CN113684277B

CN113684277B - Method for predicting ovarian cancer homologous recombination defect based on biomarker of genome copy number variation and application

Info

Publication number: CN113684277B
Application number: CN202111046756.6A
Authority: CN
Inventors: 董忠谊; 吴德华; 张萌; 王剑; 马思聪; 白雪; 谭家乐
Original assignee: Southern Hospital Southern Medical University
Current assignee: Southern Hospital Southern Medical University
Priority date: 2021-09-06
Filing date: 2021-09-06
Publication date: 2022-05-17
Anticipated expiration: 2041-09-06
Also published as: CN113684277A

Abstract

The invention relates to the field of biomedicine, in particular to a method for predicting ovarian cancer homologous recombination defects based on a biomarker of genome copy number variation and application thereof. The biomarker is CNV based on any one of chromosome fragments 5q13.2, 8q24.2 or 19q 12. The method for predicting the homologous recombination defect of the ovarian cancer by the biomarkers comprises the following steps: collecting a surgically excised sample or a tissue biopsy sample from a patient with ovarian cancer; performing DNA sequencing on the sample to obtain a corresponding sequencing file; analyzing the DNA sequencing file by using GISTIC 2.0 to obtain a CNV map of the whole tumor genome; HRD status of ovarian cancer is predicted based on CNV of chromosome fragment 5q13.2, 8q24.2 or 19q 12. The invention provides application of copy number variation CNV at the level of sub-chromosome and gene in predicting homologous recombination defects of ovarian cancer patients, so that potential benefit populations of targeted therapy are screened, and more clinical benefits are brought to HRD patients.

Description

Method for predicting ovarian cancer homologous recombination defect based on biomarker of genome copy number variation and application

Technical Field

The invention relates to the field of biomedicine, in particular to a method for predicting ovarian cancer homologous recombination defects based on a biomarker of genome copy number variation and application thereof.

Background

Homologous Recombination Defects (HRD) are a common driver of tumorigenesis and can lead to damage during DNA double strand break repair and thus to genomic instability. Homologous recombination defects are a common feature of many tumors, primarily found in ovarian cancer. Meanwhile, as a therapeutic target, the homologous recombination defect plays an important role in chemotherapy, targeted therapy and immunotherapy. Cytotoxic drugs such as platinum analogs, topoisomerase I inhibitors, topoisomerase II inhibitors, and the antimetabolite gemcitabine have been shown to be effective against tumors deficient in homologous recombination. Whereas, inhibitors of poly (A) diphosphate ribosyl polymerase (PARP) exhibit synthetic lethality when applied to cells deficient in homologous recombination. Furthermore, tumors deficient in homologous recombination have been shown to be more immunogenic than tumors without genetic defects in the homologous recombination pathway, making them potential candidates for immune checkpoint blockade. Since HRD suggests the sensitivity of tumors to cytotoxic chemotherapy and PARP inhibitor targeted therapy, accurate monitoring of HRD facilitates accurate dosing of patients with such tumors, thereby limiting drug use to potentially benefitting populations and avoiding unnecessary toxicity to non-benefitting populations.

Considerable effort has been devoted by researchers in the field of identifying and developing predictive biomarkers for HRD, however, there are currently no recognized effective assays and there is still considerable controversy over existing assays. For reasons of defective homologous recombination, biomarkers include mutations associated with genes of the homologous recombination repair pathway, such as BRCA1/2 germline and somatic mutations, methylation of the BRCA1 promoter, mutations of other genes in the homologous recombination repair pathway (including RAD51C, CHEK2, BRIP1, PTEN, etc.). In terms of the results due to homologous recombination defects, three single nucleotide polymorphism-based biomarkers were used to quantify the extent of chromosomal abnormalities, i.e., loss of heterozygosity (LOH), large-scale terminal migration (LST), and Telomere Allelic Imbalance (TAI), which were significantly correlated with BRCA1/2 status, and the composite scores of TAI, LOH, and LST were also retrospectively validated to predict response to platinum-containing neoadjuvant chemotherapy. However, false positives due to the functional proficiency of homologous recombination caused by somatic reverse mutation limit the clinical applications of existing markers, and thus there is a great need for finding more effective and accurate biomarkers in clinic.

Previous studies have shown that in cells deficient in homologous recombination, total chromosomal rearrangements and overall genomic instability are increased, which are also the major causes of copy number variation. In the case of a defect in homologous recombination, the cell either has difficulty repairing DNA damage and thus progressing to some form of programmed cell death, or attempts to repair DNA damage using a process such as non-homologous end joining and the like to create copy number variation. Array comparative genomic hybridization (aCGH) is a technique to detect tumor genomic copy number variation, and aCGH-based genomic scar analysis shows that the presence of BRCA-like aCGH signals is predictive of a preferential response to platinum drugs. In summary, previous studies have shown that there may be a correlation between CNV and HRD status of patients, but little is known about which specific genomic sites in a tumor that CNV are directly associated with HRD.

In view of the above, the present invention is particularly proposed.

Disclosure of Invention

In view of the above problems, the present invention provides a method for predicting ovarian cancer Homologous Recombination Deficiency (HRD) based on Copy Number Variation (CNV) biomarkers and application thereof.

In order to achieve the above purpose, the technical solution adopted by the invention to solve the technical problem is as follows:

a genomic copy number variation-based biomarker for predicting ovarian cancer homologous recombination defects, the biomarker being CNV based on any of chromosome fragments 5q13.2, 8q24.2 or 19q 12.

Further, the biomarker is any one of deletion of 5q13.2, amplification of 8q24.2 or amplification of 19q 12.

The invention also provides application of the biomarker in predicting homologous recombination defects of ovarian cancer patients.

The invention also provides application of the biomarker in preparing a biomarker for predicting homologous recombination defects of ovarian cancer patients.

In addition, the invention provides application of the product of the biomarker in preparing a product for predicting homologous recombination defects of ovarian cancer patients.

The invention also provides a method for predicting the homologous recombination defect of the ovarian cancer, which is characterized by adopting the biomarker for prediction.

Further, the method comprises the following steps:

s1, collecting surgical resection samples or tissue biopsy samples of ovarian cancer patients, wherein the samples comprise tumor tissues and tissues beside cancer;

s2, performing DNA sequencing on the sample obtained in the step S1 to obtain a corresponding sequencing file;

s3, analyzing the DNA sequencing file by using GISTIC 2.0 to obtain a CNV map of the whole tumor genome;

s4, predicting HRD status of ovarian cancer based on CNV of chromosome fragment 5q13.2, 8q24.2 or 19q 12: judging that the patient has a defect of homologous recombination if the GISTIC value of 5q13.2 is-1 (single copy deletion) or-2 (homozygous deletion), or the GISTIC value of 8q24.2 is 1 (low level amplification) or 2 (high level amplification), or the GISTIC value of 19q12 is 0 (diploid normal copy); a patient was judged to be free of homologous recombination defects if the GISCITC value of 5q13.2, 8q24.2 was 0 (diploid normal copy), and the GISTIC value of 19q12 was 1 (low level amplification) or 2 (high level amplification), which was calculated from GISTIC 2.0.

In addition, the invention also provides a method for predicting the homologous recombination defect of pan-cancer based on the biomarker of genome copy number variation, which mainly comprises the following steps:

s1, collecting surgical excision samples or tissue biopsy samples (including tumor tissues and tissues beside cancer) of cancer patients.

And S2, performing DNA sequencing on the sample obtained in the S1 to obtain a corresponding sequencing file.

And S3, analyzing the DNA sequencing file by using GISTIC 2.0 to obtain the CNV face of the whole tumor genome.

S4, predicting HRD state of tumor based on the amplification of chromosome segment 8q24.2 or the amplification of genes MYC and NDRG 1: judging that the patient has the homologous recombination defect if the 8q24.2 amplification or the gene MYC amplification or the gene NDRG1 amplification exists; and judging that the patient has no homologous recombination defect in other cases.

Compared with the prior art, the invention has the beneficial effects that:

1. the invention enriches the homologous recombination defect biomarkers of the sub-chromosome and gene level and also provides a new visual angle for the HRD state of a patient in the cross-tumor research. Compared with the prior art, the copy number variation has more stable prediction on homologous recombination defects and is not easily influenced by back mutation.

2. The method for predicting the homologous recombination defect of the ovarian cancer based on copy number variation, which is provided by the invention, is not reported at present, can realize specific and effective prediction and has good accuracy, and provides basis and foundation for evaluating and predicting the homologous recombination defect state of the ovarian cancer. Has important value and significance for timely taking effective clinical measures, making individualized diagnosis and treatment schemes and finally improving the survival rate of ovarian cancer patients.

Drawings

FIG. 1 is a graph of enrichment scores for genomic copy number variation, copy number differences between HRD and non-HRD groups, and copy number variation versus HRD score for example 1 of the present invention. Where a is the enrichment score for genomic copy number variation from all ovarian cancer patients from the TCGA-OV cohort, B, C is the enrichment score for CNV in HRD group (B) and non-HRD group (C), respectively, D, F, H is the difference between HRD and non-HRD group in copy number values of 8q24.2(D), 19.12(F) and 5q13.2(H) calculated by gist 2.0, E, G, I is the relationship between 8q24.2 amplification (E), 19q12 amplification (G) and 5q13.2 deletion (I) in predicting HRD/non-HRD performance, J, K, L is the relationship between copy number variation of 8q24.2(J), 19q12(K) and 5q13.2(L) and the three component indices LOH, TAI and LST of HRD score.

FIG. 2 is a validation of a method for predicting Ovarian Cancer homologous recombination defects based on copy number variation biomarkers in the Australian Ovarian Cancer Study (AOCS) cohort. Wherein A is the degree of enrichment of genome CNV in all ovarian cancer patients in the AOCS cohort, B-G is the difference between the HRD and non-HRD groups in the AOCS cohort based on BRCA1/2 mutations, BRCA1 promoter methylation and other homologous recombination repair gene mutations (including RAD51C, CHEK2, BRIP1, PTEN), and B, D, F is the copy number calculated by GISTIC 2.0 for 8q24.2(B)19q12(D) and 5q13.2 (F). C. E, G Performance in prediction of HRD/non-HRD for 8q24.2 amplification (C), 19q12 amplification (E) and 5q13.2 deletion (G).

FIG. 3 is the application of the method for predicting the defect of homologous recombination in pan-carcinoma based on copy number variation in example 2. Wherein A is the mean HRD score and HRD frequency (. gtoreq.42 is a cutoff value) of 33 tumors in the TCGA pan-cancer cohort, B is the CNV enrichment degree of 8q24.2, 19q12 amplification and 5q13.2 deletion in the 9 tumors with the highest mean HRD score, C is a violin graph of 8q24.2 copy number of HRD group and non-HRD group in all TCGA tumors, and D is the copy number of 8q24.2 copy number of HRD group and non-HRD group in 7 tumors.

FIG. 4 is the correlation of the amplification of the genes MYC and NDRG1 in example 3 with a defect in pan-cancer homologous recombination. Wherein, a is the gene amplification frequency at 8q24.2 in TCGA, B is the mutation type and frequency of MYC in different tumors, C is the linear regression plot of MYC amplification versus mean HRD score of tumors, D is the mutation type and frequency of NDRG1 in different tumors, and E is the linear regression plot of NDRG1 amplification versus mean HRD score of tumors.

FIG. 5 is a flow chart of the study design of the present invention.

Detailed Description

For better illustrating the objects, technical solutions and advantages of the present invention, the present invention will be further described with reference to the accompanying drawings and specific embodiments, but the following examples are only simple examples of the present invention and do not represent or limit the scope of the present invention, which is defined by the claims.

Example 1

Application of method for predicting ovarian cancer homologous recombination defect based on copy number variation

The protocol of the study of the invention is shown in FIG. 5, example 1 involves 587 ovarian cancer patients from the TCGA-OV cohort in the study, whose clinical characteristics and DNA sequencing data are from https:// www.cancer.gov/TCGA.

587 patients were assigned to HRD group (HRD score ≧ 42) and non-HRD group (HRD score < 42) according to HRD-score, which is the unweighted sum of Loss of heterozygosity (LOH), Large-scale terminal migration (LST), and Telomeric Allelic Imbalance (TAI).

Using the gist 2.0 to measure CNV profiles at the tumor whole genome sub-chromosomal level, segment files of the TCGA-OV cohort (recording copy numbers at various sites of the genome) were used as input to the GSITIC 2.0 to quantify the CNV status of the chromosome segments in each tumor genome, with the results shown in fig. 1A for ovarian cancer patients with significant amplification of chromosome segments 3q26.2, 8q24.2 and 19q12, and significant deletion of chromosome segments 5q13.2 and 19p 13.3.

Further comparison of the status of CNV between HRD and non-HRD groups, as shown in fig. 1B, 1C, revealed that 5q13.2 of HRD group was significantly deleted, 8q24.2 was significantly amplified, and 19q12 of non-HRD group was significantly amplified, compared to non-HRD group.

Based on the above results, the differences in copy number values of the chromosome fragments 5q13.2, 8q24.2 and 19q12 in the HRD group and non-HRD group were compared, and whether the copy number values of 5q13.2, 8q24.2 or 19q12 were significant predictors of HRD status was investigated. As a result, the amplification of 8q24.2 (FIG. 1D, p <0.0001) and deletion of 5q13.2 (FIG. 1H, p <0.0001) were significantly higher in the HRD group than in the non-HRD group, with AUC values of 67.3% (FIG. 1E) and 68.2% (FIG. 1I), respectively. Whereas the 19q12 amplification value was significantly higher for the HRD group (FIG. 1F, p <0.0001) and the AUC value was 63.7% (FIG. 1G).

In conclusion, the CNV of 5q13.2, 8q24.2 or 19q12 has the prediction potential of ovarian cancer homologous recombination defect state, and if the GISTIC value of 5q13.2 is-1 (single copy deletion) or-2 (homozygous deletion), or the GISTIC value of 8q24.2 is 1 (low level amplification) or 2 (high level amplification), or the GISTIC value of 19q12 is 0 (diploid normal copy), the patient is judged to have homologous recombination defect; patients were judged to be free of homologous recombination defects if the GISTIC value of 5q13.2, 8q24.2 was 0 (diploid normal copy) and the GISTIC value of 19q12 was 1 (low level amplification) or 2 (high level amplification) (the GISTIC value was calculated from GISTIC 2.0).

To confirm the predictive value of CNV on HRD for these three fragments, the present inventors explored the relationship between deletion or amplification of 5q13.2, 8q24.2 and 19q12 and the three components of HRD-core, LOH, TAI and LST.

The results show that the greater the degree of amplification of 8q24.2 (FIG. 1J, p <0.0001) and deletion of the 5q13.2 fragment (FIG. 1L, p <0.0001), the greater the counts of LOH, TAI and LST. Amplification of the 19q12 fragment was inversely correlated with LOH, TAI and LST, with differences mainly manifested in high-level amplification, whereas no significant differences were observed between normal diploid copies and low-level amplification. (FIG. 1K, p <0.0001), which is consistent with the results of FIGS. 1D-I.

To further clarify that deletion or amplification of the three fragments 5q13.2, 8q24.2 and 19q12 could be used as HRD-specific biomarkers for the prediction of Ovarian Cancer, we performed external validation of markers of copy number variation of the three genes in the Australian Ovarian Cancer Study (AOCS) cohort.

By analyzing the degree of enrichment of genomic CNV of all ovarian cancer patients in the AOCS cohort, we found that there was significant amplification of 8q24.2 and 19q12 in the AOCS cohort, top ranking in all copy number amplified fragments, and significant deletion of the 5q13.2 fragment (fig. 2A). In addition, we also observed 8q24.2 amplification (fig. 2B, p <0.0001), 19q12 amplification (fig. 2D, p ═ 0.0034) and 5q13.2 deletion (fig. 2F, p ═ 0.0056) of HRD groups in the AOCS cohort significantly higher than the non-HRD group, AUC of 75.0% (fig. 2C), 68.9% (fig. 2E) and 67.9% (fig. 2G), respectively

Thus, CNVs of 5q13.2, 8q24.2 or 19q12 were identified in this example as biomarkers of ovarian cancer homologous recombination defects.

Example 2

Application of biomarker for predicting pan-cancer homologous recombination defect based on copy number variation

Analysis of DNA sequencing results from 10635 patients with 33 cancer types from the TCGA pan-cancer cohort generalized the predictive role of CNVs for ovarian cancer homologous recombination defects to other cancer species.

The mean HRD score for each tumor was calculated to obtain the distribution of homologous recombination defect status among different tumors, and patients with ovarian serous cystadenocarcinoma (OV), Uterine Carcinosarcoma (UCS), lung squamous carcinoma (lucc), Sarcoma (SARC), esophageal cancer (ESCA) were found. HRD scores were mainly higher for gastric adenocarcinoma (STAD), bladder urethral carcinoma (BLCA), breast cancer (BRCA) and lung adenocarcinoma (LUAD), while those of thyroid cancer (THCA) and Acute Myeloid Leukemia (AML) patients were the lowest (fig. 3A). This is consistent with the higher frequency of homologous recombination defects in ovarian and breast cancers that have been suggested in previous studies and confirms the prevalence of homologous recombination defects.

To investigate whether CNVs could be used for homologous recombination defect detection of pan-carcinomas, CNV frequencies of 8q24.2 amplification, 19q12 amplification and 5q13.2 deletion among the 9 tumors with the highest HRD score were investigated (fig. 3B). It can be seen that of these cancers, 8q24.2 amplification is generally enriched (except for sarcomas), while 19q12 amplification occurs most frequently in ovarian and gastric carcinomas, but the 5q13.2 deletion is only present in ovarian cancers, suggesting that 8q24.2 amplification is capable of predicting homologous recombination defects across tumor types, while the 5q13.2 deletion is a characteristic feature of ovarian cancer homologous recombination defects.

Referring to the method of example 1 for grouping ovarian cancer patients, 10, 635 patients were classified into HRD group and non-HRD group according to HRD score. The chromosome 8q24.2 was studied with emphasis on the chromosome fragment and found to have a significantly higher copy number in the TCGA pan-cancer cohort in HRD patients compared to non-HRD patients (fig. 3C, p <0.0001), which was confirmed in squamous lung carcinoma (p <0.0001), urinary bladder carcinoma (p < 0.0102), adenocarcinoma lung (p < 0.0035), breast cancer (p <0.0001), and adenocarcinoma stomach (p <0.0001), respectively (fig. 3D). Together, these results provide important evidence for the predictive role of 8q24.2 CNVs on homologous recombination defects in different cancer types.

Thus, in this example 8q24.2 of CNV was confirmed as a biomarker for the defect of pan-cancer homologous recombination.

Example 3

Correlation of amplification of genes MYC and NDRG1 with defects in pan-cancer homologous recombination

The gene amplification frequency of 8q24.2 in the TCGA pan-cancer cohort was analyzed to find the correlation of the gene with HRD.

As shown in fig. 4A, the genes within 8q24.2 were amplified with higher rates in tumors, including oncogenes MYC and NDRG 1.

Given the close association of oncogenes with treatment and tumor biology, the mutation types and frequencies of MYC and NDRG1 were counted in all TCGA tumors, and it was found that the main mutation types of MYC and NDRG1 were just amplified, and that the tumors with the highest mutation frequency of MYC and NDRG1 were ovarian cancers, while the mutation frequency in thyroid cancers was lower, which is consistent with the highest HRD score of ovarian cancers and the lower HRD score of thyroid cancers in fig. 3A (fig. 4B, D).

The relationship between amplification frequency of MYC and NDRG1 and HRD score was studied using pan-carcinoma data and results showed: amplification of MYC and NDRG1 clearly correlated with HRD scores (MYC: rho-0.6809, p-0.0001; NDRG 1: rho-0.6453, p-0.0004; fig. 4C, E).

Thus, this example suggests that the amplification of MYC and NDRG1 in the 8q24.2 amplified fragment may serve as biomarkers for the defect of pan-cancer homologous recombination.

The above embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; those of ordinary skill in the art will understand that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications or substitutions are also to be considered as within the scope of the present invention.

Claims

1. Use of a reagent for detecting genomic copy number variation of a biomarker for the preparation of a product for predicting homologous recombination deficiency in an ovarian cancer patient, wherein the biomarker is a CNV based on any of the chromosome fragments 5q13.2, 8q24.2 or 19q12, and wherein the patient is judged to have homologous recombination deficiency if the gist ic value of 5q13.2 is-1 (single copy deletion) or-2 (homozygous deletion), or the gist ic value of 8q24.2 is 1 (low level amplification) or 2 (high level amplification), or the gist ic value of 19q12 is 0 (diploid normal copy); patients were judged to be free of homologous recombination defects if the GISTIC value of 5q13.2, 8q24.2 was 0 (diploid normal copy) and the GISTIC value of 19q12 was 1 (low level amplification) or 2 (high level amplification), calculated from GISTIC 2.0.