JPWO2016181912A1 - Prognostic formulas and prognostic estimation methods for lung adenocarcinoma using immune factors as indicators - Google Patents

Abstract

The present invention provides a method for predicting the prognosis of lung adenocarcinoma more easily and quickly, and a method for creating an arithmetic expression for carrying out the method. The prognostic formula creation method according to the present invention detects the expression of PD-L1, Gal-9 and XAGE1 by immunostaining for each specimen of a plurality of lung adenocarcinoma patients, and distributes the PD-L1 of the tumor in the tumor tissue First data binarized first score specified based on range and expression intensity, and second score specified based on distribution range and expression intensity of tumor Gal-9 in tumor tissue Based on 2 data and 3rd data binarized based on the presence or absence of the expression of XAGE1, lung adenocarcinoma patients are divided into good prognosis group and bad prognosis group, and the 1st score, 2nd score and 3rd data are explanatory variables The method of estimating the prognosis of lung adenocarcinoma according to the present invention is characterized in that the probabilistic estimation method of the lung adenocarcinoma according to the present invention is obtained from the obtained arithmetic expression. The third data and the first and second scores detected from by substituting into the mathematical expression and estimates by calculating the prognosis of the subject.

Description

  The present invention relates to a method for creating an arithmetic expression for predicting the prognosis of primary lung adenocarcinoma (hereinafter referred to as lung adenocarcinoma) and a method for estimating prognosis of lung adenocarcinoma using the same. More specifically, a method for creating a prognostic calculation formula based on the relationship between the expression profile in tumor tissue and the prognosis of a plurality of factors associated with the prognosis of lung adenocarcinoma, and the calculation formula for lung adenocarcinoma using the calculation formula The present invention relates to a method for estimating prognosis.

The incidence of cancer in Japan is increasing year by year, and is currently the leading cause of death. In addition to standard cancer treatments, the development of new treatments is urgent. In recent years, malignant melanoma has led the way to antibody immunotherapy for lung cancer, but the immune surveillance mechanism for lung cancer is unknown.
Conventionally, there are two methods for examining the immune surveillance mechanism in the tumor local area, examining the expression of immunosuppressive molecules and cancer antigens on the tumor side and the infiltration of immune cells on the host side. Many studies have been conducted to examine the relationship between the PD-L1 / PD-1 pathway and prognosis as an immunosuppressive molecule on the tumor side in lung cancer (Patent Document 1, Non-Patent Document 3). Regarding the relationship between the expression pattern of other immunosuppressive molecules and cancer antigens and the frequency and prognosis, for example, XAGE1 as a cancer-related protein to be measured in Patent Document 2 and Non-Patent Document 2, It has been described that Galectin-9 can be a prognostic factor for postoperative recurrence of lung cancer. On the other hand, with regard to factors on the host side and prognosis, only the relationship between various immune cell infiltration and prognosis was examined. However, a combination of a plurality of factors on the tumor side and the host side has not been studied.

Special table 2012-503984 gazette Special table 2012-515334 gazette

Yoshitomo Ozaki et al .: Journal of the Japanese Cancer Association, Vol. 65th page 645 (Aug. 28, 2006) Ohue, Y., et al.,: Clin. Cancer Res., 20, 5052-5063 (2014) Nao Kawamura et al .: Monthly Clinical Immunology and Allergy, 56, 1-6 (2011)

  It is an object of the present invention to provide a method for estimating the prognosis of lung adenocarcinoma more easily and quickly and a method for creating an arithmetic expression for carrying out the method.

In view of the above problems, the present inventors tried to estimate the prognosis of lung adenocarcinoma by examining a plurality of immune factors in lung adenocarcinoma. As a result, the present inventors have important PD-L1 (Programmed death-ligand 1) and Gal-9 (galectin-9) as tumor-side immunosuppressive factors in lung cancer, and in lung adenocarcinoma, It was found that the expression of XAGE1 having sex and the expression pattern of PD-L1 and Gal-9 molecules are associated with prognosis. These factors are not independent, and prognosis can be determined only by combining the presence or absence of XAGE1 expression, the intensity of PD-L1 expression, and the intensity of Gal-9 expression.
Furthermore, the present inventors have shown that prognosis is related to CD4 and CD8 T cell infiltration as a host-side factor, the degree of T cell infiltration, the presence or absence of XAGE1 expression, the intensity of PD-L1 expression, and the expression of Gal-9. It was found that the prognosis of lung adenocarcinoma patients can be more clearly distinguished by combining the intensities.
Based on these findings, the present inventors have further investigated and have completed the present invention.

That is, the present invention relates to the following.
[1] For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9 and XAGE1 by immunostaining,
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue Based on the second data binarized second score and the third data binarized based on the presence or absence of the expression of XAGE1, lung adenocarcinoma patients are divided into good prognosis group and bad prognosis group,
A prognostic calculation formula creation method characterized in that an arithmetic expression for dividing a good prognosis group and a bad prognosis group is obtained by multivariate analysis using the first score, the second score, and the third data as explanatory variables.
[2] For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9, XAGE1, CD4 and CD8 by immunostaining,
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue The second data binarized from the second score, the third data binarized based on the presence or absence of expression of XAGE1, and the fourth score specified based on the rate of infiltration of T cells Based on the 4 data, the lung adenocarcinoma patients are divided into a good prognosis group and a bad prognosis group, and a good prognosis group is obtained by multivariate analysis using the first score, the second score, the third data, and the fourth score as explanatory variables. A method for creating a prognostic formula, characterized in that an arithmetic formula for dividing into bad groups is obtained.
[3] Substituting the first and second scores detected from the subject's sample and the third data into the arithmetic expression obtained by the method described in [1] above, the prognosis of the subject is estimated by calculation. Prognosis estimation method for lung adenocarcinoma.
[4] Substituting the first and second scores, the third data, and the fourth score detected from the specimen of the subject into the arithmetic expression obtained by the method described in [2] above, the prognosis of the subject is determined. A prognosis estimation method for lung adenocarcinoma estimated by calculation.
[5] A method for selecting an early lung cancer patient with a poor prognosis, characterized by using the prognosis estimation method for lung adenocarcinoma described in [3] or [4] above, and for performing treatment necessary for improving the prognosis Companion diagnostic method.
[6] A method for determining the effect of a method for treating lung adenocarcinoma, comprising using the method for estimating prognosis of lung adenocarcinoma according to [3] or [4] above.
[7] The prognosis of lung adenocarcinoma according to [3] or [4] above, wherein the subject is an early lung cancer patient and the prognosis estimation method for lung adenocarcinoma is to determine the postoperative recurrence rate in the patient. Estimation method.

  ADVANTAGE OF THE INVENTION According to this invention, the system which enables estimation of prognosis by examining an immune factor in lung adenocarcinoma is provided. In clinical practice, treatment is selected according to the clinical stage and pathological stage, but by using a prognosis determination system using immune factors as an index separately from those stages, clinical early cancer Even so, patients who require aggressive multidisciplinary treatment can be selected. Such patient selection is the basis for the development of new drug discovery and treatment methods. In addition, the present invention is useful for elucidating the relationship between lung adenocarcinoma and immune factors, but weighting to immune factors is an important index for the development of new drugs such as cancer immunotherapy in lung adenocarcinoma. Further, according to the present invention, the prognosis can be determined only by examining a minute tumor tissue such as a bronchoscopic specimen or a specimen obtained by CT-guided lung biopsy. Therefore, prognosis can be determined by collecting a minute clinical sample in a patient who cannot perform whole body search of the stage, and can contribute to the determination of the subsequent treatment policy.

It is a figure which shows the result of the flow cytometry analysis of PD-L1 and Galectin-9 expression in the cell surface (Surface) and cytoplasm (Intracellular) about the shown cell line. Gray indicates results with isotype antibodies. Release from OU-LC-SK cells and PC-9 cells with or without PD-L1, Gal-9 expression (AB) and molecular targeted drug (Afatinib) treatment by IFN-γ treatment in lung cancer cell lines Gal-9 concentration (C), Gal-9 concentration (D) in serum or pleural fluid of cancer patients, and Gal-9 concentration (E) in patient serum during the course of chemotherapy. In D and E, the Gal-9 concentration is a 5-fold diluted concentration. Results of flow cytometry analysis of the expression of immunosuppressive molecules on the surface of T cells (TIL) infiltrating local cancer (AB), PD-1, Tim-3 expression in TIL, CD27 expression, Ki -67 expression, relationship with Annexin V expression (C), PD-1, Tim-3 expression (D) in XAGE1-specific CD8 clones, and when Gal-9 was added to XAGE1-specific CD8 clones at the concentrations indicated The degree of apoptosis (E) and the inhibitory effect of apoptosis by adding anti-Gal-9 (F) are shown. It is a figure which shows the immuno-staining image about PD-L1, Galectin-9, and XAGE1 in a lung cancer tissue. It is a figure which shows the relationship between the expression and prognosis of PD-L1, Gal-9, and XAGE1 in a lung cancer tissue. In each graph, the horizontal axis represents the time (month) after the operation, and the vertical axis represents the overall survival rate (%). From the results of immunostaining images of host-side factors (CD4, CD8, FoxP3) (A), scoring of expression intensity and frequency (B), and the relationship between the ratio of CD4 positive T cells and the number of samples It is a figure which shows the reference | standard (C) to convert into a value. It is a figure which shows the relationship between the number of tumor local infiltrating T cells and the number of PD-L1, Galectin-9 or XAGE1 expression and prognosis. In each graph, the horizontal axis represents the time (month) after the operation, and the vertical axis represents the overall survival rate (%). Derivation of prognostic function based on the expression of 3 tumor factors (PD-L1, Gal-9, XAGE1) in primary lung adenocarcinoma (AC), in addition to the above 3 factors, the host factor T cell infiltration It is a figure which shows derivation | leading-out (DF) of the prognosis function based on it. In A, B, D, and E, the horizontal axis represents the time after surgery (month), and the vertical axis represents the overall survival rate (%). Derivation of prognostic function based on the expression of tumor side 3 factors (PD-L1, Gal-9, XAGE1) and host cell factor T cell invasion in early stage primary lung adenocarcinoma belonging to stage IA FIG. In each figure, the horizontal axis represents the time (month) after surgery, and the vertical axis represents the overall survival rate (%). It is a figure which shows the actual prognosis of all the 120 lung adenocarcinoma patients estimated using discriminant function [Equation 5], 62 patients who belong to stage I, and 58 patients who belong to stage II-IIIA. . In each figure, the horizontal axis represents the time (month) after surgery, and the vertical axis represents the overall survival rate (%). It is a figure which shows the actual prognosis of all 120 lung adenocarcinoma patients estimated using discriminant function [Equation 6]. In the figure, the horizontal axis represents the time after surgery (month), and the vertical axis represents the overall survival rate (%). It is a figure which shows the actual prognosis of 68 patients who belong to the staging classification I estimated using discriminant function [Equation 5] by a verification cohort study. In the left figure, the horizontal axis represents the time (month) after surgery, and the vertical axis represents the overall survival rate (%). In the figure on the right, the horizontal axis represents the time (month) after surgery, and the vertical axis represents disease-free survival (%).

  In the present specification, “specimen” means a specimen derived from a plurality of lung adenocarcinoma patients used for preparing an arithmetic expression and a specimen derived from a subject to be used for prognosis estimation. The “specimen” is not particularly limited as long as it contains tumor tissue derived from lung adenocarcinoma, and includes excised specimens by surgery, bronchoscopic specimens, specimens by lung biopsy, and the like. The “specimen” may be derived from any mammal (eg, human, mouse, rat, monkey, dog, cat, cow, horse, pig, sheep, goat, rabbit, hamster, etc.), more preferably human. Mouse, rat, monkey, dog, cat, etc., particularly preferably human. The form of the specimen may vary depending on the expression analysis means used. For example, when using an immunohistochemical staining method, a tissue section thinly sectioned with a microtome or the like, or a whole-mounted tissue or individual may be used as a specimen, but preferably a tissue section is used. It is done. The tissue section may be a section in which frozen tissue is sectioned and embedded in paraffin or resin. In addition, surgically excised specimens, bronchoscopic specimens and lung biopsy specimens may be directly subjected to flow cytometry analysis, mass cytometry analysis, genetic testing, etc. without being frozen or embedded in paraffin.

  In the present specification, a “subject” is a lung adenocarcinoma patient whose prognosis is required to be estimated. The subject is usually an animal of the same species as the animal species from which the specimen used for the creation of the prognostic calculation formula according to the present invention is derived. Accordingly, the subject may be any mammal (eg, human, mouse, rat, monkey, dog, cat, cow, horse, pig, sheep, goat, rabbit, hamster, etc.), more preferably human, mouse Rats, monkeys, dogs, cats, etc., particularly preferably humans.

  In general, the prognosis of a cancer patient means prediction of the future state of cancer from a certain point of time, for example, at the time of cancer diagnosis or treatment (at the time of surgery, etc.). When the risk (possibility) of metastasis is high, the prognosis is poor, and when the risk (probability) of cancer progression, recurrence, or metastasis is not high, the prognosis is good. Although the prediction index is not specified, for example, overall survival (OS) is preferable. Here, the “total survival rate” generally means the proportion of the observed cases that can be confirmed to be alive after a specific period from the specific observation start time.

  PD-L1 (Programmed death-ligand 1) to be measured for expression in the present invention is encoded by the cluster of differentiation 274 (CD274) gene and is also known as CD274 or B7 homolog 1 (B7-H1). It is a transmembrane protein. PD-L1 is an immunosuppressive molecule on the tumor side, and is thought to play an important role in immunosuppression by binding to PD-1, which is a receptor expressed on activated T cells, B cells, and the like. . The amino acid sequence of PD-L1 is known for various animal species. For example, the sequence of human PD-L1 is available as RefSeq No. NP_001254635.

  Gal-9 (Galectin-9), whose expression is to be measured in the present invention, is a known molecule encoded by the LGALS9 gene, also known as LGALS9 or Galactoside-binding, soluble, 9, and activated T cells. Various activities such as apoptosis induction, eosinophil migration, cancer cell apoptosis induction and cell adhesion have been reported. Gal-9 is an immunosuppressive molecule on the tumor side and is considered to play an important role in immunosuppression as one of the ligands of Tim-3 which is a surface antigen such as Th1 cells. The amino acid sequence of Gal-9 is known for various animal species. For example, the sequence of human Gal-9 is available as RefSeq No. NP — 002299.2.

  XAGE1 (X Antigen Family, Member 1), whose expression is to be measured in the present invention, is a known cancer testis antigen encoded by the XAGE1 gene and also known as Cancer / Testis antigen 12.1 (CT12.1). . So far, five types of genes of XAGE1A-E have been identified, the related protein is GAGED2, and it is known that there are two types of isoforms, GAGED2a and GAGED2d. To date, four alternative splice variants of XAGE-1a, b, c, and d have been identified (Sato et al., Cancer Immunity, vol. 7, page 5 (2007)). XAGE1 whose expression is measured in the present invention may be any isoform and may be a protein encoded by any alternative splice variant, but is preferably GAGED2a (XAGE-1b protein). The amino acid sequence of XAGE1 is known for various animal species. For example, the sequence of human GAGED2a is available as RefSeq No. NM_001097594.2, and the sequence of human GAGED2d is available as RefSeq No. NM_001097596.2.

  In the present invention, the “T cell” is an immunocompetent cell as a host-side factor involved in tumor immunity, and means a CD4 positive T cell and a CD8 positive T cell. In addition, CD4 positive T cells and CD8 positive T cells may be combined to form CD3 positive T cells.

In the present invention, the detection of the expression of PD-L1, Gal-9, XAGE1, CD4 and CD8 (hereinafter collectively referred to as “factors of the present invention”) is measured for these proteins. Alternatively, it may be measured for transcripts of genes encoding these proteins.
In the case of targeting a transcript, RNA can be isolated from a biological sample according to a conventional method. General methods for extracting RNA are well known in the art and have been published in standard molecular biology textbooks such as Ausubel et al., Current Protocols of Molecular Biology, John Wiley and Sons (1997). It is disclosed. Specifically, RNA isolation can be performed according to the manufacturer's instructions using a purification kit, buffer set, and protease obtained from a manufacturer such as Qiagen.
The method for measuring the expression level of the marker gene for the transcript is not particularly limited, but Northern blotting and in situ hybridization (Parker & Barnes, Methods in Molecular Biology 106: 247-283 (1999)) RNase protection assay (Hod, Biotechniques 13: 852-854 (1992)); Reverse transcription polymerase chain reaction (RT-PCR) (Weis et al., Trends in Genetics 8: 263-264 (1992)); Real-time quantitative RT-PCR (Held et al., Genome Research 6: 986-994 (1996)); and microarray analysis methods. Microarray analysis methods can be performed with commercially available equipment, such as using Affymetrix GeneChip technology, Agilent Technologies microarray technology, or Incyte microarray technology, according to the manufacturer's instructions.

  The measurement means in the case of using a protein as a detection target is not particularly limited, but preferably an immunostaining method can be used. Specific means using the immunostaining method include immunohistochemical staining (IHC) method, flow cytometry analysis, mass cytometry analysis and the like. The antibody to be used is not particularly limited as long as it is an antibody capable of detecting each factor, such as a polyclonal antibody, a monoclonal antibody, a chimeric antibody, a single chain antibody, or a fragment generated by a Fab fragment or a Fab expression library. And a part of the antibody having antigen binding property. Commercially available antibodies can be utilized for each of the factors of the present invention. Moreover, the antibody of each factor can be prepared by a method known per se using each factor or a part thereof as an antigen.

  Immunohistochemical staining can be performed independently for each of the factors of the present invention. Each of the factors of the present invention can be immunostained according to a method known per se (for example, revised fourth edition Watanabe / Nakane enzyme antibody method, published by Interdisciplinary Planning Co., Ltd., edited by Hiroshi Nakura, Yoshiyuki Nagamura, Tsutsumi). Hiroshi, ISBN4-906514-73-X C304), Enzyme labeled antibody method, fluorescent antibody method and the like are preferably used.

  A preferred embodiment includes an immunostaining method using an enzyme antibody method from the viewpoints of sensitivity of immunostaining, detection means, and the like. In the enzyme antibody method, the factor of the present invention is detected by an antibody labeled with an enzyme. The labeling may be performed directly or indirectly using a secondary antibody. Various improvements have been made to increase the detection sensitivity, and improvements such as the ABC method using biotin-streptavidin and the tyramide method (TSA method, Tyramide Signal Amplification) using tyramide (tyramide) give extremely high sensitivity. realizable.

  As the labeling enzyme, peroxidase (HRP, horseradish peroxidase) and alkaline phosphatase (AP) are widely used, and can be preferably used in the present invention. Antibody reagents and coloring kits suitable for the enzyme antibody method are commercially available, and immunostaining can be performed according to the manufacturer's protocol attached to the reagents and kit.

  As a general protocol, cell samples that have been deparaffinized or fixed as necessary are prepared, and pretreatment (for example, depigmentation or removal of biological dyes, blocking treatment) or antigen activation (warm bath, microwave treatment, autoclaving) as necessary. Treatment, protease treatment, etc.) and reacted with the primary antibody (for example, 4 ° C. to room temperature for 0.5 to 24 hours). After completion of the reaction, the cell specimen is washed and reacted with the secondary antibody (for example, 4 ° C. to 4 ° C. After the reaction is complete, the cell specimen is washed for 5 to 120 minutes at room temperature. Next, a coloring substrate such as diaminobenzidine (DAB) is added to cause color development. For morphological observation with a microscope, counterstaining such as hematoxylin staining may be further performed.

  The immunostained sample is imaged using an optical microscope or fluorescence microscope (in the case of fluorescence staining) equipped with an imaging device, and stored in a computer as a digital image. The digital image may be two-dimensional data or three-dimensional data, but in the present invention, the two-dimensional data can be sufficiently analyzed. Image processing such as filtering may be performed before analysis.

  Image analysis of immunostained cells can be performed using image analysis software mounted on the microscope used and other commercially available software. One example of commercially available software is ScanScope (registered trademark, Aperio). Further, the image analysis can be performed by processing image data of all cells of each specimen. Classification of the immunostaining intensity of the factor of the present invention (described later) can follow standard judgment criteria by a pathologist of immunohistochemical staining.

On the other hand, when the expression of the factor of the present invention is detected by flow cytometry or mass cytometry, a cell suspension or cell suspension can be prepared from the collected lung adenocarcinoma tissue and used for analysis. Cell suspensions or cell suspensions can be prepared using techniques well known to those skilled in the art. For example, cells are dispersed by mechanical disruption or treatment with proteases such as trypsin, collagenase, elastase, and pronase and hyaluronidase. Cell suspensions or cell suspensions can be obtained. Flow cytometry analysis and mass cytometry analysis can be performed using a commercially available apparatus (for example, FACSCanto ^™ II flow cytometer, CyTOF2, etc.) according to the manufacturer's manual.

Hereinafter, the prognostic calculation formula creation method for lung adenocarcinoma of the present invention will be described in detail.
[Prognostic formula creation method]
One aspect of the prognostic formula creation method of the present invention is as follows.
For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9 and XAGE1,
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue Based on the second data binarized second score and the third data binarized based on the presence or absence of the expression of XAGE1, lung adenocarcinoma patients are divided into good prognosis group and bad prognosis group,
A prognostic formula creation method (hereinafter referred to as the present invention) characterized in that an arithmetic formula for dividing into a good prognosis group and a bad prognosis group is obtained by multivariate analysis using the first score, the second score and the third data as explanatory variables. Prognostic formula creation method 1).

  The number of the plurality of specimens is not particularly limited as long as the prognostic calculation formula according to the present invention can be created, but is, for example, 50 or more, preferably 100 or more.

Here, the first score is specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue. The method for specifying the first score is not particularly limited as long as it is performed in the same manner for each sample, but can be calculated by the following method as an example.
(Calculation method)
Based on the result of detection of PD-L1 expression according to the above-described method, the expression intensity (I: Intensity) of PD-L1 is evaluated based on the following criteria.
I: Not detected 0: Weak 1: Moderate 2: Remarkable 3
Further, the proportion of cells having intensity I in all the nucleated cells in the specimen is determined, and the distribution range (D: Distribution) is defined as follows.
D: 0% 0: 1 to 50% 1: 51 to 100% 2
The product (D × I) of the expression intensity I and the distribution range D is calculated and used as the IHC score.
The IHC score is obtained for the cell membrane and the cytoplasm, respectively, and if at least one of the cell membrane or cytoplasm has an IHC score of 3 or more, the specimen is positive.

The second score is specified based on the distribution range and expression intensity of Gal-9 in the tumor tissue. The method for specifying the second score is not particularly limited as long as it is performed in the same manner for each sample, but can be calculated by the following method as an example.
(Calculation method)
Based on the results of detection of Gal-9 expression according to the above-described method, Gal-9 expression intensity (I: Intensity) is evaluated based on the following criteria.
I: Not detected 0: Weak 1: Moderate 2: Remarkable 3
Further, the proportion of cells having intensity I in all the nucleated cells in the specimen is determined, and the distribution range (D: Distribution) is defined as follows.
D: 0% 0: 1 to 50% 1: 51 to 100% 2
The product (D × I) of the expression intensity I and the distribution range D is calculated and used as the IHC score.
The IHC score is obtained for the cell membrane and the cytoplasm, respectively, and if at least one of the cell membrane or cytoplasm has an IHC score of 3 or more, the specimen is positive.

The third data is obtained by binarizing as positive or negative based on the presence or absence of expression of XAGE1. The method for acquiring the third data is not particularly limited as long as it is performed in the same manner for each sample, but can be calculated by the following method as an example.
(Calculation method)
Based on the result of detecting the expression of XAGE1 according to the method described above, the presence or absence of XAGE1-positive cells is counted. Among all the nucleated cells analyzed, the binarized as positive when XAGE1 positive cells are 1% or more and negative when less than 1% is defined as third data.

Subsequently, based on each data obtained above, each lung adenocarcinoma patient is divided into a good prognosis group and a bad prognosis group, respectively. Prognosis can be determined using overall survival.
Specifically, as described in the examples below, for example,
(A1) A specimen in which the first data is positive, the second data is negative, and the third data is negative,
(A2) a specimen in which the first data is positive, the second data is positive, and the third data is negative; and
(A3) Classifying samples with negative first data, negative second data and negative third data into groups with good prognosis;
(B1) A specimen in which the first data is positive, the second data is negative, and the third data is positive,
(B2) a specimen in which the first data is negative, the second data is positive, and the third data is negative;
(B3) a specimen in which the first data is negative, the second data is positive, and the third data is positive;
(B4) a specimen in which the first data is positive, the second data is positive, and the third data is positive; and
(B5) Samples with negative first data, negative second data, and positive third data can be classified into groups with poor prognosis.

Subsequently, the first score, the second score, and the third data are used as explanatory variables, and by performing multivariate analysis, an arithmetic expression for dividing into a good prognosis group and a bad prognosis group, that is, a prognosis arithmetic expression (hereinafter, simply “ It may be referred to as an “arithmetic expression”.)
The method for multivariate analysis is not particularly limited, but discriminant analysis is preferably used. Discriminant analysis is a method of obtaining sample data extracted from two or more populations and examining which population the sample data belongs to when there is unknown sample data belonging to which population. In the invention, multivariate data consisting of numerical values related to PD-L1, Gal-9 and XAGE1 expression is used as explanatory variables. Such discriminant analysis techniques are known, and the most typical discriminant analysis includes discrimination by a linear discriminant function, but is not limited thereto. Moreover, it can carry out using commercially available statistical analysis software.
The arithmetic expression thus obtained can be expressed as follows.

The values of a ₀ , a ₁ , a ₂ , and a ₃ also vary depending on the number of specimens of lung adenocarcinoma patients used to create the arithmetic expression. As an example of the prognostic calculation formula created by the method of the present invention, the following calculation formulas created in examples described later can be cited.

  Since the coefficient of the first score which is an explanatory variable is positive, PD-L1 is a good prognostic factor, and the coefficient of the second score and the third data which are explanatory variables are negative, so Gal-9 and It turns out that XAGE1 is a poor prognosis factor.

Another aspect of the prognostic formula creation method of the present invention is as follows.
For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9, XAGE1, CD4 and CD8;
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue The second data binarized from the second score, the third data binarized based on the presence or absence of expression of XAGE1, and the fourth score specified based on the rate of infiltration of T cells Based on the 4 data, the lung adenocarcinoma patients are divided into a good prognosis group and a bad prognosis group, and a good prognosis group is obtained by multivariate analysis using the first score, the second score, the third data, and the fourth score as explanatory variables. 1. A prognostic formula creation method characterized by obtaining formulas that are classified into bad groups (hereinafter also referred to as prognostic formula creation method 2 of the present invention).
The first score and the first data binarized from it, the second score and the second data binarized from it, and the third data are the same as in the above “prognostic formula creation method 1 of the present invention”. get. Further, the number of specimens to be used and preferred embodiments thereof are also the same as in the above “prognostic formula creation method 1 of the present invention”.

The fourth score is identified based on the rate of T cell infiltration into the tumor tissue. The method for identifying the fourth score is not particularly limited as long as it is performed in the same manner for each sample, but can be calculated by the following method as an example.
(Calculation method)
The ratio of CD4 positive cells and the ratio of CD8 positive cells are respectively determined, and the sum of these ratios is defined as the ratio of T cell infiltration into the tumor tissue.
Specifically, based on the result of detecting the expression of CD4 according to the above-described method, the expression intensity of CD4 is classified as follows for each cell.
Expression intensity; not detected 0; weak 1: moderate 2: remarkable 3
Cells having an expression intensity of 2 or more are determined as CD4 positive cells.
For each of the plurality of specimens used, the proportion of the CD4 positive cells in all nucleated cells is determined.
Similarly, based on the result of detection of CD8 expression according to the above-described method, the expression intensity of CD8 is classified as follows for each cell.
Expression intensity; not detected 0; weak 1: moderate 2: remarkable 3
Cells with an expression intensity of 2 or more are determined as CD8 positive cells.
For each of the plurality of specimens used, the proportion of the CD8 positive cells in all nucleated cells is determined.
The sum of the ratio of CD4 positive cells and the ratio of CD8 positive cells obtained according to the above is defined as the fourth score. The results are plotted with the number of specimens on the horizontal axis and the fourth score on the vertical axis, and approximated by a cubic function. Fourth data obtained by binarizing the inflection point of the approximate curve as a cut-off value, using a sample with a positive cell frequency more than the cut-off value as a positive sample, and a sample with a positive cell frequency less than that as a negative sample And An example of the cutoff value is shown in an example (Table 4) described later.

Subsequently, based on each data obtained above, each lung adenocarcinoma patient is divided into a good prognosis group and a bad prognosis group, respectively. Prognosis can be determined using overall survival.
Specifically, as described in the examples below, for example,
(A1) A sample in which the first data is positive, the second data is negative, the third data is negative, and the fourth data is positive,
(A2) a specimen in which the first data is positive, the second data is positive, the third data is negative, and the fourth data is positive; and
(A3) Specimens in which the first data is negative, the second data is negative, the third data is negative, and the fourth data is positive are classified into a group having a good prognosis.
(B1) A sample in which the first data is positive, the second data is negative, the third data is negative, and the fourth data is negative,
(B2) A sample in which the first data is positive, the second data is positive, the third data is negative, and the fourth data is negative,
(B3) A sample in which the first data is negative, the second data is negative, the third data is negative, and the fourth data is negative,
(B4) Sample in which the first data is positive, the second data is negative, the third data is positive, and the fourth data is positive or negative,
(B5) A specimen in which the first data is negative, the second data is positive, the third data is negative, and the fourth data is positive or negative,
(B6) A specimen in which the first data is negative, the second data is positive, the third data is positive, and the fourth data is positive or negative,
(B7) a specimen in which the first data is positive, the second data is positive, the third data is positive, and the fourth data is positive or negative, and
(B8) Samples having negative first data, negative second data, positive third data, and positive or negative fourth data can be classified into groups with poor prognosis.

Subsequently, by using the first score, the second score, the third data, and the fourth score as explanatory variables and performing multivariate analysis, an arithmetic expression for dividing a good prognosis group and a bad prognosis group, that is, a prognosis arithmetic expression, Ask.
The method for multivariate analysis is not particularly limited, but discriminant analysis is preferably used. In the present invention, multivariate data consisting of numerical values related to PD-L1, Gal-9, and XAGE1 expression and the rate of T cell infiltration are used as explanatory variables. Such discriminant analysis techniques are known, and the most typical discriminant analysis includes discrimination by a linear discriminant function, but is not limited thereto. Moreover, it can carry out using commercially available statistical analysis software.
The arithmetic expression thus obtained can be expressed as follows.

The values of a ₀ , a ₁ , a ₂ , a _3, and a ₄ also vary depending on the number of specimens of lung adenocarcinoma patients used to create the arithmetic expression. As an example of the prognostic calculation formula created by the method of the present invention, the following calculation formulas created in examples described later can be cited.

  Since the coefficients of the first score and the fourth score, which are explanatory variables, are positive, infiltration of T-cells into PD-L1 and tumor tissue is a good prognostic factor, and the second score and third data as explanatory variables It can be seen that Gal-9 and XAGE1 are poor prognostic factors.

The present invention provides a method for estimating the prognosis of lung adenocarcinoma using the arithmetic expression created by the above-described prognostic arithmetic expression creating method of the present invention. Hereinafter, the prognosis estimation method for lung adenocarcinoma of the present invention will be described in detail.
[Prognosis estimation method]
One aspect of the prognosis estimation method for lung adenocarcinoma of the present invention is as follows.
The first and second scores and third data detected from the subject's sample are substituted into the arithmetic expression for the arithmetic expression obtained by using the prognostic calculation expression creating method 1 of the present invention to determine the prognosis of the subject. A prognosis estimation method for lung adenocarcinoma estimated by calculation (hereinafter also referred to as prognosis estimation method 1 of the present invention).
Here, the creation of the arithmetic expression is as described above.
The method for detecting the first and second scores and the third data from the specimen of the subject is not particularly limited as long as it is the same as the method performed when creating the arithmetic expression. The prognosis of the subject is estimated by substituting the first and second scores and the third data obtained in this way into previously obtained arithmetic expressions. When the value obtained by substituting into the arithmetic expression (hereinafter also referred to as prognostic score 1) is larger than 0, it is estimated that the prognosis is good, and when it is smaller than 0, it is estimated that the prognosis is poor.
Another aspect of the prognostic formula creation method of the present invention is as follows.
For the arithmetic expression obtained using the prognosis arithmetic expression creating method 2 of the present invention, the first and second scores, the third data, and the fourth score detected from the subject's sample are substituted into the arithmetic expression. A prognosis estimation method for lung adenocarcinoma in which the prognosis of a subject is estimated by calculation (hereinafter also referred to as prognosis estimation method 2 of the present invention).
Here, the creation of the arithmetic expression is as described above.
The method for detecting the first and second scores, the third data, and the fourth score from the specimen of the subject is not particularly limited as long as it is the same as the method performed when creating the arithmetic expression. The prognosis of the subject is estimated by substituting the first and second scores, the third data, and the fourth score obtained in this way into previously obtained arithmetic expressions. When the value (hereinafter also referred to as prognostic score 2) obtained by substituting into the arithmetic expression is larger than 0, it can be estimated that the prognosis is good, and when it is smaller than 0, it can be estimated that the prognosis is poor.
As the fourth score, the infiltration rate of CD3 positive T cells combined with CD4 positive T cells and CD8 positive T cells may be used. CD4 positive T cells and CD8 positive T cells determined without using a cut-off value Or the infiltration rate of CD3-positive T cells may be used as it is.

  The prognostic estimation method of lung adenocarcinoma of the present invention is a local progression including an early stage cancer determined to be stage IA in the conventional stage classification or an early stage cancer determined to be stage IA, IB, IIA, IIB, or IIIA. It is possible to identify patients with a poor postoperative prognosis from lung cancer patients in the stage (hereinafter referred to as “locally advanced stage”). Of course, the prognostic estimation method of lung adenocarcinoma of the present invention can also be used to estimate the prognosis of patients with stage IIIB or stage IV lung cancer. Conventionally, in stage IA, since surgery alone has a 5-year survival rate of around 90%, no post-operative additional treatment has been performed in Japan, and the prognosis is good in early or locally advanced lung cancer. Since it was difficult to discriminate, only the treatment corresponding to the stage according to each conventional judgment method was carried out, and the occurrence of patients with poor prognosis could not be prevented. The prognostic estimation method of the present invention is capable of discriminating a poor prognosis patient from a lung adenocarcinoma patient in an early stage or a locally advanced stage and determining a recurrence rate of an early lung adenocarcinoma patient, which is impossible with conventional staging. Therefore, it is possible to provide post-operative additional treatment that was not previously applied to the patient with poor prognosis, as well as opportunities for early application of multidisciplinary treatment and personalized medicine that were not previously possible. Can contribute to improving the poor prognosis of patients. Thus, since the prognosis estimation method of the present invention can provide a place for a new clinical trial for a group of patients with poor prognosis whose stage is a locally advanced stage, in particular, the treatment method can be applied. It is useful as a companion diagnostic method for selecting patients with poor prognosis of lung adenocarcinoma.

  Further, the prognostic estimation method of lung adenocarcinoma of the present invention, as described above, can estimate the poor prognosis of lung cancer that cannot be predicted from the conventional staging classification. Therefore, the treatment method determined based on the conventional staging classification Can be used as a method for determining whether or not the application is effective for the lung cancer, that is, as a method for determining the effect of a treatment method for lung cancer patients. As long as this method for determining the effect is a treatment method for lung adenocarcinoma, the lesion site of interest may be the primary lesion or the metastatic site, regardless of the method, and if the disease recurs after surgery However, in consideration of the nature of the discriminating factor used for obtaining the discriminant function, application to a therapeutic method that affects the immunosuppressive pathway is useful, and in particular, the expression of PD-L1, Gal- 9 can be advantageously used to determine the effect of a treatment method that affects one or more of the expression of 9, the expression of XAGE1, and the T cell infiltration. Furthermore, the prognostic estimation method of the present invention is useful as a method for predicting the recurrence rate of lung cancer after surgery, and is particularly useful as a method for predicting the recurrence rate of early lung adenocarcinoma. It is useful as a method of predicting the long-term recurrence rate exceeding year.

  The present invention will be described more specifically with reference to the following examples. However, these are merely examples and do not limit the present invention.

Test example 1
The expression of immunosuppressive molecules PD-L1 and Gal-9 in lung cancer cell lines was examined. The materials and methods used are shown below.
(material)
Flow cytometry antibodies used:
・ PD-L1 (clone 29E.2A3; BioLegend, San Diego, CA, USA)
・ Gal-9 (clone 9M1-3; BioLegend, San Diego, CA, USA)

measuring equipment:
Flow cytometry: FACS CantoII (BD)

Cell lines and lung cancer tissues used:
Lung adenocarcinoma (A549, PC-9, HLC-1, II-18, OU-LC-ON, OU-LC-SK)
Lung squamous cell carcinoma (Sq-1, EBC-1, RERF-LC-AI)

(Method)
Analysis of PD-L1 and Gal-9 expression in lung cancer cell lines by flow cytometry:
Cell surface
1. Lung cancer cell line is cultured in a flask with 10% FCS / RPMI for several days.
2. Remove cultured cells from flask with EDTA solution and wash with FACS buffer (1% FCS / 0.1% azide / PBS).
3. Use 1x10 ⁵ cells, use PD-L1 (clone 29E.2A3), Gal-9 (clone 9M1-3) and their respective isotype control antibodies, and determine the amount of antibody in the package insert (company data sheet). Use and stain on-ice for 20 minutes.
4. After washing with FACS buffer, measure with FACS CantoII.

Intracellular staining
1. Lung cancer cell line is cultured in a flask with 10% FCS / RPMI for several days.
2. Remove cultured cells from flask with EDTA solution and wash with FACS buffer (1% FCS / 0.1% azide / PBS).
3. Use 1 x 10 ⁵ cells and puncture the cells with the BD Cytofix / Cytoperm ^™ Fixation / Permeabilization Solution Kit.
4. Using PD-L1 (clone 29E.2A3), Gal-9 (clone 9M1-3) and their respective isotype control antibodies, use the antibody amount in the package insert (company data sheet), and on-ice 20 minutes staining.
5. After washing with FACS buffer, measure with FACS CantoII.

(result)
The results are shown in FIG.
As a result, it was confirmed that PD-L1 and Gal-9 molecules were strongly expressed in many lung cancer cell lines.

Test example 2
Further analysis was performed on the expression of PD-L1 and Gal-9 molecules in lung cancer cell lines.
Various lung adenocarcinoma cell lines (A549, PC-9, HLC-1, II-18, OU-LC-ON, OU-LC-SK) were treated with IFN-γ, and the cell surface and cells before and after treatment Changes in the expression of PD-L1 and Gal-9 were examined. Expression analysis was performed by flow cytometry in the same manner as in Test Example 1.
FIG. 2A shows the results of expression analysis in II-18 cells, and FIG. 2B shows changes in expression by IFN-γ treatment in the tested cell lines. As a result, PD-L1 expression on the cell surface was increased by IFN-γ treatment, but Gal-9 expression was not changed.

Next, the concentration of Gal-9 in the culture supernatant after the treatment with the molecular target drug was measured using the lung cancer cell line PC-9 which was successful with the molecular target drug and OU-LC-SK which was not successful. The molecular target drug treatment was performed by culturing the cells with 10 nM Afatinib for 96 hours.
As a result, it was revealed that Gal-9 was released into the supernatant when the tumor was disintegrated with the molecular target drug (FIG. 2C). A molecularly targeted drug releases Gal-9, a substance that strongly suppresses immunity, even if the tumor partially collapses. Therefore, this means that a combination therapy of a molecular target drug and an antibody drug targeting Gal-9 and Tim-3 is important.
In fact, the pleural effusion of cancer patients (similar to local cancer) has high Gal-9 concentration (Fig. 2D), and it was confirmed that there are patients whose Gal-9 is detected in the blood during the course of chemotherapy. (FIG. 2E; 3 out of 10 cases showed a difference of more than standard deviation).

Test example 3
Expression analysis of immunosuppressive molecules (PD-1, TIM-3, BTLA, LAG-3) on the surface of T cells (TIL) that infiltrate the cancer site was performed. Analysis was performed by flow cytometry in the same manner as described in Test Example 1 using the following antibodies.
・ PD-1 (Anti-CD279-PE / Cy7 clone EH12.2H7, Biolegend)
・ TIM-3 (Anti-Tim3-APC clone F38-2E2, eBioscience)
・ BTLA (Anti-CD272-PE clone MH26, Biolegend)
・ LAG-3 (Anti-CD223-FITC clone 17B4, Enzo)
As a result, in both CD4 positive cells and CD8 positive cells, PD-L1 receptor PD-1 and Gal-9 receptor Tim-3 are highly expressed (FIGS. 3A-B). It was shown that the PD-1 / PD-L1 and Tim-3 / Gal-9 pathways are important as immunosuppressive molecular pathways in lung adenocarcinoma.

Furthermore, the relationship between the expression of PD-1 and Tim-3 in cancer local cells and the degree of differentiation (CD27), proliferation ability (Ki-67), and apoptosis (Annexin V) was examined.
As a result, there were many reports that Tim-3 + PD-1 + T cells were exhausted so far, but CD27 was maintained (because it decreased when terminally differentiated), and Ki-67 was high. Annexin V was high (FIG. 3C). This reflects that T cell proliferation ability is enhanced and cell death is induced at a high frequency. It is important to maintain these activated cells for cancer treatment. So far, clinical trials with PD-1 and PD-L1 antibodies have been reported, and the main prevention of apoptosis is the PD-1 + Tim-3- cell population. 3+ cells are immune to apoptosis. Therefore, it is suggested that antibody drugs targeting Gal-9 and Tim-3 are important.

Next, CD8 + T cells specific for the cancer antigen XAGE1 were established. When the expression of PD-1 and TIM-3 on the cell surface was examined by flow cytometry, it was found that Tim-3 was highly expressed (FIG. 3D).
When Gal-9 was added to the T cells (1 × 10 ⁴ cells) at the concentration shown in the figure and cultured for 8 hours, the cells were apoptotic (FIG. 3E). In addition, when T cells (1 x 10 ⁴ cells) were used and Gal-9 (300 pg / ml) was added with anti-Gal-9 antibody (1 μg / ml) and cultured for 12 hours, apoptosis was observed. Inhibited (FIG. 3F).

Test example 4
The results of Test Example 1-3 showed that the PD-1 / PD-L1 and Tim-3 / Gal-9 pathways are important as immunosuppressive molecular pathways in lung adenocarcinoma. Therefore, we investigated the relationship between the presence of these molecules and the prognosis, including the XAGE1 antigen, which is known to have a prognosis related to immune response in lung adenocarcinoma.
(Materials and methods)
Lung cancer tissues used: 361 (Tissue Micro Array LK400-1, LK400-2: 194 lung adenocarcinoma, 112 squamous cell carcinoma, 12 small cell lung cancer, 11 other tissue types, metastatic lung tumor 32 cases)
Provided by Pathology Research Institute, Inc. (Toyama, Japan)

Immunostaining antibodies used:
・ Rabbit anti-PD-L1 Ab (Catalog number: 4059 ProSci incorporated Poway, CA, USA)
・ Mouse anti-human Galectin-9 mAb (clone 9M1-3 GalPharma Co., Ltd, Kagawa, Japan)
・ XAGE1 monoclonal antibody (USO 9-13: prepared by the inventors)

Immunostaining (XAGE1):
1. Deparaffinize 5 μm paraffin sections.
2. Antigen activation is performed in a microwave using 10 mmol / L citrate buffer (pH 6.0) using a pressure cooker for 10 minutes.
3. Inactivate endogenous peroxidase in 0.3% H ₂ O ₂ for 5 minutes.
4. Overnight at 4 ° C using 2 μg / mL of XAGE1 monoclonal antibody.
5. After washing, color was developed with 3,3'-diaminobenzidine containing H ₂ O ₂ following streptavidin-biotin complex (SimpleStain MAX-PO kit; Nichirei, Tokyo, Japan).

Other immunostaining (by Pathology Institute):
1. Deparaffinization operation
2. Antigen activation Antigen activation treatment solution KN9 (Code: KN-09001, Institute of Pathology) treated at 95 ° C for 40 minutes, cooled at room temperature for 20 minutes
3. Endogenous peroxidase treatment
10% room temperature with 3% hydrogen peroxide
4. Wash several times with distilled water, soak in KN Buffer (Code: KN-09002, Pathology Laboratory) for 5 minutes at room temperature to equilibrate
5. Primary antibody reaction (room temperature)
6. Wash several times with KN Buffer
7. Secondary antibody reaction (room temperature 30 minutes)
Envision + Dual Link HRP labeling / polymer reagent (Dako: K4061)
8. Wash several times with KN Buffer
9. DAB coloring [Dako DAB +] (room temperature 10 minutes)
10. Wash with distilled water
11. Hematoxylin (stained nucleus) Room temperature 3 minutes
12. Washing, dehydration, clearing, sealing

Scoring:
・ Immunostaining
XAGE1:
If the number of positive cells in the tissue is 1% or more, it is determined as positive.
PD-L1, Gal-9:
After performing immunostaining, the staining intensity I (Intensity) and range D (Distribution; how many cells of intensity I are present) of the cell membrane and cytoplasm are measured using 2 sections.
Intensity I: no expression 0, weak 1, moderate 2, marked 3
Range D: 0% 0, 1-50% 1, 51-100% 2
Immunostaining score IHC Score = D x I
For the cell membrane, the higher of the two sections is taken as the cell membrane score of the tissue. Similarly, in the cytoplasm, the higher of the two sections is taken as the cell membrane score of the tissue.

Positive test:
XAGE1:
If the number of positive cells in the tissue is 1% or more, it is determined as positive 1;
PD-L1, Gal-9:
A positive score of 3 or more for each of the cell membrane score and cytoplasm score obtained by immunostaining. If at least one of the cell membrane or cytoplasm is positive, the specimen is determined to be positive.

analysis:
For the survival analysis, IBM SPSS Statistics 19 for Windows (IBM, New York, NY) was used.

(result)
FIG. 4 shows immunostained images with anti-PD-L1 antibody, anti-Gal-9 antibody or anti-XAGE1 antibody. PD-L1, Gal-9 and XAGE1 expression was observed in lung cancer tissues.
The results of scoring and positive determination for PD-L1 and Gal-9 are shown in Tables 1 and 2 below.

  The results of positive determination for XAGE1 are shown in Table 3 below.

Furthermore, the result obtained about the relationship between the expression of PD-L1, Gal-9, and XAGE1 and prognosis is shown in FIG.
In lung cancer, PD-L1 and Gal-9 are important as immunosuppressive factors on the tumor side. Furthermore, in lung adenocarcinoma, the immunogenic XAGE1 expression and the expression pattern of PD-L1 and Gal-9 molecules are prognostic. I found it involved.

Test Example 5
Since the relationship between the tumor side factor and the prognosis was clarified by Test Example 4, the host side factor and the prognosis were examined next. That is, the relationship between the number of immunocompetent cells infiltrating in lung cancer and the prognosis was examined. Furthermore, we investigated the relationship between prognosis by combining XAGE1 expression, which is a factor on the tumor side, which is strongly associated with prognosis, and the expression patterns of PD-L1 and Gal-9 molecules.
(Materials and methods)
Lung cancer tissue used:
We used 120 cases of lung cancer tissue with prognostic and clinical information available.

Immunostaining:
Immunostaining of XAGE1, PD-L1, and Gal-9 was performed as described above in Test Example 4.
Immunostaining of host-side factors was performed in the same manner as PD-L1 and Gal-9 immunostaining using the following antibodies.
・ CD4 (abcam ab133616) dilution 1: 100 30 minutes reaction ・ CD8 (BIOSB BIOSB5173) dilution 1:50 30 minutes ・ Foxp3 (abcam ab20034) dilution 1: 100 30 minutes

Scoring:
Scoring of XAGE1, PD-L1, and Gal-9 expression was performed in the same manner as in Test Example 4.
CD4 and CD8 expression was measured as% by the Pathology Laboratory using the Scan scope system, and staining intensity 2+ or higher was determined as positive cells.

Positive test:
The positive determination of XAGE1, PD-L1, and Gal-9 expression was performed in the same manner as in Test Example 2.
The positive determination of CD4 and CD8 was measured by the Pathology Laboratory using the Scan scope system as%, staining intensity 2+ or higher was regarded as positive cells, and data from 120 cases of lung adenocarcinoma were used. An approximate curve (cubic function) of 120 cases of data was created, and the positive cell frequency above the inflection point was defined as a positive sample, and the positive cell frequency less than that was defined as a negative sample.

analysis:
For the survival analysis, IBM SPSS Statistics 19 for Windows (IBM, New York, NY) was used.

(result)
Criteria for binarization based on immunostained images of host-side factors (A), scoring and frequency of expression intensity (B), and plotting relationship between CD4 positive T cell ratio and number of specimens (C) Is shown in FIG. Based on the inflection points in the approximate curve, the following cutoff values were obtained for each parameter of T cells.

  Furthermore, the results shown in FIG. 7 were obtained for the relationship between PD-L1, Gal-9 or XAGE1 expression and T cell infiltration and prognosis. There was a strong correlation between PD-L1, Gal-9 or XAGE1 expression and tumor local infiltrating T cells and prognosis.

Example 1
Based on the above findings, an attempt was made to create an arithmetic expression that enables estimation of prognosis using tumor tissue specimens obtained from a plurality of lung cancer patients.
In the experiment, 120 lung adenocarcinoma specimens that were positively tested for XAGE, PD-L1, Gal-9 expression and T cell infiltration in Test Example 5 were used. The 120 specimens were classified into the following 8 groups.
a. PD-L1 positive, Gal-9 negative, XAGE negative (n = 21)
b. PD-L1 positive, Gal-9 negative, XAGE positive (n = 8)
c. PD-L1 negative, Gal-9 positive, XAGE negative (n = 11)
d. PD-L1 negative, Gal-9 positive, XAGE positive (n = 2)
e. PD-L1 positive, Gal-9 positive, XAGE negative (n = 19)
f. PD-L1 positive, Gal-9 positive, XAGE positive (n = 5)
g. PD-L1 negative, Gal-9 negative, XAGE negative (n = 36)
h. PD-L1 negative, Gal-9 negative, XAGE positive (n = 18)
The relationship between each group and prognosis is shown in FIG. 8A. It was found that the prognosis can be determined by combining the expression of three molecules that are factors on the tumor side. That is, it was found that PD-L1 expression is a good prognosis, and Gal-9 and XAGE1 expression are poor prognostic factors. The above groups a, e, and g are divided into groups with good prognosis (cluster A; n = 76), and groups b, c, d, f, and h are divided into groups with poor prognosis (cluster B; n = 44). It was also found that it can be done (FIG. 8B). Based on these, we performed discriminant analysis using PD-L1 total score (first score), Gal-9 total score (second score), and the presence of XAGE1 expression (third data) as explanatory variables, and the prognosis The calculation formula to divide into good group and bad group was obtained. Discriminant analysis was performed using IBM SPSS Statistics 19 for Windows (IBM, New York, NY). As a result, the following discriminant function was obtained.

Furthermore, in addition to the above-described three molecular expressions that are factors on the tumor side, in addition to the host cell factor T cell infiltration, an attempt was made to create an arithmetic expression that enables estimation of prognosis. The percentage of T cell infiltration was determined by the sum of% of CD4 positive cells and% of CD8 positive cells.
120 specimens were i) included in cluster A and positive for T cell infiltration (n = 52), ii) included in cluster A and negative for T cell infiltration (n = 24), iii) included in cluster B and The cells were classified into four groups: T cell infiltration positive (n = 32), iv) included in cluster B and T cell infiltration negative (n = 12). The relationship between each group and prognosis is shown in FIG. 8D. It can also be seen that the above group i) can be further divided into groups with good prognosis (cluster C; n = 52), and groups ii), iii) and iv) can be further divided into groups with poor prognosis (cluster D; n = 68). (FIG. 8E). Based on these, PD-L1 total score (first score), Gal-9 total score (second score), presence or absence of XAGE1 expression (third data), T cell infiltration rate (fourth score) Was used as an explanatory variable, and an arithmetic expression was obtained to divide into a good prognosis group and a bad prognosis group. Discriminant analysis was performed using IBM SPSS Statistics 19 for Windows (IBM, New York, NY). As a result, the following discriminant function capable of determining the prognosis more clearly was obtained.

  Table 5 shows the standardized coefficient value of each explanatory variable in the above-described cluster fractionation, and Table 6 shows the results of univariate and multivariate Cox regression analysis using various factors.

  The coefficients shown in Table 5 indicate the weighting of immune factors. Therefore, when PD-L1 is used as an index, in Cluster C and D, T cell infiltration is about 2 times better in prognosis, Gal-9 is about 1.5 times worse in prognosis, and XAGE1 expression is about 2 times worse in prognosis.

Example 2
<Test method>
Using tumor tissue specimens obtained from 52 early stage lung adenocarcinoma patients classified in stage IA among 120 lung adenocarcinoma patients available for prognostic information and clinical information used in Example 1 By the same method as in Example 1, even in the case of early cancer, as in Example 1, using PD-L1 expression, Gal-9 expression, XAGE1 expression and T cell infiltration as identification factors, prognosis It was confirmed whether or not it can be estimated.
FIG. 9 shows the relationship between the same class of clusters used in Example 1 and the prognosis.
<Result>
In the case of lung cancer at stage IA, cluster classification similar to that in Example 1 is possible, and it has been found that classification can be made when the prognosis is good or poor.
This result shows that by using the prognosis estimation method of the present invention, it is possible to identify a group of patients with a poor prognosis among early stage cancer patients of stage IA, which have been conventionally difficult to identify.
Furthermore, the prognostic estimation method of the present invention provides an opportunity for additional treatment or multidisciplinary treatment, which has been impossible in the past, for a group of patients with poor prognosis in stage IA lung cancer patients who are not subjected to additional treatment after surgery. It shows that it can be used as a companion diagnostic method for providing.

Example 3
<Test method>
The discriminant function obtained in Example 1 after surgery for lung cancer using the tumor tissue specimen obtained from lung adenocarcinoma patients for whom the prognostic information and clinical information of 120 people used in Example 1 are available [Equation 5] The prognostic estimation was performed using the results of the actual follow-up, all of them (Pathological Stage I-IIIA), 62 patients with early stage lung adenocarcinoma (Pathological Stage I) classified as stage I, and FIG. 10 shows a group of patients with locally advanced lung adenocarcinoma (Pathological Stage II-IIIA) in the stage. Further, FIG. 11 shows the results of estimating the prognosis of the 120 lung cancer patients using the discriminant function [Equation 6] obtained in Example 1 and actually tracking the prognosis. The follow-up period after surgery was 5 years, which is generally used as a standard period for determining whether there is no recurrence after lung cancer surgery.
<Result>
As is clear from FIG. 10, it was confirmed that the prognosis of any stage of lung adenocarcinoma can be estimated by using [Equation 5].
Further, comparing the survival rates of the patient groups of FIG. 10 and FIG. 11 analyzing the same patient group, the survival rate of the good prognosis group (Cluster A) of FIG. 10 is 55.8%, whereas the good prognosis group of FIG. In (Cluster C), the survival rate is as high as 62.0%, so it was confirmed that the prognosis can be estimated with higher accuracy by using [Equation 6] with host factors added as immune factors. It was.
These results are based on the prognostic estimation method using the arithmetic expression obtained based on [Equation 1] of the present invention, as well as patients with stage II-IIIA lung adenocarcinoma, and those that have been conventionally difficult to identify. It shows that patients with poor prognosis among patients with early stage I cancer can be identified. Furthermore, these results show that the prognosis estimation method using the arithmetic expression obtained based on [Equation 3] can estimate the prognosis of lung adenocarcinoma more accurately than using the arithmetic expression obtained based on [Equation 1]. Show.

Example 4
<Test method>
Monoclonal Rabbit anti-human PD-L1 antibody (clone SP-142 Spring Bioscience Corporation, Pleasanton, Calif.) Was used as an antibody for immunostaining PD-L1, diluted 100-fold, and monoclonal as an antibody for immunostaining T cells. Rabbit anti-human CD3 antibody (clone: SP-7, Nichirei Biosciences Inc., Tokyo, Japan) diluted 100-fold, without using cut-off, using T cell infiltration rate as the fourth score In the same manner as in Example 1, except for the university hospital that collected the data of the lung adenocarcinoma patient group used in Example 1, the staging classification I for which prognostic information and clinical information are available is available. For a tumor tissue specimen obtained from a patient with lung adenocarcinoma, a verification cohort study was performed to estimate the prognosis using [Equation 5] obtained in Example 1, and postoperative survival (Overall survival) and disease free survival (Disease-free survival) was followed. The results are shown in FIG. In recent years, the 10-year survival rate has been announced in addition to the 5-year survival rate used as a standard for normal lung cancer healing as the overall cancer cooperative survival rate published by the National Cancer Center (Adult Diseases) Center Council. According to this, it has been clarified that even in early lung adenocarcinoma, there is a death from the original disease even after 5 years from the operation. Therefore, the prognostic estimation method of the present invention is effective for prognosis over a long period of time. In this study, 10 years after surgery was set as the follow-up period to confirm the sex.
<Result>
As is clear from FIG. 12, it was confirmed that the prognosis of the lung adenocarcinoma patient group at other facilities can be estimated by using [Equation 5] as in the case of Example 3. In addition, it was confirmed that all patients with early lung cancer determined to have a good prognosis were ill even after 10 years from 5 years after surgery.
This result has shown that the prognostic estimation method of this invention has versatility. Furthermore, it has been clarified that the prognosis estimation method of the present invention is extremely useful as a method for predicting the recurrence rate over a long period of time in patients with early lung adenocarcinoma.

  The results of Examples 3 and 4 not only indicate that the prognostic estimation method of the present invention can be applied to lung cancer patients regardless of the stage or implementation facility, but usually additional treatment is performed after surgery. It is also very useful for discriminating patients with poor prognosis among lung cancer patients with no stage I stage, including postoperative treatment methods for those patients with poor prognosis and setting of inspection frequency necessary for early detection of recurrence. It is a useful tool for deciding treatment strategies. Furthermore, since the prognosis estimation method of the present invention can reliably discriminate poor prognosis patients in early lung adenocarcinoma patients, the prognosis improvement for patients who have been judged as poor prognosis in early lung cancer patients which has not been conventionally performed It is useful as a companion diagnostic method to improve the survival rate by providing additional treatment necessary for medical care and providing opportunities for multidisciplinary treatment or personalized medicine that were impossible in the past. .

  ADVANTAGE OF THE INVENTION According to this invention, the system which can determine the prognosis of lung adenocarcinoma is provided. In clinical practice, treatment is selected according to the clinical stage and pathological stage, but by using a prognosis determination system using immune factors as an index separately from those stages, clinical early cancer Even so, patients who require aggressive multidisciplinary treatment can be selected. Further, according to the present invention, the prognosis can be determined only by examining a minute tumor tissue such as a bronchoscopic specimen or a specimen obtained by CT-guided lung biopsy. Therefore, prognosis can be determined by collecting a minute clinical sample in a patient who cannot perform whole body search of the stage, and can contribute to the determination of the subsequent treatment policy.

  This application is filed in Japanese Patent Application No. 2015-096013 (filing date: May 8, 2015), Japanese Patent Application No. 2015-212393 (filing date: Oct. 28, 2015), and Japanese Patent Application No. (Application date: March 16, 2016), the contents of which are incorporated in full herein.

Claims (7)

For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9 and XAGE1 by immunostaining,
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue Based on the second data binarized second score and the third data binarized based on the presence or absence of the expression of XAGE1, lung adenocarcinoma patients are divided into good prognosis group and bad prognosis group,
A prognostic calculation formula creation method characterized in that an arithmetic expression for dividing a good prognosis group and a bad prognosis group is obtained by multivariate analysis using the first score, the second score, and the third data as explanatory variables. For each specimen of multiple lung adenocarcinoma patients,
Detecting the expression of PD-L1, Gal-9, XAGE1, CD4 and CD8 by immunostaining,
First data binarized from the first score specified based on the distribution range and expression intensity of tumor PD-L1 in the tumor tissue, and specified based on the distribution range and expression intensity of tumor Gal-9 in the tumor tissue The second data binarized from the second score, the third data binarized based on the presence or absence of expression of XAGE1, and the fourth score specified based on the rate of infiltration of T cells Based on the 4 data, the lung adenocarcinoma patients are divided into a good prognosis group and a bad prognosis group, and a good prognosis group is obtained by multivariate analysis using the first score, the second score, the third data, and the fourth score as explanatory variables. A method for creating a prognostic formula, characterized in that an arithmetic formula for dividing into bad groups is obtained.   A lung adenocarcinoma that estimates the prognosis of the subject by computation by substituting the first and second scores and the third data detected from the subject's specimen for the computation obtained by the method according to claim 1 Prognosis estimation method.   The first and second scores, the third data, and the fourth score detected from the subject's sample are substituted into the arithmetic expression obtained by the method according to claim 2, and the prognosis of the subject is estimated by calculation. Prognosis estimation method for lung adenocarcinoma.   A companion for selecting a lung cancer patient with a poor prognosis from early stage to locally advanced stage and performing a treatment necessary for improving the prognosis, characterized by using the prognosis estimation method for lung adenocarcinoma according to claim 3 or 4. Diagnosis method.   A method for determining the effect of a method for treating lung adenocarcinoma, wherein the method for estimating prognosis of lung adenocarcinoma according to claim 3 or 4 is used.   The prognosis estimation method for lung adenocarcinoma according to claim 3 or 4, wherein the test subject is an early lung cancer patient, and the prognosis estimation method for lung adenocarcinoma is for determining the postoperative recurrence rate in the patient.