WO2020085937A1

WO2020085937A1 - Test system for predicting effectiveness of cancer patient treatment using the preparation bevacizumab (avastin)

Info

Publication number: WO2020085937A1
Application number: PCT/RU2018/000704
Authority: WO
Inventors: Антон Александрович БУЗДИН; Максим Игоревич СОРОКИН; Андрей Владимирович ГАРАЖА; Даниил Михайлович НИКИТИН
Original assignee: Общество С Ограниченной Ответственностью "Онкобокс"
Priority date: 2018-10-24
Filing date: 2018-10-24
Publication date: 2020-04-30

Abstract

The proposed technical solution relates to the field of personalized medicine for oncological diseases, and specifically to test systems used to evaluate the efficacy of therapy using a drug having bevacizumab as an active substance. The solution utilizes the construction of a classifier that provides a prognosis of the clinical efficacy of the use of bevacizumab for individual cancer patients, dividing said patients into responsive and unresponsive groups. A mathematical formula for calculating a gene sum is the basis for all calculations being executed to help divide the patients into the groups of those who are responsive to and those who are not responsive to the drug therapy being conducted with bevacizumab. The present invention makes it possible to increase the degree of reliability with which the clinical efficacy of using bevacizumab to treat oncological diseases in patients is evaluated.

Description

Test system for predicting treatment outcome

cancer patients with bevacizumab (avastin)

Technical field

The proposed technical solution relates to test systems used in personalized molecular and postgenomic medicine to diagnose a particular disease, or to evaluate the effectiveness of a particular therapy for this disease.

State of the art

There are a limited number of platforms that use particular types of large-scale genetic profiling to advise doctors and patients. An example is the Caris Molecular Intelligence system (Russell et al. 2014; Green et al. 2014; Popovtzer et al. 2015; Vigneswaran et al. 2016). The use of the platform is based on the analysis of a limited spectrum of mutations with previously shown clinical significance, as well as on the profiling of patient biosamples on the immunohistochemical panel for determining protein tumor markers. However, this system does not use the results of studies of large-scale data on gene expression. Regarding the prediction of the activity of drugs that block the activity of endothelial growth factor VEGF-bevacizumab and other drugs, the following inventions can be cited:

1) US2011 / 0117083 A1, publ. May 19, 2011. “Biological markers for determining the response of patients to VEGF antagonists.” This patent measured gene expression before and after treatment with bevacizumab on various tumors of mice and human tumors implanted in mice. In various cases, 204 and 260 genes were identified that significantly reduced their expression level after treatment with bevacizumab. The main drawback of this work is that the obtained gene sets are in no way related to the degree of response to treatment, which was not measured in the course of patent studies. In addition, the murine model system limits the applicability of the proposed set of genes in humans.

2) US8163499 B2, publ. April 24, 2012. “Definition of a medication for treating breast cancer using microarrays with antibodies.” In this patent, the optimal drug was selected based on the levels of activation of various signaling pathways in a cancer tumor sample based on immunomicrometric data. The test system suggests treating the patient with that drug, the target signaling pathway of which is activated to the greatest degree, and bevacizumab is among the proposed drugs. The disadvantage of this approach is that the system does not take into account the known differences in the levels of activation of certain biological markers (genes or pathways) in responders and non-responders to this therapy. In other words, this patent proposes a simple the logic is “one way, one cure”, while changes in the gene networks of a cancerous tumor during therapy are often significantly more complex (Xu and Wang, 2015).

3) The Angiopredict project (htps: //www.angiopredict.com/) aims to comprehensively predict the response of patients with metastatic colon and rectal cancer to bevacizumab. The invention uses some biological markers at the level of the genome structure (10 single nucleotide polymorphisms and 10 regions with variations in the number of copies (htps: //cordis.europa.eu/result/rcn/186952 en.html). Published in WO2017182656A1, 2017-10- 26, “Methods and Therapies for VEGF”) predicting patients' response to bevacizumab.

The latest development is closest in technical essence to the present invention. However, studies of gene expression in this case did not lead to the creation of a predictive gene signature that allows the separation of responders and non-responders. In addition, the latter invention does not provide for predicting the response to bevacizumab complex therapy, while the invention published in this application provides for predicting the response to bevacizumab according to gene expression, including in the regimen of complex therapy. In addition, the test system proposed in the present invention is applicable to colorectal cancer and glioblastoma, that is, to different types of cancer.

Currently, the expected clinical result for patients with cancer is based on subjective definitions of the clinical and pathological characteristics of the tumor. For example, doctors decide on appropriate surgical procedures and adjuvant therapy based on the stage, class, and degree of necrosis of the organ affected by the tumor. Standard clinical criteria alone have a limited ability to accurately assess a patient's prognosis. To date, in clinical practice, there are no effective methods for predicting the effectiveness of existing antitumor drugs for a particular patient, which would take into account the peculiarities of the molecular imbalance formed during the development of a particular tumor. The development of a personalized approach to the treatment of cancer based on molecular changes in the patient’s body is an urgent task, and this invention is intended to expand the range of approaches used to solve this problem.

SUMMARY OF THE INVENTION

The technical task of the present invention was to develop new approaches to the treatment of cancer using a drug based on bevacizumab, based on profiling gene expression in the tissue of a tumor biopsy. This problem is solved using a test system that allows you to predict the effectiveness of the response to a drug containing bevacizumab as an active substance (for example, avastin), for patients with different types of cancer according to data obtained on different gene expression analysis platforms. This test system is a method of analyzing an expression signature consisting of 157 marker and 6 control human genes. In some embodiments of the invention, the test system allows predicting the response of patients with glioblastoma and colon and rectal cancer to both monotherapy with bevacizumab and complex therapy with bevacizumab and non-targeted drugs (the FOLFOX / FOLFIRI complex from calcium oxalinate, fluorouracil and oxaliplatin or irinotecan). In other embodiments of the invention, the test system successfully (AUO0.8) predicts the response of patients with colon and rectal cancer to monotherapy and complex therapy with bevacizumab according to data obtained on microarrays and using new generation sequencing platforms. In some other embodiments of the invention, the test system reliably predicts (AUC = 0.7) the response of patients with glioblastoma to complex therapy with bevacizumab in combination with radiation therapy and treatment with temozolomide according to gene expression data obtained using new generation sequencing.

This problem is solved by creating a method for determining the clinical effectiveness of the use of bevacizumab as part of a certain type of therapy for the treatment of cancer in a patient, which includes the following stages: (a) the expression levels of the following genes are measured in a patient’s tumor tissue sample: SLC46A2, PPP1R13L, IFT27, COLCA1, HUS1B , APOBEC3D, TCAF2, SEPT10, BCL2L11, MORN5, RNF144B, ELK3, TMS04, HRH2, MXI, SELENON, CDH24, DEFB119, SLAMF9, ADAM15, CBX4, KBTBD2, EMPZ, SLC5A3, TX1, TX1, TX1, TX1, TX1, TXT, TXT , ELFN2, SZRD1, BCKDK, OSBPL7, COL6A1, CCDC177, PIP4K2B, MTMR3, PDLIM7, OGDH, TGFB1, FM02, CYB5R3, DOCK3, EN01, INPP5A, ABHD11, RASSF5, MFH313, 3AP1, RRAS, wherein expression levels are measured using one of the following experimental platforms: a device for sequencing total RNA, a microchip, or a device for reverse transcription and polymerase chain reaction; (b) the expression levels of the following genes are measured in the same patient’s tumor tissue using the same experimental platform: HARS, SHROOM2, RCN1, RPL26L1, MRPL22, HSPA9, PH AX, UBE2D2, RAD50, LCORL, PRELID1, KDM3B, SLC22A15, KIAA02 , VAC, ZMPSTE24, KIF3A, SLC25A46, HYPK, SAP30L, UQCRQ, METTL1, NOP58, ERAP1, ZCCHC10, SWSAP1, RPS24, UQCRH, FAM53C, CPSF6, HSP90B1, WDR36, HMX2, LSNP1, HNP1, DRP1, HNP1, HNP1, HNP1, HNP1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1 , NUCB2, NLN, STIL, NDUFS5, HNRNPA1P10, BOLA3, KIF21A, MCPP2, RARS, AACS, RTN4IP1, SFXN1, EPM2AIP1, OSM, ZMAT2, PCBD2, NLK, NOP14, PPPX, TNPC, TNPC2, PPCX, TNC, PPC2, PPC2, PPC2, PPC2 , TOE1, MTKhZ, ARL15, NDUFS6, PSMB10, DDX55, BBS10, MTERF2, DHX37, KRR1, MAGOH, FASTKD3, CYP4F2, CCNE1, NSUN2, CNPY2, CCDC127, UNC5CL,

HNRNPA1, SNX7, RAPGEF6, LYSMD4, SHH, JADE1, LYAR, BRWD3, TRUB2, GTPBP8, KDM4A, C12or † 65, PPAN, OTULIN, SLC12A2, MRPS30, HMGCR, MPV17L, HINT 1, SEC24A, SRPRB; (c) the expression levels of the following normalization genes are measured in the same patient’s tumor tissue sample using the same experimental platform: VCP, DIABLO, PSMB2, POLR2C, GAPDH, ASTV; (d) calculate the value of the gene sum, which is a combination of the expression levels of the genes measured previously in (a), (b) and (c) so that the expression levels of the genes measured in (a) are involved in the calculation of the gene sum with a positive sign in the logarithm of the absolute expression level, the expression levels of the genes measured in (b) are involved in the calculation of the negative sum of the gene sum in the logarithm of the absolute expression level, and the expression levels of the genes measured in (c) are used to normalize the expression levels of the genes measured in (a) and in ( b) before using the values of the gene sum in the calculation; (e) define bevacizumab as a clinically effective agent in a certain type of therapy for treating an oncological disease in a patient when the gene sum value exceeds a threshold value predetermined for the experimental platform used, the oncological disease type and the type of therapy used.

This problem is also solved by creating a method for treating an oncological disease in a patient, comprising administering bevacizumab as part of a certain type of therapy to this patient in the case when this patient has been defined as responding to bevacizumab using a method that includes the following steps: (a) measure expression levels of the following genes in the patient’s tumor tissue sample: SLC46A2, PPP1R13L, IFT27, COLCA1, HUS1B, APOBEC3D, TCAF2, SEPT10, BCL2L11, MORN5, RNF144B, ELK3, TMC04, HRH2, MXI, SELENAM9, SELENAMB, CD119, SELENF9 CBX4, KBTBD2, EMRZ, SLC5A3, FYTTD1, TBX19, SIAH2, RNF26, ECEL1, N OXC9, ELFN2, SZRD1, BCKDK, OSBPL7, COL6A1, CCDC177, PIP4K2B, MTMR3, PDLIM7, OGDH, TGFB1, FM02, CYB5R3, DOCK3, EN01, INPP5A, ABHD11, RASSN3, MAPP5A, ABHD11, RASSF3, CD3FAP, PAP5A, ABHD11, RASSF5 wherein the measurement of expression levels is carried out using one of the following experimental platforms: a device for sequencing total RNA, a microchip or a device for reverse transcription and polymerase chain reaction; (b) the expression levels of the following genes are measured in the same patient’s tumor tissue using the same experimental platform: HARS, SHROOM2, RCN1, RPL26L1, MRPL22, HSPA9, PH AX, UBE2D2, RAD50, LCORL, PRELID1, KDM3B, SLC22A15, KIAA02 , VAC, ZMPSTE24, KIF3A, SLC25A46, HYPK, SAP30L, UQCRQ, METTL1, NOP58, ERAP1, ZCCHC10, SWSAP1, RPS24, UQCRH, FAM53C, CPSF6, HSP90B1, WDR36, HMX2, LSNP1, HNP1, DRP1, HNP1, HNP1, HNP1, HNP1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1, N1 , NUCB2, NLN, STIL, NDUFS5, HNRNPA1P10, BOLA3, KIF21A, MCPP2, RARS, AACS, RTN4IP1, SFXN1, EPM2AIP1, OSM, ZMAT2, PCBD2, NLK, NOP14, PPP2, SNPP, TNP2, SNP, TNP2, SNP

EKHOSZ, FANCB, SOKH5A, TOE1, MTKhZ, ARL15, NDUFS6, PSMB10, DDX55, BBS10, MTERF2, DHX37, KRR1, MAG OH, FASTKD3, CYP4F2, CCNE1, NSUN2, CNPY2, CCDC127, UNC5CL,

HNRNPA1, SNX7, RAPGEF6, LYSMD4, SHH, JADE1, LYAR, BRWD3, TRUB2, GTPBP8, KDM4A, C12orf65, PPAN, OTULIN, SLC12A2, MRPS30, HMGCR, MPV17L, HINT 1, SB24A; (c) the expression levels of the following normalization genes are measured in the same patient’s tumor tissue sample using the same experimental platform: VCP, DIABLO, PSMB2, POLR2C, GAPDH, ASTV; (d) calculate the value of the gene sum, which is a combination of the expression levels of the genes measured previously in (a), (b) and (c) so that the expression levels of the genes measured in (a) are involved in the calculation of the gene sum with a positive sign in the logarithm of the absolute expression level, the expression levels of the genes measured in (b) are involved in the calculation of the negative sum of the gene sum in the logarithm of the absolute expression level, and the expression levels of the genes measured in (c) are used to normalize the expression levels of the genes measured in (a) and in ( b) before using the values of the gene sum in the calculation; (e) define the patient as responding to bevacizumab when the value of the gene sum exceeds a threshold value predetermined for the experimental platform used, the type of cancer and the type of therapy used.

In some embodiments of the invention, these methods are characterized in that the threshold value is determined by calculating the share of false positive results and the share of false negative results of determining the effectiveness of bevacizumab in patients with a known response status to bevacizumab and minimizing the value of the following amount: 3 * the share of false positive results + the share of false negative results.

The threshold value determined in the methods of the present invention, using a group of patients with a known response status for bevacizumab therapy, depends on the type of cancer, the type of tumor tissue, the experimental platform and the type of therapy (therapeutic regimen).

In some embodiments of the invention, these methods are characterized in that the gene expression levels are measured using the Affymetrix Human Genome U133 Plus 2.0 Array microchip platform, the type of cancer is colon and rectal cancer, the therapy used is bevacizumab monotherapy and the threshold value is -10, 92. In some embodiments of the invention, these methods are characterized in that the gene expression levels are measured using the Affymetrix Human Genome U 133 Plus 2.0 Array microchip platform, the type of cancer is colon and rectal cancer, the therapy used is combination therapy with bevacizumab (bevacizumab with the FOLFOX complex / FOLFIRI from calcium oxalinate, fluorouracil and oxaliplatin or irinotecan) and the threshold value is -17.16. In some embodiments of the invention, these methods are characterized in that gene expression levels are measured using the lllumina HumanRef-8 WG-DASL v3.0 microchip platform, the type of cancer is colon and rectal cancer, the therapy used is bevacizumab monotherapy and the threshold value is -17.46. In some embodiments of the invention, these methods are characterized in that gene expression levels are measured using a new generation sequencing platform lllumina HiSeq 2500, the type of cancer is glioblastoma, the therapy used is combination therapy with bevacizumab (bevacizumab with radiation and temozolomide) and the threshold value is - 104.00.

Studies providing the initial data for this method can be multi-mix genetic profiling: high-performance profiling of gene expression at the mRNA level using hybridization on microarrays, high-performance sequencing of a new generation (using methods known in the art for sequencing a complete transcript or genome-wide total RNA sequencing), RT-PCR (reverse transcription polymerase chain reaction) in real time. The present invention improves the reliability of determining the clinical efficacy of bevacizumab for the treatment of solid tumors in patients.

Brief Description of Drawings

FIG. 1- Gene sums for patients with colon and rectal cancer, bevacizumab monotherapy, Asymetrix Human Genome U133 Plus 2.0 Array platform.

FIG. 2- Receiver performance curve for gene sums of patients with colon and rectal cancer, bevacizumab monotherapy, Affymetrix Human Genome U133 Plus 2.0 Array platform. The value of the AUC classifier is 0.816. The p value for the Student t-test is 0.064.

FIG. 3- The values of the gene sum for patients with colon and rectal cancer, complex therapy with bevacizumab (with the combination of FOLFOX / FOLFIRI), Affymetrix Human Genome U133 Plus 2.0 Array platform.

FIG. 4- The values of the gene sum for patients with colon and rectal cancer, monotherapy with bevacizumab, platform lllumina HumanRef-8 WG-DASL v3.0.

FIG. 5- Receiver performance curve for gene sums of patients with colon and rectal cancer, bevacizumab monotherapy, platform lllumina HumanRef-8 WG-DASL v3.0. The AUC of the classifier is 0.893. The p value for the Student t-test is 0.009.

FIG. 6- The values of the gene sum for patients with glioblastoma, combination therapy with bevacizumab (in combination with temozolomide and radiation therapy), lllumina HiSeq 2500 platform.

FIG. 7- Receiver performance curve for gene sums of patients with glioblastoma, combination therapy with bevacizumab (in combination with temozolomide and radiation therapy), lllumina HiSeq 2500 platform. Classifier AUC value is 0.692. The p-value for the Student t-test is 0.391.

Detailed Disclosure of Invention

In the description of the present invention, the terms “includes” and “including” are interpreted as meaning “includes, inter alia,”. These terms are not intended to be construed as “consists of only”. Unless defined separately, technical and scientific terms in this application have standard meanings generally accepted in the scientific and technical literature.

The use of bevacizumab or the introduction of bevacizumab as part of a certain type of therapy in this invention means the use or administration of a drug containing bevacizumab as an active ingredient, as well as additionally containing excipients (inactive substances), for example, salts, stabilizers, acidity regulators, etc. Examples such drugs are, for example, Avastin and Avegra drugs authorized in the Russian Federation.

The proposed test system is a set of genes that together predict the effectiveness of the response of the cancer patient to the drug bevacizumab (for example, avastin), including in combination with other types of treatment. To use the test system, a set of data on gene expression obtained for a sample of a cancer tumor on a new generation microchip or sequencing experimental platform is required. For each gene, a normalized expression level (signal) is required minus the background in the case of microchipping and a normalized number of reads mapped onto a given gene for sequencing of a new generation. For normalization, a set of genes with a stable expression level (hereinafter, household genes) is used, as described below.

In particular, but not limited to the above examples, the present invention allows to predict the effectiveness of bevacizumab therapy according to gene expression obtained on the following experimental platforms for the following types of therapy:

1) Colorectal cancer, bevacizumab monotherapy, Affymetrix Human Genome U133 Plus 2.0 Array microchip platform.

2) Colorectal cancer, bevacizumab monotherapy, lllumina HumanRef-8 WG-DASL v3.0 microchip platform.

3) Colorectal cancer, combination therapy with bevacizumab (with the FOLFOX / FOLFIRI complex of calcium oxalinate, fluorouracil and oxaliplatin or irinotecan), Affymetrix Human Genome U133 Plus 2.0 Array microchip platform.

4) Glioblastoma, combination therapy with bevacizumab (with radiation and temozolomide), a new generation sequencing platform lllumina HiSeq 2500.

The invention works with data obtained on the above and other platforms as follows.

1) A sample of the tumor tissue of the patient, for which it is required to predict the effectiveness of bevacizumab therapy, is processed in accordance with the manufacturer's instructions of the used experimental platform for determining gene expression. In particular, processing may include the following steps: a) fixation of the sample by freezing or using formalin with further fixation of the sample by pouring it into a paraffin block, b) isolation of total RNA or mRNA fraction with an accessible set of reagents, c) construction of a DNA library for gene analysis expression on a selected experimental platform by converting RNA into complementary DNA followed by PCR amplification.

2) The resulting DNA library is subjected to experimental analysis using this platform, as a result, an array of raw gene expression data is generated, according to the recommendations of the manufacturer of the experimental platform.

3) Raw data of gene expression are filtered from noise and low quality signals by programs provided for this by manufacturers of the corresponding experimental platform used to determine gene expression. This stage can be combined with the next one within the framework of some software packages (for example, the Affy microchip data analysis software package, which combines normalization and background subtraction).

4) Where possible, the purified gene expression data is normalized using the programs provided for the data of a particular platform. For each gene, a normalized signal is required (the number of reads of the RNA sequencing library mapped to the gene, or the normalized signal minus the background in case of microchipping). After normalization, the data of gene expression should be absolute, non-logarithmic values proportional to the concentration of transcripts (mRNA) for each compared gene in the analyzed biological samples. The value obtained at the output after normalization should be proportional to the absolute value of gene expression (i.e., the concentration of mRNA copies of this gene in the test tissue). If the software package used to process the data of a specific platform produces logarithmic data, it is necessary to potentiate the obtained expression values (for example, the DESeq2 software package for normalizing a new generation of sequencing data generates absolute values of expression levels, whereas the Affy microarray data analysis software package for normalizing and subtracting the background by the RMA method gives the logarithm of the absolute values of expression on the basis of 2).

5) Normalized and noise-free gene expression data for each marker gene are additionally normalized to the geometric mean expression levels of control genes - genes with a stable expression level (housekeeping genes).

In particular, within the framework of this invention, the following housekeeping control genes can be used for normalization, gene names are given according to the HUGO Gene Nomenclature Committee (HGNC - Genome Nomenclature Committee of the European Bioinformatics Institute, https://www.qenenames.org):

VCP, DIABLO, PSMB2, POLR2C, GAPDH, ACT.

Hereinafter in the description of the present invention, the indicated gene names mean both the gene itself, the known sequence of which is given in the National Center for Biotechnological Information (NCBI, (https://www.ncbi.nlm.nih.gov/) under the corresponding number, and to allelic variants of this gene (isoforms) present in the patient’s genomes, for example, the “VCP” or “human VCP” gene refers to the gene encoding valosin containing protein, the sequence of which is given in the National Center for Biotechnological Information (NCBI) under the number NM_007126 (https: // www .ncbi.nlm.nih.gov / nuccore / NM 007126).

The sequences of these control genes are located in the National Center for Biotechnological Information (NCBI) under the following numbers:

6) After a double normalization, a value is calculated that predicts the patient's response to therapy (hereinafter referred to as the sum total). The gene sum is calculated as the sum of the logarithms of the absolute expression levels according to claim 5), where each logarithmic member is assigned a sign depending on whether the expression level of the gene is increased or decreased in the calibration data sets for cancer tissues of patients responding to bevacizumab therapy compared to patients with non-responders to bevacizumab therapy. The following is a set of genes used to calculate the gene amount.

Genes whose expression level is increased in respondents compared to non-responders; participate in the calculation of the gene sum with a positive sign in the logarithm of the absolute level of expression, gene names are given according to the nomenclature of the HUGO Gene Nomenclature Committee (HGNC);

SLC46A2, PPP1R13L, IFT27, COLCA1, HUS1B, APOBEC3D, TCAF2, SEPT10, BCL2L11, MORN5, RNF144B, ELK3, TMC04, HRH2, MHI, SELENON, CDH24, DEFB119, SLAMFB2, EM15BD5, MB, TXAM, CBAM, DBM, TXAM TBX19, SIAH2, RNF26, ECEL1, HOXC9, ELFN2, SZRD1, BCKDK, OSBPL7, COL6A1, CCDC177, PIP4K2B, MTMR3, PDLIM7, OGDH, TGFB1, FM02, CYB5R3, DOCK3, INF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, ENDF3, END1, 02 PLEKHH3, CD6, PIK3AP1, RRAS. The sequences of these genes are located at the National Center for Biotechnological Information (NCBI) under the following numbers:

Genes whose expression level is reduced in responders compared to non-responders; participate in the calculation of the negative gene sum in the logarithm of the absolute level of expression, gene names are given according to the HUGO Gene Nomenclature Committee (HGNC) nomenclature; HARS, SHROOM2, RCN1, RPL26L1, MRPL22, HSPA9, PH AX, UBE2D2, RAD50, LCORL,

PRELID1 KDM3B SLC22A15 HNRNPA1P33, DNAJC21, UBLCP1, DROSHA, NUCB2, NLN, STIL, NDUFS5, HNRNPA1P10, BOLA3, KIF21A, MCTP2, RARS, AACS, RTN4IP1, SFXN1, EPM2AIP1, OSM, PPL, PPM, MPL, PCP, MPL, PPM, PPM, MPP PRDM4, EX0C3, FANCB, C0X5A, T0E1, MTX3, ARL15, NDUFS6, PSMB10, DDX55, BBS10, MTERF2, DHX37, KRR1, MAGOhl, FASTKD3, CYP4F2, CCNE1, NSUN2, CNPCL7, UNCN5, CNPY7, CNC, RNC, UNCN5, CN5, CN5, UNCN, CCD LYSMD4, SHH, JADE1, LYAR, BRWD3, TRUB2, GTPBP8, KDM4A, C12orf65, PPAN, OTULIN, SLC12A2, MRPS30, HMGCR, MPV17L, HINT1, SEC24A, SRPRB.

The sequences of these genes are located at the National Center for Biotechnological Information (NCBI) under the following numbers:

Thus, the calculation of the gene amount according to the present invention is carried out according to the formula (1):

/ Level of expression \ / Level of expression

log! gene i / SSU) - log l geneau / SSU

where the sum over i is the sum over the genes whose expression level is increased in respondents compared to non-responders; the sum for j is the sum for genes whose expression level is lowered in respondents compared to non-responders; SSU - The geometric mean of expression levels of housekeeping genes.

7) After calculating the gene amount, its obtained values are compared with the threshold value inherent in a specific experimental platform, a specific disease, and a specific type of therapy. If the patient’s gene amount exceeds a threshold, the invention predicts an effective patient response to therapy. If the gene sum is below a threshold, the invention predicts a lack of an effective response to therapy. Below examples of threshold values of the gene sum for some experimental platforms and treatment regimens including treatment with bevacizumab are given.

To successfully predict the response to various bevacizumab therapy options on various experimental platforms (for example, Affymetrix Human Genome U133 Plus 2.0 Array, lllumina HumanRef-8 WG-DASL v3.0 and lllumina HiSeq 2500), both a set of genes for calculating the gene sum and threshold value of the gene amount.

The development of methods for analyzing gene expression data leads to the emergence of new experimental platforms. The proposed test system is also applicable for working with data from existing experimental platforms not mentioned in the examples, as well as platforms under development and platforms that will be created in the future. To apply the invention to any of the experimental platforms for determining gene expression and determining the characteristic threshold value of the gene sum, the following sequence of actions is provided:

A) Based on the experimental data obtained on the analyzed platform, an array of patient data of a certain nosology (colon and rectal cancer, glioblastoma, etc.) with a known response status for a specific type of bevacizumab therapy is created.

B) The resulting array of gene expression data is processed in accordance with the recommendations of the manufacturer of the platform for cleaning raw data from noise and primary normalization, similar to paragraphs 3) -4) in the description above.

B) For each patient, the expression level of each gene is normalized to the geometric mean of the expression levels of the household genes, similar to paragraph 5) in the description above ..

D) For each patient, the amount of the gene sum is calculated.

E) For a set of values of the gene sum of patients with a known response status, the AUC value is calculated (for a graph of the receiver operating characteristics).

E) Upon receipt of AUC values exceeding 0.8, it is possible to use the proposed gene signature according to paragraph 6) in the description above for this platform with this type of therapy with bevacizumab in relation to the studied nosology.

G) Based on the calculated gene amounts for patients of this nosology with a known outcome of therapy with bevacizumab of this type, the threshold value of the gene amount is selected. The threshold value is selected in such a way that the frequency of falsely identified non-responders (for which this therapy was actually successful) was three times lower than the frequency of falsely identified responders (for whom disease progression occurred).

3) During the application of the invention, for a patient of this nosology who needs to predict the likelihood of the effectiveness of the studied type of therapy, gene expression data are analyzed and the gene amount is calculated in accordance with paragraph 6.

I) A comparison is made of the gene amount obtained for an individual patient with the threshold value obtained in step G). If the patient’s gene amount exceeds a threshold, the invention predicts an effective patient response to therapy. If the gene sum is below a threshold, the invention predicts a lack of an effective response to therapy.

Unlike the analogs given above, the present invention is based on significant differences in gene expression between responders and non-responders to various types of bevacizumab therapy. The response to bevacizumab can manifest itself to varying degrees. According to the international clinical standard RECIST (http://chemoth.com/recist), the response of a cancer patient to therapy is classified into four types:

1) Full answer - all tumors disappeared after treatment.

2) Partial response - the largest diameter tumor has decreased by more than 30%.

3) Stabilization - no significant changes in the number and size of tumors were observed.

4) Progression of the disease - an increase in the largest diameter tumor by more than 20%.

The proposed system considers a patient with a disease progression to be a non-responder, while all other types of response are considered a successful response to therapy.

The following examples of the method are given in order to disclose the characteristics of the present invention and should not be construed as in any way limiting the scope of the invention.

Examples of the invention for data obtained on some experimental platforms.

Example 1. Evaluation of the response to bevacizumab for data obtained using affymetrix Human Genome U133 Plus 2.0 Array microarrays

1) Processing of biological material and loading onto a microchip.

A tumor sample of a patient with rectal cancer, for which it is necessary to predict the effectiveness of bevacizumab therapy, was frozen in liquid nitrogen, then homogenized and used to isolate RNA using the RNeasy Mini Kit reagent kit (QIAGEN, Chatswort.CA) according to the manufacturer's instructions. The resulting RNA preparation was converted into a cDNA form, then a cDNA library for hybridization was loaded onto a microchip and hybridized in accordance with the manufacturer's instructions, then the measurement of gene expression signals was carried out.

2) Processing and normalization of raw gene expression data.

The raw data obtained after analysis of the sample on a microchip and issued by Affymetrix software in the format of CEL files was cleaned of noise and normalized using the RMA software package. In the event that several oligonucleotide probes on the microchip corresponded to one gene, the signal for this gene was calculated as the geometric mean of the signals on each probe. After normalization with the RMA packet, the data was potentiated on base 2 to obtain the absolute value of the signal. Further, the gene signals for each sample were normalized to the geometric mean of the signals of the housekeeping genes. After double normalization, the normalized expression logarithms were summarized - the gene sum was calculated, which was -5.6 for this sample. This value was compared with a threshold value to predict the patient's response to bevacizumab.

3) Comparison with threshold values.

The choice of a threshold value for comparison with the gene sum was made for a particular type of therapy. The proposed test system allows you to predict the answer, including to:

For) monotherapy with bevacizumab and

36) for combination therapy with bevacizumab with the FOLFOX / FOLFIRI complex of calcium oxalinate, fluorouracil and oxaliplatin or irinotecan.

For monotherapy with bevacizumab, on this experimental platform, the threshold value of the gene amount is -10.92. For combination therapy with bevacizumab, on this experimental platform, the threshold value of the gene amount is -17.16.

4) Prediction of the patient's response status to bevacizumab. After choosing a threshold value, the obtained gene sum for an individual sample (-5.6) was compared with threshold values. The gene sum turned out to be higher than both threshold values; therefore, an effective response to both of the above types of therapy containing bevacizumab was predicted for this patient.

Example 2. Evaluation of the response to bevacizumab for data obtained using microarrays lllumina HumanRef-8 WG-DASL v3.0

1) Processing of biological material and loading onto a microchip.

A tumor sample of a patient with colon cancer, for which it was necessary to predict the effectiveness of bevacizumab monotherapy, was fixed with formalin and paraffinized, then homogenized and RNA was isolated using the RNeasy FFPE Kit (Qiagen, Hilden, D) according to the manufacturer's instructions.

The resulting RNA preparation was converted into a cDNA form, then a cDNA library for hybridization was loaded onto a microchip and hybridized in accordance with the manufacturer's instructions, then the measurement of gene expression signals was carried out.

2) Data processing and calculation of the gene amount. The values of the signals obtained on the microchip were subjected to a quantile normalization procedure (A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bolstad BM, Irizarry RA, Astrand M, Speed TP. Bioinformatics. 2003 Jan 22; 19 (2 ): 185-93.), Then the gene signals for each sample were normalized to the geometric mean of the signals of the housekeeping genes. After double normalization, the gene sum was calculated for the logarithms of the normalized expression values. This value (-26.9) was compared with a threshold value to predict the patient's response to bevacizumab.

3) Prediction of the patient's response status to bevacizumab. After calculating the gene sum, the obtained gene sum was compared with a threshold value, which for the given experimental system and this type of therapy was -17.46. The gene sum turned out to be lower than the threshold value; therefore, for this patient, a lack of response to therapy was predicted.

Example 3. Evaluation of the response to bevacizumab for data obtained using the new generation sequencing platform lllumina HiSeq 2500.

1) Processing of biological material.

A sample of a patient’s glioblastoma tumor, for which it was necessary to predict the effectiveness of combination therapy with bevacizumab (bevacizumab in combination with temozolomide and radiation therapy), was fixed with formalin and paraffinized, then homogenized and RNA was isolated using the RNeasy FFPE Kit (Qiagen, Hilden) and Deparaffinization Solution (Qiagen ,, Hilden, D) according to manufacturers recommendations. The library was prepared for sequencing using the Ovation Human FFPE RNA-seq Library System, manufactured by Nugen. The amount of RNA to create the library was determined according to the instructions of the manufacturer of the Ovation Human FFPE RNA-seq Library System kit.

2) Processing raw gene expression data and calculating the gene amount

The readings in raw fastq files were mapped onto the human genome using the STAR charting program, using the version of the genome for GRCh38 mapping. In other applications of the present invention, it is also possible to map to other versions of the assembly of the human genome. At the same time, the total number of uniquely mapped readings should be at least 1.2 million. Further, the obtained data on the number of readings mapped to each gene was normalized using the R package DESeq2 programs and normalized readings per gene were used to calculate the gene amount, which was -63.5. This value was compared with a threshold value to predict the patient's response to bevacizumab.

3) Prediction of the patient's response status to bevacizumab. After calculating the gene amount for the patient’s sample, the result was compared with a threshold value, which for the type of disease, the therapy used and the experimental platform is -104.00. Since the gene amount for a particular biosample was above the threshold value, this patient was predicted as a responder.

Example 4. Evaluation of the response to bevacizumab for data obtained using a different gene expression profiling platform.

The development of experimental methods leads to the emergence of new platforms for the analysis of gene expression. The proposed test system is also applicable to work with data on gene expression obtained both on platforms under development and on platforms not yet developed. At the same time, the following requirements are imposed on the experimental platform for the applicability of the present invention:

1) The platform must provide for each measured gene expression signals proportional to the concentration of mRNA of the gene or its protein product in the analyzed tissue.

2) Among the genes whose expression is measured using this platform, all genes for which the gene sum is calculated in the framework of the present invention should be present.

3) Among the genes whose expression is measured in this platform, all the genes of the household must be present, according to which the data are normalized in the framework of the present invention.

To work with the new platform, the following sequence of actions is provided:

A) Based on the available experimental data obtained on this platform, an array of gene expression data of patients of a certain nosology (for example, colon and rectum cancer and glioblastoma) with a known status of response to bevacizumab therapy is created, so that the array contains at least 14 patients , of which at least 4 respondents and at least 4 non-responders. Response status is determined by RECIST clinical standards. For the successful operation of the invention, nonresponders should be considered patients with the progression of the disease, then how other types of response (disease stabilization, partial and complete response) are considered a successful response to therapy.

B) The resulting array of gene expression data is processed in accordance with the recommendations of the platform manufacturer (in the general case, it is cleaning the raw data from noise and initial normalization).

C) For each biosample, an additional normalization of the expression levels of each gene to the geometric mean expression levels of the control genes of the household is carried out.

D) For each biosample, the amount of the gene sum is calculated.

E) For the available data on the values of the gene sum for patients with a known response status, the AUC value is calculated (for the receiver performance curve).

E) Upon receipt of AUC values> 0.8, it is possible to use the proposed gene signature for a given platform with this type of bevacizumab therapy for a given nosology.

G) Based on the calculated gene amounts for patients of this nosology with a known outcome of bevacizumab therapy, the threshold value of the gene amount is selected. For a given threshold, all patients whose gene amounts are higher than this value are identified as responders, while all other patients are considered non-responders. The threshold value is selected so that the frequency of falsely identified non-responders is three times lower than the frequency of falsely identified responders.

3) For a patient of this nosology who needs to predict the effectiveness of bevacizumab therapy, the data are analyzed and the gene amount is calculated in accordance with paragraphs B) -G) ·

I) A comparison is made of the resulting gene amount with a threshold value. If the patient’s gene amount exceeds a threshold, a positive response status is predicted for the patient. If the gene sum is lower or equal to the threshold value, a negative response status is predicted for the patient.

Example 5. Work of the invention with data obtained in patients with a known status of a clinical response to bevacizumab therapy.

5.1. Affymetrix Human Genome U133 Plus 2.0 Array microchip, colorectal cancer, bevacizumab monotherapy.

MATERIALS AND METHODS

1) Patients and therapy. For this example, we used the data of patients with colon and rectal cancer, published in the GEO database (https: //www.ncbi. nlm.nih.gov/geo/query/acc.cgi?acc=GSE19862). Data taken from dataset with identification number GSE19862. Patients with metastatic or recurrent colon or rectal cancer underwent bevacizumab monotherapy, after which the response status was measured according to RECIST standards and confirmed using computed tomography or magnetic resonance imaging. The data array includes 7 responders and 7 non-responders.

2) Processing of biological material for the platform. Tumor samples taken before treatment were frozen in liquid nitrogen immediately after collection, then lysed and RNA preparations were isolated using the RNeasy Mini Kit reagent kit (Qiagen, Chatswort.CA) in accordance with the manufacturer's instructions. The resulting RNA libraries were processed in accordance with the instructions of the microchip platform manufacturer. To create biotinylated cRNA libraries, 8 μg of isolated RNA was used.

3) Processing of gene expression data and calculation of the gene amount. The raw data obtained after analyzing the sample on a microchip were cleaned of noise and normalized using the Affy package (programming language R, the method of cleaning and normalizing RMA). In the case when several oligonucleotide samples on a microchip corresponded to one gene, the signal for this gene was calculated as the geometric mean of the signals for each of the samples. In the case when one sample corresponded to several genes, this value was used to calculate the expression value of each gene. After normalization with the RMA packet, the data was potentiated on base 2 in order to obtain the absolute value of the signal. Further, the gene signals for each sample were normalized to the geometric mean of the signals of the housekeeping genes. After double normalization, the normalized logarithms of expression were summed (gene sum was calculated).

4) Prediction of the patient's response status to bevacizumab. After calculating the gene amount, the obtained gene amount was compared with a threshold value of -10.92. If the gene sum turned out to be higher than the threshold value, this patient was predicted as a responder, otherwise the patient was predicted as a non-responder.

RESULTS

1) Calculation of the gene amount. The calculated gene sums for patients, as well as the known and predicted response status for bevacizumab monotherapy, are shown in Table 1.

Table 1. The values of the gene amount and the status of the response to bevacizumab monotherapy (known and predicted) according to the Affymetrix Human Genome U133 Plus 2.0 Array platform.

The distribution of gene sums for patients indicating their known response status to bevacizumab is shown in Figure 1. 2) Evaluation of the success of the gene amount as a value predicting the response of patients to bevacizumab. For the calculated values of the gene sum and known statuses of the patients' response to bevacizumab, a receiver performance curve was constructed (Figure 2). The AUC of the gene sum as a classifier of the known response status is 0.816. The p value for the student t-test between the sample of responders and non-responders is 0.064.

3) Prediction of response status. After predicting the response status of patients, a system error table was constructed (Table 2).

Table 2. Error table for the data of patients undergoing bevacizumab monotherapy according to the Affymetrix Human Genome U133 Plus 2.0 Array platform. In the cells of the table, the number of patients in this group.

5.2. Microchip Affymetrix Human Genome U133 Plus 2.0 Array, colorectal cancer, combination therapy with bevacizumab.

MATERIALS AND METHODS

1) Patients and therapy. For this example, we used the data of patients with colorectal cancer published in the GEO database (https: //www.ncbi. Nlm.nih.gov/geo/query/acc.cgi?acc=GSE72970).flaHHbie taken from dataset with identification number GSE72970 (Del Rio et al. 2017). Patients with colon or rectal cancer underwent combination therapy with bevacizumab (with the FOLFOX / FOLFIRI complex of calcium oxalinate, fluorouracil and oxaliplatin or irinotecan), after which the response status was measured according to RECIST standards. The data array includes 12 responders and 4 non-responders.

2) Processing of biological material for the platform. Tumor samples taken before treatment were frozen in liquid nitrogen immediately after collection, then lysed and RNA preparations were isolated using the RNeasy Mini Kit reagent kit (QIAGEN, Chatswort, CA) in accordance with the instructions of the kit manufacturer. The resulting RNA libraries were processed in accordance with the instructions of the microchip manufacturer.

3) Processing of data received from the platform and calculation of the gene amount.

The raw data obtained after analysis of the sample on a microchip were cleaned from noise and normalized with the Affy package (R programming language, function for cleaning and normalizing RMA). In the case when several oligonucleotide samples on a microchip corresponded to one gene, the signal for this gene was calculated as the geometric mean of the signals on each of the samples. In the case when one sample corresponded to several genes, this value was used to calculate the expression value of each gene. After normalization with the RMA packet, the data was potentiated on base 2 in order to obtain the absolute value of the signal. Further, the gene signals for each sample were normalized to the geometric mean of the signals of the housekeeping genes. After double normalization, the normalized logarithms of expression were summed (gene sum was calculated).

4) Prediction of the patient's response status to bevacizumab. After calculating the gene amount, the obtained gene amount was compared with a threshold value of -17.16. If the gene amount turned out to be higher than the threshold value, this patient was predicted as a responder, otherwise the patient was predicted as a non-responder to bevacizumab therapy.

RESULTS

1) Calculation of the gene amount. The calculated gene sums for the patients, as well as the known and predicted response statuses for the combination therapy with bevacizumab, are shown in Table 3.

Table 3. Values of the gene amount and response status for combination therapy with bevacizumab (known and predicted) according to the Affymetrix Human Genome U133 Plus 2.0 Array platform.

The distribution of gene sums for patients indicating their known response status to bevacizumab is shown in Figure 3.

2) Evaluation of the success of the gene amount as a value predicting the response of patients to bevacizumab. Due to the fact that in this case the predicted response statuses completely coincide with the known ones, the AUC value of the gene sum as a classifier of the known response status is 1. The p value for the Student t-test between the sample of responders and non-responders is 1, 88 * 10 ^L -6.

3) Prediction of response status. After predicting the response status of patients, a system error table was constructed (Table 4).

Table 4. Error table for the data of patients undergoing bevacizumab combination therapy according to the Affymetrix Human Genome U133 Plus 2.0 Array platform. In the cells of the table, the number of patients in this group.

5.3. Microchip lllumina HumanRef-8 WG-DASL v3.0, colon cancer, monotherapy with bevacizumab.

MATERIALS AND METHODS

1) Patients and therapy. For this example, we used the data of colon cancer patients published in the GEO database (https: //www.ncbi. Nlm.nih.gov/geo/query/acc.cgi?acc=GSE53127). Data is taken from a dataset with identification number GSE53127 (Pentheroudakis et al., 2014). Patients with metastatic colon cancer underwent bevacizumab monotherapy, after which the response status was measured according to RECIST standards. The data array includes 14 responders and 4 non-responders.

2) Processing of biological material for the platform. Tumor samples taken prior to treatment were fixed with formalin and paraffinized, then homogenized and RNA was isolated using the RNeasy FFPE Kit (Qiagen, Hilden, D) according to the kit manufacturer's instructions. The resulting RNA libraries were processed in accordance with the instructions of the microchip manufacturer.

3) Processing of primary gene expression data. The values of the signals obtained on the microchip were quantile normalized, then the gene signals for each sample were normalized to the geometric mean of the signals of the housekeeping genes. After double normalization, the gene amount was calculated for each sample.

4) Prediction of the patient's response status to bevacizumab. After calculating the gene amount, it was compared with a threshold value of -17.46. If the gene sum turned out to be higher than the threshold value, this patient was predicted as a responder, otherwise the patient was predicted as a non-responder.

RESULTS

1) Calculation of the gene amount. The calculated gene sums for the patients, as well as the known and predicted response statuses for the combination therapy with bevacizumab, are shown in Table 5.

Table 5. Values of the gene amount and status of the response to bevacizumab monotherapy (known and predicted) according to the platform lllumina HumanRef-8 WG-DASL v3.0.

The distribution of gene sums for patients indicating their known response status to bevacizumab is shown in Figure 4.

2) Evaluation of the success of the gene amount as a value predicting the response of patients to bevacizumab. For the calculated values of the gene sum and known statuses of the patients' response to bevacizumab, a receiver performance curve was constructed (Figure 5). The AUC of the gene sum as a classifier of the known response status is 0.893. The p-value for the student t-test between the sample of responders and non-responders is 0.009.

3) Prediction of response status. After predicting the response status of patients, a system error table was constructed (Table 6).

Table 6. Error table for the data of patients undergoing combination therapy with bevacizumab according to the platform lllumina HumanRef-8 WG-DASL v3.0. In the cells of the table, the number of patients in this group.

5.4. New generation sequencing platform lllumina HiSeq 2500, glioblastoma, bevacizumab combination therapy.

MATERIALS AND METHODS

1) Patients and therapy. For this example, we used the data of glioblastoma patients published in the GEO database (https: //www.ncbi. Nlm.nih.gov/geo/query/acc.cgi?acc=GSE79671). Data is taken from a dataset with identification number GSE79671 (Urup et al., 2017). Patients underwent combination therapy with bevacizumab (in combination with temozolomide and radiation therapy), after which the response status was measured according to RECIST standards. The data array includes 7 responders and 13 non-responders.

2) Processing of biological material for platforms. Tumor samples taken before treatment were fixed with formalin and paraffinized, then homogenized and RNA preparations were isolated using the RNeasy FFPE Kit reagent kit (Qiagen, Hilden, D) and Deparaffinization Solution (Qiagen, Hilden, D) in accordance with the manufacturer's instructions. The libraries were prepared for sequencing using the Ovation Human FFPE RNA-seq chain-specific system Library System, manufactured by Nugen. To create the libraries, 250 ng of total RNA was used.

3) Processing data received from the platform. The readings in raw fastq files were mapped to the reference genome of the GRCh38 version by the STAR charting program. Further, the obtained data on the number of readings mapped to each gene was normalized using the DESeq2 R-package and the normalized readings on the gene were used to calculate the gene sum. Normalization to the geometric mean of the levels of gene expression of the household in this case was not carried out.

4) Prediction of the patient's response status to bevacizumab. After calculating the gene amount, it was compared with a threshold value of -104.00. If the gene sum turned out to be higher than the threshold value, this patient was predicted as a responder, otherwise the patient was predicted as a non-responder.

The calculation of the gene amount. The calculated gene sums for the patients, as well as the known and predicted response statuses for the combination therapy with bevacizumab, are shown in Table 7.

Table 7. Values of the gene amount and status of the response to bevacizumab monotherapy (known and predicted) according to the lllumina HiSeq 2500 platform.

The distribution of gene sums for patients indicating their known response status to bevacizumab is shown in Figure 6.

5) Assessment of the success of the gene amount as a value predicting the response of patients to bevacizumab. For the calculated values of the gene sum and known statuses of the patients' response to bevacizumab, a receiver performance curve was constructed (Figure 7). The AUC of the gene sum as a classifier of the known response status is 0.692. The p-value for the student t-test between the sample of responders and non-responders is 0.391.

6) Prediction of response status. After predicting the response status of patients, a system error table was constructed (Table 8).

Table 8. Error table for the data of patients undergoing combination therapy with bevacizumab according to the lllumina HiSeq 2500 platform data. In the table cells, the number of patients in this group.

Thus, the proposed test system is capable of working with different platforms for obtaining experimental gene expression data (for example, Affymetrix Human Genome U133 Plus 2.0 Array and lllumina HumanRef-8 WG-DASL v3.0 microchips and the new generation sequencing platform lllumina HiSeq 2500), with different types of cancer (colon and rectal cancer and glioblastoma) and with different types of therapy containing bevacizumab (monotherapy and two options for combination therapy). It is possible to use the system to work with new and experimental platforms. Although the invention has been described with reference to the disclosed embodiments, it should be apparent to those skilled in the art that the specific cases described in detail are for illustrative purposes only and should not be construed as limiting the scope of the invention in any way. It should be understood that various modifications are possible without departing from the gist of the present invention.

Claims

CLAIM

1. A method for determining the clinical efficacy of bevacizumab as part of a specific type of therapy for the treatment of cancer in a patient, comprising the following stages:

(a) measure the expression levels of the following genes in a patient’s tumor tissue sample: SLC46A2, PPP1R13L, IFT27, COLCA1, HUS1B, APOBEC3D, TCAF2, SEPT10, BCL2L11, MORN5, RNF144B, ELK3, TMC04, HRH2, MEFI, CD11, MEF9, SLAMF9, ADAM15, CBX4, KBTBD2, EMRZ, SLC5A3, FYTTD1, TBX19, SIAH2, RNF26, ECEL1, NOXC9, ELFN2, SZRD1, BCKDK, OSBPL7, COL6A1, CCDC177, PDL3, FM1, BM1 DOCK3, EN01, INPP5A, ABHD11, RASSF5, MAP3K13, MFN2, PLEKHH3, CD6, PIK3AP1, RRAS, and expression levels are measured using one of the following experimental platforms: a total RNA sequencing device, a microchip, or a reverse transcription and polymerase device chain reaction;

(b) expression levels of the following genes are measured in the same patient’s tumor tissue using the same experimental platform: HARS, SHROOM2, RCN1, RPL26L1, MRPL22, HSPA9, PHAX, UBE2D2, RAD50, LCORL, PRELID1, KDM3B, SLC22A15, KIAA02 VAC, ZMPSTE24, KIF3A, SLC25A46, HYPK, SAP30L, UQCRQ, METTL1, NOP58, ERAP1, ZCCHC10, SWSAP1, RPS24, UQCRH, FAM53C, CPSF6, HSP90B1, WDR36, HMX2, LSMP1, RN1, NR1, NR1, NR1, NR1, N1, N1, N1, N, N1, N1 NUCB2, NLN, STIL, NDUFS5, HNRNPA1P10, BOLA3, KIF21A, MCPP2, RARS, AACS, RTN4IP1, SFXN1, EPM2AIP1, OSM, ZMAT2, PCBD2, NLK, NOP14, PPPCOCH, TNPCO, CHP, CHP, CHP, CHP, CHP, CHM TOE1, MTKhZ, ARL15, NDUFS6, PSMB10, DDX55, BBS10, MTERF2, DHX37, KRR1, MAGOH, FASTKD3, CYP4F2, CCNE1, NSUN2, CNPY2, CCDC127, UNC5CL, HNRNPA1, SNX1, JDF1, SNX1, SNX1, JDF1, SNX1, SNX7, JX7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx1, Jx1, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Sem7 BRWD3, TRUB2, GTPBP8, KDM4A, C12or † 65, P PAN, OTULIN, SLC12A2, MRPS30, HMGCR, MPV17L, HINT1, SEC24A, SRPRB;

(c) the expression levels of the following normalization genes are measured in the same patient’s tumor tissue sample using the same experimental platform: VCP, DIABLO, PSMB2, POLR2C, GAPDH, ASTV;

(d) calculate the value of the gene sum, which is a combination of the expression levels of the genes measured previously in (a), (b) and (c) so that the expression levels of the genes measured in (a) are involved in the calculation of the gene sum with a positive sign in the logarithm of the absolute expression level, the expression levels of the genes measured in (b) are involved in the calculation of the negative gene sum in the logarithm of the absolute expression level, and the expression levels of genes measured in (c) are used to normalization of expression levels of genes measured in (a) and in (b) before using the calculated values of the gene sum;

(e) define bevacizumab as a clinically effective agent in a certain type of therapy for treating an oncological disease in a patient when the gene sum value exceeds a threshold value predetermined for the experimental platform used, the oncological disease type and the type of therapy used.

2. A method of treating an oncological disease in a patient, comprising administering bevacizumab as part of a certain type of therapy to this patient in the case when the patient has been determined to be responding to bevacizumab using a method that includes the following steps:

(b) expression levels of the following genes are measured in the same patient’s tumor tissue using the same experimental platform: HARS, SHROOM2, RCN1, RPL26L1, MRPL22, HSPA9, PHAX, UBE2D2, RAD50, LCORL, PRELID1, KDM3B, SLC22A15, KIAA02 VAC, ZMPSTE24, KIF3A, SLC25A46, HYPK, SAP30L, UQCRQ, METTL1, NOP58, ERAP1, ZCCHC10, SWSAP1, RPS24, UQCRH, FAM53C, CPSF6, HSP90B1, WDR36, HMX2, LSMP1, RN1, NR1, NR1, NR1, NR1, N1, N1, N1, N, N1, N1 NUCB2, NLN, STIL, NDUFS5, HNRNPA1P10, BOLA3, KIF21A, MCPP2, RARS, AACS, RTN4IP1, SFXN1, EPM2AIP1, OSM, ZMAT2, PCBD2, NLK, NOP14, PPPXCO, TNPX, CHP, CHP, CHP, CHP, CHP, CHP, CHM TOE1, MTKhZ, ARL15, NDUFS6, PSMB10, DDX55, BBS10, MTERF2, DHX37, KRR1, MAGOH, FASTKD3, CYP4F2, CCNE1, NSUN2, CNPY2, CCDC127, UNC5CL, HNRNPA1, SNX1, JDF1, SNX1, SNX1, JDF1, SNX1, SNX7, JX7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx1, Jx1, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Jx7, Sem7 BRWD3, TRUB2, GTPBP8, KDM4A, C12orf65, PPA N, OTULIN, SLC12A2, MRPS30, HMGCR, MPV17L, HINT1, SEC24A, SRPRB; (c) expression levels of the following normalization genes are measured in the same patient’s tumor tissue sample using the same experimental platform: VCP, DIABLO, PSMB2, POLR2C, GAPDH, ASTV

(d) calculate the value of the gene sum, which is a combination of the expression levels of the genes measured previously in (a), (b) and (c) so that the expression levels of the genes measured in (a) are involved in the calculation of the gene sum with a positive sign in the logarithm of the absolute expression level, the expression levels of the genes measured in (b) are involved in the calculation of the negative sum of the gene sum in the logarithm of the absolute expression level, and the expression levels of the genes measured in (c) are used to normalize the expression levels of the genes measured in (a) and in ( b) before using the values of the gene sum in the calculation;

(e) define the patient as responding to bevacizumab when the value of the gene sum exceeds a threshold value predetermined for the experimental platform used, the type of cancer and the type of therapy used.

3. The method according to claim 1, in which the threshold value is determined by calculating the share of false positive results and the share of false negative results of determining the effectiveness of bevacizumab in patients with a known response status to bevacizumab and minimizing the value of the following amount: 3 ^{* the} share of false positive results + the share of false negative results.

4. The method according to claim 2, in which the threshold value is determined by calculating the share of false positive results and the share of false negative results of determining the effectiveness of bevacizumab in patients with a known response status to bevacizumab and minimizing the value of the following sum: 3 ^{* the} share of false positive results + the share of false negative results.

5. The method according to p. 3, in which the measurement of gene expression levels is performed using the Affymetrix Human Genome U133 Plus 2.0 Array microchip platform, the type of cancer is colon and rectal cancer, the therapy used is bevacizumab monotherapy and the threshold value is -10, 92.

6. The method according to p. 3, in which the measurement of gene expression levels is performed using the Affymetrix Human Genome U133 Plus 2.0 Array microchip platform, the type of cancer is colon and rectal cancer, the treatment used is the combination therapy with bevacizumab (bevacizumab with the FOLFOX / complex FOLFIRI from calcium oxalinate, fluorouracil and oxaliplatin or irinotecan) and the threshold value is -17.16.

7. The method according to claim 3, in which the measurement of gene expression levels is carried out using the lllumina HumanRef-8 WG-DASL v3.0 microchip platform, the type of cancer is colon cancer, the therapy used is bevacizumab monotherapy and the threshold value is -17 , 46.

8. The method according to claim 3, in which the measurement of gene expression levels is carried out using the new generation sequencing platform lllumina HiSeq 2500, the type of cancer is glioblastoma, the therapy used is combination therapy with bevacizumab (bevacizumab with radiation and temozolomide) and the threshold value is - 104.00.