WO2020137076A1 - Method for predicting susceptibility of cancer to parp inhibitors, and method for detecting cancer having homologous recombination repair deficiency - Google Patents

Abstract

The purpose of the present invention is to provide a method for predicting the susceptibility of a target cancer to PARP inhibitors. The present invention provides a method for predicting the susceptibility of a cancer to PARP inhibitors. The method includes a step for creating a mutation catalog by performing a sequence analysis on nucleic acid samples obtained from cancer cells. The presence of a cancer that is susceptible to PARP inhibitors is indicated by the contribution of BRCA signatures to the mutation catalog being greater than the contribution of Age signatures to the mutation catalog and by an absence of cyclin E1 amplification.

Description

Method for predicting cancer susceptibility to PARP inhibitor and method for detecting cancer having homologous recombination repair deficiency Reference to related applications

This application enjoys the benefit of the priority of Japanese Patent Application No. 2018-248619 (filing date: December 28, 2018), which is a prior Japanese application, and the entire disclosure content thereof is incorporated herein by reference. Be part of.

The present invention relates to a method for predicting cancer susceptibility to PARP inhibitors and a method for detecting cancer having homologous recombination repair failure.

In recent years, molecularly targeted drugs that target changes in genes and proteins that occur specifically in cancer cells have been developed. Since such a molecular targeting drug targets cancer cells and does not damage normal cells, it has attracted attention as an anticancer agent with reduced side effects. In order to administer a molecular target drug, it is necessary not only to diagnose a cancer type but also to detect a gene mutation or a protein change at the molecular level.

Olaparib, which is a poly(ADP-ribose)polymerase 1: PARP inhibitor that was developed targeting homologous recombination repair deficiency (HRD) among molecular target drugs, is 2014. Approved by the US Food and Drug Administration (FDA) for recurrent ovarian cancer with a germline BRCA mutation. PARP is an enzyme involved in repairing single-strand breaks and double-strand breaks in DNA, and repair of double-strand breaks is performed by homologous recombination (HR). On the other hand, the BRCA1/2 gene has an important role for HR in DNA double-strand break repair. When a PARP inhibitor acts on cells having a BRCA1/2 mutation, DNA repair by homologous recombination is inhibited, resulting in cell death and an antitumor effect.

It has been reported that in serous ovarian cancer, germline BRCA1/2 mutations account for about 20%, whereas HRD accounts for about 50% (Non-patent Document 1). Since PARP inhibitors are drugs that target HRD, it is desirable that they be widely administered to patients with cancers that have HRD, without being limited to subjects with germline BRCA1/2 mutations. However, until now, there was not an appropriate biomarker for evaluating HRD without excess or deficiency.

It is an object of the present invention to provide a method for predicting the susceptibility of a target cancer to a PARP inhibitor and a method for detecting a cancer having a homologous recombination repair deficiency.

The present inventors have recently carried out mutation signature analysis based on whole exon sequence analysis on cases of serous ovarian cancer, and found that the contribution of BRCA signature to the mutation catalog was greater than that of Age signature in 77% of cases. Found. The present inventors also set a biomarker (MSBM) indicating the presence of a cancer having a homologous recombination repair deficiency, performed total exon sequence analysis and nonnegative matrix factorization (NMF), etc. Was found to be 68%. The present inventors further carried out MSBM determination using an oncogene panel test, and found that MSBM positivity can be an index of homologous recombination repair failure and sensitivity of PARP inhibitors. The present inventors have also found that the contribution of 30 types of mutation signatures to the mutation catalog can be calculated by destructStructs analysis, and MSBM determination can be performed based on the analysis result. The present invention is based on these findings.

According to the present invention, the following inventions are provided.
[1] A method for predicting cancer susceptibility to a PARP inhibitor, comprising:
Comprising the step of performing a sequence analysis on a nucleic acid sample obtained from a cancer cell to create a mutation catalog, wherein (A)(a1) the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog. And (a2) the absence of cyclin E1 amplification indicates the presence of a cancer sensitive to a PARP inhibitor.
[2] The method according to [1] above, wherein the cancer cell is a cell obtained from a cancer patient.
[3] The method of the above-mentioned [1] or [2], wherein the cancer is a cancer that may exhibit a homologous recombination repair abnormality.
[4] The method according to any one of [1] to [3] above, wherein the PARP inhibitor is olaparib, lucaparib, niraparib, beriparib or tarazoparib.
[5] The method according to any of [1] to [4] above, wherein the sequence analysis comprises a sequence determination procedure and a sequence data analysis procedure.
[6] The method according to any one of [1] to [5] above, wherein somatic gene mutations are identified by the sequence analysis according to the 96 types of substitution classification shown in FIG. 1, and a mutation catalog is prepared.
[7] The method according to any of [1] to [6] above, wherein the sequence analysis is performed by whole exon sequence analysis.
[8] The method according to any of [1] to [6] above, wherein the sequence analysis is performed by an oncogene panel test.
[9] The following steps (1) to (3):
(1) Preparing an Age signature and a BRCA signature,
[2] above, further comprising (2) evaluating the contribution of the Age and BRCA signatures to the mutation catalog and (3) evaluating the superiority of the contribution of the BRCA signature to the contribution of the Age signature. The method according to any one of 8].
[10] In the step (1), mutation signature analysis is carried out on a plurality of cancer patients suffering from cancers whose susceptibility is predicted, and the mutation rate of substitution mutations of the Age signature and/or the BRCA signature is specified to obtain the Age signature and The method according to [9] above, wherein a BRCA signature is prepared.
[11] The method according to [9] above, wherein in the step (2), the contribution is evaluated by calculating an Age signature score and a BRCA signature score for the mutation catalog.
[12] The Age signature score is represented by the following formula (1):

(In the above formula, n=96, i means the i-th substitution mutation in the 96 types of substitution classification shown in FIG. 1, and Mi is the mutation rate of the i-th substitution mutation in the mutation catalog. Or the number of mutations, and KAi is the mutation rate of the i-th substitution mutation in the Age signature.)
The method according to [11] above, which is calculated by:
[13] The BRCA signature score is represented by the following formula (2):

(In the above formula, n=96, i means the i-th substitution mutation in the 96 types of substitution classification shown in FIG. 1, and Mi is the mutation rate of the i-th substitution mutation in the mutation catalog. Or the mutation number, KBi is the mutation rate of the i-th substitution mutation of the BRCA signature.)
The method according to [11] above, which is calculated by:
[14] In the step (3), the ratio A(%) of the BRCA signature score to the sum of the Age signature score and the BRCA signature score is calculated to evaluate the superiority of the contribution of the BRCA signature. 9] and the method according to any of [11] to [13].
[15] In the step (3), C to T with respect to the sum of the score of the substitution mutation other than the substitution of C to T (BRCA signature score) and the score of the substitution mutation of C to T (Age signature score) The ratio B (%) of the scores of substitution mutations other than the substitution to is evaluated to evaluate the superiority of the contribution of the BRCA signature. [9] and [11] to [13] above. Method.
[16] The method according to [14] above, wherein when the ratio A (%) exceeds 50%, it is determined that the condition (a1) is satisfied.
[17] The method according to [15] above, wherein when the ratio B (%) exceeds 50%, it is determined that the condition (a1) is satisfied.
[18] The following steps (4) to (6):
(4) Providing a mutant signature including at least an Age signature and a BRCA signature,
[5] further comprising (5) evaluating the contribution of each mutation signature to the mutation catalog, and (6) evaluating the superiority of the contribution of the BRCA signature to the contribution of the Age signature. The method according to any one of [8].
[19] (7) The method according to [18] above, which further comprises evaluating the contribution of the BRCA signature to the entire mutation signature.
[20] The method according to [18] above, wherein the mutation signature includes all or part of 30 mutation signatures in the COSMIC database.
[21] The method according to [18] above, wherein in the step (5), the contribution degree of the mutation signature to the mutation catalog is evaluated by the destructStructs analysis.
[22] In the step (6), by calculating the ratio C (%) of the contribution of the BRCA signature to the sum of the contribution of the Age signature and the contribution of the BRCA signature, the contribution of the BRCA signature is superior. The method according to [18] above, wherein the sex is evaluated.
[23] In the step (6), by calculating the ratio C (%) of the contribution of the BRCA signature to the sum of the contribution of the Age signature and the contribution of the BRCA signature, the contribution of the BRCA signature is superior. Sex, and
In the step (7), the contribution of the BRCA signature is evaluated by calculating the ratio D(%) of the contribution of the BRCA signature to the sum (1) of the contributions of the mutant signatures. The method described.
[24] The method according to [22] above, wherein when the ratio C (%) exceeds 50%, it is determined that the condition (a1) is satisfied.
[25] When the ratio C (%) exceeds 50% and the ratio D (%) exceeds k% (k=20 to 30), it is determined that the above (a1) is satisfied, [23] The method described in.
[26] comprising the step of determining that the cancer cell contains a cancer cell population having sensitivity to a PARP inhibitor when both (a1) and (a2) of the condition (A) are satisfied. The method according to any one of 1] to [25].
[27] further comprising the step of performing an exhaustive methylation analysis on the subject test sample, wherein (B) methylation of the promoter region of BRCA1 indicates the presence of a cancer sensitive to a PARP inhibitor, The method according to any one of [1] to [26].
[28] A method for detecting cancer having homologous recombinant repair deficiency (HRD), comprising:
Comprising the step of performing a sequence analysis on a nucleic acid sample obtained from a cancer cell to create a mutation catalog, wherein (A)(a1) the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog. And (a2) the absence of cyclin E1 amplification indicates the presence of a cancer having a homologous recombination repair deficiency.
[29] The method further comprising the step of performing an exhaustive methylation analysis on the test sample of interest, wherein (B) methylation of the promoter region of BRCA1 indicates the presence of a cancer having a homologous recombination repair deficiency. ] The detection method of description.
[30] A method for treating a cancer patient with a PARP inhibitor, which comprises carrying out the method according to any of [1] to [27] above to obtain a cancer patient having a cancer sensitive to the PARP inhibitor. Identify or perform the method according to [28] or [29] above to identify a cancer patient having a cancer with a homologous recombination repair deficiency, and then administer an effective amount of the PARP inhibitor to the patient. A method of treatment comprising:
[31] A therapeutic agent for cancer comprising a PARP inhibitor as an active ingredient, wherein the treatment is carried out by the method according to [30] above.

The prediction methods [1] to [27] and the detection methods [28] and [29] above may be referred to as the “method of the present invention” below.

The method of the present invention is advantageous in that it can detect a cancer having a homologous recombination repair deficiency without excess or deficiency.

FIG. 1 is a diagram showing substitution classifications of 6 classes of 96 types of somatic gene substitution mutations in human cancer. Each class includes 16 trinucleotide sequences of 4 kinds of nucleotides one base before the mutated base and 4 kinds of nucleotides one base after the mutated base, and substitution mutations of 6 classes are classified into 96 kinds of trinucleotide sequences. FIG. 2 is a diagram showing a mutation signature analysis performed on a fresh frozen sample of 78 cases of ovarian serous cancer in Example 1(1) using whole exon sequence analysis. (A) shows the relationship between the number of signatures by non-negative matrix factorization (NMF) and stability and swing width. (B) shows the Age signature and BRCA signature extracted by integrating the gene mutations (mutation catalog) of 78 cases of ovarian serous cancer and performing non-negative matrix factorization (NMF). Each signature is represented by a 96 substitution classification. The abscissa indicates the mutant type (96 types of substitution classification, see FIG. 1), and the ordinate indicates the rate of mutation of the specific mutant type. Mutational signatures are displayed based on the trinucleotide frequency of the human genome. FIG. 3 is a diagram showing the contribution of the mutation signature to the mutation catalog of each case by whole exon sequence analysis of fresh frozen specimens of 78 cases of serous ovarian cancer in Example 1(2). Each bar graph shows each case, and (A) and (B) arranged each case correspondingly. (A) shows the results of analysis of the contribution of BRCA signature and Age signature to the mutation catalog of each case by non-negative matrix factorization (NMF). The vertical axis represents the number of mutations. The ratio of each signature score to the sum of the BRCA signature score and the Age signature score was calculated for each case. (B) The figure sorted based on the ratio of the signature score when the number of mutations in each case in (A) is 100% is shown. The cases in which the BRCA signature was dominantly contributed were 76.9% (60 cases out of 78 cases), and the average mutation number was 95.7. FIG. 4 is a diagram showing the contribution of the mutation signature to the mutation catalog of each case in 78 cases of TCGA, ICGC, and high-grade ovarian cancer in Example 1(2). Each vertical line indicates each case, and the cases are arranged in correspondence with each other in (A) and (B). (A) shows the results of analysis of the contribution of the BRCA signature and the Age signature to the somatic mutation of each case by nonnegative matrix factorization (NMF). The vertical axis represents the number of mutations. The ratio of contribution of each signature to the sum of the ratio of contribution of BRCA signature and the ratio of contribution of Age signature was calculated in each case. (B) The figure sorted based on the ratio of the contribution of each signature when the number of mutations in each case in (A) is set to 100% is shown. 67.5% (54 out of 80 cases, average mutation number: 92.5) in ICGC and 63.3% (200 out of 316 cases, average mutation number) in TCGA had a significant contribution of BRCA signature. Was 57.4). FIG. 5 shows that in Example 2(2), whole exon sequence analysis, non-negative matrix factorization (NMF), exhaustive methylation analysis and analysis of cyclin E1 amplification were performed on fresh frozen samples of 78 cases of ovarian serous cancer, FIG. 3 is a diagram showing the results (signatures that contribute to dominance, presence/absence of methylation of the BRCA1 promoter region, and presence/absence of cyclin E1 amplification). As a result of the determination by MSBM, 68% of cases were determined to be MSBM positive. Gray boxes represent cases in which the contribution of the BRCA signature was predominant, cases with methylation of the BRCA1 promoter region or cases without the cyclin E1 amplification. FIG. 6 shows a mutation catalog of each case by the oncogene panel test in Example 3. Case X was positive for gBRCA1/2 mutation and positive for MSBM. FIG. 7 shows a mutation catalog of each case by the oncogene panel test in Example 3. Case Y was negative for gBRCA1/2 mutation and positive for MSBM. FIG. 8 shows a mutation catalog of each case by the oncogene panel test in Example 3. Case Z was negative for gBRCA1/2 mutation and negative for MSBM. FIG. 9 is a diagram showing a comparison between the results of NMF analysis in Example 1(2) and the results of destructSigs analysis in Example 4 for 78 cases of TCGA, ICGC, and high-grade ovarian cancer. (A) shows the results of analysis of the contribution of the BRCA signature and the Age signature to the somatic mutation of each case by nonnegative matrix factorization (NMF). The ratio of contribution of each signature to the sum of the ratio of contribution of BRCA signature and the ratio of contribution of Age signature was calculated for each case, and based on the ratio of contribution of each signature when the mutation number of each case was 100%. FIG. 4B shows a sorted view. (B) shows the results of analyzing the contributions of the BRCA signature and the Age signature to the somatic mutation in each case by destructStructs. Since only the contributions of the BRCA signature and the Age signature are displayed, and the contributions of the signatures other than these two types are not displayed, the total contribution is not 1 in each case. FIG. 10 is a scatter diagram showing the contribution of the BRCA signature by the NMF analysis in Example 1(2) and the destructStructs analysis in Example 4 for 78 cases of TCGA, ICGC and high-grade ovarian cancer. FIG. 11 is a scatter diagram showing the contribution of the BRCA signature by the NMF analysis in Example 1(2) and the destructStructs analysis in Example 4 for 78 cases of TCGA, ICGC and high-grade ovarian cancer. FIG. 12 is a scatter diagram showing the contribution of the BRCA signature by the NMF analysis in Example 1(2) and the destructStructs analysis in Example 4 for 78 cases of high-grade ovarian cancer.

Detailed explanation of the invention

In the present invention, the “PARP inhibitor” refers to a drug that inhibits the function of poly(ADP ribose) polymerase-1, and examples thereof include olaparib, lucaparib, niraparib, beriparib, and tarazoparib.

In the present invention, the cancer cell can be a cell obtained from a cancer patient. Although the frequency is rare, germline mutations in the BRCA gene are widely found in solid cancers (Mandelker D. et.al., JAMA. 318:825-835(2017)). It can be excluded cancer or solid cancer. In a preferred embodiment of the present invention, the cancer may be a cancer that may exhibit a homologous recombination repair abnormality, and examples of such cancer include ovarian cancer (particularly ovarian serous cancer, intraclass). High-grade ovarian cancer such as membrane cancer, carcinosarcoma, mixed cancer, and undifferentiated cancer), breast cancer, prostate cancer, and pancreatic cancer.

In the present invention, “susceptibility” means a response to a standard dose of a drug or a standard treatment procedure.

In the present invention, the “Age signature” is a mutation signature showing a correlation with aging in cancer cells. In the present invention, the “BRCA signature” is a mutant signature showing a correlation with a homologous recombination repair abnormality due to factors including BRCA1/2 mutation and BRCA1/2 inactivation in cancer cells.

In the present invention, the “mutation signature” is a characteristic mutation pattern or profile extracted from somatic gene substitution mutations in a plurality of human cancers, and includes 6 classes of 96 substitution classifications (C→A, C→G, C→T, T→A, T→C, T→G, each of which is classified according to 16 types of base-substituted trinucleotides (see FIG. 1), and the mutation type is the horizontal axis, and the ratio of mutations contributing to a specific mutation type Can be displayed on the vertical axis (Alexandrov LB et al., Nature, 500:415-421 (2013)).

In the present invention, the “mutation catalog” refers to 6 classes of 96 somatic gene substitution mutations identified for a test nucleic acid sample (C→A, C→G, C→T, T→A, T→C). , T→G, respectively, according to 16 types of base-substituted trinucleotides (see FIG. 1). The classified or organized substitution mutations can be expressed by a mutation rate (%) (ratio to all substitution mutations) or the number of mutations.

The method of the present invention includes the step of performing sequence analysis on a nucleic acid sample obtained from cancer cells. The sequence analysis includes DNA sequence analysis, RNA sequence analysis, whole exon sequence analysis, and whole genome sequence analysis. In the present invention, the susceptibility of a target cancer to a PARP inhibitor can be predicted, and a cancer having a homologous recombination repair deficiency. Can be used without particular limitation as long as it can be detected. Sequence analysis may also be carried out by an oncogene panel test, and as an oncogene panel test that can be used in the present invention, Todai OncoPanel (University of Tokyo Hospital), MSK-IMPACT, Foundation One Are listed. It should be noted that when the present invention is approved as a drug in relation to the whole exon sequence analysis or the whole genome sequence analysis, the present invention is advantageous because it can be positioned as an in vitro diagnostic agent capable of accurately determining the sensitivity of a PARP inhibitor.

In the present invention, the sequence analysis may consist of a sequence determination procedure and a sequence data analysis procedure, for example, the sequence determination procedure is performed on a nucleic acid sample, and the sequence data is obtained based on the sequence data obtained by the sequence determination procedure. Analysis procedures can be performed.

The sequencing procedure can include the steps of sequencing the nucleic acid sample obtained from the subject and obtaining sequence data for the nucleic acid sample. Sequencing can be performed, for example, using a next generation sequencer.

The sequence data analysis procedure can include a step of analyzing the obtained sequence data and obtaining a desired analysis result. The analysis of sequence data can include analysis for creating a mutation catalog, analysis for determining condition (A) or condition (B), and signature analysis.

The analysis for creating the mutation catalog can be performed as follows, for example. That is, somatic gene mutations are identified in a test nucleic acid sample by sequence data analysis, and the identified somatic gene mutations are classified or organized according to the 96 types of substitution classifications shown in FIG. 1 to create a mutation catalog of test nucleic acid samples. can do. Here, the identification of the somatic gene mutation is performed by comparing the sequence data of normal tissues of cancer patients (for example, white blood cells in blood, surgical/biopsy tissue specimens of non-tumor portion) with tumor tissues of cancer patients (for example, cancer cells). ) It can be performed by comparison with the sequence data. Such comparison of sequence data and identification of somatic gene mutations can be performed according to a known algorithm (for example, Totokietetal., Nat Genet.2014Dec;46(12):1267-73). The identified mutations include base substitutions, deletions, insertions and additions, copy number changes, inversions, translocations, and fusion genes.

In the analysis for determining the condition (a1) among the conditions (A), it is determined whether the mutation catalog obtained for the case to be evaluated is similar to the Age signature or the BRCA signature, that is, the contribution of the Age signature. The degree of predominance of the contribution of the BRCA signature to the can be analyzed. Such an analysis can be performed using a known algorithm for pattern recognition such as non-negative matrix factorization (NMF), or can be performed as in the following (1) to (3). ..

(1) Preparation of Age signature and BRCA signature In (1), the mutation rate of substitution mutation of Age signature and/or BRCA signature is prepared according to the 96 types of substitution classification shown in FIG. For example, a mutation signature analysis is performed in advance on nucleic acid samples of a plurality of cancer patients suffering from a cancer that predicts susceptibility, and the mutation rate of substitution mutations of the Age signature and/or the BRCA signature for the 96 types of substitution classification shown in FIG. Can be specified. The mutation signature analysis can be performed, for example, as described in Alexandrov LB et al., Nature, 500:415-21 (2013), in which the mutation signature analysis is performed by non-negative matrix factorization (NMF). be able to. Moreover, it is preferable to use the same method as the analysis method of the mutation catalog for the signature analysis. For example, when the mutation catalog is created by the total exon sequence analysis, the mutation signature analysis can also be performed by the total exon sequence analysis.

In (1), instead of carrying out the mutation signature analysis, the mutation rate of the substitution mutation of the known Age signature and BRCA signature may be used (for example, Alexandrov LB et al., Nature, 500:415-421). (2013)). Here, the mutation rate can be rephrased as a weighting (coefficient) for each substitution classification that characterizes the Age signature and the BRCA signature. Further, the mutation rate obtained by the mutation signature analysis or the known mutation rate can be used as it is in (2) below, but a part or all of the mutation rate may be arbitrarily changed to make a better judgment. May be.

(2) Evaluation of Contribution of Age Signature and BRCA Signature to Mutation Catalog In (2), each contribution of Age signature and BRCA signature to the mutation catalog of the test nucleic acid sample is evaluated. The Age signature contribution and the BRCA signature contribution can be analyzed by non-negative matrix factorization (NMF). Evaluation of the Age signature contribution and the BRCA signature contribution can also be performed by calculating the Age signature score and the BRCA signature score for the mutation catalog. Specifically, the Age signature score and the BRCA signature score are the mutation rates of the substitution mutations obtained in (1) above for the substitution mutation amounts (mutation rate or number) constituting the mutation catalog for all 96 substitution mutations. It can be calculated by multiplying by and calculating the product, and obtaining the total value of the products obtained for 96 types of substitution mutations.

For example, the Age signature score can be calculated as follows.

(In the above formula, n=96 (that is, i=1 to 96), i is the i-th substitution mutation of the 96 substitution classes shown in FIG. 1, and Mi is the mutation catalog The mutation rate or the number of mutations of the i-th substitution mutation, and KAi is the mutation rate of the i-th substitution mutation of the Age signature.)

The BRCA signature score can be calculated as follows, for example.

(In the above formula, n=96 (that is, i=1 to 96), i is the i-th substitution mutation of the 96 substitution classes shown in FIG. 1, and Mi is the mutation catalog The mutation rate of the i-th substitution mutation, KBi is the mutation rate of the i-th substitution mutation of the BRCA signature.)

(3) Evaluation of superiority of contribution of BRCA signature to contribution of Age signature In (3), superiority of contribution of BRCA signature to contribution of Age signature is evaluated. Specifically, which signature is superior to the mutation catalog is evaluated based on the Age signature score and the BRCA signature score calculated in (2) above. For example, the ratio A (%) of the BRCA signature score to the sum of the Age signature score and the BRCA signature score can be calculated, and the superiority can be evaluated based on the ratio A. When the obtained ratio A exceeds 50%, it can be determined that the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog, that is, the condition (a1) is satisfied.

When the sequence analysis is performed by whole exon sequence analysis or whole genome sequence analysis, the mutation rate of Age signature and BRCA signature is the Age signature described in Alexandrov LB et al., Nature, 500:415-421 (2013). And the BRCA signature, or a signature analysis may be performed to set the Age signature and the BRCA signature and the mutation rate may be determined based on it.

When the sequence analysis is performed by an oncogene panel test, signature analysis can be performed to set the Age signature and the BRCA signature, and the mutation rate can be determined based on that, or other than the substitution mutation from C to T Substitution mutations (corresponding to substitution mutations i=1 to 32 and 49 to 96 in FIG. 1) reflect the BRCA signature, and substitution mutations C to T (corresponding to substitution mutations i=33 to 48 in FIG. 1) ) May reflect the Age signature, the Age signature score and the BRCA signature score may be calculated. In the latter case, the Age signature score is expressed by the above formula (1) (where i=33 to 48, KAi=1, and i=1 to 32 and 49 to 96, KAi=0. The BRCA signature score can be calculated by the equation (2) (where i=1 to 32 and 49 to 96, KBi=1, and i=33 to 48, KBi=1. =0). For example, the sum of the score of the substitution mutation other than the substitution of C to T (BRCA signature score) and the score of the substitution mutation of C to T (Age signature score) of substitution mutations other than the substitution of C to T The ratio B (%) of the scores can be calculated, and the superiority can be evaluated based on the ratio B. When the obtained ratio B exceeds 50%, it can be determined that the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog, that is, the condition (a1) is satisfied.

The analysis for determining the condition (a1) of the condition (A) can also be performed as in the following (4) to (6).

(4) Preparation of Mutation Signature In (4), a plurality of mutation signatures including at least an Age signature and a BRCA signature are prepared according to the 96 types of substitution classification shown in FIG. The lower limit of the number of mutation signatures prepared in (4) can be 10 or 20, and the upper limit of the number can be 70, 60, 50 or 40. These lower limit values and upper limit values can be arbitrarily combined, and the range of the above number can be, for example, 10 to 70 kinds or 20 to 40 kinds.

As the mutation signature including at least the Age signature and the BRCA signature, a set of known mutation signatures can be used, and, for example, 30 kinds of mutations registered in the COSMIC (Catalogue of somatic mutations in cancer) database of the UK Sanger Institute can be used. All or part of the mutation signature (COSMIC version 2), and among the mutation signatures (COSMIC version 3) registered in the COSMIC database of the institute, SBS (Single Base Substitution) signature (67 types), Doublet Base Substitution ( All or part of the DBS) signature (11 types) or Small Insertion and Deletion (ID) signature (17 types) can be used.

(5) Evaluation of Contribution of Age Signature and BRCA Signature to Mutation Catalog In (5), the contribution of each mutation signature to the mutation catalog of the test nucleic acid sample is evaluated. For example, when 30 mutation signatures are prepared, the contribution is evaluated by calculating the optimum weight (contribution) for each of the 30 mutation signatures so that the mutation profile of the mutation catalog can be reconstructed. be able to. Such a calculation of the degree of contribution can be performed using a known program, for example, by the destructStructs analysis (Rosenthal et al., Genome Biology, 17:31 (2016)).

(6) Evaluation of superiority of contribution of BRCA signature to contribution of Age signature In (6), superiority of contribution of BRCA signature to contribution of Age signature is evaluated. Specifically, which signature is superior to the mutation catalog is evaluated based on the contribution of the Age signature and the contribution of the BRCA signature calculated in (5) above. For example, the ratio C(%) of the contribution of the BRCA signature to the sum of the contribution of the Age signature and the contribution of the BRCA signature can be calculated, and the superiority can be evaluated based on the calculated ratio. When the obtained ratio C exceeds 50%, it can be determined that the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog, that is, the condition (a1) is satisfied. ..

In the analysis for determining the condition (a1) of the conditions (A), the following (7) can be further performed in addition to the above (4) to (6).
(7) Evaluation of Contribution of BRCA Signature to Overall Mutant Signature In (7), evaluation of contribution of BRCA signature to entire mutant signature. Specifically, based on the contribution of the BRCA signature calculated in (5) above, the ratio of the BRCA signature to the entire prepared mutation signature is evaluated. For example, the ratio D(%) of the contribution of the BRCA signature to the sum of the contributions of all prepared mutation signatures (100%) can be calculated, and the contribution of the BRCA signature can be evaluated based on the ratio D. .. When the ratio C obtained in (6) exceeds 50% and the ratio D obtained in (7) exceeds k% (k=15 to 40, preferably 20 to 30), the mutation It can be determined that the contribution of the BRCA signature to the catalog exceeds the contribution of the Age signature to the mutant catalog, that is, the condition (a1) is satisfied.

In the analysis for determining the condition (a2) of the condition (A), the chromosome copy number can be analyzed using the probe (SNP) of the genomic region of cyclin E1 as an index. It can be determined that the condition (a2) is satisfied when the copy number of the cyclin E1 on the chromosome number is 2 or less.

In the analysis for determining condition (B), comprehensive methylation analysis can be performed to identify the methylation of the promoter region of BRCA1 in the test sample. The method of exhaustive methylation analysis is publicly known and can be carried out, for example, according to The Cancer Genome Atlas Research Network, Nature, 474: 609-615 (2011).

In the present invention, when the following condition (A) is satisfied (that is, both conditions (a1) and (a2) are satisfied), the target cancer may be a cancer sensitive to a PARP inhibitor. It can be determined that the cancer is highly likely or that the cancer has a homologous recombination repair deficiency.
Condition (A): the contribution of the BRCA signature to the (a1) mutation catalog exceeds the contribution of the Age signature to the mutation catalog, and (a2) the absence of cyclin E1 amplification.

In the present specification, the characteristic or profile of the condition (A) is referred to as Mutational Signature-based BioMarker (MSBM), and at least the condition (A) is satisfied with MSBM positive, and the condition (A) is not satisfied with MSBM negative. , There are things to say. When the condition (A) is satisfied, that is, when it is determined that both the condition (a1) and the condition (a2) are satisfied as a result of the sequence analysis, the MSBM positive can be determined. When the analysis for determining the condition (a1) is performed by the steps (4) to (6), the step (7) is further performed and the condition (A) is set from the viewpoint of accurate determination. The following conditions can be set.
Condition (A): (a1) The contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog, and the contribution of the BRCA signature to the entire mutation signature exceeds k% (k=15-40), And (a2) absence of cyclin E1 amplification

In the present invention, the following condition (B) can be supplementarily used, and when the following condition (B) is satisfied, the target cancer is highly likely to be a cancer having sensitivity to a PARP inhibitor. Alternatively, it can be determined that the cancer is highly likely to have a homologous recombination repair deficiency.
Condition (B): methylation of the promoter region of BRCA1

When the condition (B) is satisfied, that is, when the result of the comprehensive methylation analysis indicates that the promoter region of BRCA1 is methylated, the condition (A) is satisfied, and the MSBM positive condition is satisfied. Can be determined. In this sense, the profile of condition (B) is sometimes referred to as MSBM.

In the present invention, the determination of MSBM positive/negative does not depend on the analysis result of gBRCA mutation. Therefore, even if the subject is negative for gBRCA mutation 1/2, it can be determined to be positive for MSBM if the condition (A) is satisfied.

The condition (a1) “dominance of BRCA signature over Age signature” is defined as the ratio (%) of the sum of Age signature scores obtained for nucleic acid samples and the sum of BRCA signature scores to the sum of BRCA signature scores. It can be determined based on the above conditions, and when the ratio exceeds 50%, it can be determined that the condition (a1) is satisfied.

Condition (a2) "absence of cyclin E1 amplification" can be determined to satisfy condition (a2) when the copy number of cyclin E1 is 2 or less, and when the number of chromosomes is greater than 2. Can be determined not to satisfy the condition (a2).

In the method of the present invention, a mutation signature analysis is performed in advance on nucleic acid samples of a plurality of cancer patients suffering from a cancer that predicts susceptibility, and the substitution mutations of Age signature and BRCA signature of the 96 types of substitution classification shown in FIG. 1 are analyzed. The mutation rate can be specified. That is, according to the method of the present invention, an Age signature and a BRCA signature can be set for each type of sequence analysis, and an optimal inspection method can be constructed. For example, a signature analysis is performed in a specific cancer gene panel test to prepare an Age signature and a BRCA signature, and the PARP inhibitor according to the method of the present invention is used by using the Age signature and the BRCA signature prepared in advance in the oncogene panel test. It is possible to accurately and simply predict the susceptibility of the cancer to the cancer and to accurately and easily detect the cancer having a homologous recombination repair deficiency (HRD). When performing signature analysis in the cancer gene panel test, archive samples may be used.

According to the present invention, there is provided a method for treating a cancer patient with a PARP inhibitor, which comprises performing the prediction method of the present invention to identify a cancer patient having a cancer sensitive to the PARP inhibitor, or the present invention Is provided to identify a cancer patient having a cancer having a homologous recombination repair deficiency, and then to administer an effective amount of a PARP inhibitor to the patient. According to the present invention, there is also provided a cancer therapeutic agent comprising a PARP inhibitor as an active ingredient, wherein the above-mentioned treatment is performed by the procedure of the therapeutic method of the present invention. .. In the treatment method of the present invention, identifying a cancer patient having a cancer sensitive to a PARP inhibitor or identifying a cancer patient having a cancer having a homologous recombination repair deficiency is the predictive method of the present invention. And the detection method. A therapeutically effective amount of a PARP inhibitor can be the standard dose of that drug. For example, for olaparib, 600 mg (which may be appropriately reduced depending on adverse events) can be used as the effective daily dose for adults (oral administration), for example, 300 mg once a day (300 mg twice a day). (×2 times/day) (depending on adverse events, it may be gradually reduced to 250 mg×2 times/day, 200 mg×2 times/day, etc.). According to the therapeutic method of the present invention and the therapeutic agent of the present invention, since a PARP inhibitor can be administered to a cancer patient who is highly likely to respond to the PARP inhibitor, the cancer patient (particularly, the cancer patient exhibiting HRD) can be administered. Is advantageous in that it can provide treatment options suitable for

The present invention will be described more specifically based on the following examples, but the present invention is not limited to these examples.

Example 1: Mutation signature analysis by whole exon sequence analysis (1) Mutation signature analysis in high-grade ovarian cancer Fresh frozen samples of cancer cells and normal cells (blood) in 78 cases of high-grade ovarian cancer (serous ovarian cancer) All exon sequence analysis was carried out (Center for Advanced Science and Technology Research, University of Tokyo), and mutational signature analysis (Mutational signature) was performed as described in Alexandrov LB et al., Nature, 500:415-21 (2013). The mutation catalog of each case in all exon sequence analysis was integrated, and clustering was performed using nonnegative matrix factorization (NMF, Lee DD and Seung HS, Nature, 401(6755):1999). Stability and Reconstruction Error based on (mutation signature number) were calculated (Sawada et al., Gastroenterology, 150:1171-1182(2016), Rosales RA et al., Bioinformatics, 33:8- 16 (2017), Patch et al., Nature, 521:489-94 (2015)). Here, the higher the stability and the smaller the swing range, the more stable the clustering is.

The results were as shown in Figure 2. From the result of FIG. 2A, the value of the swing width decreases as the number of signatures increases, but the stability is 0.5 when the number of signatures is 3, whereas the stability is close to 1.0 when the number of signatures is 2, A signature number of 2 resulted in significantly higher stability. Since the larger the ratio of stability to the fluctuation range is, the more stable it is, the validity of dividing into two signatures (Age signature and BRCA signature) was shown. Also, from the results of FIG. 2B, characteristic patterns of gene mutations in each signature such as C to T substitution in Age signature and uniform base substitution mutation in BRCA signature (Alexandrov LB et al., Nature, 500) :415-421(2013)) was confirmed.

(2) Judgment of a signature that contributes to high-grade ovarian cancer by using whole-exon sequence analysis About 78 cases of high-grade ovarian cancer (serous ovarian cancer), each case by whole-exon sequence analysis in (1) above The contributions of the BRCA and Age signatures were analyzed for each of the variant catalogs based on non-negative matrix factorization (NMF). Specifically, the number of mutations in each base substitution of somatic gene mutations in each case was identified. Here, the base substitution can be classified into 6 classes (C→A, C→G, C→T, T→A, T→C and T→G), and 5′ and 3′ of each substitution base. Since each class contains 16 trinucleotide sequences, it can be classified into 96 base-substituted trinucleotides as a whole (Serena Nki-Zainal et al., Cell 149:979-993 (2012). )). The number of mutated trinucleotide sequences (mutation number) and the mutation rate were identified according to 96 types of substitution classification for each case, and a mutation catalog was prepared and scored. The scoring is performed by calculating the product of the mutation rate and the mutation rate of the mutated trinucleotide sequence in the Age signature or BRCA signature (see FIG. 2B) calculated in (1) above for each substitution classification. went. The sum of products calculated in relation to the Age signature or the BRCA signature was used as the Age signature score or the BRCA signature score. For each case, the ratio (%) of the Age signature score or the BRCA signature score to the sum of the Age signature score and the BRCA signature score was calculated. Further, the ratios were calculated and sorted in the same manner for the mutation catalogs in TCGA and ICGC, which are public data sets.

The results were as shown in Figures 3 and 4. It was confirmed that 76.9% of cases had a dominant contribution of the BRCA signature to the contribution of the Age signature. It was also confirmed that the group in which the contribution of the BRCA signature was dominant had a significantly larger number of mutations (p<0.0001) than the group in which the contribution of the Age signature was dominant (see FIG. 3 ). Non-negative matrix factorization (NMF) in TCGA or ICGC confirmed that the cases in which the contribution of the BRCA signature was dominant were 63.3% and 67.5%, respectively (see FIG. 4).

Example 2: Biomarker (MSBM) setting in whole exon sequence analysis (1) MSBM setting In serous ovarian cancer, amplification of cyclin E1 (Cyclin E1; CCNE1) is positively correlated with Age signature (Etemadmoghadam, D et al., Proc Natl Acad Sci USA, 110:9489-94 (2013), Patch AM et al., Nature, 521:489-94 (2015)). Therefore, “the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog and the amplification of cyclin E1 is absent” is set as the condition (A), and the profile corresponding to the condition (A) is obtained. , A biomarker for predicting susceptibility to PARP inhibitors targeting a homologous recombination repair abnormality and a biomarker for detecting a homologous recombination repair abnormality. In addition, it is known that methylation of the promoter region of BRCA1 causes BRCA1 inactivation and leads to abnormal homologous recombination repair (The Cancer Genome Atlas Research Network, Nature, 474: 609-615 (2011), Moschetta M et al., Ann Oncol, 8:1449-55 (2016)). Therefore, “having methylation of the promoter region of BRCA1” was set as the condition (B).

Regarding condition (A), further condition (a1): “the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog” and condition (a2): “absence of amplification of cyclin E1”. Condition A is satisfied when both condition (a1) and condition (a2) are satisfied. Condition (a1) is calculated by total exon sequence analysis and non-negative matrix factorization (NMF), specifically, the BRCA signature score and Age signature score are calculated according to the method described in Example 1(2) for the mutation catalog of each case. Then, it was determined that the condition (a1) was satisfied when the ratio of the BRCA signature score to the sum of the BRCA signature score and the Age signature score exceeded 50%. Regarding the condition (a2), the chromosome copy number was analyzed by examining a probe (SNP) in the genomic region of cyclin E1, and it was determined that the condition (a2) was satisfied when the chromosome copy number of cyclin E1 was 2 or less ( Ohshima, et al., Sci Rep, 7:641 (2017)).

Regarding condition (B), an exhaustive methylation analysis (OuchiK et al., Cancer Sci., 106:1722-9 (2015)) was performed. Specifically, based on a methylation array (Infinium HumanMethylation450 BeadChip, manufactured by Illumina), 6 to 12 methylation probes in the promoter region of BRCA1 are extracted, and clustering is performed based on the methylation value. It was divided into a hypermethylated group and a hypomethylated group. The hypermethylated group was determined to satisfy the condition (B).

(2) Judgment by MSBM in serous ovarian cancer Judgment by MSBM was carried out in 78 cases of high-grade ovarian cancer (serous ovarian cancer) according to the method and criteria described in (1) above. The result was as shown in FIG. It was shown that 68% of 78 high-grade ovarian cancers (ovarian serous cancer) were MSBM positive.

Example 3: Setting of MSBM in oncogene panel test (1) Determination of signature that contributes to mutation catalog predominantly All exons of 78 cases of high-grade ovarian cancer (serous ovarian cancer) performed in Example 2 (2) Following the MSBM determination by sequence analysis and non-negative matrix factorization (NMF), 7 cases were subjected to MSBM determination using an oncogene panel test. Specifically, in 7 cases of ovarian serous cancer (including serous peritoneal cancer), DNA was extracted from the tumor tissue and blood collected from the patient, and sequenced and sequenced using Todai OncoPanel (University of Tokyo Hospital). Data analysis was performed and a mutation catalog for each case was created. The mutation catalog was created based on the 96 types of substitution classification shown in FIG. 1 and the mutation rate for each substitution mutation.

Furthermore, germline BRCA (germline BRCA; gBRCA) mutations are known databases such as ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) and COSMIC (http://www.sanger). .ac.uk/genetics/CGP/cosmic/) and OncoKB (https://oncokb.org/#/) were used to determine pathological significance. Specifically, when it was determined to be "pathogenic" or "highly likely" in the database, it was determined to be germline BRCA mutation positive. In the case of determination of VUS (unclear significance), germline BRCA mutation was positive.

(2) Results The mutation catalog of each case and the results of gBRCA mutation analysis were classified into X, Y and Z, and the results are shown in FIGS. 6 to 8. No cyclin E1 amplification was observed in any of the cases (data not shown).

Classification X
FIG. 6 shows the results of classification X. Analysis of gBRCA mutations revealed that case X1 (uterine body ovarian cancer poorly differentiated endometrioid adenocarcinoma), case X2 (lymph node ovarian cancer lymph node metastasis, same patient as case X1) and case X3 (spleen ovarian cancer Seeding to the splenic hilum) was positive for gBRCA1/2 mutation. Therefore, it was predicted that the contribution of the BRCA signature to the mutation catalog of cases X1, X2 and X3 would exceed the contribution of the Age signature. In any of the mutation catalogs of case X1, case X2, and case X3, the score of Age signature when the substitution of C to T is defined as the Age signature (the above formula (1) (where i=33 to 48) Is KAi=1 and KAi=0 when i=1 to 32 and 49 to 96. The same applies in Example 3) and substitutions other than C to T substitution (from C to Substitution of A, C to G, T to A, T to C and T to G, and so on (the above formula (2) (where i=1 to 32 and 49 to 96, KBi= 1 and i=33 to 48 when KBi=0). The same applies to the sum of the following) and the score of substitution other than C to T substitution exceeds 50%. Was confirmed. From the above results, it was suggested that substitutions other than C to T substitution could be set as the BRCA signature. In other words, in the Todai OncoPanel examination, it was found that the contribution of the BRCA signature (corresponding to substitutions other than C to T substitution) exceeds the contribution of Age signature (corresponding to C to T substitution), which indicates homologous recombination failure. It is considered that it can be used as an index of cancer.

In addition, cases X1, X2, and X3 were positive for gBRCA1/2 mutation according to the analysis of gBRCA mutation. From these results, it can be said that the cases of classification X are “gBRCA1/2 mutation positive” and “MSBM positive”. In case X3, it was confirmed that the progression-free survival was maintained by the administration of the PARP inhibitor olaparib. From the above, it was confirmed that cancer having a homologous recombination repair abnormality (HRD) can be detected by the determination by MSBM, and the sensitivity to a PARP inhibitor can be predicted.

Classification Y
FIG. 7 shows the results of case Y. In both case Y1 (ovarian cancer, metastatic breast cancer) and case Y2 (ovarian cancer, metastatic breast cancer), calculation was performed in the same manner as in the method described in classification X, and the BRCA signature (except for the substitution of C to T) was calculated. It was confirmed that the BRCA signature score for the sum of the (corresponding to substitution) score and the Age signature (corresponding to C to T substitution) score was more than 50%. Furthermore, from the results of classification X and classification Y, it was suggested that the analysis in Todai OncoPanel may have a tendency that the BRCA signature has a substitution pattern of T to C. As described above, it was confirmed that the BRCA signature and the Age signature in the Todai Onco Panel can be determined in more detail by accumulating future cases.

Moreover, according to the analysis of the gBRCA mutation, cases Y1 and Y2 were negative for the gBRCA1/2 mutation. From these results, it was confirmed that the cases of classification Y were “gBRCA1/2 mutation negative” and “MSBM positive”. From the above, it was confirmed that even if the gBRCA1/2 mutation is negative, a cancer having a homologous recombination repair abnormality (HRD) can be detected by the determination by MSBM, and the sensitivity to a PARP inhibitor can be predicted.

Classification Z
FIG. 8 shows the results of case Z. In both case Z1 (ovarian cancer recurrence) and case Z2 (peritoneal cancer after bilateral adnexectomy) Age signature (corresponding to C to T substitution) score and BRCA signature (other than C to T substitution) It was confirmed that the Age signature score with respect to the sum with the score was more than 50%. Here, case Z1 was resistant to platinum drug, which is the main drug against ovarian cancer. The sensitivity of platinum preparations has been shown to be high in cases with HRD and is known to correlate with the sensitivity of PARP inhibitors (De Picciotto N et al., Crit Rev Oncol Hematol., 101:50. -9 (2016)) (Indications for the PARP inhibitor olaparib in Japan are limited to platinum-sensitive recurrence). Therefore, it can be said that the case Z1 is highly likely to have low sensitivity to the PARP inhibitor olaparib. Moreover, according to the analysis of the gBRCA mutation, cases Z1 and Z2 were negative for the gBRCA1/2 mutation. These are classifications in which substitutions other than C to T substitutions in the Todai OncoPanel reflect the BRCA signature and are indicators of homologous recombination repair abnormality (HRD) when the contribution of the BRCA signature exceeds the contribution of the Age signature. Consistent with the X and classification Y considerations. From these results, it was confirmed that the cases of classification Z were “gBRCA1/2 mutation negative” and “MSBM negative”.

Example 4: Setting of MSBM using destructStructs and determination using it (1) destructStructs analysis a Whole exon sequence analysis DNA was extracted from tumor tissue (FFPE) and normal tissue (peripheral blood) of the case, and exome (Exome). The exon sequence was concentrated using a capture kit (Agilent SureSelect Human All Exon V6 (S07604514), Illumina), and a library was prepared according to the attached protocol. The sequence was obtained at 151 base pair ends using NextSeq (Illumina). The acquired data was converted into a fastq file for each sample based on the index sequence added for each library and stored in the storage in the computer used for the subsequent data processing. The fastq file is used when storing a nucleotide sequence such as DNA and its quality score together in one file, and is a format for storing nucleotide sequence data output from a next-generation sequencer or the like.

B. Mapping to the reference genome The fastq file obtained in the above step (a) is used to analyze the human genome sequence using bwa-mem (genome analysis algorithm implemented in mapping software bwa (http://www.liheng.org)). (Human reference genome sequence: GRCh38).

C. Mutation detection Mapping results (bam file) are processed with the mutation detection program karkinos (obtained at https://github.com/genome-rcast/karkinos) by combining the tumor part and normal part of the same case. , Created a list of somatic mutations.

Frequency data of 96 base substitution pattern One base substitution and one base before and after the substitution were extracted from the obtained mutation list, and the frequency was aggregated for each base substitution pattern. Specifically, by aligning the +/- strands of the genomic DNA into the strands whose bases before the change are C or T (for example, CGT>CAT is counted as the same as ACG>ATG). , 96 types of substitutions and mutation rates for each substitution mutation were totaled to create a mutation catalog.

Estimating mutation signatures Using the mutation frequencies of the 96 mutation patterns (that is, mutation catalogs) obtained in the above item d as input, deconstructStrigs, which is a program for estimating the weight of mutation signatures, is used to extract 30 existing signatures. The contribution was estimated. The set of 30 types of mutation signatures (version 2, March 2015) is the 30 types of mutation signatures registered in the COSMIC database (https://cancer.sanger.ac.uk/cosmic/signatures_v2) of the UK Sanger Research Institute. (COSMIC version 2) was used. Of the 30 types of signatures, the contributions of the BRCA signature (Signature 3 in the COSMIC database) and the AGE signature (Signature 1 in the COSMIC database) were analyzed in detail.

deconstructSigs is a program that estimates the weight (contribution) of the existing mutation signature for each sample, compares the mutation profile reconstructed from the calculated weight with the original input, and judges the appropriateness of the estimation ( Rosenthal et al., Genome Biology, 17:31 (2016)). In this example, the frequency of 96 mutation patterns of individual tumor samples [T: 1x96] (in parentheses indicates a matrix: matrix size (row x col). The same applies hereinafter) and existing mutation signature matrix. Using [S:30×96] as an input, the contribution degree Wi (i=1 to 30) of each signature was calculated. In order to determine the weight W that optimally recreates T, one initial signature Si closest to the mutation profile of the tumor sample of interest is selected from existing mutation signatures, and the tumor mutation profile in which Si is reconstructed is selected. W was determined to be the only signature that contributes to Next, a weight (contribution) matrix that minimizes the reconstruction error calculated as T-(SW) was determined by performing iterative calculation while changing the weight matrix [W].

(2) Comparison of Signature Contribution in NMF Analysis and destructStructs Analysis A method 78 cases of TCGA, ICGC and high-grade ovarian cancer were analyzed by the destructStrucs described in (1) above. The results of the analysis by destructStructs were compared with the results of the analysis (NMF analysis) based on the non-negative matrix factorization (NMF) described in Example 1(2).

B result The result was as shown in FIG. From the results of FIG. 9, it was confirmed that the contributions of Age signature and BRCA signature generally showed the same tendency in the NMF analysis and the destructStructs analysis. On the other hand, for example, in the NMF analysis, when the contribution ratio of the BRCA signature is 0.5 (see the vertical line in FIG. 9), the BRCA signature is dominant in the NMF analysis (>0.5), but the destructSigs analysis is performed. In cases where the BRCA signature is not dominant (<0.2), or conversely, in cases where the BRCA signature is not dominant in the NMF analysis (<0.5), the BRCA signature is not so low (>0.3) in the destructSigs analysis. , NMF analysis and destructStructSs analysis confirmed that there were a certain number of cases in which the signature contribution tendencies did not match.

(3) Comparison of contribution of BRCA signature in NMF analysis and destructStructs analysis a. Method For 78 cases of TCGA, ICGC and high-grade ovarian cancer, the destructStrucs analysis described in (1) above was performed. The contribution of the BRCA signature (Signature 3 in the COSMIC database) obtained by the destructStructs analysis and the contribution rate of the BRCA signature obtained by the NMF analysis described in Example 1 (2) are shown in a scatter diagram, and the contribution of the BRCA signature is determined by the NMF analysis. The threshold value of the BRCA signature of the destructStrigs analysis (Signature 3 in the COSMIC database) corresponding to the threshold value (0.5) was determined.

B. Results The results were as shown in FIGS. 10 and 11. From the result of FIG. 10, the threshold value of the destructStructs analysis corresponding to the threshold value (0.5) determined by the NMF analysis was 0.25. From the results in FIG. 11, it was shown that 66% of the cases were superior to the BRCA signature when the threshold value (0.5) of NMF analysis was used. Further, when the threshold value of the destructStructs analysis (0.25) was used, it was shown that the contribution of the BRCA signature was 0.25 or more in 54% of the cases.

(4) Setting method of biomarker (MSBM) in destructStructs analysis A method of destructStrigs described in (1) above was performed on 78 cases of high-grade ovarian cancer. Specifically, with respect to the mutation catalog of each case, the contribution of 30 kinds of mutation signatures including the BRCA signature and the Age signature was calculated according to the method described in (1) above, and the contribution of the BRCA signature was 0.25( 25%) or more and the condition (a1) is satisfied when the contribution of the BRCA signature exceeds the contribution of the AGE signature, except that the condition and the condition described in Example 2(1) are satisfied. (A) (Condition (a2): "Amplification of cyclin E1 is absent" is included).

B. Results The resultStructSigs analysis showed that the contribution of the BRCA signature exceeded the contribution of the AGE signature in 73% (57/78) of the cases, and the mutation in 73% (57/78) of the cases. It was shown that the contribution of the BRCA signature to the entire signature is 0.25 (25%) or more. In addition, in 73% (57/78), the contribution of the BRCA signature to the entire mutation signature was 0.25 (25%) or more, and the contribution of the BRCA signature exceeded the contribution of the AGE signature. there were. Furthermore, the contribution of the BRCA signature to the entire mutation signature is 0.25 (25%) or more, and the contribution of the BRCA signature exceeds the contribution of the AGE signature, and there is no cyclin E1 amplification. 68% (53/78) cases were confirmed to exist. That is, it was shown that, out of 78 cases of high-grade ovarian cancer (serous ovarian cancer), 68% of cases were MSBM positive.

Here, the germline HRD mutation (HRD genes' germline mutation; including germline BRCA1/2 mutation, RAD51C mutation and RAD51D mutation) is high-grade ovarian cancer (serous ovarian cancer) 78 Where 23 of the cases exist (Fig. 12), 18 cases exist in the cases where the contribution of the BRCA signature to the entire mutation signature is 0.25 (25%) or more (78%), and the contribution of the BRCA signature. There were 13 cases (57%) in which the degree exceeded the contribution of the AGE signature. From these results, it can be determined that the condition (a1) is satisfied when the contribution of the BRCA signature by the destructStructs analysis exceeds the contribution of the AGE signature, but for more accurate determination, the contribution of the BRCA signature is Is 0.25 (25%) or more and the contribution of the BRCA signature exceeds the contribution of the AGE signature, it can be determined that the condition (a1) is satisfied. From the above, the possibility that a cancer having a homologous recombination repair abnormality (HRD) can be detected by the determination by MSBM based on the destructStructs analysis and the sensitivity to a PARP inhibitor can be predicted was shown.

Claims (31)

1. A method of predicting cancer susceptibility to a PARP inhibitor comprising:
   Comprising the step of performing a sequence analysis on a nucleic acid sample obtained from a cancer cell to create a mutation catalog, wherein (A)(a1) the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog. And (a2) the absence of cyclin E1 amplification indicates the presence of a cancer sensitive to a PARP inhibitor.
2. The method according to claim 1, wherein the cancer cells are cells obtained from a cancer patient.
3. The method according to claim 1 or 2, wherein the cancer is a cancer that may exhibit a homologous recombination repair abnormality.
4. The method according to any one of claims 1 to 3, wherein the PARP inhibitor is olaparib, lucaparib, niraparib, beriparib or tarazoparib.
5. The method according to any one of claims 1 to 4, wherein the sequence analysis comprises a sequence determination procedure and a sequence data analysis procedure.
6. The method according to any one of claims 1 to 5, wherein somatic gene mutations are identified by the sequence analysis according to the 96 types of substitution classification shown in Fig. 1 and a mutation catalog is prepared.
7. The method according to any one of claims 1 to 6, wherein the sequence analysis is performed by whole exon sequence analysis.
8. The method according to any one of claims 1 to 6, wherein the sequence analysis is performed by an oncogene panel test.
9. The following steps (1) to (3):
   (1) Preparing an Age signature and a BRCA signature,
   9. The method of claim 1, further comprising (2) assessing the contribution of the Age and BRCA signatures to the mutation catalog, and (3) assessing the superiority of the contribution of the BRCA signature to the contribution of the Age signature. The method according to any one of claims.
10. In the step (1), a mutation signature analysis is performed on a plurality of cancer patients suffering from cancers that predict susceptibility, and the Age signature and the BRCA signature are determined by specifying the mutation rate of the substitution mutation of the Age signature and/or the BRCA signature. The method according to claim 9, which is prepared.
11. The method according to claim 9, wherein in the step (2), the contribution is evaluated by calculating the Age signature score and the BRCA signature score for the mutation catalog.
12. The Age signature score is represented by the following formula (1):
    

    

    (In the above formula, n=96, i means the i-th substitution mutation in the 96 types of substitution classification shown in FIG. 1, and Mi is the mutation rate of the i-th substitution mutation in the mutation catalog. Or the number of mutations, and KAi is the mutation rate of the i-th substitution mutation in the Age signature.)
    The method according to claim 11, calculated by:
13. The BRCA signature score is represented by the following formula (2):
    

    

    (In the above formula, n=96, i means the i-th substitution mutation in the 96 types of substitution classification shown in FIG. 1, and Mi is the mutation rate of the i-th substitution mutation in the mutation catalog. Or the mutation number, KBi is the mutation rate of the i-th substitution mutation of the BRCA signature.)
    The method according to claim 11, calculated by:
14. The superiority of the contribution of the BRCA signature is evaluated by calculating the ratio A (%) of the BRCA signature score to the sum of the Age signature score and the BRCA signature score in the step (3). 14. The method according to any one of 13 to 13.
15. In the step (3), the substitution of C to T with respect to the sum of the substitution mutation score (BRCA signature score) other than the substitution of C to T and the substitution mutation score of C to T (Age signature score) The method according to any one of claims 9 and 11 to 13, wherein the superiority of the contribution of the BRCA signature is evaluated by calculating the ratio B (%) of the scores of substitution mutations other than.
16. The method according to claim 14, wherein it is determined that the condition (a1) is satisfied when the ratio A (%) exceeds 50%.
17. The method according to claim 15, wherein it is determined that the condition (a1) is satisfied when the ratio B (%) exceeds 50%.
18. The following steps (4) to (6):
    (4) Providing a mutant signature including at least an Age signature and a BRCA signature,
    9. The method according to claim 1, further comprising (5) evaluating the contribution of each of the mutation signatures to the mutation catalog, and (6) evaluating the superiority of the contribution of the BRCA signature to the contribution of the Age signature. The method according to any one of 1.
19. 19. The method of claim 18, further comprising (7) assessing the contribution of a BRCA signature to the overall mutation signature.
20. The method according to claim 18, wherein the mutation signature includes all or a part of 30 kinds of mutation signatures (COSMIC version 2) in the COSMIC database.
21. The method according to claim 18, wherein in the step (5), the contribution degree of the mutation signature to the mutation catalog is evaluated by the destructStructs analysis.
22. In the step (6), the contribution ratio of the BRCA signature is evaluated by calculating the ratio C (%) of the contribution ratio of the BRCA signature to the sum of the contribution ratio of the Age signature and the contribution ratio of the BRCA signature. 19. The method of claim 18, wherein
23. In the step (6), the contribution ratio of the BRCA signature is evaluated by calculating the ratio C (%) of the contribution ratio of the BRCA signature to the sum of the contribution ratio of the Age signature and the contribution ratio of the BRCA signature. And
    The contribution of the BRCA signature is evaluated by calculating the ratio D(%) of the contribution of the BRCA signature to the sum (1) of the contributions of the mutant signatures in the step (7). the method of.
24. The method according to claim 22, wherein when the ratio C (%) exceeds 50%, it is determined that the condition (a1) is satisfied.
25. The method according to claim 23, wherein when the ratio C (%) exceeds 50% and the ratio D (%) exceeds k% (k=15 to 40), it is determined that (a1) is satisfied. ..
26. The method according to any one of claims 1 to 25, which comprises determining that the cancer cell contains a cancer cell population susceptible to a PARP inhibitor when both (a1) and (a2) of the condition (A) are satisfied. The method according to any one of 1.
27. The method according to claim 1, further comprising the step of performing an exhaustive methylation analysis on the test sample of interest, wherein (B) methylation of the promoter region of BRCA1 indicates the presence of a cancer sensitive to a PARP inhibitor. 27. The method according to any one of 26.
28. A method for detecting cancer having homologous recombinant repair deficiency (HRD), comprising:
    Comprising the step of performing a sequence analysis on a nucleic acid sample obtained from a cancer cell to create a mutation catalog, wherein (A)(a1) the contribution of the BRCA signature to the mutation catalog exceeds the contribution of the Age signature to the mutation catalog. And (a2) the absence of cyclin E1 amplification indicates the presence of a cancer having a homologous recombination repair deficiency.
29. 29. The method of claim 28, further comprising the step of performing an exhaustive methylation analysis on the test sample of interest, wherein (B) methylation of the promoter region of BRCA1 indicates the presence of a cancer with a homologous recombination repair deficiency. Detection method.
30. A method for treating a cancer patient with a PARP inhibitor, wherein the method according to any one of claims 1 to 27 is performed to identify a cancer patient having a cancer sensitive to the PARP inhibitor, Alternatively, comprising performing the method of claim 28 or 29 to identify a cancer patient having a cancer with homologous recombination repair deficiency, and then administering to said patient an effective amount of a PARP inhibitor. Method of treatment.
31. A therapeutic agent for cancer comprising a PARP inhibitor as an active ingredient, wherein the treatment is carried out by the method according to claim 30.
    
    

Images