JP2011004743A - Method for deciding efficacy of infliximab medicinal effect in patient with rheumatoid arthritis - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for efficiently and effectively discriminating the presence or absence of the efficacy of an infliximab administering on a gene level by analyzing the gene expression profile of a blood sample.SOLUTION: The method for discriminating the efficacy of the infliximab medicinal effect in a patient with rheumatoid arthritis using one or more indicators selected from a plurality of indicators includes: a process for measuring the expression level of one or more genes selected from a specific group in blood of the patient with rheumatoid arthritis; a process for analyzing the measured expression level, using the gene expression profile prepared beforehand; and a process for discriminating the efficacy of infliximab medicinal effect in the patient with rheumatoid arthritis based on the analysis result.

Description

  The present invention relates to a method for determining the efficacy of infliximab efficacy in patients with rheumatoid arthritis using one or more indices selected from a plurality of indices. Specifically, the present invention relates to infliximab biopharmaceuticals in patients with rheumatoid arthritis based on a gene expression profile drawn from the blood collected from purified rheumatoid arthritis and measured and standardized. The present invention relates to a method for determining the presence or absence of a medicinal effect.

  To date, various drugs have been researched and developed for rheumatoid arthritis, an autoimmune disease. In particular, TNFα, a type of tumor necrosis factor (TNF), which is a type of cytokine secreted from macrophages and lymphocytes, is located upstream of the inflammatory cascade and is greatly involved in the development of cellular immunity and the development of autoimmune diseases. It has been elucidated, and biological preparations that exert therapeutic effects by selectively inhibiting TNFα have been developed.

  Infliximab, one of its biologics, is a monoclonal antibody drug based on a chimeric organism of human immunoglobulin and mouse immunoglobulin, and directly targets the TNFα receptor, which plays a central role in the immune response. A new biological product. For this reason, while it is a high-price drug, it acts in a complex immune mechanism, so it may not be effective at all or may be effective once due to genetic characteristics and living environment factors of the patient. There has been a clinical report of a phenomenon called secondary ineffectiveness that loses its efficacy after a certain period of time and then recurs.

  Regarding the use of infliximab in clinical settings, since infliximab is a relatively expensive drug for rheumatoid arthritis, it has been used in the judgment of clinicians in cases where other drugs do not work. It was rarely prescribed. In addition, the effectiveness is determined only by measuring the white blood cell count and biochemical characteristic values of serum by conducting a normal blood test, and observing and judging the results based on the measurement results. (See Non-Patent Documents 1 to 3). In addition, TNF inhibitors such as infliximab are known to have various side effects that increase various infectious diseases, particularly tuberculosis, because they attenuate host immunity.

  On the other hand, DNA chip technology has been developed and introduced into medical care, so that it is possible to obtain comprehensive gene expression profiles at the molecular level in a target for pathological biopsy tissues and blood samples. . Therefore, it is possible to observe the expression network of multiple key genes as a whole, and basic technologies for diagnosing the progression of disease states and the effectiveness of drugs at the molecular level represented by gene expression have been accumulated. (See Non-Patent Document 4, Patent Documents 1 and 2).

  As a diagnosis at the molecular level for drugs, there are reported examples of non-patent documents 5 to 8 and US provisional applications 61 / 269,642 regarding prediction of drug efficacy at the molecular level of infliximab.

JP 2007-135465 A JP 2007-135466 A

Tsutomu Takeuchi “Current Status and Problems of Rheumatoid Drug Treatment”, Pharmacology Journal, Folia Pharmacol. Jpn. 2007, 129, 182-185. Hiroyuki Matsubara, Yasuhisa Hana, “Experience of using biologics in rheumatoid arthritis”, Chubu Disaster Reduction Vol. 48; No. 2: (2005). 297-298. Daihei Kurita, Hiroyuki Miyake, Masashi Mizuno “Small Experience with Infliximab Use” Chubu Disaster Report Vol. 48; No. 2: (2005). 299-300. Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri MA, Bloomfield CD, Lander ES. "Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. "Science. 1999 Oct 15; 286 (5439): 531-537. Tanino M, Matoba R, Nakamura S, Kameda H, Amano K, Okayama T, Nagasawa H, Suzuki K, Matsubara K, Takeuchi T. "Prediction of efficacy of anti-TNF biologic agent, infliximab, for rheumatoid arthritis patients using a comprehensive transcriptome analysis of white blood cells. "Biochem Biophys Res Commun. 2009 Sep 18; 387 (2): 261-5. Motohiko Tanino, Koichi Amano, Tsutomu Takeuchi “Prediction of effects after administration from gene expression pattern of whole blood RNA before administration of infliximab” molecular rheumatic treatment Vol.2 no.4 p34-37 2009. Hideto Kameda, Motohiko Tanino, Tsutomu Takeuchi "Which biological product is not predictable before use?" Molecular Rheumatoid Treatment Vol.2 no.2 p23-25 2009. Motohiko Tanino, Tsutomu Takeuchi “Individualized Rheumatoid Arthritis for Remission by RNA Check” Gene Medicine MOOK10 p315-320 2008. Medical Dou Co., Ltd.

  However, with regard to the reversal of drug efficacy by administration of infliximab, since the mechanism of action of infliximab in the body is complicated, there is currently no diagnostic method for the effectiveness at the molecular level or gene expression profile level.

  In particular, if the presence or absence of infliximab can be predicted in advance by examining the gene expression profile in the patient's blood before administering infliximab to the patient, the drug price is high for patients with infliximab ineffective The disadvantage of selecting a drug can be avoided. At the same time, clinicians will be able to select treatment strategies that are more appropriate for individual patients by considering different treatment measures, such as increasing or decreasing the dose of infliximab or switching to other drugs. .

  The present invention has been made in view of such circumstances, and provides a method for effectively and effectively discriminating whether or not infliximab administration is effective by analyzing a gene expression profile in a blood sample. The purpose is to do.

  In order to objectively determine the efficacy of infliximab administration, the present inventors can easily obtain it as a specimen and determine the mechanism of action of infliximab in vivo via a molecular level change called a gene expression network. We focused on blood, which is thought to be comprehensively expressed. And, the present inventors, in the analysis using the statistical method of the gene group expressed in the blood in the sample in which infliximab is administered and not, and in the sample in which infliximab is effective and ineffective, As a result of comprehensive analysis of mRNA expression of tens of thousands of genes, gaining new knowledge on genes whose expression level changes in relation to the presence or absence of infliximab drug efficacy in patients, and earnest research and experiments, The invention has been completed.

  As a result of intensive studies to solve the above-mentioned problems, the inventors of the present invention have a group of genes listed in Tables 1 to 3 below, which are genes in blood that change specifically in patients who are effective with infliximab. However, it has been found that it can be used to determine the effectiveness of infliximab administration in patients with rheumatoid arthritis.

  Specifically, according to the first main aspect of the present invention, there is a method for determining the efficacy of infliximab drug efficacy in rheumatoid arthritis patients using one or more indices selected from a plurality of indices. The step of measuring the expression level of one or more genes selected from Table 1, Table 2, or Table 3 in the blood of a rheumatoid arthritis patient, and the measured expression level is a gene expression prepared in advance. There is provided a method comprising: analyzing using a profile; and determining effectiveness of infliximab drug efficacy in the rheumatoid arthritis patient based on the analyzed result.

  According to such a configuration, by analyzing a specific gene expression profile from purified blood of a patient with rheumatoid arthritis, before or after administration of infliximab, the presence or absence of a medicinal effect can be discriminated as a material for determining whether or not administration is possible. It is possible to provide clinical diagnosis information. In addition, through this, the effect that rheumatoid arthritis patients can receive more effective medical care can be expected.

  Further, according to an embodiment of the present invention, in such a method, the determining step is selected from the plurality of indices when infliximab is administered to the rheumatoid arthritis patient based on the analysis result. Predicting the presence or absence of a medicinal effect determined by at least one index value.

  According to another embodiment of the present invention, as described above, when predicting the presence or absence of a drug effect determined by at least one index value selected from a plurality of indices, the serum CRP concentration is used as the index. In this case, the step of measuring measures the expression level of one or more genes selected from Table 1, and the step of determining is based on the result of the analysis. In addition, it may be predicted that there is a drug effect or no drug effect determined by the serum CRP concentration when infliximab is administered to the rheumatoid arthritis patient.

  In this case, the determining step is effective when it is predicted that the serum CRP concentration is 0.3 mg / dl or less when infliximab is administered to the rheumatoid arthritis patient based on the analysis result. It is preferable that it is determined that there is no medicinal effect when it is predicted to be greater than 3 mg / dl.

  According to another embodiment of the present invention, as described above, when predicting the presence or absence of a medicinal effect determined by at least one index value selected from a plurality of indexes, the ACR value is used as the index. In this case, the step of measuring measures the expression level of one or more genes selected from Table 2, and the step of determining is based on the result of the analysis. In addition, it may be predicted that there is a drug effect or no drug effect determined by the ACR value when infliximab is administered to the rheumatoid arthritis patient.

  In this case, the step of determining is ACR0 when the ACR value is predicted to be ACR20, ACR50, or ACR70 when infliximab is administered to the rheumatoid arthritis patient based on the analysis result, and is ACR0. When it is predicted, it is preferable to determine that there is no medicinal effect.

  Furthermore, according to another embodiment of the present invention, as described above, when predicting the presence or absence of a medicinal effect determined by at least one index value selected from a plurality of indices, the DAS28 value is used as the index. In this case, the step of determining predicts the presence or absence of a drug effect determined by the DAS28 value when infliximab is administered to the rheumatoid arthritis patient based on the analysis result. It may be.

  In this case, the determining step is effective when it is predicted that the DAS28 value when infliximab is administered to the rheumatoid arthritis patient is 4.1 or less based on the analysis result, and is larger than 4.1. When it is predicted, it is preferable to determine that there is no medicinal effect.

  Further, when the DAS28 value is used as the index, the determining step is based on the analysis result, and the DAS28 value when infliximab is administered to the rheumatoid arthritis patient is 3.2 or less, and It is effective when it is predicted that the value obtained by subtracting the DAS28 value when infliximab is administered to the rheumatoid arthritis patient from the current DAS28 value of the rheumatoid arthritis patient is larger than 1.2, and when it is predicted to be otherwise It may be determined that there is none.

  According to yet another embodiment of the present invention, as described above, when predicting the presence or absence of a medicinal effect determined by at least one index value selected from a plurality of indices, the step of determining Predicts the efficacy or non-efficacy determined by the index value 14 weeks after administration of infliximab to the rheumatoid arthritis patient.

  According to another embodiment of the present invention, in such a method, the blood of the rheumatoid arthritis patient may be collected before infliximab is administered to the rheumatoid arthritis patient, Alternatively, it may be collected after infliximab is administered to the rheumatoid arthritis patient.

  According to a second main aspect of the present invention, there is provided an array used in the method as described above, which encodes one or more genes selected from Table 1, Table 2, or Table 3. There is provided an array, wherein probes that hybridize with at least a part of a base sequence under stringent conditions are immobilized at different positions on a solid support.

  It should be noted that the features of the present invention other than those described above and the remarkable actions and effects will become clear to those skilled in the art with reference to the following embodiments and drawings of the present invention.

FIG. 1 is a flowchart for obtaining a highly discriminating gene group according to an embodiment of the present invention. FIG. 2 is a table showing Prediction Strength (PS) values for each predicted gene before administration of infliximab by the Weighted Vote method in the learning set according to an embodiment of the present invention. FIG. 3 is a table showing a predicted high discriminant gene group before administration of infliximab determined in a learning set and a prediction accuracy evaluation based thereon, according to an embodiment of the present invention. FIG. 4 is a table showing PS values for each predicted gene by the Weighted Vote method in the verification set using the predicted high discrimination gene group before administration of infliximab determined in the learning set in an embodiment of the present invention. FIG. 5 is a table showing discrimination results in the verification set according to an embodiment of the present invention. FIG. 6 is a table showing verification results using CRP, ACR, or DAS 28 in an embodiment of the present invention.

Hereinafter, an embodiment and an example according to the present invention will be described with reference to the drawings.
As described above, the method for determining the effectiveness of infliximab drug efficacy in rheumatoid arthritis patients according to the present embodiment is one or more selected from Table 1, Table 2, or Table 3 in the blood of rheumatoid arthritis patients. The step of measuring the expression level of the gene, the step of analyzing the measured expression level using a gene expression profile prepared in advance, and the effectiveness of infliximab drug efficacy in the rheumatoid arthritis patient based on the analysis result And a step of discriminating.

  Here, in one embodiment of the present invention, each of the above genes has an expression level that changes in relation to the presence or absence of infliximab medicinal effect, and a sample group before administration of infliximab is used as described in Examples below. It is a gene group obtained as a result of exhaustive analysis of gene expression profiles used as a learning set. In this analysis, a plurality of indices are used to determine the efficacy / invalidity of the drug effect in rheumatoid arthritis patients.

  In the present specification, “gene” serves as an index for evaluating the state or action of a certain object, and here refers to a gene-related substance that correlates with the expression level of a certain gene. For example, the gene itself includes mRNA that is a transcript, peptide that is a translation, protein that is the final product of gene expression, and the like. In addition, “expression level” of a gene means a standardized expression level of the gene unless otherwise specified, and the expression level of the gene at the transcription level (in the case of a transcript or the like) or the translation level ( In the case of polypeptides, proteins, etc.).

  In the present embodiment, the array used in the method for determining the efficacy of infliximab drug efficacy in rheumatoid arthritis patients is at least a part of bases encoding one or more genes selected from Tables 1 to 3. Probes that hybridize with the sequence under stringent conditions are fixed at different positions on the solid support.

  The array can be replaced with a PCR plate if necessary. In this case, the probes are arranged at different positions of wells on the plate.

  As an index that can be used to determine whether a specific drug is therapeutically effective or ineffective for patients with rheumatoid arthritis, first, serum CRP (C- reactive protein) concentration. The normal value of serum CRP value is 0.3 mg / dl. In the present embodiment, when the serum CRP value is predicted to be 0.3 mg / dl or less by analysis of the gene expression profile, it is effective, and when it is predicted to be greater than 0.3 mg / dl, there is no effect. Although it is discriminated, this discrimination reference value can be appropriately changed according to the experimental conditions, and is not limited to this numerical value. For example, when the analysis of the gene expression profile predicts that the serum CRP value is 0.5 mg / dl or less, there is a drug effect, and when it is predicted that the serum CRP value is more than 0.5 mg / dl, the drug CRP value is not effective. Effective if expected to be 0.4 mg / dl or less, not effective if expected to be greater than 0.4 mg / dl, predicted to have a serum CRP value of 0.2 mg / dl or less If the drug is predicted to be greater than 0.2 mg / dl, the drug is not effective, or if the serum CRP value is predicted to be 0.1 mg / dl or less, the drug is effective. It is also possible to determine that there is no medicinal effect when it is predicted to be larger.

  The second index that can be used to determine whether a specific drug is therapeutically effective or ineffective for patients with rheumatoid arthritis is the pathology of rheumatoid arthritis established by the American College of Rheumatology. ACR improvement standard, which is a classification standard of the degree. The ACR improvement standard is defined by seven items that are ACR core sets. The ACR core set consists of (1) number of tender joints, (2) number of swollen joints, (3) assessment of disease by the patient, (4) general assessment of disease activity by the patient, (5) disease activity by the physician. It consists of seven items: general assessment, (6) physical function assessment by patient, and (7) acute phase reactant. ACR improvement criteria are, for example, that after treatment (after drug administration compared to before drug administration), both (1) and (2) are improved by 20% or more, and (3) to (7) ) Is improved by 20% or more, it is determined that “ACR standard is improved by 20% (ACR20)”. Similarly, if the ACR improves by 50% as well as 70%, the criteria ACR50, ACR70 are used. In the present embodiment, it is preferable to determine that there is no drug effect when the ACR is predicted to be ACR0 by analysis of the gene expression profile, and that it is effective when it is predicted to be ACR20, ACR50, or ACR70. The discrimination reference value can be appropriately changed according to the experimental conditions, and is not limited to this. For example, when the ACR is predicted to be ACR0 or ACR20 by analysis of the gene expression profile, it may be determined that there is no drug effect, and when it is predicted that the ACR is ACR50 or ACR70, the drug effect is determined. In the present specification, when “ACR” is simply indicated, it means “ACR improvement standard” unless otherwise specified.

  Furthermore, DAS28 is mentioned as a 3rd parameter | index which can be used in order to discriminate | determine whether a specific chemical | medical agent is therapeutically effective or ineffective with respect to a rheumatoid arthritis patient. DAS28 narrowed down the joints to be evaluated to 28 joints in order to make it easy to use DAS (dissease activity score), which is the evaluation method recommended by EULAR (European League Against Rheumatism). Is. In DAS28, for 28 joints, (1) number of joints with tenderness, (2) number of joints with swelling, (3) 1 hour value of blood sedimentation, or CRP (mg / dl), (4) The activity of the disease is evaluated from the evaluation value. Note that DAS28-ESR is used when blood sedimentation is used, and DAS28-CRP is used when CRP values are used. In the present embodiment, as the DAS 28, any of DAS28-ESR and DAS28-CRP can be used, but DAS28-CRP is preferable. In this embodiment, in the case of DAS28-CRP, there is a drug effect when DAS28-CRP is predicted to be 4.1 or less by analysis of the gene expression profile, and when it is predicted to be greater than 4.1, the drug effect Although it is determined that there is none, the determination reference value can be changed as appropriate according to the experimental conditions, and is not limited to this value. For example, it may be determined that there is a drug effect when DAS28-CRP is predicted to be 2.7 or less by analysis of the gene expression profile, and that there is no drug effect when it is predicted that the DAS28-CRP is greater than 2.7. In addition, in the case of DAS28-ESR, it is determined that there is a drug effect when DAS28-ESR is predicted to be 5.1 or less by analysis of the gene expression profile, and it is determined that there is no drug effect when it is predicted to be greater than 5.1. However, the determination reference value can be appropriately changed according to the experimental conditions, and is not limited to this value. For example, when DAS28-ESR is predicted to be 3.2 or less, it may be determined that there is a drug effect, and when it is predicted that DAS28-ESR is greater than 3.2, it may be determined that there is no drug effect.

  In addition, when using DAS28 to determine whether a specific drug is therapeutically effective or ineffective in patients with rheumatoid arthritis, the EULAR improvement criteria can also be used. In the EULAR improvement criteria, By combining two DAS28 values after treatment (before drug administration versus after drug administration), the therapeutic effect is evaluated in three stages: good response, moderate response, and no response. Specifically, for example, when the post-treatment DAS28 value is 3.2 or less and the value obtained by subtracting the post-treatment DAS28 value from the pretreatment DAS28 value is greater than 1.2, the drug efficacy is obtained. It can be determined that there is none. In addition, when the post-treatment DAS28 value is 3.2 or less and the value obtained by subtracting the post-treatment DAS28 value from the pretreatment DAS28 value is greater than 0.6, or the post-treatment DAS28 value is 5.1 or less, and When the value obtained by subtracting the post-treatment DAS 28 value from the pre-treatment DAS 28 value is greater than 0.6, or the post-treatment DAS 28 value is greater than 5.1, and the value obtained by subtracting the post-treatment DAS 28 value from the pre-treatment DAS 28 value is 1. It may be determined that there is a medicinal effect when it is larger than 2, and there is no medicinal effect in other cases.

  Further, in this embodiment, as described later, based on the analysis result of the gene expression profile of the rheumatoid arthritis patient, the efficacy or non-efficacy determined from the DAS28 value when infliximab is administered is predicted. Therefore, for example, as a result of gene expression profile analysis, when infliximab is administered to a patient with rheumatoid arthritis, the DAS28 value is 3.2 or less, and the current DAS28 value of the rheumatoid arthritis patient is administered to the patient with rheumatoid arthritis. When the value obtained by subtracting the DAS28 value in the case of the above is predicted to be greater than 1.2, it can be determined that the drug is effective, and in other cases, the drug is not effective. In this case, the “current DAS28 value” comprehensively indicates the DAS28 value of the patient before treatment (before drug administration), and preferably, when collecting blood of a rheumatoid arthritis patient according to the present invention. DAS28 value.

  Note that the index that can be used to determine whether a specific drug is therapeutically effective or ineffective for patients with rheumatoid arthritis is not limited to the above-mentioned index, and is acceptable in clinical practice. Any index may be used as long as the validity / invalidity can be determined with certainty.

  Next, a method for selecting a gene group used to determine the effectiveness of infliximab drug efficacy in a rheumatoid arthritis patient according to an embodiment of the present invention will be described. In this embodiment, the gene group (discriminating gene group) used for discriminating the efficacy of infliximab drug efficacy in rheumatoid arthritis patients is the high discrimination gene list according to one embodiment of the present invention as shown in FIG. It can be obtained according to the flowchart for obtaining. In summary, samples for discriminating gene determination analysis and verification before administration of infliximab are purified, and a gene expression profile is prepared by a comprehensive DNA microarray. Then, for the discriminant gene determination analysis, using a sample with known clinical information on whether or not infliximab is effective or ineffective, a gene whose expression level is different between the two groups is analyzed using Mann-Whitney U test statistics. The discriminant gene list is created in descending order of the importance of the signal and the signal for noise regarding discrimination. For the partial set including the entire gene list, for example, it is confirmed that there is no overfitting by the multiple cross-validation method, the method disclosed in Japanese Patent Application No. 2007-230142, etc. Self-evaluation, and further, accuracy evaluation is performed on the profile data for verification before and after administration, and an optimal gene list for determining whether or not infliximab is effective is obtained at a molecular level.

  Specifically, first of all, classify samples as “effective” or “not effective” according to the index for determining the efficacy of infliximab in patients with rheumatoid arthritis as described above, and make a 2-group discrimination To search for a gene list showing a high discrimination power. The acquired gene expression level data is preferably normalized in advance by a standardization method usually used in the field of statistics such as Quantile Normalization or the field of molecular biology. In the specification of the present application, the term “profile” refers to data after performing the normalization process unless otherwise specified.

  In the present embodiment, a sample (including samples before and after administration) used for analyzing genes to obtain a discriminating gene group is a learning set, and the obtained discriminating gene group actually has a high discriminating ability. Samples (including samples before and after administration) used for verifying whether or not are used as a verification set. For any sample set, it is preferable to prepare a sample before administration of infliximab and a sample after a certain period of administration of infliximab, and practice the flowchart described below. As the fixed period, for example, an arbitrary period such as 54 weeks, 22 weeks, 14 weeks, 6 weeks, or the like can be selected. For example, when the serum CRP value is used as the above-mentioned index, the normal threshold value is determined depending on whether the normal threshold value (0.3 mg / dl or less) is reached for each subject during a certain period after administration (after 14 weeks). It can be classified as having a medicinal effect when it reaches the limit, and as having no effect when it does not reach the limit.

  The discriminating gene group is obtained from gene expression profile data before administration of infliximab. In the flowchart described below, the processing is common except that it is a gene expression profile when using any of serum CRP concentration, ACR, or DAS28 value as an index for discrimination. Further, S101 to S106 shown in the description of each step below indicate steps S101 to S106 in FIG. In addition, a commercially available RNA extraction kit such as PAX Gene Kit can be used for collecting RNA from a blood sample of a subject.

  First, the entire profile data is divided into “medicine effective” and “no drug effective” subjects as described above, and, for example, a Mann-Whitney U test is performed between the two groups (S101). A non-parametric statistical test such as Wilcoxon rank sum test can also be used.

  Next, the respective partial data are sorted in the descending order of the statistical value of the Mann-Whitney U test (S102). As described above, “with medicinal effect” means that, for example, when the serum CRP concentration is 0.3 mg / dl or less after 14 weeks of administration, the ACR is ACR20, ACR50, and ACR70, and the DAS28 value is 4. 1 or less and “no effect” can be 14 weeks after administration if serum CRP concentration is greater than 0.3 mg / dl, ACR is ACR0, DAS28 value is greater than 4.1 .

  Then, in step S103, n is sequentially taken from the beginning of the sorted list, and this n is set as the cumulative number of genes. Using the partial data of the cumulative gene number n, classification prediction of all samples in the learning set is performed to determine whether the answer is correct or incorrect (S103). Preferably, the weighted vote method (see Non-Patent Document 4) can be used for the configuration of the discriminator, but the discriminant function is not limited to this, and for example, k nearest neighbor (k-NN) The method, support vector machine (SVM), Artificial Neural Network (ANN), Fisher linear classifier, hierarchical clustering, principal component analysis (PCA), and the like can also be used.

Subsequently, the correct or incorrect answer values determined for each sample number and cumulative gene count are aggregated over all partial data, the accuracy list is evaluated, and the discrimination rate defined below is defined for each cumulative gene count. On the other hand, n which obtains the target accuracy is obtained (S104).
・ Specificity (spec.): (Number of correct samples without efficacy) / (total number of samples without efficacy)
Sensitivity (sens.): (Number of correct samples with medicinal properties) / (total number of samples with medicinal properties)
・ Accuracy (acc.): (Number of correct samples with or without medicinal effect) / (total number of samples)
Ppv. : (Number of correct samples with medicinal properties) / (number of judgment samples with medicinal properties)
-Npv. : (Number of correct samples without medicinal effect) / (number of judgment samples without medicinal effect)

  In addition, although all of the above numerical values are not independent variables, it is an effective clinical diagnosis index that all these indices exceed 50% in percentage and values as close as possible to 100% are obtained. In addition, an increase in the number of genes can be expected to improve the discrimination rate, but on the other hand, an overfitting occurs due to an excessive increase in the number of genes, and it is predicted that the discrimination rate will decrease as the number of genes increases again after saturation. Is done. Therefore, by searching for the number of genes that cause a peak with a maximum discrimination rate in the results of actual data, the most reasonable number of genes can be determined as the optimal gene list.

  Here, since the numerical value obtained by each step described above includes the possibility of overfitting, it is necessary to determine the validity with respect to the same number of samples and the same number of genes. Therefore, for example, by using the method disclosed in Japanese Patent Application No. 2007-230142, clinical information (in this case, label data of the determination group of the presence or absence of a drug effect) is randomly shuffled and shown in steps S101 to S104. At least once that an accuracy of significantly greater than 50% has not been achieved due to false fit predictions against random classification criteria in the same scale profile. Preferably, confirmation may be performed over a plurality of times (S105).

  Subsequently, in step S106, an optimal gene list is determined for the number of genes obtained in the above steps. For this purpose, an average rank across all partial profiles is obtained for each gene from the partial profiles, using the information on the gene rank at the time of sorting in step S102. From the order of consensus, those obtained with the number of genes described above are finally set as the optimal gene list (S106).

  The optimal gene list obtained by the above steps needs to be verified again. In this revalidation, the sample is divided into a number of times, divided into training and verification, and is included in the mccv (Montecarlo Cross Validation) method and, if possible, the analysis using the learning set described above. It is desirable to prepare a large number of new samples and perform prospective verification described later.

  Hereinafter, the present invention will be described in more detail with reference to examples, but the present invention is not limited to these examples. In particular, in this example, RNA was purified from 42 blood samples for analysis and 56 samples for analysis before administration of infliximab using PAX Gene Kit, and expression profiles were comprehensively analyzed using an Agilent DNA microarray. Although it has been acquired, a method usually used in the field of molecular biology can be adopted for collecting a sample.

(Basic verification of prediction data before infliximab administration)
FIG. 2 shows the Prediction Strength (PS) value by the higher weighted vote method obtained in steps S101 to S104 in the data of 42 samples before administration of infliximab, and FIG. 3 shows the discrimination rate. For the definition of various discrimination rates, see spec. Sens. Acc. , Ppv. , Npv. Is as described above, and other definitions will be described later. According to this, when serum CRP concentration is used as an indicator of the presence or absence of medicinal effect, accuracy of acc. It can be seen that the (accuracy) peak is achieved. Moreover, although not shown, when an ACR value is used, a peak is reached at 27 genes. In these cases, in the overconformance check in step S105, the accuracy did not exceed 55%, and no sign of overconformity was observed.

(The optimal gene list for prediction data before infliximab administration)
FIG. 2 and FIG. 3 show the prediction before administration of infliximab when the serum CRP concentration after 14 weeks of administration obtained from step S101 to step S106 of the data of 42 samples before administration described above is used as an indicator of efficacy / non-efficacy The optimal gene list of the data is shown.

(Prospective verification of optimal gene list before infliximab administration)
FIG. 4 shows a PS value obtained by the weighted vote method obtained from a verification set that is a completely unknown prospective sample different from the learning set, using the above-described optimal gene list. FIG. 5 shows the prediction result. As the discriminant function, the weighted vote used in the configuration of the discriminator described above is similarly used. At this time, when another discriminant function is used in the configuration of the discriminator, it is preferable to use the same discriminant function also in the verification.

  As a result, the prediction accuracy of the optimal gene list of the infliximab pre-administration prediction data is estimated to be 62.5% using the entire optimal gene list, indicating the effectiveness.

(Prospective verification of optimal gene list using ACR and DAS28)
FIG. 6 shows the verification results when ACR and DAS 28 are used. As shown here, even when ACR and DAS28 are used, the accuracy of infliximab validity / invalidity obtained from the optimal gene list for each index is 63.6% and 75.0%, respectively. It can be seen that it is.

(References and explanations in each figure)
Below, the code | symbol in each figure and the calculation formula about each code | symbol are demonstrated.

(Figure 1)
Each reference numeral in FIG. 1 is the description of each step described above and the following.

S101: Genes with significant differences in RNA expression profiles with “drug efficacy” and “no drug efficacy” are extracted by Mann-Whitney U test S102: Sorted in order of significant difference in expression profiles S103: Classification prediction of learning set samples S104: Accumulation of prediction accuracy and achievement of target accuracy determination of gene size S105: Check for possibility of over-matching of gene size S106: Determination of high discrimination gene list

(Figure 2)
201: Cumulative gene. In one embodiment according to the present invention, the genes are sorted in the order of significance according to the Mann-Whitney U test, and are used as cumulative genes in the weighted vote method.
202: Prediction Strength (PS) value for each patient The PS value is obtained by the following equation.
Prediction Strength value = Σi (Weighted Vote value) / Σi | Weighted Vote value |
Here, the Weighted Vote value is
Weighted Vote value = (x_i− (ave_e_i + ave_n_i) / 2) SNR
And the SNR is
SNR = | ave_e_i-ave_n_i | / (sigma_e_i + sigma_n_i)
Defined by In this equation, each value is determined using the following variables for each subscript “_i” representing the i-th gene.
X is an expression value of each sample, ave_e is an average value of profile intensity of a group having medicinal effect, ave_n is an average value of profile intensity of a group having no medicinal effect, and sigma_e is a standard deviation of profile intensity of a group having medicinal effect Sigma_n is a standard deviation of profile intensity of a group having no drug effect. Here, a negative value indicates “with drug effect” and a positive value indicates “without drug effect”. It should be noted that positive / negative is a problem in the definition of an expression and can be defined in reverse. For example, it can be seen that the patient 3 in FIG.

(Figure 3)
301: Cumulative gene. In one embodiment according to the present invention, the genes are sorted in the order of significance according to the Mann-Whitney U test, and are used as cumulative genes in the weighted vote method.
302 ... Flag for determination. Each flag is as follows.
-TP (True Positive): Number of samples that were predicted to have medicinal effects and actually had medicinal effects-FP (False Positive): Number of samples that were predicted to have medicinal effects and had no medicinal effects-TN (True Negative): No effect Number of samples that were predicted and actually had no medicinal effect FN (False Negative): Number of samples that were predicted to have no medicinal effect and had medicinal properties 303... (Specificity): (number of correct samples without medicinal effects) versus (total number of samples without medicinal effects) percentage (Sensitivity): (number of correct samples with medicinal properties) versus (total number of samples with medicinal properties) percentage acc. (Accuracy): (number of correct samples with or without medicinal effects) versus (total number of samples) percentage ppv. (Positive Predictive Value): (number of correct samples with medicinal effects) vs. (number of determination samples with medicinal effects) percentage npv. (Negative Predictive Value): (number of correct samples without efficacy) vs. (total number of samples without efficacy) percentage

(Fig. 4)
401 ... Cumulative gene. In one embodiment according to the present invention, the genes were sorted in the order of significance by Mann-Whitney U test and used as cumulative genes in the Weighted Vote method (rank determined in the learning set).
402: Prediction Strength (PS) value. In addition, PS value is calculated | required with the following formula | equation.
Prediction Strength (PS) value = Σi (Weighted Vote value) / Σi | Weighted Vote value |
Here, the Weighted Vote value is
Weighted Vote value = (x_i− (ave_e_i + ave_n_i) / 2) SNR
And the SNR is
SNR = | ave_e_i-ave_n_i | / (sigma_e_i + sigma_n_i)
Defined by In this equation, each value is determined using the following variables for each subscript “_i” representing the i-th gene.
X is the expression value of each sample. The following variables use data from the learning set.
Ave_e is the mean value of the profile intensity of the medicinal group. Ave_n is the mean value of the profile intensity of the ineffective group. Sigma_e is the standard deviation of the profile intensity of the medicinal group. Sigma_n is the group having no medicinal effect. Standard deviation of profile intensity

(Fig. 5)
501: Cumulative gene. In one embodiment according to the present invention, the genes are sorted in the order of significance according to the Mann-Whitney U test, and are used as cumulative genes in the weighted vote method.
502... Determination flag. Each flag is as follows.
-TP (True Positive): Number of samples that were predicted to have medicinal effects and actually had medicinal effects-FP (False Positive): Number of samples that were predicted to have medicinal effects and had no medicinal effects-TN (True Negative): No effect Number of samples that were predicted and actually had no medicinal effect FN (False Negative): The number of samples that were predicted to have no medicinal effect and had medicinal properties 503... (Specificity): (number of correct samples without medicinal effects) versus (total number of samples without medicinal effects) percentage (Sensitivity): (number of correct samples with medicinal properties) versus (total number of samples with medicinal properties) percentage acc. (Accuracy): (number of correct samples with or without medicinal effects) versus (total number of samples) percentage ppv. (Positive Predictive Value): (number of correct samples with medicinal effects) vs. (number of determination samples with medicinal effects) percentage npv. (Negative Predictive Value): (number of correct samples without efficacy) vs. (total number of samples without efficacy) percentage

(Fig. 6)
601: Clinical information. It is a clinical index that estimates the effects of infliximab. In FIG. 6, CRP, ACR, and DAS 28 are shown.
602... Threshold value of “with medicinal effect” and “no medicinal effect” for each clinical information. In the present embodiment, “with medicinal properties” means that 14 weeks after administration, when the serum CRP concentration is 0.3 mg / dl or less, ACR is ACR20, ACR50, or ACR70, and DAS28 value is 4.1 or less. “No effect” means that after 14 weeks of administration, the serum CRP concentration is greater than 0.3 mg / dl, the ACR is ACR0, or the DAS28 value is greater than 4.1.
603... Determination flag. Each flag is as follows.
-TP (True Positive): Number of samples that were predicted to have medicinal effects and actually had medicinal effects-FP (False Positive): Number of samples that were predicted to have medicinal effects and had no medicinal effects-TN (True Negative): No effect Number of samples that were predicted and actually had no medicinal effect FN (False Negative): The number of samples that were predicted to have no medicinal effect and had medicinal properties 603... (Specificity): (number of correct samples without medicinal effects) versus (total number of samples without medicinal effects) percentage (Sensitivity): (number of correct samples with medicinal properties) versus (total number of samples with medicinal properties) percentage acc. (Accuracy): (number of correct samples with or without medicinal effects) versus (total number of samples) percentage ppv. (Positive Predictive Value): (number of correct samples with medicinal effects) vs. (number of determination samples with medicinal effects) percentage npv. (Negative Predictive Value): (number of correct samples without efficacy) vs. (total number of samples without efficacy) percentage

(Molecular biology experiment method)
Hereinafter, experimental methods and substances used in one embodiment according to the present invention and their definitions will be described. In the present embodiment, the following experimental method is used, but the same result can be obtained by using other experimental methods.
(Quantification of gene expression)
In the present embodiment, the gene expression level is measured (quantified) by, for example, a DNA chip, microarray method, real-time PCR, Northern blotting, dot blotting, quantitative RT-PCR (quantitative reverse transcription-polymerization chain reaction). It can be performed by measuring the amount of mRNA by various molecular biological techniques such as the method.

  The primer used in the quantitative RT-PCR method is not particularly limited as long as it can specifically detect a gene marker, but an oligonucleotide consisting of 12 to 26 bases is preferable. The base sequence is determined based on the sequence information of each human gene. And the primer which has the determined arrangement | sequence can be produced using a DNA synthesizer, for example.

  On the other hand, when a gene translation product (polypeptide) or final product (protein) is used as a quantification target, for example, a Western blot method or an ELISA method using a polyclonal antibody or a monoclonal antibody specific to the gene, etc. The expression level of the gene can be measured by the above method, but it is not limited to these methods, and various methods such as RIA (radioimmunoassay) and EIA (enzyme immunoassay) are used. Can be used.

  In one embodiment according to the present invention, “gene” refers to genetic information represented by a base sequence such as RNA or DNA. Also included are orthologous genes that are conserved among species such as humans, mice, and rats. The gene may function not only as a protein-encoding protein but also as RNA or DNA. A gene generally encodes a protein according to its base sequence, but a protein having a biological function equivalent to that protein (for example, a homologue (such as a homolog or a splice variant), a mutant or a derivative) You may code. For example, a protein encoding a protein whose base sequence is slightly different from the protein indicated by the base sequence based on genetic information, and whose base sequence hybridizes with a complementary sequence of the base sequence based on the genetic information. There may be.

  “RNA” is a concept including not only single-stranded RNA but also single-stranded RNA having a sequence complementary thereto and double-difference RNA composed of these. The concept also includes total RNA, mRNA, and rRNA.

  “DNA chip” and “DNA array” have a structure in which probe DNA is arranged on a substrate, and measure the expression of a plurality of genes by hybridization. This includes not only optically measuring the expression level but also electrical output of the expression level. As the “DNA chip”, for example, Whole Human Genome (trademark) manufactured by Agilent or GeneChip (trademark) manufactured by Affymetrix can be used. As the “DNA array”, CodeLink Expression Bioarray (trademark) manufactured by Amersham Biosciences can be used. The DNA array includes not only a DNA microarray but also a DNA macroarray.

  The “expression level” is a concept including not only a value obtained by directly measuring the expression level of a gene but also a value converted by a predetermined calculation or statistical method. In addition, “gene expression level”, “expression signal”, “gene expression signal”, “expression signal value”, “gene expression signal value”, “gene expression data”, “expression data”, etc. It is synonymous to indicate the value to be reflected.

  “Gene expression” refers to an aspect of gene expression in a living body expressed by the expression level of a gene, both when expressed by the expression level of one gene and when expressed by the expression level of a plurality of genes Is included. “Expression” is also synonymous with the expression of gene expression in a living body.

  In addition, it cannot be overemphasized that this invention can be deform | transformed variously, It is not limited to one Embodiment mentioned above, A various deformation | transformation is possible in the range which does not change the summary of invention.

Claims (13)

A method for determining the efficacy of infliximab efficacy in patients with rheumatoid arthritis using one or more indicators selected from a plurality of indicators,
Measuring the expression level of one or more genes selected from Table 1, Table 2, or Table 3 in the blood of a rheumatoid arthritis patient;
Analyzing the measured expression level using a gene expression profile prepared in advance;
Determining the efficacy of infliximab efficacy in the rheumatoid arthritis patient based on the analyzed results;
A method characterized by comprising: The method of claim 1, wherein
The step of determining predicts the presence or absence of a medicinal effect determined by at least one index value selected from the plurality of indices when infliximab is administered to the rheumatoid arthritis patient based on the analysis result. A method, characterized in that The method of claim 2, wherein
The step of measuring is to measure the expression level of one or more genes selected from Table 1.
The step of determining predicts the presence or absence of a drug effect determined by the serum CRP concentration when infliximab is administered to the rheumatoid arthritis patient based on the analyzed result. . The method of claim 3, wherein
The discriminating step is effective when it is predicted that the serum CRP concentration is 0.3 mg / dl or less when infliximab is administered to the rheumatoid arthritis patient based on the analysis result, and 0.3 mg / dl A method characterized by discriminating that it is not effective when predicted to be larger. The method of claim 2, wherein
The step of measuring is to measure the expression level of one or more genes selected from Table 2.
The determining step is based on the result of the analysis, and predicts the presence or absence of a drug effect determined by an ACR value when infliximab is administered to the rheumatoid arthritis patient. The method of claim 5, wherein
Based on the result of the analysis, the step of determining was predicted to be ACR0 when the ACR value was predicted to be ACR20, ACR50, or ACR70 when infliximab was administered to the rheumatoid arthritis patient, and to be ACR0. A method characterized in that it is determined that there is no medicinal effect. The method of claim 2, wherein
The step of measuring is to measure the expression level of one or more genes selected from Table 3.
The determining step is based on the result of the analysis, and predicts the presence or absence of a drug effect determined by the DAS28 value when infliximab is administered to the rheumatoid arthritis patient. The method of claim 7, wherein
The step of determining is based on the result of the analysis and has a drug effect when it is predicted that the DAS28 value is 4.1 or less when infliximab is administered to the rheumatoid arthritis patient, and is predicted to be greater than 4.1. A method characterized in that it is determined that there is no medicinal effect. The method of claim 7, wherein
The step of determining, based on the analysis result, has a DAS28 value of 3.2 or less when infliximab is administered to the rheumatoid arthritis patient, and is based on the current DAS28 value of the rheumatoid arthritis patient. When it is predicted that the value obtained by subtracting the DAS28 value when infliximab is administered is greater than 1.2, it is determined to be effective, and when it is predicted to be other than that, it is determined that there is no effect. Method. The method of claim 2, wherein
The determining step is for predicting the presence or absence of a drug effect determined by an index value after 14 weeks of administration of infliximab to the rheumatoid arthritis patient. The method of claim 1, wherein
The method is characterized in that the blood of the rheumatoid arthritis patient is collected before infliximab is administered to the rheumatoid arthritis patient. The method of claim 1, wherein
The method is characterized in that the blood of the patient with rheumatoid arthritis is collected after infliximab is administered to the patient with rheumatoid arthritis. An array for use in the method of claim 1, comprising:
Probes that hybridize under stringent conditions with at least a portion of the base sequence encoding one or more genes selected from Table 1, Table 2, or Table 3 are located at different positions on the solid support. An array characterized by being fixed.

Images