Cervical cancer recurrence related gene, cervical cancer recurrence prediction model and system

Abstract

The invention relates to the technical field of biology, in particular to a cervical cancer relapse related gene, a cervical cancer relapse prediction model and a cervical cancer relapse prediction system. The system comprises a memory, a burying device and a computer program which is stored on the memory and can run on a processor; when the processor executes the computer program, the following steps are implemented: comparing somatic cells of a recurrent group of the cervical cancer patients with somatic cells of a non-recurrent group of the cervical cancer patients, and screening out target mutant genes meeting preset conditions; inputting the target mutant gene into a cervical cancer recurrence prediction model, and determining a recurrence-free correlation score; and performing cervical cancer recurrence prediction on the cervical cancer patient according to the recurrence-free correlation score. According to the application, mutation conditions of DCHS2, DNAH10, RYR1 and WDFY4 in tumor tissues of the cervical cancer patient are detected, and the cervical cancer recurrence prediction model is constructed, so that the recurrence risk of the cervical cancer patient is predicted through the recurrence-free correlation score, and certain help is provided for prognosis evaluation of the cervical cancer patient.

Description

Background Art

[0002] Cervical cancer (CC) is the second most common cancer of the female reproductive system, with an estimated 604,127 new cases and 341,831 deaths in 2020. Despite improved treatments for CC, approximately 6.4%-18% of patients with early-stage disease relapse after initial treatment, and the recurrence rate for patients with locally advanced CC (LACC) further increases to 8%-40%. Recurrent CC is mostly discovered within two years after initial treatment, and the prognosis is often poor due to limited treatment options. On the other hand, delayed recurrence detection due to irregular follow-up strategies, which leads to delayed treatment, may also lead to a poor prognosis for CC. Therefore, it is crucial to predict the clinical outcomes of CC in advance, conduct early intervention, and improve the prognosis of CC.

[0003] Previous studies have identified certain clinicopathological parameters (such as tumor stage, histological type, tumor size, and lymphovascular space invasion), radiomic features, and transcriptomic features as predictors of cervical cancer recurrence and prognosis; however, the accuracy of these markers is not always satisfactory in clinical practice. With the rapid development of NGS technology, the importance of mutation-based biomarkers in predicting recurrence and prognosis of various cancers has been increasingly recognized, because genetic mutations play a crucial role in the development of cancer. Another rationale for using genetic mutations as cancer biomarkers is that these mutations are able to generate new antigens, thereby inducing tumor-specific immune responses and leading to tumor regression. Therefore, tumor mutation burden (TMB) was initially identified as a predictor of patient response to immunotherapy and has also been used to predict clinical outcomes in patients with various solid tumors, including CC.

[0004] However, there are obviously relatively few studies on cervical cancer in this area. Currently, there is no clinically applied and commercialized cervical cancer prognosis recurrence risk assessment system in the world to guide the prediction of cervical cancer recurrence. Summary of the invention

[0005] In order to overcome the problem that there is no recurrence risk assessment system for cervical cancer prognosis to guide the prediction of cervical cancer recurrence, the present invention provides cervical cancer recurrence-related genes, a cervical cancer recurrence prediction model and system.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides genes associated with cervical cancer recurrence, wherein the genes associated with cervical cancer recurrence include mutant genes DCHS2, DNAH10, RYR1 and WDFY4.

[0007] In a second aspect, the present invention provides a cervical cancer recurrence prediction model, the calculation formula is as follows:

[0008]

[0009] RRS stands for recurrence-free relative score, M i Including mutant genes DCHS2, DNAH10, RYR1 and WDFY4, b i represents the preset regression coefficient of the i-th mutant gene, and n represents the number of mutant genes.

[0010] In a third aspect, the present invention provides a cervical cancer recurrence prediction model, including a memory, a processor, and a computer program stored in the memory and executable on a processor. When the processor executes the computer program, the following steps are implemented:

[0011] Compare the somatic cells of the recurrent cervical cancer patients with the somatic cells of the non-recurrent cervical cancer patients to screen out target mutation genes that meet the preset conditions;

[0012] Inputting the target mutant gene into a cervical cancer recurrence prediction model to determine a score associated with no recurrence;

[0013] Cervical cancer recurrence is predicted for cervical cancer patients according to the recurrence-free related score.

[0014] The beneficial effect of the present invention is as follows: the present application detects the mutation status of DCHS2, DNAH10, RYR1 and WDFY4 in the tumor tissues of cervical cancer patients and constructs a cervical cancer recurrence prediction model, thereby predicting the recurrence risk of cervical cancer patients through the recurrence-free related score output by the cervical cancer recurrence prediction model, thereby providing certain help for the prognosis evaluation of cervical cancer patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0016] Figure 1 This is a flow chart of a processor executing a computer program in a cervical cancer recurrence prediction system according to an embodiment of the present invention;

[0017] Figure 2 The DFS curves of four core mutation genes and cervical cancer patients;

[0018] Figure 3 The OS curves of four core mutation genes and cervical cancer patients;

[0019] Figure 4 This is the probability diagram of each mutant gene appearing in the relapse group and the non-relapse group in the TCGA CESC cohort;

[0020] Figure 5 is the correlation coefficient between the four core mutation genes and cervical cancer recurrence;

[0021] Figure 6 For RRShigh and RRS low DFS curve of patients with cervical cancer at P < 0.0001;

[0022] Figure 7 For RRS high and RRS low OS curve of patients with cervical cancer at P = 0.0086;

[0023] Figure 8 This is the distribution diagram of the recurrence-free correlation score of each mutated gene in the TCGA CESC cohort;

[0024] Fig. 9 For RRS high and RRS low OS curve of patients with cervical cancer at P = 0.044. DETAILED DESCRIPTION

[0025] The following examples are provided to further explain and supplement the present invention and do not constitute any limitation to the present invention.

[0026] The following describes the cervical cancer recurrence-related genes, cervical cancer recurrence prediction model and system according to the embodiments of the present invention in conjunction with the accompanying drawings.

[0027] The present invention provides cervical cancer recurrence-related genes, which include mutant genes DCHS2, DNAH10, RYR1 and WDFY4.

[0028] The screening process of cervical cancer recurrence-related genes in this embodiment is as follows:

[0029] A total of 95 patients with cervical cancer were included, of whom 67 had recurrence after surgery and 28 had complete remission.

[0030] DNA was extracted from paraffin samples of tumor tissue and paired adjacent tissues from patients with cervical cancer using the QIAamp DNA FFPE Tissue Kit (QIAGEN) according to the manufacturer's instructions. DNA degradation and contamination were monitored on a 1% agarose gel; DNAAssayKit, using The DNA concentration was detected by 3.0 Fluorometer (Invitrogen, USA). The total amount of DNA in each sample was 0.2 μg, which was used as the input material for DNA library preparation. The sequencing library was prepared using Illumina (NEB, USA) Ultra TMDNA library preparation kit was used according to the manufacturer's recommendations, and index codes were included in each sample. Genomic DNA samples were fragmented to 350 bp by sonication, and then the DNA fragments were end-polished, a-tailed, and ligated with full-length adapters for Illumina sequencing, followed by further PCR amplification. PCR products were purified using the AMPureXP system (Beckman Coulter, Beverly, USA) and amplified using The DNA concentration was determined using a 3.0 Fluorometer (Invitrogen, USA), the size distribution of the library was analyzed using NGS3K / Caliper, and quantified using real-time PCR (3 nM).

[0031] After the above sample extraction and sequencing were completed, the adapters and low-quality sequences were removed by cutadapt software (Kechin et al. 2017), and then the remaining sequences were aligned with the human reference genome (GRCh37) using BWA software (Li and Durbin 2009) using default parameters. The SAM file was converted to a BAM file and then sorted using Samtools software (Li et al. 2009), and repeated sequences were marked using Picard software. BAM files were locally adjusted using GATK software (McKenna et al. 2010). Somatic mutation calls were performed by GATK somatic MuTect2 (Benjamin et al. 2019) with a set of normal samples, and then filtered by MuTect2 using FilterMutectCalls. The retained somatic mutations were annotated using ANNOVAR (Wang et al. 2010). Then, high-confidence somatic mutations were determined according to the following inclusion criteria: (i) somatic mutations within exons; (ii) the coverage of tumor and normal samples was more than 10 times; (iii) the variant allele frequency in tumor samples was greater than 5%; (iv) somatic mutations were not recorded in the dbSNP database (Sherry et al. 2001).

[0032] After the above tumor mutation gene analysis, a total of 588 somatic mutations with a variant allele frequency of more than 5% were screened by the Fisher exact test method, and then 16 mutant genes were found to have significant differences (P < 0.05, P-value) between the recurrence group and the non-recurrence group by univariate Cox regression analysis. Then, LASSO regression analysis was performed on the 16 mutant genes, which can both narrow the range of candidate genes and prevent model overfitting, optimizing the selection of candidate mutant genes. This approach generated 11 candidate mutant genes, which were further selected as genes with a burst frequency of more than 5% in cervical cancer in the TCGA CESC dataset to allow for a wider range of applicability of these genes.

[0033] In summary, these analysis results identified DCHS2, DNAH10, RYR1 and WDFY4 as four core mutated genes.

[0034] The present invention provides a cervical cancer recurrence prediction model, the calculation formula of the cervical cancer recurrence prediction model is as follows:

[0035]

[0036] RRS stands for recurrence-free relative score, M i Including mutant genes DCHS2, DNAH10, RYR1 and WDFY4, b i represents the preset regression coefficient of the i-th mutant gene, and n represents the number of mutant genes.

[0037] In this embodiment, the four mutant genes DCHS2, DNAH10, RYR1 and WDFY4 selected above are used to construct the recurrence-free correlation score, wherein the preset regression coefficients are 1.10, 1.21, 0.99 and 1.41 respectively. The cervical cancer recurrence prediction model can obtain RRS=1.10*DCHS2+1.21*DNAH10+0.99*RYR1+1.41*WDFY4. In addition, the R software package "survminer". The surv_cutpoint() function uses the maximum rank statistic selected from the 'maxstat' R package, i.e., the preset threshold, to divide each sample into RRS high and RRS low , where the preset threshold is 0.99, and RRS high Patients with RRS low Patients had better disease-free and overall survival.

[0038] The present invention provides a cervical cancer recurrence prediction system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following is achieved: Figure 1 Steps shown:

[0039] S1. Compare the somatic cells of the recurrent cervical cancer patients with those of the non-recurrent cervical cancer patients to screen out target mutant genes that meet the preset conditions.

[0040] S2. Inputting the target mutant gene into a cervical cancer recurrence prediction model to determine a score associated with no recurrence.

[0041] S3. Predicting the recurrence of cervical cancer in patients with cervical cancer based on the recurrence-free related score.

[0042] In this embodiment, the mutation status of DCHS2, DNAH10, RYR1 and WDFY4 in the tumor tissues of cervical cancer patients is detected, and a cervical cancer recurrence prediction model is constructed. The recurrence risk of cervical cancer patients is predicted by the recurrence-free correlation score output by the cervical cancer recurrence prediction model, so as to provide certain help for the prognosis assessment of cervical cancer patients.

[0043] Optionally, the processor is specifically configured to:

[0044] The somatic cells of the recurrent group of cervical cancer patients were compared with those of the non-recurrent group of cervical cancer patients, and the first initial mutation gene was screened by the Fisher exact test method;

[0045] Univariate Cox regression analysis was used to screen the second initial mutation gene from the first initial mutation gene;

[0046] The third initial mutation gene was screened out from the second initial mutation gene by LASSO regression analysis method;

[0047] From the third initial mutant genes, mutant genes whose cervical cancer mutation frequency is greater than a preset value are screened out as target genes.

[0048] In this embodiment, the preset value is set according to actual conditions. Preferably, in this embodiment, the preset value is 5%.

[0049] In this embodiment, the Fisher's Exact Test is a method used in statistics to test whether two categorical variables are independent (i.e., correlated). This test is particularly suitable for situations where the sample size is small or the data distribution does not meet the progressive requirements. The Fisher's Exact Test method can be used as an initial screening to screen out the first initial mutant gene.

[0050] The univariate Cox regression analysis in this embodiment is a statistical method commonly used in survival analysis, which is mainly used to study the relationship between a single variable (explanatory variable) and the time to the occurrence of an event (such as death, disease recurrence, etc.). In the univariate Cox regression analysis, the P value (P-value) is an important statistic used to evaluate the significance of the relationship between the independent variable (explanatory variable) and the survival time (dependent variable). In this embodiment, the P value is 0.05, that is, the first initial mutation gene with P < 0.05 is used as the second initial mutation gene.

[0051] The LASSO (Least Absolute Shrinkage and Selection Operator) regression analysis method in this embodiment is a linear regression method widely used in statistics and machine learning. This method can not only perform variable selection, but also regularize the regression coefficient, which can not only narrow the range of candidate genes, but also prevent model overfitting, optimize the selection of candidate mutant genes, and thus screen out the third initial mutant gene from the second initial mutant gene.

[0052] Optionally, the processor is specifically configured to:

[0053] If the non-recurrence-related score is greater than a preset threshold, it is judged that the probability of cervical cancer recurrence is low. If the non-recurrence-related score is less than or equal to the preset threshold, it is judged that the probability of cervical cancer recurrence is high.

[0054] In this embodiment, the recurrence probability of cervical cancer can be intuitively observed through the non-recurrence related score, thereby providing certain help for the prognosis evaluation of cervical cancer patients.

[0055] The R package "survminer" was used in this example. The surv_cutpoint() function uses the maximum rank statistic selected from the 'maxstat' R package, i.e., a preset threshold of 0.99, to divide each sample into RRS high or RRS low .

[0056] Optionally, the cervical cancer recurrence prediction is verified by a cervical cancer recurrence prediction system:

[0057] In this example, 67 patients with recurrent CC and 28 matched patients without recurrent CC were included. 67 patients who met the following criteria were assigned to the recurrence group: (i) diagnosed with cervical cancer according to the 2009 International Federation of Gynecology and Obstetrics (FIGO) staging system; (ii) tissue specimens; (iii) tumor recurrence diagnosis based on computed tomography and magnetic resonance imaging, with or without histological confirmation. 28 matched patients who met the above (i) and (ii) criteria and had no tumor recurrence within three years were randomly assigned to the non-recurrence group. All patients were followed up after surgery. DFS (Disease Free Survival) and OS (Overall Survival) were defined as the time interval from the date of surgery to the date of first diagnosis of recurrence, and the time from the date of surgery to death or the end of follow-up, respectively. Surviving patients were reviewed at the end of follow-up.

[0058] After initial evaluation, specimens from 67 recurrent CC patients and 28 non-recurrent CC patients were eligible for WES. The genomic and clinical data of these patients were then screened to identify candidate mutation genes that were closely associated with clinical outcomes, and their prognostic value was subsequently validated using mutation data from TCGA CESC patients (N = 289).

[0059] Preliminary evaluation showed that samples from 67 patients with recurrent cervical cancer and 28 matched non-recurrent cervical cancer patients met the requirements for WES and mutation analysis. Univariate Cox analysis of these patients showed the effects of different clinical characteristics and gene mutations on DFS (Table 1), and multivariate analysis showed the effects of different clinical characteristics and gene mutations on DFS (Table 2). It is worth noting that no statistically significant differences were found between the two groups in these parameters, although patients with advanced cervical cancer appeared more frequently in the recurrent group compared with the non-recurrent group (19.4% vs. 7.1%, chi-square test, P = 0.217).

[0060] Table 1

[0061] variable HR (95% Cl) p-value age 0.997(0.967-1.027) 0.823 Pathological type, adenocarcinoma 1.295(0.638-2.629) 0.475 Stage, late stage 2.252(1.382-3.672) 0.001 Degree of differentiation, high school differentiation 0.750(0.464-1.214) 0.242 Vascular invasion is 1.539(0.758-3.127) 0.233 Lymph node metastasis, yes 1.446(0.877-2.383) 0.148 Parauterine invasion, yes 2.721(1.279-5.790) 0.009 Neoadjuvant chemotherapy is 1.459(0.892-2.388) 0.133 RRS, high 0.242(0.118-0.495) ＜0.001 DCHS2 mutations 0.310(0.076-1.270) 0.104 DNAH10 mutation 0.616(0.151-2.519) 0.500 RYR1 mutation 0.459(0.167-1.264) 0.132 WDFY4 mutation 0.379(0.093-1.548) 0.176

[0062] Table 2

[0063] variable HR (95% Cl) p-value Stage, late stage 1.735(1.033-2.914) 0.037 Parauterine invasion, yes 2.029(0.920-4.475) 0.079 RRS, high 0.263(0.128-0.543) ＜0.001

[0064] like Figure 2 As shown in the figure, the disease-free survival (DFS) of patients carrying four mutated genes (mutated mutated genes) was significantly better than that of patients without mutated genes (wildtype native genes). Figure 3As shown, when overall survival (OS) was analyzed, it was also found that CC patients with DCHS2 or RYR1 mutations had significantly better OS than those without mutations, while there was no significant difference between patients with DNAH10 and WDFY4 mutations and those without mutations.

[0065] like Figure 4 As shown in Figure 2, these four mutated genes mainly appeared in the non-relapse group, such as Figure 5 As shown, the correlation coefficients with cervical cancer recurrence are all positive. Figure 2-3 It can be seen that the clinical outcomes of patients carrying these four gene mutations are better than those without mutations, suggesting that these four mutant genes are candidate biomarkers for good prognosis in patients with cervical cancer. Therefore, a cervical cancer recurrence prediction model was established based on these four mutant genes to accurately predict the prognosis of patients with cervical cancer. The recurrence-free score (RRS) was established using the mutation status and coefficient of variation of each gene, and patients were divided into RRS high (N=26) and RRS low (N=69), with a cutoff RRS score of 0.99. In fact, patients carrying any of the four mutated genes fall within the RRS high The recurrence rate (84.06%) of the group was also much higher than that of the RRS low group (34.62%) (P < 0.0001; Figure 6 ). In addition, if Figure 6 and Figure 7 As shown, RRS high The patients' DFS and OS were significantly better than those of RRS. low patients (DFS and OS, P < 0.0001 and P = 0.0086, respectively). Importantly, these differences were even more pronounced than those between patients with and without these four mutated genes. In summary, the results of the study showed that the cervical cancer recurrence prediction model established based on the core mutated genes that differ between recurrent and non-recurrent cervical cancer patients can predict the clinical outcomes of cervical cancer patients.

[0066] In order to further evaluate the prognostic value of the cervical cancer recurrence prediction model for cervical cancer patients, 289 cervical cancer patients in the TCGA database were used as a validation cohort. Similar to the cohort study results of this example, mutations in the DCHS2, DNAH10, RYR1, and WDFY4 genes in the TCGA CESC cohort were also significantly concentrated in the RRS high ( Figure 8 ). Since the TCGA database does not include information related to CC recurrence, only OS was used for prognostic analysis. Although each of the four mutated genes could not distinguish OS in cervical cancer patients, RRS high Can effectively identify patients with better OS ( Fig. 9; P = 0.044), confirming the cohort study results of this embodiment. Therefore, these results show that the cervical cancer recurrence prediction model can be extended to other cohorts or other populations in predicting the prognosis of cervical cancer patients.

[0067] Those skilled in the art know that the present invention can be implemented as a system, method or computer program product. Therefore, the present disclosure can be specifically implemented in the following forms, namely: it can be complete hardware, it can be complete software (including firmware, resident software, microcode, etc.), or it can be a combination of hardware and software, generally referred to as "circuit", "module" or "system" herein. In addition, in some embodiments, the present invention can also be implemented in the form of a computer program product in one or more computer-readable media, and the computer-readable medium contains computer-readable program code. Computer-readable storage media can be, for example, but not limited to - electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above.

[0068] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0069] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A gene associated with cervical cancer recurrence, characterized in that: The cervical cancer recurrence-related genes include mutant genes DCHS2, DNAH10, RYR1 and WDFY4.

2. A cervical cancer recurrence prediction model, characterized in that: The calculation formula of the cervical cancer recurrence prediction model is as follows: RRS stands for recurrence-free relative score, M i Including mutant genes DCHS2, DNAH10, RYR1 and WDFY4, b i represents the preset regression coefficient of the i-th mutant gene, and n represents the number of mutant genes.

3. The cervical cancer recurrence prediction model according to claim 2, characterized in that: The preset regression coefficient is: When M i When it is DCHS2, the default regression coefficient is 1.10; When M i When it is DNAH10, the default regression coefficient is 1.21; When M i When it is RYR1, the default regression coefficient is 0.99; When M i For WDFY4, the default regression coefficient is 1.

41.

4. A cervical cancer recurrence prediction system, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Compare the somatic cells of the recurrent cervical cancer patients with the somatic cells of the non-recurrent cervical cancer patients to screen out target mutation genes that meet the preset conditions; Inputting the target mutant gene into a cervical cancer recurrence prediction model to determine a score associated with no recurrence; Cervical cancer recurrence is predicted for cervical cancer patients according to the recurrence-free related score.

5. The system according to claim 4, characterized in that The processor is specifically used for: The somatic cells of the recurrent group of cervical cancer patients were compared with those of the non-recurrent group of cervical cancer patients, and the first initial mutation gene was screened by the Fisher exact test method; Univariate Cox regression analysis was used to screen the second initial mutation gene from the first initial mutation gene; The third initial mutation gene was screened out from the second initial mutation gene by LASSO regression analysis method; From the third initial mutant genes, mutant genes whose cervical cancer mutation frequency is greater than a preset value are screened out as target genes.

6. The system according to claim 5, characterized in that The preset value is 5%.

7. The system according to claim 4, characterized in that The processor is specifically used for: If the non-recurrence-related score is greater than a preset threshold, it is judged that the probability of cervical cancer recurrence is low. If the non-recurrence-related score is less than or equal to the preset threshold, it is judged that the probability of cervical cancer recurrence is high.

8. The system according to claim 7, characterized in that The default threshold is 0.99.

Images