit. The concentration of the obtained gDNA was determined by employing a Qubit fluorometer (Qubit® dsDNA BR Assay Kit) and purity was measured by a spectrophotometer. The DNA exhibited a minimum concentration of 50 ng/µl and had an OD 260/280 ratio ranging from 1.8 to 2.0. To evaluate DNA integrity, a 1% agarose gel electrophoresis was performed.

Covaris randomly fragmented the gDNA to produce gDNA fragments up to 200 bp in length, which were then subjected to three enzymatic processes: endrepair, atailing, and adapter ligation. The DNA libraries were purified with Agencourt Hieff NGSTM DNA Selection Beads and PCR was performed. The sequencing library was prepared using the Oncology Multi-Gene Variant Assay, following the guidelines provided by the manufacturer.Commercial hybridization probes (https //www.roche.com) specifically designed for the human DNA regions of interest were used to enrich the target sequences. Afterwards, the resulting hybrid was purified, amplified, and assessed. Structural variations were annotated using the Refseq release 96 (http://www.ncbi.nlm.nih.gov/refseq/) and the Gencode gene annotation library (https://www.gencodegenes.org/). All variations exhibited a confidence level exceeding 99%, along with an average sequencing depth surpassing 500X. The annotation of structural variations was performed utilizing the Refseq release 96 (http //www.ncbi.nlm.nih.gov/refseq/) and the Gencode gene annotation library ((https://www.gencodegenes.org/). The 57 gene mutations of 150 CRC patients were analyzed according to Origin 2022, including mutation type, mutation number, and mutation frequency, and the histogram was drawn. Circos plots were generated using R software, including the distribution, mutation frequency, medication, and relationship of the 57 genes on chromosomes.

Common detection gene analysis

R software was used to analyze the genes detected in all patients and draw a waterfall map using the ComplexHeatmap and ColorBrewer packages. Through the academic analysis of Xiantao, the mutation sites of the drugable gene and the corresponding therapeutic drugs were analyzed, and the split lollipop map was drawn.Pharmacogenomics Knowledgebase (PharmGKB; http://www.pharmgkb.org/) for single nucleotide polymorphisms (SNPs) and OncoKB for single nucleotide variants (SNVs) were used to analyze actionable mutations. Therapeutic drugs are divided into four grades (A, B, C, and D) according to the PharmGKB website. The significance of each grade is as follows: A) Annotations are based on guidelines recognized by the Medical Association or conclusions/annotations recognized by certain major health systems are based on multiple repeated studies, such as the pharmacodynamic relationship is likely to be meaningful. B) Annotations are based on some studies with statistically significant differences/Annotations are based on a number of repeated studies, but some studies may not have a significant statistical difference or may have a small number of samples. C) Notes were based on only one study with significant differences (not repeated) or on multiple studies with no apparent pharmacodynamic correlation. D) Annotation is based on a small number of cases, non-authoritative studies, or in vitro molecular function studies.

The genes of more than 40 patients were defined as common detection genes, and 36 common detection genes of CRC were screened out. The mutation of 36 common CRC detection genes in 150 CRC patients was analyzed by Origin 2022, including mutation type, mutation number, and mutation frequency, and the histogram was drawn.

Validation of 36 CRC common detection genes

Mutation data from 590 CRC patients were downloaded from the TCGA database (https://www.cancer.gov/ccg/research/genome-sequencing/tcga), and the Sangerbox online tool (http://sangerbox.com/index.html) was used to analyze the mutation rate of 36 CRC common detection genes in the TCGA database. The ggplot2 package of R software was used to draw the fitting curve of the mutation rate (abscissa) in NGS and the mutation rate (ordinate) in the TCGA database.

Functional enrichment analysis of 36 CRC common detection genes

Annotating Gene Ontology (GO) terms and analyzing pathway enrichment using KEGG were performed using the DAVID v6.8 online tool (https://david-d.ncifcrf.gov/). The functional terms included biological processes, cellular components, and molecular functions. P values less than 0.05 were considered statistically significant .

Construction PPI network and analysis of mutual exclusivity

An online tool called STRING (https://string-db.org/) was used to analyze the PPI network. The list of official gene names was imported into STRING and the species was set to Homo sapiens. PPI network modules were screened using the MCODE plug-in in Cytoscape (v3.6.2, www.cytoscape.org) with the following parameters: Degree cutoff, 2; Node score cutoff, 0.2; Kcore, 2; and Max depth, 100. Hub genes were defined with the MCC algorithm in the cytoHubba plug-in. In the cBioPortal database, mutual exclusivity analysis was performed on genes .

Analysis of immune infiltration of 36 CRC common detection genes

Eight algorithms (TIMER, EPIC, MCP-COUNTER, CIBERSORT, CIBERSORT-ABS, QUANTISEQ, XCELL, and TIDE) in the TIMER2.0 database (http://timer.cistrome.org/) were used to analyze immune infiltration of CD8 + T cells, CD4 + T cells, regulatory T cells (Tregs), B cells, neutrophils, macrophages, dendritic cells (DC), natural killer cells (NK), and myeloid-derived suppressor cells (MDSCs) in 36 CRC common detection genes. First, the total score of all existing algorithms of each immune cell is set to 1, and the proportion of statistically significant algorithms among the existing algorithms is calculated. Finally, the proportion of 9 types of immune cells is added to obtain the score for the immune infiltration of each gene. The infiltration degree of 28 types of immune cells including CD8 ⁺ T cells, CD4 ⁺ T cells, Treg cells, B cells, neutrophils, macrophages, and DC cells in CRC was calculated using the GSVA package in R software, and the boxplot was drawn. Using the Venny online tool (https://bioinfogp.cnb.csic.es/tools/venny), genes with immune infiltration scores greater than 2 screened from the TIMER2.0 website were intersected with genes with statistical differences in more than 5 immune-infiltrating cells screened by the ssGSEA method, that is, common immune-related genes were obtained.

Construction and evaluation of an immune-related prognostic risk model

Clinical and survival data of 590 CRC patients from the TCGA database were downloaded. Univariate Cox risk regression analysis on immune-related genes and clinicopathologic characteristics was performed using SPSS 22.0 software, and significant variables (bilateral P < 0.05) were screened, and then multivariate Cox risk regression analysis was performed. The confidence interval (CI) was calculated at the 95% level. The prognostic risk model was constructed using the multivariate Cox risk regression coefficient, and the risk score was calculated. The calculation formula was: risk score = β1Exp1 + β2Exp2 +... + βxExpx (βi is the coefficient value, Expx is the gene expression). 590 patients with CRC were divided into two groups based on their median risk scores. The Kaplan-Meier analysis and Log-rank tests were used to evaluate the difference between the two groups, and P 0.05 was considered statistically significant. The pheatmap package of R language was used to draw the risk factor linkage map, and the differences in the distribution of survival time, survival status, and ALK gene mutation between the low-risk group and the high-risk group of CRC were analyzed. In order to evaluate the accuracy of the prognostic risk model, the area under the receiver operating characteristic curve (ROC) was calculated using the survivvalroc package in the R software. The rms package of R software was used to draw the nomogram, and the 1-year, 3-year, and 5-year survival probabilities of CRC patients were predicted by the combination of clinical stage, T classification, and ALK gene.

Follow-up and prognostic evaluation

According to the NCCN guidelines for follow-up of rectal and colon cancer, a physical examination was performed every 3–6 months; serum CEA was detected once; CT of the chest, abdominal, and pelvic was performed once a year; and fiber colonoscopy was performed once a year. For patients with persistently elevated serum CEA, CT and colonoscopy were performed in a timely manner to determine whether distant metastasis and/or local recurrence were present. The follow-up period was 49–54 months. From October 2022 to October 2023, 112 patients with CRC were followed up by telephone, mainly to observe survival time and survival status. From the date of diagnosis to the date of follow-up or death, survival time was calculated. Patients were divided into low-risk and high-risk groups based on the immune-related prognostic risk model. Log-rank tests were used to evaluate the difference between the two groups based on the Kaplan-Meier survival curves.

Correlation analysis between the immune-related prognostic risk model and MSI

From the TCGA database, clinicopathological data including clinical stage, T classification, survival data, and mutations of the ALK gene were downloaded for 92 patients with CRC and MSI when they had MSI testing results. According to the stability of MSI, patients were divided into MSI-H group and MSS group. According to the immune-related prognostic risk model, the risk scores of 92 CRC patients were calculated, and the rank sum test of SPSS 22.0 software was used to compare whether there was a statistical difference between the two groups.

Statistical analysis

A statistical analysis was conducted with SPSS software (version 19; SPPS, Inc.). All metric and normally distributed data are presented as mean ± standard deviation; In categorical data, nonnormally distributed data are presented as medians (25-75th percentiles) and n (%). The multivariate Cox analysis was screened for significant variables using the univariate Cox regression analysis (two-sided P 0.05). The Kaplan-Meier survival curve was created and analyzed using the log-rank test and the Kaplan-Meier method.

Clinicopathological characteristics of the patients with CRC

The clinicopathological features of 150 patients with CRC are shown in Table 1. There were 91 male patients (60.67%) and 59 female patients (39.33%). The age of the patients ranged from 31 to 77 years, with a median age of 46 years. All 150 patients were adenocarcinoma, of which low, medium, and high differentiated adenocarcinoma accounted for 21.33% (32/150), 68.67% (103/150), and 10.00% (15/150), respectively. According to the AJCC clinical staging (stage) criteria, the proportions of patients with stage I/II, stage III, and stage IV were 14.67% (22/150), 26.00% (39/150), and 59.33% (89/150), respectively. In TNM staging, the proportion of patients with T1, T2, T3 and T4 was 2.00% (3/150), 7.33% (11/150), 57.33% (86/150) and 33.33% (50/150), respectively. The proportions of patients with N0, N1, N2 and N3 were 33.33%(5/150), 27.33% (41/150), 36.67% (55/150) and 2.66% (4/150), respectively. The proportion of patients with stage M0 and M1 was 46.67% (70/150) and 53.33% (80/150), respectively. In addition, in the detection of serum tumor markers in 150 CRC patients, the proportion of patients with elevated CEA, CA125 and CA199 levels was 60.00% (90/150), 18.70% (28/150) and 48.00% (72/150), respectively.

Mutational analysis

A total of 57 gene panels were detected in 150 CRC patients by NGS, including 7 mutation types. These were single-nucleotide polymorphisms (SNPs), single nucleotide variations (SNVs), insertion-deletion (Indels), copy-number variation (CNV), fusion, complex, and insertion mutation (INS). As shown in Fig. 1A, SNV and Indels were the most common mutation types in CRC patients. SNV mutations occurred in 42 genes, such as TP53 (90/150), KRAS (66/150), APC (45/150), PIK3CA (36/150), FBXW7 (28/150), PTEN (15/150), KIT (14/150), and EGFR (10/150). The cumulative mutation frequency was as high as 92.67% (cumulative mutation frequency = mutation frequency/total frequency). Indels mutations occurred in 11 genes, including APC (26/150), TP53 (15/150), RB1 (6/150), TERT (6/150), ALK (5/150) and FBXW7 (5/150), with a cumulative mutation frequency of 47.33%. The 12 genes panel of TP53, APC, PTEN, BRAF, ALK, KRAS, TERT, FBXW7, TSC1, EGFR, RB1, and CTNNB1 showed both types of SNV and Indels mutations. The other five mutation types are less variable in CRC patients. The cumulative mutation frequencies of Complex, CNV, and fusion were 23.33%, 8.67%, and 8.00%, respectively. INS mutation occurred only in TP53 (1/150) and APC (1/150) genes and the cumulative mutation frequency was 4.67%. SNP variation occurred only in the PMS2 (1/150) gene, with a cumulative variation frequency of 0.67%.

As shown by the outermost circle in Fig. 1B, 12 of the 57 CRC genes are distributed on chromosomes 1 and 9, whereas the other genes are mainly found on chromosomes 3, 4, and 17. Among the 57 CRC genes, TP53 with the highest SNV frequency was located on chromosome 17, and APC with the highest frequency of Indels was located on chromosome 5. Figure 1B shows that among the 57 CRC genes, there are 39 (red) genes with targeted drug therapy and 18 (blue) genes with chemotherapeutic drugs. The histogram of the innermost circle showed the mutation frequency of the 57 CRC genes. In addition, a total of 729 gene network interactions among 57 CRC genes were identified by the String database, of which only one gene G0S did not interact with any other gene.

Common detection of gene mutational analysis

There were 12 gene panels detected in 150 patients. As shown in Fig. 2A, the mutation frequency of TP53 was the highest, and mutations occurred in 112 patients, with a mutation frequency of 74.67%(112/150). The second was the APC gene ranked second with a mutation frequency of 66.67%(100/150). The others were KRAS 44.67%(67/150), PIK3CA 24.67%(37/150), FBXW7 23.33%(35/150), PTEN 16%(24/150), BRAF 7.33%(11/150), NRAS6% (9/150), ERBB 24.67%(7/150), MAP2K 12%(3/150), PDGFRA 1.33%(2/150) and HRAS 1.33%(2/150).

At the same time, five types of mutations were observed in these 12 genes, namely SNV, Indels, CNV, fusion, and complex. In addition, among the 12 mutant genes, there are 6 drug-usable genes, namely KRAS, PIK3 CA, PTEN, BRAF, NRAS, and TP53. Figure 2B shows the mutation sites of these six drug-resistant genes and the corresponding therapeutic drugs for each site. The analysis results show that the genes with the most drug-resistant mutation sites are TP53 and PIK3CA, with TP53 having six exons EX4, EX5, EX6, EX7, EX8, and EX10 mutation sites. PIK3CA had six exons EX2, EX5, EX10, EX12, EX14, and EX21 mutations, but their medication levels were relatively low. The TP53 therapeutic drug was the class C targeted drug Adavosertib; the therapeutic drugs of PIK3CA were the class B targeted drug Alpelisib and the class D targeted drugs cetuximab, panitumumab, and chemotherapeutic drug 5-Fu in combination with irinotecan. Although there were not many mutation sites of KRAS, NRAS, and BRAS, KRAS had three exons EX2, EX3, and EX4 site mutations, and one EX4 site had a SNP mutation. In NRAS, there were three exons EX2, EX3, and EX4 mutations. There are three exons EX11, EX15, and EX18 mutations in BRAF, but their medication grades are class A targeted drugs cetuximab, panitumumab, etc., suggesting that the mutations of these three genes have important clinical significance in the relationship between the efficacy of targeted drugs. In addition, PTEN has three exons EX4, EX7, and EX8 site mutations, which are treated with class C and class A drugs. The Class D targeted drugs are cetuximab and panitumumab; two whole exon mutations were treated with Class C targeted drugs cetuximab and panitumumab.

To screen out the common detection genes of CRC, we sorted 57 CRC genes according to the number of patients detected. As shown in Fig. 2C, genes that detect more than 40 patients are defined as common detection genes. Finally, 36 CRC common detection genes were screened out: TP53, APC, KRAS, EGFR, PIK3CA, FBXW7, PTEN, BRAF, NRAS, ERBB2, MAP2K1, HRAS, PDGFRA, NTRK1, KIT, MTHFR, ROS1, RB1, GSTP1, XRCC1, CTNNB1, ALK, KDR, SMAD4, MET, NQO1, TERT, VHL, JAK1, STK11, TSC1, TSC2, AKT1, PMS2, MSH6 and MSH2.

Comparison with the TCGA database

To assess the universality of 36 common detections of gene panel in patients with CRC, the mutational status of the clinical cohort was compared with the TCGA CRC cohort. The clinical characteristics of 590 patients were obtained, including sex, age at diagnosis, histologic types, clinical stage, T classification, N classification, M classification, and overall survival status (Fig. 3A). To verify whether the mutation frequency of 36 CRC common detection genes in 150 CRC patients matched the mutation data in the TCGA database, we generated a scatter plot based on the mutation rate of the two cohorts. As shown in Fig. 3B, the R² = 0.8611 of the fitting curve is approximately close to the R² = 1 of the ideal curve, indicating that the mutation frequency of 36 CRC common detection genes is highly consistent in the two cohorts, suggesting that the 36 mutation genes screened according to the NGS data are credible.

Function annotation and PPI network analysis of 36 CRC common detection genes

GO and KEGG pathway analyses were performed to annotate the 36 CRC genes. The biological process category contained 977 terms, of which the major terms were ‘negative regulation process of apoptosis’, ‘Negative regulation of cell proliferation’ and ‘positive regulation process of kinase’ (Fig. 4A). The top enriched terms for cellular components (Fig. 4B) and molecular functions (Fig. 4C) were ‘cytosol’ and ‘protein binding’, respectively. In the analysis of KEGG pathway, many pathways related to Oncology were observed, including ‘pathways in cancer’, ‘PI3K-AKT signaling pathway’, and ‘Hepatocellular carcinoma’(Fig. 4D).

Additionally, to explore the potential relationships among the 36 CRC common gene detections, a PPI network was constructed. As shown in Fig. 5A, there were 36 nodes and 358 edges in the PPI network. A significant module had 22 nodes connected by 214 edges with a score of 20.381 (Fig. 5B). In this module, the top 10 hub genes were identified according to the MCC ranking, including TP53, PTEN, PIK3CA, KRAS, CTNNB1, NRAS, HRAS, ERBB2, BRAF and MSH6 (Fig. 5C).

The level of immune infiltration of 36 CRC common detection genes

To evaluate the relationship between 36 CRC common detection genes and infiltration by immune cell, the GSVA package was run in R for immune cell enrichment. As shown in Fig. 6A,S1,S2, the genes with more than 5 types of immune infiltrating cells were statistically different. Finally, 27 genes were obtained: STK11, HRAS, RB1, NRAS, GSTP1, TP53, NQO1, JAK1, MSH2, MTHFR, EGFR, KDR, XRCC1, APC, KRAS, TSC1, PMS2, KIT, ALK, PDGFRA, NTRK1, BRAF, PTEN, CTNNB1, FBXW7, MAP2K1 and AKT1.

At the same time, we used 8 algorithms from the TIMER2.0 database (TIMER, EPIC, MCP-COUNTER, CIBERSORT, CIBERSORT-ABS, QUANTISEQ, XCELL, TIDE). The correlation between the mutation status of 36 CRC common detection genes and the infiltration level of 9 immune cells: CD8⁺T cells, CD4⁺T cells, Treg cells, B cells, neutrophils, macrophages, DC cells, NK cells, and MDSC cells was analyzed. As shown in Fig. 6B, the round representative gene mutation was positively correlated with the level of immune cell infiltration, whereas the diamond was a negative correlation. Finally, 16 mutant genes with an immune infiltration score greater than 2 were screened: BRAF, MTHFR, APC, TSC1, MSH6, PMS2, TP53, JAK1, CTNNB1, EGFR, ALK, FBXW7, KDR, KRAS, KIT and TSC2.

Venny online tool was used to intersect 16 genes screened in the TIMER2.0 database with 27 genes screened by the ssGSEA method. Finally, 14 common immune-related genes were obtained: TP53, JAK1, MTHFR, EGFR, KDR, APC, KRAS, TSC2, PMS2, KIT, ALK, BRAF, CTNNB1 and FBXW7 (Fig. 6C).

Establishment of an immune-related prognostic risk model

Clinical data from 590 CRC patients and the mutation data of 14 immune-related genes were downloaded from the TCGA database. SPSS 22.0 software was used to perform univariate Cox risk regression analysis for14 immune-related genes and clinicopathological features. The results showed that among the 14 immune-related genes, ALK was statistically significant (P = 0.009). Clinical stage (P < 0.001), T classification (P < 0.001), N classification (P < 0.001), and M classification (P < 0.001) were statistically significant, which were related to the risk of death (Fig. 7A). Subsequent multivariate Cox regression analysis showed that clinical stage (P < 0.001), T classification (P = 0.015), and ALK (P = 0.003) were independent prognostic risk factors (Fig. 7B). Finally, the prognostic risk model was established according to the multivariate Cox risk regression coefficient, h (t,x)/h₀ (t) = EXP (0.835stage + 0.878T + 1.094ALK).

Performance evaluation of the prognostic risk model for immune-related genes

The risk scores of 590 CRC patients were calculated according to the immune prognostic risk model, and the patients were divided into high- and low-risk groups. As shown in Fig. 8A, the survival time of CRC patients in the high-risk group was significantly shorter than that in the low-risk group (P = 0.001). The risk scores of risk factors were visualized by Xiantao Academic. The 590 CRC patients were ranked from low to high based on the risk scores and divided into a low-risk group (left) and a high-risk group (right). The survival time and outcome of the patients in the high- and low-risk groups, and the expression heatmap of ALK mutant genes in the high- and low-risk groups (Fig. 8B). The results showed that there were more ALK gene mutations in the high-risk group, and the prognosis of the high-risk group was worse than that of the low-risk group. To improve the accuracy of the risk score in predicting the prognosis of patients and to facilitate clinical application, we used the clinical stage, T classification, and ALK gene to draw a Cox regression nomogram. By summingthe scores of the variables included in the modeland then drawing a vertical line at the total score, it intersects with the three lines representing the predicted survival probability, and the corresponding values of the intersections are the 1-year, 3-year, and 5-year relapse free survival (RFS) of the predicted individual (Fig. 8C). For example, a CRC patient with clinical stage IV, T3, and ALK gene mutation had a total score of approximately 53, and his 1-year, 3-year, and 5-year RFS were 81.00%, 28.50%, and 10.50%, respectively.

To evaluate the specificity and sensitivity of the prognostic risk model, we plotted the ROC curve and calculated the area under the ROC curve (AUC). As shown in Fig. 8D, the AUC of the risk model, clinical stage, T classification, and ALK gene were 0.543,0.768,0.663, and 0.536, respectively. Because the AUC was greater than 0.5, the model was considered to have good specificity and sensitivity. The above assessment of immune-related prognostic risk suggests that late clinical stage, late T classification, and ALK gene mutation will increase the risk of death and the prognosis of patients is poor.

Application of the immune-related prognostic risk model in prognosis evaluation of CRC patients

A total of 112 CRC patients were followed up. Patients were divided into a high-risk group and a low-risk group based on the risk score, including 76 cases in the high-risk group and 36 cases in the low-risk group. The Kaplan-Meier survival curve (Fig. 9A) and log-rank test analysis (P = 0.001) were drawn by SPSS version 2.0. The results showed that the survival time of patients in the low-risk group was significantly longer than that in the high-risk group, indicating that the immune-related prognostic risk model we constructed can be used to predict the prognosis of CRC patients.

The immune-related prognostic risk model is associated with MSI in CRC patients

The clinical data of 92 CRC patients with MSI detection were downloaded from the TCGA database. According to the stability of MSI, patients were divided into MSI-H group and MSS group, including 23 patients in MSI-H group and 69 patients in MSS group. The risk score of 92 patients was calculated according to the immune prognosis risk model. The risk scores of MSI-H group and MSS group were compared by rank sum test. The results showed that the risk score of CRC patients in the MSI-H group was significantly higher than that in the MSS group, and there was a statistical difference (P = 0.020). According to the 2023 Chinese Society of Clinical Oncology (CSCO) guidelines for the diagnosis and treatment of colorectal cancer, immune checkpoint inhibitors are recommended as grade III for the first-line palliative treatment of MSI-H CRC patients. The results suggest that the immune-related prognostic risk model can be used to predict the efficacy of immunotherapy (Fig. 9B).

In this study, we detected the mutations of 57 genes in CRC patients based on NGS. These 57 genes are all cancer-driven genes, and their screening meets the following three conditions: i) Genes that are clinically significant using the NCCN guideline as a reference. The NCCN guideline is a clinical practice guidefor various malignant tumors published annually by the United States National Comprehensive Cancer Network. It is recognized and followed by clinicians around the world. It is an internationally recognized ‘gold standard’ for clinical practice norms in the oncology community. The target genes recommended in the NCCN guidelines for the treatment of CRC drugs are often the key to molecular targeted therapy [21]. ii) At the same time, the screening of genes in this study also referred to databases for some molecular targeted therapies, such as TTD, CKB, OncoKB database, etc. Dedicated to drug targets, TTD provides information on therapeutic protein and nucleic acid targets, targeted diseases, pathway information, and corresponding drugs. The CKB database focuses on datasets derived from genes from clinical patient samples, and then performs a large-scale analysis of cancer genome variations, which contributes greatly to precision medicine; oncoKB is a comprehensive and accurate oncology knowledge base maintained by the Memorial Sloan Kettering (MSK) Cancer Center. It includes professional guidelines and recommendations from the FDA, NCCN, or American Cancer Society (ACS), as well as scientific literature, treatment strategies, consensus reached by oncologists or cancer societies, and reference literature research. OncoKB currently contains detailed information on 554 cancer gene-specific changes, as well as treatment information for level 1 FDA approval, level 2 standard care, level 3 clinical evidence and biological evidence. iii) In addition, some of the genes selected in this study have been reported extensively for clinical research [22].

SNV is called single nucleotide mutation, which refers to the mutation of nucleotide sequences in the genome. This mutation is one of the most common and easiest types of genetic variation to find and analyze. The cause of SNV is mainly due to an incorrect copy in the process of DNA replication. SNV is closely related to the structure and function of microorganisms and is essential for maintaining microbial homeostasis in the gut. In recent years, studies have shown that microbial SNV can be used to diagnose many diseases and is a novel and valuable biomarker for early screening of tumors [23]. In the study to detect potential targets for poor prognosis in cetuximab treatment, Gu et al. [24]analyzed the expression profiles of 416 genes in 33 CRC patients by NGS, and found a total of 411 somatic mutations, including 320 SNV mutations, 38 CNVs, and 58 Indels. This mutation type had a higher number of SNVs than other mutations. The most frequently mutant gene was TP53, followed by APC. In addition, Tang et al. [25] analyzed 24 CRC cell-specific mutant genes by whole-exome sequencing of CRC single cells and verified their role in cell migration. Whole-exome sequencing data showed that 2636 somatic SNVs and 456 somatic Indels were identified in 7 CRC patients, and the mutation frequency of SNVs was more than 50% of the other mutation types. In this study, 57 genes were analyzed for mutation types in 150 CRC patients. SNVs and Indels were the most common mutation types in CRC patients. The cumulative mutation frequency of SNV was as high as 92.67%, and the cumulative mutation frequency of Indels was 47.33%. In addition, 12 genes showed both SNV and Indels mutation types. Among the 57 genes detected, 12 genes were detected in all 150 CRC patients : TP53, APC, KRAS, PIK3CA, FBXW7, PTEN, BRAF, NRAS, ERBB2, MAP2K1, PDGFRA, and HRAS. According to the relevant guidelines and practical recommendations at home and abroad, mutations in the corresponding targets of CRC-related genes (including APC, MMR, RAS and BRAF, etc.) can be used to guide medication. For example, the synergistic effect of APC deletion and KRAS mutation leads to the formation of CRC drug resistance, resulting in poor survival rate of APC/KRAS mutant CRC patients and increasing the difficulty of treatment [26–27]. In this study, we found that TP53, PIK3 CA, PTEN, BRAF, KRAS, and NRAS are the drug-resistant genes among the above 12 genes. There are no effective therapeutic drugs for the remaining 6 genes. Among the six druggable genes, the druggable grades for TP53 and PIK3CA mutation targets were all grade C, which is due to the fact that the US FDA has not approved any targeted drugs for TP53; there are no PIK3CA inhibitors on the market, especially gene therapy, targeted tumor vaccines and anticancer drugs for TP53 and PIK3CA mutations are in the early stage of clinical trials. The mutation sites of the RAS gene account for the vast majority of cases of CRC, but drugs that directly target the RAS gene are still under investigation. Because RAS genes are involved in the regulation of multiple signaling pathways, such as the interaction between RAS and receptor tyrosine kinases (EGFR family, etc.), the EGFR family is located upstream of RAS, and inactivation of these receptor tyrosine kinases may decrease RAS activation. Therefore, it can be used to treat CRC patients with RAS gene mutations by inhibiting EGFR family targets such as BRAF and EGFR [28]. In addition, the therapeutic effect can also be achieved by inhibiting the MAPK pathway downstream of the RAS gene, such as the use of RAF inhibitors, MEK inhibitors and ERK inhibitors. One can even treat RAS-mutated CRC tumors using inhibitors of the PI3K pathway, because RAS can activate the PI3K and MAPK pathways and there is some feedback regulation mechanism. However, it should be noted that inhibition of one pathway may lead to compensatory activation of another pathway, so the combination of MAPK and PI3K pathways maybe inhibited, and the therapeutic effect on CRC patients is better, but the side effects also increase [29]. At present, further research on the drug target of RAS gene is still needed [30].

We defined 36 genes of more than 40 CRC patients as common detection genes of CRC and compared them with the gene mutation data of CRC patients downloaded from the TCGA database to construct a fitting curve. It was confirmed that the mutation frequency of these 36 common detection genes of CRC in 150 clinical patients was recognition consistent with the trend of mutation frequency trend in the TCGA database, indicating that the reliability of the common detection genes of CRC screened by NGS was high. Functional annotations were then performed for 36 common CRC detection genes. Enrichment analysis suggested that the occurrence of CRC may be related to kinases in the PI3K-AKT signaling pathway and protein serine / threonine / tyrosine kinase activity. Different clinical features, biological pathways, and copy number changes may cause activation of multiple kinase signals, and then related phosphorylation cascades. Therefore, in the treatment of CRC targeted drugs, there are mainly two categories of kinase inhibitors that mainly inhibit receptor tyrosine kinase targets [31–33].

Besides tumor cells, tumor tissue also contains stroma, fibroblasts, and immune cells, the cells consists of the tumor micro-environment (TME), and we will infiltrate into the tumor lymphocytes called tumor infiltrating lymphocytes (TILs). TILs have a function role for tumors. On the one hand, these immune cells can play a role in immune surveillance and prevent tumor progression. On the other hand, TILs may also allow tumors to escape, thereby promoting tumor progression [34]. Therefore, understanding tumor immune infiltration is crucial for deciphering immunotherapy mechanisms, identifying predictive biomarkers, and discovering new therapeutic targets. Mlecnik B et al. [35] analyzed the relationship between immune infiltration in the tumor center and the invasive margin of the tumor and the frequency of recurrence in 599 cases of stage I to IV CRC specimens. The results showed that the growth, metastasis, and diffusion of primary CRC tumors were related to the decrease in the density of immune T cells in the tumor. Assessment based on the type and density of TILs could be a strong prognostic factor, even better than the separate TNM staging prediction criteria. In addition, Angelova M et al. [36] used the mutant gene data set of 598 CRC patients in the TCGA database to identify the relationship between TILs subsets and MSI and methylation status, and proved that MSI-H can lead to T cell activation and infiltration, and vice versa. It is shown that tumor lymphocyte infiltration can not only help to determine tumor progression and improve the effect of immunotherapy, but also help to formulate the best clinical treatment plan. Therefore, this study compared immune cell infiltration using the TIMER2.0 immune infiltration database and ssGSEA enrichment analysis. After overlapping the two methods, 14 genes were foundto be associated with immune infiltration of CRC patients. On this basis, an immune-related prognostic risk model was constructed, which is an important basis for the risk model used to evaluate the efficacy of immunotherapy in CRC patients.

We analyzed the mutations of 14 immune-related genes in CRC patients through the TCGA database and combined them with clinicopathological data for Cox risk regression analysis to construct an immune-related prognostic risk model. Univariate and multivariate Cox risk regression analysis revealed that the poor prognosis of CRC patients might be related to ALK gene mutation. We divided 112 CRC patients into a high-risk group and a low-risk group by the median of risk score. A Kaplan-Meier curve was constructed according to the actual follow-up. It can be seen that the survival time of patients in the low-risk group was significantly longer than that the high-risk group. Numerous publications have reported that CRC patients have gene rearrangements encoding receptor tyrosine kinases, such as rearrangements of ALK, ROS1, RET, and NTRK1-3 tyrosine kinases, which will lead to the fusion of kinase domains with active transcription genes. Expression of some related proteins in the kinase domain of the receptor tyrosine kinaseincrease, leading to the activation of the downstream pathway of the tyrosine kinase [37–38]. When ALK gene rearrangement occurs, it may activate downstream signaling pathways : MAPK/MEK/ERK, PI3K/AKT and RAS/STAT3 [31]. Rearrangement of tyrosine kinases such as ALK is highly likely to occur in CRC patients who do not respond to treatment with EGFR monoclonal antibodies. This explains the limited benefit of EGFR monoclonal antibodies in some RAS/BRAF wild-type CRC patients. Therefore, for CRC patients with tyrosine kinase gene rearrangement such as ALK, factors associated with the gene rearrangement such as ALK should also be considered in addition to the appropriate targeted therapy [38]. At the same time, a large proportion of CRC patients with gene rearrangements such as ALK have MSI-H status, which provides more treatment options for future combination therapy (targeted therapy combined with immunotherapy for ALK) [33].

According to the NCCN Oncology Clinical Practice Guidelines published in 2023, the two subtypes of pMMR/MSS and dMMR/MSI-H are officially used as global classification criteria for the diagnosis and treatment of colon cancer, which means that mismatch repair gene/microsatellite status has officially become selection criterion for the treatment strategy for all patients with early, middle and advanced colon cancer. The treatment of CRC, immune checkpoint therapy is mainly used in patients with dMMR or MSI-H severe mutations, whereas it is less effective for tumors with MSI-L [39–42]. MSI-H/dMMR is a predictor of immunotherapy efficacy in pan cancers, including CRC. The MSI-H phenotype is closely related to the associated high-density tumor-infiltrating lymphocytes. This is primarily because CRC patients with MSI-H often exhibit cytotoxic immune infiltration triggered by new antigens, which contributes to their response to immunotherapy [43]. This link may be derived from its own anti-tumor immune response. When the MMR system is defective, the immune microenvironment selectively upregulates certain immune checkpoints including PD-L1 [44]. PD-L1 can bind and activate PD-1 on the surface of T cells, inhibit the proliferation and activation of T cells, induce their apoptosis, and further make tumor cells escape immune surveillance [45]. When the interaction of PD-1 / PD-L1 is blocked, the inhibitory effect of tumor cells on T cells can be eliminated, and eventually the infiltration of tumor T cells is achieved, and the anti-tumor effect of the immune system is enhanced. The mechanism described above may be one of the most important reasons why MSI-H CRC patients are more likely to benefit from immunosuppressive therapy. It is speculated that MSI may increase tumor mutation burden (TMB) and response to immune checkpoint inhibitors [46–48]. In this study, the clinicopathological data of 92 CRC patients were downloaded from the TCGA database. The CRC patients were divided into the MSS group and the MSI-H group. The risk scores of each of the two groups were compared. It can be observed that the risk score of MSS patients was significantly lower than that of the MSI-H group, suggesting that the immune-related prognostic risk model in this study may have certain significance in evaluating the efficacy of immunotherapy.

In conclusion, this study detected gene mutations in clinical CRC patients by NGS and found that ALK gene mutations were associated with prognosis and immunotherapy of CRC tumors. The immune-related prognostic risk model constructed in combination with clinicopathological features can be used to predict the prognosis and efficacy of immunotherapy of CRC patientsand provide new ideas and directions for the treatment of clinical CRC patients. In the follow-up study, we will collect the specimens of CRC patients treated with immunotherapy, and further evaluate the application of the immune prognostic risk model in the efficacy of immunotherapy. At the same time, further research is needed to elucidate the potential mechanism.

CRC : colorectal cancer

EGFR : epidermal growth factor receptor

HER2 : human epidermal growth factor receptor-2

VEGF : vascular endothelial growth factor

VEGFR-1/2/3 : vascular endothelial growth factor receptor-1/2/3

PDGF : platelet-derived growth factor

RET : Ret proto-oncogene tyrosine-protein kinase receptor Ret

NTRK1/2/3 : neurotrophic receptor kinase 1/2/3

PDGFR : platelet-derived growth factor receptors

TIE-2 : angiopoietin-1 receptor

c-kit : mast/stem cell growth factor receptor Kit

ROS1 : c-ros sarcoma proto-oncogene 1-receptor tyrosine kinase

BRAF : vrafmurine sarcoma viral oncogene homolog B

MEK1/2 : mitogen-activated protein kinase 1/2

PIK3CA : phosphatidylinositol-4,5-bisphosphate 3-kinase

NGS : next generation sequencing

CTLA-4 : cytotoxic T lymphocyte-associated antigen-4

PD-1 : programmed death receptor-1

PD-L1 : programmed death receptor-1 ligand

dMMR : mismatch repair deficiency

MSI-H : microsatellite instability-high

MSS : microsatellite stability

pMMR : mismatch repair incomplete

MSI-L : microsatellite instability-low

ssGSEA : single-sample gene set enrichment analysis

CEA : carcinoembryonic antigen

CA125 : carbohydrate antigen 125

CA199 : carbohydrate antigen 199

AJCC : American Joint Committee on Cancer

gDNA : genetic material

SNPs : single nucleotide polymorphisms

SNVs : single nucleotide variants

Indels : insertion-deletion

CNV : opy-number variation

INS : insertion mutation

Tregs : regulatory T cells

DC : endritic cells

NK : natural killer cells

MDSCs : myeloid-derived suppressor cells

CI : confidence interval

ROC : receiver operating characteristic curve

AUC : area under the ROC curve

CSCO : Chinese Society of Clinical Oncology

ACS : American Cancer Society

TME : the tumor micro-environment

TILs : tumor infiltrating lymphocytes

TMB : tumor mutation burden

Acknowledgements

We thank the patients who provided us the biological samples used in that study.

Funding

This study was supported by the present study was funded by the Natural Science Foundation of Hebei Province, China (No. 236Z7701G), the National Natural Science Foundation of China (No. 81703001), the Natural Science Foundation of Hebei Province (No. H2021406021), Hebei Province Key Research and Development Plan Projects (No. 19277783D), Chengde Medical University Discipline Construction Funds, Hebei Province Health Commission Government-funded Training Program for Outstanding Clinical Medical Talents.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. TCGA dataset analyzed for this study can be found in the cBioPortal database (https://www.cbioportal.org/study/summary?id=crc_msk_2017%2Crectal_msk_2019).

Author information

Authors’contributions

All authors contributed to the design of the study, drafted the manuscript and will analyse the clinical data. All authors have read and approved the final manuscript.

Ethics approval and consent to participate

This study was approved by the Research Ethics Review Committee of Chengde Medical University (Hebei, China; approval no. 202215). All participants provided written informed consent. All methods were carried out in accordance with relevant guidelines and regulations i.e. Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests

Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023,73(1): 17-48. https://doi.org/10.3322/caac.21763.
Yang Y, Wang HY, Chen YK, et al. Current status of surgical treatment of rectal cancer in China[J].Chin Med J (Engl). 2020,133(22): 2703-2711. https://doi.org/10.1097/CM9.0000000000001076.
Li C, Sun YD, Yu GY, et al. Integrated omics of metastatic colorectal cancer[J]. Cancer Cell. 2020,38(5): 734-747. https://doi.org/10.1016/j.ccell.2020.08.002.
CÁCERES W, CRUZ-CHACÓN A. Renal cell carcinoma: molecularly targeted therapy[J]. PR Health Sci J. 2011,30(2):73-77.
HOORENS A, JOURET-MOURIN A, SEMPOUX C, et al. Accurate KRAS mutation testing for EGFR targeted therapy in colorectal Cancer: emphasis on the key role and responsibility of pathologists[J]. Acta Gastroenterol Belg. 2010,73(4): 497-503.
LIVRE A, BACHET JB, BOIGE V, et al. KRAS mutations as an in- dependent prognostic factor in patient swithadvanced colorectal cancer treated with cetuximab[J]. J clin oncol. 2008,26(3): 374-379.https://doi.org/10.1200/JCO.2007.12.5906.
STINTZING S, MODEST DP, ROSSIUS L, et al. FOLFIRI plus cetuximab versus FOLFIRI plus bevacizumab for metastatic colorectal cancer (FIRE-3): a post-hoc analysis of tumour dynamics in the final KRAS wild-type subgroup of this randomised open-label phase 3 trial[J]. Lancet Oncol. 2016,17(10): 1426-1434. https://doi.org/10.1016/S1470-2045(16)30269-8.
KHATTAK MA, MARTIN H, DAVIDSON A, et al. Role of first-line anti-epidermal growth factor receptor therapy compared with anti-vascular endothelial growth factor therapy in advanced colorectal cancer: a metaanalysis of randomized clinical trials[J]. Clin Colorectal Cancer. 2015,14(2): 81-90. https://doi.org/10.1016/j.clcc.2014.12.011.
Giantonio BJ, Catalano PJ, Meropol NJ, et al. Bevacizumab in combination with oxaliplatin,fluorouracil,and leucovorin (FOLFOX4) for previously treated metastatic colorectal cancer: results from the Eastern Cooperative Oncology Group Study E3200[J]. J Clin Oncol. 2007,25(12): 1539-1544. https://doi.org/10.1200/JCO.2006.09.6305.
QIN S, LI A, YI M, et al. Recent advances on antiangiogenesis receptor tyrosine kinase inhibitors in cancer therapy[J]. J Hematol Oncol. 2019,12(1): 27. https://doi.org/10.1186/s13045-019-0718-5.
HUGHES PE, CAENEPEEL S, WU LC. Targeted therapy and checkpoint immunotherapy combinations for the treatment of cancer[J]. Trends Immunol. 2016,37(7): 462-476. https://doi.org/10.1016/j.it.2016.04.010.
Ganesh K, Stadler ZK, Cercek A, Mendelsohn RB, Shia J, Segal NH, Diaz LA Jr. Immunotherapy in colorectal cancer:rationale,challenges and potential. Nat Rev Gastroenterol Hepatol. 2019,(6): 361-375. https://doi.org/10.1038/s41575-019-0126-x.
Siegel RL, Wagle NS, Cercek A, Smith RA, Jemal A. Colorectal cancer statistics. CA Cancer J Clin. 2023,73(3): 233-254. https://doi.org/10.3322/caac.21772.
Cann CG, LaPelusa MB, Cimino SK, Eng C. Molecular and genetic targets within metastatic colorectal cancer and associated novel treatment advancements. Front Oncol. 2023,13: 1176950. https://doi.org/10.3389/fonc.2023.1176950.
FARKONA S, DIAMANDIS EP, BLASUTIG IM. Cancer immunotherapy:the beginning of the end of cancer [J]. BMC Med. 2016,14: 73. https://doi.org/10.1186/s12916-016-0623-5.
AGARWAL MB. Is cancer chemotherapy dying[J]. Asian J Transfus Sci. 2016,10(Suppl1): S1-7. https://doi.org/10.4103/0973-6247.182735.
BRAHMER JR, DRAKE CG, WOLLNERI, et al. Phase I study of single-agentanti-programmed death-1 (MDX-1106) in refractory solid tumors: safety,clinical activity,pharmacodynamics,and immunologic correlates[J]. J Clin Oncol. 2010,28(19): 3167-3175. https://doi.org/10.1200/JCO.2009.26.7609.
QUINTANILHA J, GRAF RP, FISHER VA, et al. Comparative Effectiveness of Immune Checkpoint Inhibitors vs Chemotherapy in Patients With Metastatic Colorectal Cancer With Measures of Microsatellite Instability,Mismatch Repair,or Tumor Mutational Burden[J]. JAMA Netw Open. 2023,6(1): e2252244. https://doi.org/10.1001/jamanetworkopen.2022.5394.
Formslag CR, Zhao L, Heslin AJ, Lewis CC, Miller CW, Bai Q, Wakefield MR, Fang Y. The past,present,and future of immunotherapy for colorectal cancer. Med Oncol. 2023,40(3): 95. https://doi.org/10.1007/s12032-023-01967-1.
Mui M, Kong JCH, Guerra GR, Behrenbruch CC, Michael M, Heriot AG. Immunotherapy in colorectal cancer: a beacon of hope. ANZ J Surg. 2023,93(1-2): 15-16. https://doi.org/10.1111/ans.18142.
NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines). Colon Carcinoma. Version 1.2023
Martínez-Jiménez F, Muiños F, Sentís I, Deu-Pons J, Reyes-Salazar I, Arnedo-Pac C, Mularoni L, Pich O, Bonet J, Kranas H, Gonzalez-Perez A, Lopez-Bigas N. A compendium of mutational cancer driver genes. Nat Rev Cancer. 2020,20(10): 555-572. https://doi.org/10.1038/s41568-020-0290-x.
Yang L, Wang S, Lee JJ, Lee S, Lee E, Shinbrot E, Wheeler DA, Kucherlapati R, Park PJ. An enhanced genetic model of colorectal cancer progression history. Genome Biol. 2019,20(1): 168. https://doi.org/10.1186/s13059-019-1782-4.
Mei Z, Shao YW, Lin P, Cai X, Wang B, Ding Y, Ma X, Wu X, Xia Y, Zhu D, Shu Y, Fu Z, Gu Y. SMAD4 and NF1 mutations as potential biomarkers for poor prognosis to cetuximab-based therapy in Chinese metastatic colorectal cancer patients. BMC Cancer. 2018,18(1): 479. https://doi.org/10.1186/s12885-018-4298-5.
Tang J, Tu K, Lu K, Zhang J, Luo K, Jin H, Wang L, Yang L, Xiao W, Zhang Q, Liu X, Ge XY, Li G, Zhou Z, Xie D. Single-cell exome sequencing reveals multiple subclones in metastatic colorectal carcinoma. Genome Med. 2021,13(1): 148. https://doi.org/10.1186/s13073-021-00962-3.
Wong CC, Wu JL, Ji F, Kang W, Bian X, Chen H, Chan LS, Luk STY, Tong S, Xu J, Zhou Q, Liu D, Su H, Gou H, Cheung AH, To KF, Cai Z, Shay JW, Yu J. The cholesterol uptake regulator PCSK9 promotes and is a therapeutic target in APC/KRAS-mutant colorectal cancer. Nat Commun. 2022,13(1): 3971. https://doi.org/10.1038/s41467-022-31663-z.
Poloznikov A, Nikulin S, Bolotina L, Kachmazov A, Raigorodskaya M, Kudryavtseva A, Bakhtogarimov I, Rodin S, Gaisina I, Topchiy M, Asachenko A, Novosad V, Tonevitsky A, Alekseev B. 9-ING-41, a Small Molecule Inhibitor of GSK-3β, Potentiates the Effects of Chemotherapy on Colorectal Cancer Cells. Front Pharmacol. 2021,12: 777114. https://doi.org/10.3389/fphar.2021.777114.
Wong CC, Wu JL, Ji F, Kang W, Bian X, Chen H, Chan LS, Luk STY, Tong S, Xu J, Zhou Q, Liu D, Su H, Gou H, Cheung AH, To KF, Cai Z, Shay JW, Yu J. The cholesterol uptake regulator PCSK9 promotes and is a therapeutic target in APC/KRAS-mutant colorectal cancer. Nat Commun. 2022,13(1): 3971. https://doi.org/10.1038/s41467-022-31663-z.
Hao Y, He B, Wu L, Li Y, Wang C, Wang T, Sun L, Zhang Y, Zhan Y, Zhao Y, Markowitz S, Veigl M, Conlon RA, Wang Z. Nuclear translocation of p85β promotes tumorigenesis of PIK3CA helical domain mutant cancer. Nat Commun. 2022,;13(1): 1974. https://doi.org/10.1038/s41467-022-29585-x.
Moore AR, Rosenberg SC, McCormick F, Malek S. RAS-targeted therapies: is the undruggable drugged? Nat Rev Drug Discov. 2020,19(8): 533-552. https://doi/.org/10.1038/s41573-020-0068-6.
Okano S, Yamashiro Y, Onagi H, Sasa K, Hayashi T, Takahashi M, Sugimoto K, Sakamoto K, Yao T, Saito T. Tyrosine kinase alterations in colorectal cancer with emphasis on the distinct clinicopathological characteristics. Histopathology. 2023,83(5): 733-742. https://doi.org/10.1111/his.15015.
Coant N, García-Barros M, Zhang Q, Obeid LM, Hannun YA. AKT as a key target for growth promoting functions of neutral ceramidase in colon cancer cells. Oncogene. 2018,37(28):3852-3863. https://doi.org/10.1038/s41388-018-0236-x
Mulkidjan RS, Saitova ES, Preobrazhenskaya EV, Asadulaeva KA, Bubnov MG, Otradnova EA, Terina DM, Shulga SS, Martynenko DE, Semina MV, Belogubova EV, Tiurin VI, Amankwah PS, Martianov AS, Imyanitov EN. ALK, ROS1, RET and NTRK1-3 Gene Fusions in Colorectal and Non-Colorectal Microsatellite-Unstable Cancers. Int J Mol Sci. 2023,24(17): 13610. https://doi.org/10.3390/ijms241713610.
Tang T, Huang X, Zhang G, Hong Z, Bai X, Liang T. Advantages of targeting the tumor immune microenvironment over blocking immune checkpoint in cancer immunotherapy. Signal Transduct Target Ther. 2021,6(1): 72. https://doi.org/10.1038/s41392-020-00449-4.
Mlecnik B, Tosolini M, Kirilovsky A, Berger A, Bindea G, Meatchi T, Bruneval P, Trajanoski Z, Fridman WH, Pagès F, Galon J. Histopathologic-based prognostic factors of colorectal cancers are associated with the state of the local immune reaction. J Clin Oncol. 2011,29(6): 610-8. https://doi.org/10.1200/JCO.2010.30.5425.
Angelova M, Charoentong P, Hackl H, Fischer ML, Snajder R, Krogsdam AM, Waldner MJ, Bindea G, Mlecnik B, Galon J, Trajanoski Z. Characterization of the immunophenotypes and antigenomes of colorectal cancers reveals distinct tumor escape mechanisms and novel targets for immunotherapy. Genome Biol. 2015,16(1): 64. https://doi.org/10.1186/s13059-015-0620-6.
Lee J, Kim HC, Hong JY, Wang K, Kim SY, Jang J, Kim ST, Park JO, Lim HY, Kang WK, Park YS, Lee J, Lee WY, Park YA, Huh JW, Yun SH, Do IG, Kim SH, Balasubramanian S, Stephens PJ, Ross JS, Li GG, Hornby Z, Ali SM, Miller VA, Kim KM, Ou SH. Detection of novel and potentially actionable anaplastic lymphoma kinase (ALK) rearrangement in colorectal adenocarcinoma by immunohistochemistry screening. Oncotarget. 2015,6(27): 24320-32. https://doi.org/10.18632/oncotarget.4462.
Houang M, Toon CW, Clarkson A, Sioson L, de Silva K, Watson N, Singh NR, Chou A, Gill AJ. ALK and ROS1 overexpression is very rare in colorectal adenocarcinoma. Appl Immunohistochem Mol Morphol. 2015,23(2): 134-8. https:// doi.org/10.1097/PAI.0000000000000025.
Aurelien Marabelle , Dung T Le, Paolo A Ascierto, et al. Efficacy of Pembrolizumab in Patients With Noncolorectal High Microsatellite Instability/Mismatch Repair-Deficient Cancer: Results From the Phase II KEYNOTE-158 Study[J]. J Clin Oncol. 2020; 38(1): 1-10.https:// doi.org/10.1200/JCO.19.02105.
Zhang X, Wu T, Cai X, Dong J, Xia C, Zhou Y, Ding R, Yang R, Tan J, Zhang L, Zhang Y, Wang Y, Dong C, Li Y. Neoadjuvant Immunotherapy for MSI-H/dMMR Locally Advanced Colorectal Cancer: New Strategies and Unveiled Opportunities. Front Immunol. 2022,13: 795972. https://doi.org/10.3389/fimmu.2022.795972.
Qin Y, Liu Y, Xiang X, Long X, Chen Z, Huang X, Yang J, Li W. Cuproptosis correlates with immunosuppressive tumor microenvironment based on pan-cancer multiomics and single-cell sequencing analysis. Mol Cancer. 2023,22(1): 59. https:// doi.org/10.1186/s12943-023-01752-8.
Chen B, Scurrah CR, McKinley ET, Simmons AJ, Ramirez-Solano MA, Zhu X, Markham NO, Heiser CN, Vega PN, Rolong A, Kim H, Sheng Q, Drewes JL, Zhou Y, Southard-Smith AN, Xu Y, Ro J, Jones AL, Revetta F, Berry LD, Niitsu H, Islam M, Pelka K, Hofree M, Chen JH, Sarkizova S, Ng K, Giannakis M, Boland GM, Aguirre AJ, Anderson AC, Rozenblatt-Rosen O, Regev A, Hacohen N, Kawasaki K, Sato T, Goettel JA, Grady WM, Zheng W, Washington MK, Cai Q, Sears CL, Goldenring JR, Franklin JL, Su T, Huh WJ, Vandekar S, Roland JT, Liu Q, Coffey RJ, Shrubsole MJ, Lau KS. Differential pre-malignant programs and microenvironment chart distinct paths to malignancy in human colorectal polyps. Cell. 2021,184(26): 6262-6280.e26. https://doi.org/10.1016/j.cell.2021.11.031.
FDA grants nivolumab accelerated approval for MSI-H or dMMR colorectal cancer. [EB/OL]. [2020-09-19]. https://www.fda.gov/drugs/resources-information-approved-drugs/fda-grants-nivolumab-accelerated-approval-msi-h-or-dmmr-colorectal-cancer.
Derks S, Liao X, Chiaravalli AM, et al. Abundant PD-L1 expression in Epstein-Barr Virus-infected gastric cancers[J]. Oncotarget, 2016,7(22): 4052-4052. https://doi.org/10.18632/oncotarget.9076.
Dung T Le , Jennifer N Durham , Kellie N Smith, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade[J]. Science. 2017,357(6349): 409-413. https://.doi.org/10.1126/science.aan6733.
Derks S, Liao X, Chiaravalli AM, et al. Abundant PD-L1 expression in Epstein-Barr Virus-infected gastric cancers[J]. Oncotarget, 2016,7(22): 4052-4052. https://.doi.org/10.18632/oncotarget.9076.
Oliveira AF, Bretes L, Furtado I. Review of PD-1/PD-L1 Inhibitors in Metastatic dMMR/MSI-H Colorectal Cancer. Front Oncol. 2019,9:396. https://.doi.org/10.3389/fonc.2019.00396.
Xin H, Zhou C, Wang G, Liu Y, Zhang J, Liu Y, Li B, Zhang J, Su M, Li Z, Wang G. Heterogeneity of PD-L1 expression and CD8 lymphocyte infiltration in metastatic colorectal cancer and their prognostic significance. Heliyon. 2023,9(2):ve13048. https:// doi.org/10.1016/j.heliyon.2023.e13048.

Table 1 is available in the Supplementary Files section.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
01 Feb, 2024
Reviews received at journal
06 Jan, 2024
Reviewers agreed at journal
05 Jan, 2024
Reviewers invited by journal
26 Nov, 2023
Editor assigned by journal
26 Nov, 2023
Editor invited by journal
20 Nov, 2023
Submission checks completed at journal
20 Nov, 2023
First submitted to journal
11 Nov, 2023

You are reading this latest preprint version

An immune-related prognostic risk model for colorectal cancer was constructed based on next-generation sequencing

Status:

Version 1

Abstract

Objective

Methodology:

Results

Conclusion

Figures

Introduction

Materials and Methods

Patients and specimens

Design and general performance of the CRC 57 gene NGS panel

Common detection gene analysis

Validation of 36 CRC common detection genes

Functional enrichment analysis of 36 CRC common detection genes

Construction PPI network and analysis of mutual exclusivity

Analysis of immune infiltration of 36 CRC common detection genes

Construction and evaluation of an immune-related prognostic risk model

Follow-up and prognostic evaluation

Correlation analysis between the immune-related prognostic risk model and MSI

Statistical analysis

Results

Clinicopathological characteristics of the patients with CRC

Mutational analysis

Common detection of gene mutational analysis

Comparison with the TCGA database

Function annotation and PPI network analysis of 36 CRC common detection genes

The level of immune infiltration of 36 CRC common detection genes

Establishment of an immune-related prognostic risk model

Performance evaluation of the prognostic risk model for immune-related genes

Application of the immune-related prognostic risk model in prognosis evaluation of CRC patients

The immune-related prognostic risk model is associated with MSI in CRC patients

Discussion

Conclusion

Abbreviations

Declarations

References

Table

Additional Declarations

Supplementary Files

Status:

Version 1