Method and apparatus for predicting therapeutic response to immune checkpoint inhibitors

Abstract

The present invention relates to a method or apparatus for predicting the therapeutic response to immune checkpoint inhibitors, and according to one embodiment of the method, the therapeutic response of cancer patients to immune checkpoint inhibitors can be predicted more accurately, and the treatment method can be determined.

Description

Technical Field

[0001] Relates to a method and apparatus for predicting a therapeutic response to an immune checkpoint inhibitor.

Background Art

[0002] Tumor cells act on host immunity in several ways to avoid immune defense in the tumor microenvironment. This phenomenon is generally called "cancer immune escape". One of the most important components in such a system is the immune checkpoint co-signal mediated by the PD-1 (Programmed cell death protein 1) receptor and its ligand PD-L1 (Programmed death ligand).

[0003] In recent years, cancer immunotherapy, particularly immune checkpoint inhibitors, has been developed and used in cancer treatment as a therapy for activating the human immune system to kill cancer cells. The most widely used immune checkpoint inhibitors include inhibitors using monoclonal antibodies against PD-1 or PD-L1. These inhibitors suppress the binding of PD-1 on the surface of T cells and PD-L1 on the surface of cancer cells, and remove cancer cells by activating immune cells such as T cells. Immune checkpoint inhibitors such as monoclonal antibodies targeting PD-1 or PD-L1 using such a mechanism of action are regarded as a new standard treatment for various cancer treatments.

[0004] Inhibitors using monoclonal antibodies against PD-1 or PD-L1 are effective only against cancer cells that express PD-L1. However, even among cancers that express PD-L1, it has been reported that more than 50% of cancer patients do not respond to inhibitors against PD-1 or PD-L1, and the reason for this remains unknown. Therefore, in order to effectively treat cancer with immune checkpoint inhibitors targeting PD-1 or PD-L1, there is a growing need to select patients who show a high therapeutic response to these immune checkpoint inhibitors.

[0005] Therefore, the inventors have developed a method for predicting the therapeutic response to immune checkpoint inhibitors using a computing device. [Overview of the project] [Problems that the invention aims to solve]

[0006] The objective is to provide a method and apparatus for predicting the therapeutic response to immune checkpoint inhibitors. Furthermore, it is also to provide a computer-readable recording medium storing a program for executing the above method on a computer. The technical problems to be solved are not limited to those described above, and other technical problems may exist. [Means for solving the problem]

[0007] A method for predicting a therapeutic response to an immune checkpoint inhibitor according to one embodiment includes the steps of: determining the density of tumor-infiltrating lymphocytes (TILs) in a pathological image at a first time point using a machine learning model; determining the density of tumor-infiltrating lymphocytes in a pathological image at a second time point using the machine learning model; determining the fold change of tumor-infiltrating lymphocytes using the densities of tumor-infiltrating lymphocytes in the pathological image at the first time point and the fold change of tumor-infiltrating lymphocytes in the pathological image at the second time point; and predicting a therapeutic response to an immune checkpoint inhibitor in a cancer patient based on the fold change of tumor-infiltrating lymphocytes.

[0008] A method for providing information to predict a therapeutic response to an immune checkpoint inhibitor in another aspect includes the steps of: determining the density of tumor-infiltrating lymphocytes (TIRs) in a pathological image at a first time point using a machine learning model; determining the density of tumor-infiltrating lymphocytes in a pathological image at a second time point using the machine learning model; determining the fold change of tumor-infiltrating lymphocytes using the densities of tumor-infiltrating lymphocytes in the pathological image at the first time point and the fold change of tumor-infiltrating lymphocytes in the pathological image at the second time point; and predicting a cancer patient's therapeutic response to an immune checkpoint inhibitor based on the fold change of tumor-infiltrating lymphocytes.

[0009] Another embodiment of a computer-readable recording medium includes a recording medium that stores a program for causing a computer to perform the above-described method.

[0010] A computing device in yet another embodiment includes at least one memory and at least one processor connected to the memory and configured to execute at least one computer-readable program contained in the memory, wherein the at least one program uses a machine learning model to determine the density of tumor-infiltrating lymphocytes (TILs) in a pathological image at a first time point, uses the machine learning model to determine the density of tumor-infiltrating lymphocytes in a pathological image at a second time point, uses the density of tumor-infiltrating lymphocytes in the pathological image at the first time point and the density of tumor-infiltrating lymphocytes in the pathological image at the second time point to determine the fold change of tumor-infiltrating lymphocytes, and includes instructions for predicting the therapeutic response of a cancer patient to an immune checkpoint inhibitor based on the fold change of tumor-infiltrating lymphocytes.

[0011] Another aspect involves using a machine learning model to determine the density of tumor-infiltrating lymphocytes (TILs) in a pathological image at a first time point.

[0012] The present invention provides a method for treating cancer, comprising the steps of: determining the density of tumor-infiltrating lymphocytes in a pathological image at a second time point using the machine learning model; determining the rate of change (fold change) of tumor-infiltrating lymphocytes using the density of tumor-infiltrating lymphocytes in the pathological image at a first time point and the density of tumor-infiltrating lymphocytes in the pathological image at a second time point; predicting the treatment response of a cancer patient to an immune checkpoint inhibitor based on the rate of change of tumor-infiltrating lymphocytes; and, if it is determined that the cancer patient will respond to the immune checkpoint inhibitor based on the treatment response prediction results, administering the immune checkpoint inhibitor to the patient. [Effects of the Invention]

[0013] Using a method and apparatus for predicting the therapeutic response to immune checkpoint inhibitors according to one embodiment, it is possible to more accurately predict the therapeutic response of cancer patients to immune checkpoint inhibitors and determine the appropriate treatment method. [Brief explanation of the drawing]

[0014] [Figure 1] This figure illustrates an example of a computing system that generates analysis results for pathological images according to one embodiment. [Figure 2] This is a diagram illustrating an example of a computing system for predicting therapeutic responses to immune checkpoint inhibitors according to one embodiment. [Figure 3] This figure illustrates an example of a processor according to one embodiment analyzing pathological images. [Figure 4] This flowchart illustrates an example of a method for predicting the therapeutic response to an immune checkpoint inhibitor according to one embodiment. [Figure 5] This schematic diagram illustrates the process of analyzing tumor cells collected during surgery before chemoradiotherapy (pre-CRT, B1), after chemoradiotherapy (CRT) / before nivolumab administration (B2), after three doses of nivolumab (B3), and after five doses of nivolumab in patients with locally advanced rectal cancer (LARC), using an AI-based pathology slide image analyzer. [Figure 6] This graph shows the density and fold change of intratumoral tumor-infiltration lymphocytes (iTILs) measured using an AI-based pathology slide image analyzer before and after chemoradiotherapy / nivolumab administration. [Figure 7]This graph shows the density of stromal tumor-infiltrating lymphocytes (sTILs) and the rate of change in density (fold change) measured using an AI-based pathology slide image analyzer before and after chemoradiotherapy / nivolumab administration. [Modes for carrying out the invention]

[0015] The terminology used in the embodiments has been selected to the greatest extent possible from currently widely used and common terms, although this may change depending on the intent of the articulators working in the art, case law, the emergence of new technologies, etc. In some cases, the applicant has arbitrarily selected terms, in which case their meaning will be described in detail in the relevant explanatory section. Therefore, the terms used in the specification should not be merely names of terms, but should be defined based on the meaning of those terms and the overall content of the specification.

[0016] Throughout this specification, when a part is said to "include" a component, this means, unless otherwise stated, that it may include other components rather than excluding them. Furthermore, terms such as "~unit," "~module," and "~part" as used herein mean a unit that performs at least one function or operation, which may be implemented in hardware or software, or in a combination of hardware and software.

[0017] According to one embodiment of the present specification, a "module" or a "unit" can be implemented by a processor and a memory. The "processor" should be broadly interpreted to include a general-purpose processor, a central processing unit (CPU), a GPU (Graphic Processing Unit), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, and the like. In some environments, the "processor" can also refer to an on-demand semiconductor (ASIC), a programmable logic device (PLD), a field programmable gate array (FPGA), and the like.

[0018] In the present specification, "each of a plurality of As" or "each of multiple As" can refer to each of all the components included in the plurality of As, or can refer to each of some of the components included in the plurality of As. For example, each of a plurality of tumor cells can refer to each of all the tumor cells included in the plurality of tumor cells, or can refer to each of some of the tumor cells included in the plurality of tumor cells. Similarly, each of a plurality of immune cells can refer to each of all the immune cells included in the plurality of immune cells, or can refer to each of some of the immune cells included in the plurality of immune cells.

[0019] In the present specification, "similar" can include all meanings of being the same or similar. For example, two pieces of information being similar can mean that the two pieces of information are the same or similar to each other.

[0020] In the present specification, an "instruction" can refer to a component of a computer program as a series of instructions grouped based on a function and to be executed by a processor.

[0021] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, the embodiments can be implemented in various different forms and are not limited to the examples described in this specification.

[0022] FIG. 1 is a diagram for explaining an example of a computing system 10 that generates analysis results of a pathological image 20 according to an embodiment.

[0023] Referring to FIG. 1, the computing system 10 can receive the pathological image 20 and generate analysis results 30 for the pathological image 20. Here, the analysis results 30 and / or medical information generated based on the analysis results 30 can be used to predict the treatment response of a cancer patient to an immune checkpoint inhibitor.

[0024] In FIG. 1, the computing system 10 is shown as one computing device, but is not limited thereto, and the computing system 10 can be configured to perform distributed processing of information and / or data via a plurality of computing devices. Further, although FIG. 1 does not show a storage system communicable with the computing system 10, the computing system 10 may be configured to be connected or communicable with one or more storage systems. The computing system 10 can be any computing device used to generate the analysis results 30 for the pathological image 20. Here, the computing device can refer to any type of device having a computing function, and for example, may be a notebook, a desktop, a laptop, a server, a cloud system, etc., but is not limited thereto.

[0025] A storage system configured to communicate with the computing system 10 may be a device or cloud system for storing and managing various data related to pathological image analysis tasks. To efficiently manage the data, the storage system can store and manage various data using a database. Here, various data may include any data related to pathological image analysis. For example, various data may include pathological images 20 and histological information (histological components) regarding the type, location, and state of cells, tissues, and / or structures contained in the pathological images 20. Furthermore, various data include clinical information such as patient age, menopausal status, clinical T stage (Clinical_T), breast imaging reporting and data system (BIRADS), number of tumors, tumor size, lymph node enlargement (e.g., Node_Enlargement), results of biopsy to confirm the status of estrogen receptors in cancer tissue (e.g., Biopsy_ER), results of biopsy to confirm the status of progesterone receptors in cancer tissue (e.g., Biopsy_PR), results of biopsy to determine whether the tumor is HER2-positive (e.g., Biopsy_HER2), results of evaluation of whether the tumor was completely removed by surgery after a given cancer treatment (e.g., pCR_final), pathology type, homologous recombination deficiency (HRD), etc. It may further include Factors.

[0026] The computing system 10 can receive pathological images 20 obtained from the human tissue of a patient who is a target for predicting the therapeutic response to an immune checkpoint inhibitor. Such pathological images 20 can be received via a communicable storage medium (e.g., a hospital system, a local / cloud storage system, etc.). The computing system 10 can analyze the received pathological images 20 to generate analysis results 30 for the pathological images 20. Here, the pathological images 20 may include histological components relating to at least one patch contained in the image.

[0027] As used herein, the term "patch" can refer to a small region within a pathological image. For example, a patch may include a region corresponding to a semantic object extracted by segmenting a pathological image. Alternatively, a patch may refer to a combination of pixels associated with histological information generated by analyzing a pathological image.

[0028] The term "histological components" as used herein may include characteristics or information relating to cells, tissues, and / or structures within the human body contained in a pathological image. Here, characteristics relating to cells may include cytologic features such as the nucleus and cell membrane. The histological components may refer to histological information relating to at least one patch contained in a pathological image inferred through a machine learning model. On the other hand, histological components can be obtained as a result of an annotation task performed by an annotator. The term "annotation" means the task of tagging histological components to a data sample or the tagged information (i.e., annotations) itself. The term annotation may be used interchangeably with terms such as tagging and labeling in the art.

[0029] According to one embodiment, the computing system 10 can extract histological information, which is a characteristic of the cells, tissues, and / or structures within the human body of a target patient, by analyzing the pathological image 20. Specifically, the computing system 10 can extract histological information about at least one patch contained in the pathological image 20 by analyzing (e.g., inference) the pathological image 20 using a machine learning model. For example, the histological information may include, but is not limited to, information about the cells within the patch (e.g., tumor cells, lymphocytes, macrophages, dendritic cells, fibroblasts, endothelial cells, etc.) (e.g., the number of specific cells, information about the tissue in which the specific cells are located). Here, tumor cells can refer to cells that continue to proliferate excessively, disregarding the cell growth cycle, and malignant tumor cells that penetrate (invade) surrounding tissues and spread and grow (metastasize) to distant tissues can be called cancer cells.

[0030] According to one embodiment, the computing system 10 can detect the expression of biomarkers in cells contained in a pathological image 20, or extract information about biomarker expression from the pathological image 20. For example, the pathological image 20 may include an H&E-stained image. Furthermore, the computing system 10 can extract information about the expression levels of biomarkers of interest from the stained pathological image 20. For example, the computing system 10 can derive the density of tumor-infiltrating lymphocytes (TILs) from the pathological image. The computing system 10 can also use the density of tumor-infiltrating lymphocytes to calculate the fold change of tumor-infiltrating lymphocytes and predict the therapeutic response of cancer patients to immune checkpoint inhibitors based on the fold change of tumor-infiltrating lymphocytes.

[0031] Other examples of histological information extracted from a patch by the computing system 10 include information about the tissue within the patch (e.g., cancer area, cancer epithelial, cancer stroma, normal epithelial, normal stroma, necrosis, fat, background, etc.). Further examples of information extracted from a patch by the computing system 10 include, but are not limited to, tumor-infiltrating lymphocyte (TIL) density, macrophage (MP) density, fibroblast (FB) density, endothelial cell (EC) density, microsatellite instability (MSI status), tumor mutational burden (TMB), tumor proportion score (TPS), and combined positive score (CPS). The information extracted from the patch is not limited to the examples above and may include any histological information that can be quantified by the patch, such as cellular instability, cell cycle, and biological function.

[0032] As described above, the analysis results 30 generated by the computing system 10 and / or the medical information generated and output based on the analysis results 30 can be used to predict and / or output the results of treatment response to immune checkpoint inhibitors. Furthermore, by utilizing patient clinical information related to pathology slides received from an accessible external system as additional input data, it is possible to infer and / or output predicted results of treatment response to immune checkpoint inhibitors in cancer patients.

[0033] The following describes an example of how the computing system 10 analyzes pathological images, with reference to Figures 2 to 4.

[0034] Figure 2 is a diagram illustrating an example of a computing system for predicting the therapeutic response to an immune checkpoint inhibitor according to one embodiment.

[0035] The computing system 10 described above can correspond to a computing system 200. Referring to Figure 2, the computing system 200 includes a processor 210 and memory 220. For convenience of explanation, only the components relevant to the present invention are shown in Figure 2. Therefore, in addition to the components shown in Figure 2, other general-purpose components can be further included in the computing system 200. For example, the computing system 200 may include, but is not limited to, at least one of a server device and a cloud device. As another example, the computing system 200 may consist of one or more server devices. As yet another example, the computing system 200 may consist of one or more cloud devices. As yet another example, the computing system 200 can consist of a server device and a cloud device configured and operating together. It will also be apparent to those ordinary skill in the art related to the present invention that the processor 210 and memory 220 shown in Figure 2 can be implemented in separate devices.

[0036] The processor 210 can process computer program instructions by performing basic arithmetic, logical, and input / output operations. Instructions may be provided from memory 220 or external devices (e.g., a server). The processor 210 can also provide overall control over the operation of other components included in the computing system 200.

[0037] The processor 210 may be implemented as an array of multiple logic gates, or as a combination of a general-purpose microprocessor and memory storing a program that can be executed by this microprocessor. For example, the processor 210 may include a general-purpose processor, a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a controller, a microcontroller, a state machine, etc. In some environments, the processor 210 may also include an on-demand semiconductor (ASIC), a programmable logic device (PLD), a field-programmable gate array (FPGA), etc. For example, the processor 210 may also refer to a combination of processing devices such as a combination of a digital signal processor (DSP) and a microprocessor, a combination of multiple microprocessors, a combination of one or more microprocessors coupled with a digital signal processor (DSP) core, or any other combination of such configurations.

[0038] The processor 210 analyzes pathological images. The analysis of pathological images includes a process of dividing or detecting specific tissues and / or cells in various tissues and cells expressed on pathological slides and then converting them into information useful for making medical decisions. Here, the process of converting into information useful for making medical decisions may mean extracting features that can be used to classify the disease pathology of a patient, diagnose cancer, establish a cancer treatment plan, prescribe anticancer drugs, or predict the likelihood of developing cancer, and in particular may include predicting the therapeutic response to immune checkpoint inhibitors in cancer patients.

[0039] For example, the processor 210 can use at least one machine learning model to derive the density of tumor-infiltrating lymphocytes (TILs) from pathological images.

[0040] Tissue samples from cancer patients can be taken from them at any point in the treatment process, including chemoradiotherapy, and pathological images may be generated from such tissue samples. For example, tissue samples can be obtained from the patient at at least one of the following points in time: before chemoradiotherapy, after chemoradiotherapy / before administration of immune checkpoint inhibitors, after three doses of immune checkpoint inhibitors, or at the time of surgery after five doses of immune checkpoint inhibitors, and pathological images can be generated based on such tissue samples.

[0041] Furthermore, the processor 210 can use at least one machine learning model to calculate the fold change of tumor-infiltrating lymphocytes using the density of tumor-infiltrating lymphocytes.

[0042] As a result of the analysis, if the rate of change in the density of tumor-infiltrating lymphocytes is above a predetermined value, the pathology slide image can be classified as a case in which the cancer patient responds to an immune checkpoint inhibitor. Specifically, the response of the cancer patient to an immune checkpoint inhibitor may mean that the cancer patient has good drug sensitivity and treatment prognosis to the immune checkpoint inhibitor.

[0043] On the other hand, the processor 210 can derive the proportion of PD-L1-positive tumors from pathological images using at least one machine learning model. The proportion of PD-L1-positive tumors may be calculated as a TPS (tumor proportion score) or a CPS (combined positive score). The TPS can be defined as the number of PD-L1-positive tumor cells divided by the total number of tumor cells, and the CPS can be defined as the number of PD-L1-positive tumor cells and immune cells divided by the total number of tumor cells.

[0044] As a result of the analysis, if the proportion of PD-L1-positive tumors is above a predetermined value, the processor 210 can output a result classifying the pathology slide images as cases in which cancer patients have good drug sensitivity to immune checkpoint inhibitors and a good treatment prognosis.

[0045] Memory 220 may include a non-temporary, readable storage medium from any computer. For example, memory 220 may include a permanent mass storage device such as RAM (random access memory), ROM (read-only memory), a disk drive, an SSD (solid-state drive), or flash memory. Alternatively, a permanent mass storage device such as ROM, an SSD, flash memory, or a disk drive may be a separate, independent persistent storage device distinct from memory. Memory 220 may also store an operating system (OS) and at least one program code.

[0046] These software components can be loaded from a computer-readable storage medium separate from memory 220. Such a separate computer-readable storage medium may be a storage medium that can be directly connected to the computing system 200 and may include, for example, computer-readable storage media such as floppy drives, disks, tapes, DVD / CD-ROM drives, and memory cards.

[0047] On the other hand, although not shown in Figure 2, the computing system 200 may further include a display device. Alternatively, the computing system 200 may be connected to an independent display device by wired or wireless means, and data may be sent and received from each other. For example, pathological images, pathological slide images, analysis information of pathological slide images, medical information, and additional information based on medical information can be provided to the user via the display device.

[0048] Figure 3 illustrates an example of a processor analyzing pathological images according to one embodiment. The following describes an example of processor 210 analyzing pathological images with reference to Figure 3. All of the examples described below with reference to Figure 3 can be applied to the deriving of tumor-infiltrating lymphocyte (TIL) density, the calculation of tumor-infiltrating lymphocyte density fold change 330, and / or the percentage of PD-L1-positive tumors by processor 210.

[0049] Referring to Figure 3, the processor 210 can analyze the pathological image 310 using the machine learning model 320. For example, the analysis result of the pathological image 310 can be derived as the rate of change in the density of tumor-infiltrating lymphocytes 330.

[0050] Here, the rate of change in tumor-infiltrating lymphocyte density 330 may be the rate of change calculated by comparing the density measured from the first time point pathology slide obtained from the patient before chemoradiotherapy (CRT) with the density measured from the second time point pathology slide obtained from the patient after chemoradiotherapy (CRT).

[0051] For example, the processor 210 can use a machine learning model 320 to output the detection results in the form of layers representing the tissue in the pathological image 310. In this case, the machine learning model 320 can be trained to detect regions in the pathological image 310 that correspond to the tissue in the reference pathological slide images, using training data that includes multiple reference pathology slide images and multiple reference label information.

[0052] Furthermore, the processor 210 can classify multiple tissues represented in the pathological image 310. Specifically, the processor 210 can classify the pathological image 310 into at least one of the following: cancer area, cancer stroma area, necrosis area, and background area.

[0053] However, the examples of the processor 210 classifying at least some of the regions represented in the pathological image 310 are not limited to those described above. In other words, the processor 210 can classify at least one region represented in the pathological image 310 into multiple categories according to various criteria, not limited to the four types of regions described above (cancer region, cancer matrix region, necrotic region, and background region). At least one region represented in the pathological image 310 can be classified into multiple categories according to pre-set criteria or criteria set by the user.

[0054] The processor 210 can then analyze the pathological image 310 and classify the multiple cells represented in the pathological image 310.

[0055] First, the processor 210 analyzes the pathological image 310 to detect cells from the pathological image 310 and can output the detection results in the form of layers representing the cells.

[0056] The processor 210 can output detection results in the form of layers representing cells in the pathological image 310 using the machine learning model 320. In this case, the machine learning model 320 can be trained to detect the location and type of cells in the reference pathological slide images within the pathological image 310 using training data that includes multiple reference pathological slide images and multiple reference label information.

[0057] The processor 210 can then classify the multiple cells represented in the pathological image 310. For example, the processor 210 can classify the multiple cells expressed in the pathological image 310 into at least one of the following: tumor cells, lymphocytes, macrophages, fibroblasts, endothelial cells, and other cells.

[0058] However, the examples of how the processor 210 classifies the cells represented in the pathological image 310 are not limited to those described above. In other words, the processor 210 can classify the cells represented in the pathological image 310 into multiple categories according to various criteria, not limited to the six types of cells described above (i.e., tumor cells, lymphocytes, macrophages, fibroblasts, endothelial cells, and other cells). The cells represented in the pathological image 310 can be grouped into multiple categories according to pre-set criteria or user-defined criteria.

[0059] For example, the processor 210 can determine at least one of the following values ​​for a pathological image 310: the density of intra-tumoral tumal-infiltrating lymphocytes (iTILs) within the tumor, the density of sTILs within the matrix, the density of sTILs within the tumor microenvironment (TME), or the stromal TIL score.

[0060] In this specification, the density of tumor-infiltrating lymphocytes within a tumor refers to the amount of immune cells within the tumor. The density of tumor-infiltrating lymphocytes within the matrix refers to the density of tumor-infiltrating lymphocytes in the connective tissue or matrix surrounding the tumor. The density of tumor-infiltrating lymphocytes within the tumor microenvironment refers to the density of total tumor-infiltrating lymphocytes in the tumor microbial environment, indicating how well lymphocytes, which are immune cells, are distributed throughout the tumor tissue. In this specification, the density of tumor-infiltrating lymphocytes within the tumor microenvironment may be used interchangeably with tTIL (total TIL).

[0061] For example, the processor 210 can use a machine learning model to detect cancerous regions and cancerous matrix regions within pathological images. The density of tumor-infiltrating lymphocytes within the tumor microenvironment can be determined based on the area of ​​the tumor microenvironment, including the cancerous regions and cancerous matrix regions, and the number of tumor-infiltrating lymphocytes contained within the tumor microenvironment.

[0062] For example, the processor 210 can divide the pathological image 310 into grids of a predetermined size. The processor 210 can determine the immunophenotype (IP) of each grid. Examples of immunophenotypes, though not limited to these, may include immuno-inflamed, immuno-excluded, and immuno-desert phenotypes. The processor 210 can determine the immunophenotype of a grid as immuno-inflamed if the density of tumor-infiltrating lymphocytes in the cancerous area is above a first threshold. The processor 210 can determine the immunophenotype of a grid as immuno-excluded if the density of tumor-infiltrating lymphocytes in the cancerous area is below the first threshold, and the density of tumor-infiltrating lymphocytes in the cancer matrix area is above a second threshold. The processor 210 can determine the immunophenotype of a grid as an immune exclusion phenotype if the density of tumor-infiltrating lymphocytes in the cancer region is less than a first threshold and the density of tumor-infiltrating lymphocytes in the cancer matrix region is less than a second threshold.

[0063] For example, the processor 210 can determine a score for the immunoactive phenotype of the pathological image 310 based on the immunophenotype of each grid. Based on the score for the immunoactive phenotype, the processor 210 can output the responsiveness to an immune checkpoint inhibitor. If the score for the immunoactive phenotype is less than a preset cutoff value, the processor 210 can determine that the responsiveness to the immune checkpoint inhibitor is negative. If the score for the immunoactive phenotype is equal to or greater than the preset cutoff value, the processor 210 can determine that the responsiveness to the immune checkpoint inhibitor is positive.

[0064] As an example, the processor 210 can derive the percentage of PD-L1-positive tumors from pathological images at a specific time point. The processor 210 can derive the percentage of PD-L1-positive tumors as a TPS (tumor proportion score) or CPS (combined positive score) value. The TPS can be defined as the number of PD-L1-positive tumor cells divided by the total number of tumor cells, and the CPS can be defined as the number of PD-L1-positive tumor cells and immune cells divided by the total number of tumor cells. The processor 210 can derive the percentage of PD-L1-positive tumor cells out of all tumor cells from pathological images at a specific time point, and if the derived percentage is greater than or equal to a preset cutoff value, it can be determined that the response to an immune checkpoint inhibitor is positive.

[0065] Here, machine learning model 320 refers to a statistical learning algorithm implemented based on the structure of a biological neural network, or a structure that executes that algorithm.

[0066] For example, machine learning model 320 can represent a model that has problem-solving ability by having nodes, which are artificial neurons that form a network by synaptic connections, like a biological neural network, repeatedly adjust the weights of the synapses and learn to reduce the error between the correct output corresponding to a particular input and the inferred output. For example, machine learning model 320 can include any probabilistic model used in artificial intelligence learning methods such as deep learning, neural network models, etc.

[0067] For example, the machine learning model 320 can be implemented as a multilayer perceptron (MLP) consisting of multiple layers of nodes and connections between them. The machine learning model 320 according to this embodiment can be implemented using one of various artificial neural network model structures, including MLPs. For example, the machine learning model 320 can consist of an input layer that receives input signals or data from the outside, an output layer that outputs output signals or data corresponding to the input data, and at least one hidden layer located between the input layer and the output layer that receives signals from the input layer, extracts characteristics, and transmits them to the output layer. The output layer receives signals or data from the hidden layer and outputs them to the outside.

[0068] Therefore, the machine learning model 320 can learn to receive one or more pathological images 310 and extract information about one or more objects (e.g., cells, tissues, structures, etc.) contained in the pathological images 310 and / or biomarker expression information.

[0069] Figure 4 is a flowchart illustrating an example of a method for analyzing pathological images according to one embodiment. The method shown in Figure 4 consists of steps processed chronologically by the computing systems 10, 200, or processor 210 shown in Figures 1 and 2. Therefore, even if the details are omitted below, the above-mentioned details regarding the computing systems 10, 200, or processor 210 shown in Figures 1 and 2 can also be applied to the method shown in Figure 4.

[0070] Referring to Figure 4, in step 410, the processor 210 uses a machine learning model to determine the density of tumor-infiltrating lymphocytes (TILs) in the pathological image at a first time point; in step 420, it uses the same machine learning model to determine the density of tumor-infiltrating lymphocytes in the pathological image at a second time point; in step 430, it uses the densities of tumor-infiltrating lymphocytes in the pathological image at the first time point and the densities of tumor-infiltrating lymphocytes in the pathological image at the second time point to determine the fold change rate of tumor-infiltrating lymphocytes; and in step 440, it predicts the treatment response of cancer patients to immune checkpoint inhibitors based on the fold change rate of tumor-infiltrating lymphocytes. Depending on the fold change rate of tumor-infiltrating lymphocytes, the treatment response of cancer patients to immune checkpoint inhibitors may vary.

[0071] The method for predicting the therapeutic response to the aforementioned immune checkpoint inhibitors will be described in more detail below.

[0072] One embodiment provides a method for predicting a therapeutic response to an immune checkpoint inhibitor (ICI), comprising the steps of: determining the density of tumor-infiltrating lymphocytes (TIMs) in a pathological image at a first time point using a machine learning model; determining the density of tumor-infiltrating lymphocytes in a pathological image at a second time point; determining the fold change of tumor-infiltrating lymphocytes using the densities of tumor-infiltrating lymphocytes in the pathological image at the first time point and the fold change of tumor-infiltrating lymphocytes in the pathological image at the second time point; and predicting a therapeutic response to an immune checkpoint inhibitor in a cancer patient based on the fold change of tumor-infiltrating lymphocytes.

[0073] In one embodiment, a method for predicting a therapeutic response to an immune checkpoint inhibitor (ICI) may include the steps of: deriving the density of tumor-infiltrating lymphocytes (TILs) from pathological images; calculating the fold change of tumor-infiltrating lymphocytes using the density of TILs; and predicting a cancer patient's therapeutic response to an immune checkpoint inhibitor based on the fold change of tumor-infiltrating lymphocytes.

[0074] In this specification, the term "pathological image" refers to a scanned image of a pathological slide that has been fixed and stained through a series of chemical processing steps for microscopic observation of tissues or other materials removed from the human body.

[0075] The pathological image may include a whole slide image (WSI) containing high-resolution images of all slides, and may include at least one of H&E (Hematoxylin & Eosin) stained slides or immunohistochemistry (IHC) stained slides. The pathological image may also refer to a portion of a high-resolution whole slide image, for example, one or more patches. Furthermore, the pathological image may refer to a digital image obtained by scanning a pathology slide using a digital scanner, and may contain information about cells, tissues, and / or structures within the human body. On the other hand, the pathological image may include one or more patches, and histological information may be applied to one or more patches through annotation (e.g., tagging). In this specification, “pathological image” may be used interchangeably with “pathological image,” “pathology slide image,” “tissue slide image,” “whole slide image (WSI),” etc. Furthermore, in this specification, “pathological image” may also refer to “at least some regions included in a pathological image.”

[0076] As used herein, the term "tumor microenvironment (TME)" may refer to the environment consisting of tissues and cells surrounding a tumor. For example, the tumor microenvironment may include, but is not limited to, the tumor cells themselves, as well as the surrounding blood vessels, extracellular matrix (surrounding tissue), immune cells, and inflammatory mediators.

[0077] As used herein, the term "tumor-infiltrating lymphocyte (TIL)" refers to lymphocytes that are concentrated within or around tumor cells and possess specificity that allows them to attack tumors more effectively than peripheral blood lymphocytes (PBMCs) present in the blood. In one embodiment, the tumor-infiltrating lymphocyte may be intratumoral tumor-infiltrating lymphocyte (iTIL) and / or stromal tumor-infiltrating lymphocyte (sTIL). Specifically, the tumor-infiltrating lymphocyte may include at least one of intratumoral tumor-infiltrating lymphocyte (iTIL) or stromal tumor-infiltrating lymphocyte (sTIL).

[0078] In this specification, the term "tumor cell (TC)" may refer to a cell that continues to proliferate excessively, disregarding the cell growth cycle. In particular, malignant tumor cells that invade surrounding tissues and spread and grow in distant tissues (metastasize) can be called cancer cells.

[0079] As used herein, the term "immune checkpoint inhibitor" may mean a substance that inhibits the mechanism by which cancer cells prevent T cell activation. The immune checkpoint inhibitor can be any substance capable of suppressing the function of immune checkpoint proteins, and may be a protein, compound, natural product, DNA, RNA, peptide, etc. Preferably, it may be an antibody, more preferably a monoclonal antibody, and even more preferably a human antibody, humanized antibody, or chimeric antibody.

[0080] In one embodiment, the immune checkpoint inhibitor may include nivolumab. For example, the immune checkpoint inhibitor may consist of, but is not limited to, proteins, compounds, natural products, DNA, RNA, peptides, or combinations thereof that can suppress any of the functions of PD-L1, PD-1, CTLA4, PD-L2, LTF2, LAG3, A2aR, TIGIT, TIM-3, B7-H3, B7-H4, VISTA, CD47, BTLA, KIR, and IDO. Specifically, the immune checkpoint inhibitor may be any of the following selected from the group consisting of anti-PD-L1 antibody, anti-PD-1 antibody, anti-CTLA4 antibody, anti-PD-L2 antibody, LTF2 regulatory antibody, anti-LAG3 antibody, anti-A2aR antibody, anti-TIGIT antibody, anti-TIM-3 antibody, anti-B7-H3 antibody, anti-B7-H4 antibody, anti-VISTA antibody, anti-CD47 antibody, anti-BTLA antibody, anti-KIR antibody, anti-IDO antibody, and combinations thereof, but is not limited to these.

[0081] The term "treatment response in cancer patients" as used herein may include pathological complete response (pCR), responsiveness of cancer patients to immune checkpoint inhibitors, etc. For example, "responding to immune checkpoint inhibitors" may mean "high drug sensitivity to immune checkpoint inhibitors" or "a favorable prognosis with the administration of immune checkpoint inhibitors."

[0082] As used herein, the term "pathological complete response (pCR)" refers to a state in which a tumor mass is present but no cancer cells are found through postoperative tissue examination. For example, pathological complete response may mean, but is not limited to, the absence of invasive cancer in human tissue as a result of anti-cancer chemotherapy or radiotherapy. Specifically, pathological complete response can mean a state in which all or at least some of the tumor cells present in human tissue have been removed as a result of anti-cancer treatment, and generally, the better the pathological complete response (pCR), the longer the patient's survival period may be.

[0083] In this specification, the term "biomarker" can refer to a marker that can be measured from an objective standpoint, such as a normal or pathological state or the degree of response to a drug.

[0084] In this specification, the term “machine learning model” or “machine learning model” can mean a structure of a computer algorithm that learns from data to discover patterns, predict, or make decisions. Such machine learning models can generally learn on training data and then make predictions or classifications on new data. For example, such machine learning models may include any model used to infer an answer to a given input.

[0085] According to one embodiment, the machine learning model may include an artificial neural network model comprising an input layer, a plurality of hidden layers, and an output layer, where each layer may contain one or more nodes. For example, the machine learning model may be trained to infer histological information about a pathological image and / or at least one patch contained in the pathological image. In this case, histological information generated through annotation work can be used to train the machine learning model. As another example, the machine learning model may be trained to infer the responsiveness of a cancer patient to treatment based on interaction scores, characteristics of at least one of cells, tissues, or structures in the pathological image, and / or clinical information about the patient. Furthermore, the machine learning model may include weights associated with a plurality of nodes contained in the machine learning model. The weights may include any parameters associated with the machine learning model.

[0086] In this specification, machine learning models may refer to artificial neural network models, and artificial neural network models may refer to machine learning models. The machine learning models described herein may be models trained using various learning methods. For example, various learning methods such as supervised learning, unsupervised learning, and reinforcement learning may be used in this disclosure, but are not limited to these.

[0087] As used herein, the term “learning” may refer to any process of modifying the weights included in a machine learning model using at least one patch, interaction score, histological information, and / or clinical information. According to one embodiment, learning may refer to the process of modifying or updating the weights associated with a machine learning model using at least one patch and histological information through one or more forward propagations and backward propagations.

[0088] In one embodiment, a method for predicting a therapeutic response to an immune checkpoint inhibitor (ICI) may further include the step of detecting cancerous regions and cancer matrix regions in a pathological image using a machine learning model, wherein the density of tumor-infiltrating lymphocytes may be determined based on the area of ​​the tumor microenvironment including the cancerous regions and cancer matrix regions and the number of tumor-infiltrating lymphocytes contained in the tumor microenvironment.

[0089] In one embodiment, a method for predicting the therapeutic response to an immune checkpoint inhibitor (ICI) may include a first time point pathology image comprising a whole slide image (WSI) of the cancer patient taken before chemoradiotherapy (CRT), and a second time point pathology image comprising a second time point pathology image comprising a WSI of the cancer patient taken after chemoradiotherapy and before administration of the immune checkpoint inhibitor.

[0090] In one embodiment, the step of predicting the therapeutic response of the cancer patient to an immune checkpoint inhibitor may include determining that the cancer patient will respond to the immune checkpoint inhibitor if the rate of change in tumor-infiltrating lymphocyte density is 1.8 or higher. Specifically, a response from the cancer patient to an immune checkpoint inhibitor may mean that the cancer patient has good drug sensitivity and a good prognosis to the immune checkpoint inhibitor.

[0091] As used herein, the term "favorable treatment outcome" may mean a high survival rate for cancer patients in response to treatment and follow-up after cancer diagnosis. Specifically, it may mean the absence or reduction of invasive cancer in the patient's tissues, and a low or absent likelihood of tumor recurrence, in response to treatments such as chemotherapy, radiotherapy, administration of immune checkpoint inhibitors, and chemoradiotherapy.

[0092] In one embodiment, the procedure may further include the steps of deriving the percentage of PD-L1-positive tumor cells among all tumor cells from a pathological image at a first time point, and predicting the therapeutic response to an immune checkpoint inhibitor using the percentage of PD-L1-positive tumor cells.

[0093] In one embodiment, the percentage of PD-L1-positive tumor cells may be calculated as a TPS or CPS value.

[0094] In this specification, the term TPS (tumor proportion score) is a useful indicator for predicting the effect of PD-1 inhibitors and is defined as the number of PD-L1-positive tumor cells divided by the total number of tumor cells. The TPS can reflect the expression level and distribution of PD-L1.

[0095] The term "CPS" as used herein is a useful indicator for predicting the effectiveness of PD-L1 inhibitors and is defined as the number of PD-L1-positive tumor cells and immune cells divided by the total number of tumor cells. The CPS (combined positive score) may also take into account the number of PD-L1-positive immune cells.

[0096] In one embodiment, the predicting step may include determining that the patient will respond to an immune checkpoint inhibitor if the proportion of PD-L1-positive tumors among all tumor cells is higher than 1%. Specifically, a response from the cancer patient to an immune checkpoint inhibitor may mean that the patient has good drug sensitivity and treatment prognosis to the immune checkpoint inhibitor.

[0097] As used herein, the term "cancer" means a physiological condition in an animal characterized typically by abnormal or uncontrolled cell growth. Such cancer may be associated, for example, with metastasis, interference with normally functioning surrounding cells, release of cytokines or other secretory products at abnormal levels, suppression or increase of inflammatory or immunological responses, neoplasia, premalignant, malignancy, or invasion of surrounding or distant tissues or organs, such as lymph node invasion.

[0098] In one embodiment, the cancer may be a solid tumor. Specifically, the solid tumor may be any one selected from the group consisting of lung cancer, skin cancer, stomach cancer, gastrointestinal cancer, intestinal cancer, colorectal cancer, colon cancer, pancreatic cancer, liver cancer, thyroid cancer, uterine cancer, cervical cancer, ovarian cancer, testicular cancer, prostate cancer, breast cancer, and oral cancer, but is not limited to these.

[0099] In one embodiment, the cancer may have microsatellite stability (MSS) or microsatellite instability (MSI-High).

[0100] In this specification, the term "microsatellite" refers to a condition in which 1 to 10 short DNA sequences are repeatedly arranged throughout the entire gene, and may refer to a number of short DNA regions that tend to differ from one individual to another.

[0101] As used herein, the term "microsatellite instable (MSI)" may refer to a phenomenon in which mutations in microsatellites are not repaired by mismatch repair (MMR) protein gene defects, resulting in changes in the number of repeats and a length that differs from that of normal cells. This microsatellite instable is characterized by a high mutation rate and the production of frameshift peptide neoantigens, increasing the number of lymphocytes surrounding and infiltrating tumors, and creating a highly immunogenic environment.

[0102] As used herein, the term "mismatch repair (MMR)" may refer to the function of a particular protein in recognizing and repairing mismatches that occur due to errors in DNA replication.

[0103] In this specification, the term "mismatch repair deficient (dMMR)" may refer to a condition in which the mismatch repair (MMR) function is unable to operate properly, resulting in persistent errors in the replicated DNA base sequence. dMMR is caused by germline or somatic mutations that inactivate mismatch repair proteins (MMR proteins), or by promoter hypermethylation. dMMR can lead to numerous mutations, causing cancer, and can also contribute to increased tumor antigen expression and active immune responses in cancerous tumors. Therefore, dMMR can be used as a marker to predict the response to immunotherapy.

[0104] One embodiment provides a method for treating cancer, comprising the steps of: using a machine learning model to determine the density of tumor-infiltrating lymphocytes (TILs) in a pathological image at a first time point; using the machine learning model to determine the density of tumor-infiltrating lymphocytes in a pathological image at a second time point; using the density of tumor-infiltrating lymphocytes in the pathological image at the first time point and the density of tumor-infiltrating lymphocytes in the pathological image at the second time point to determine the fold change of tumor-infiltrating lymphocytes; predicting the treatment response of a cancer patient to an immune checkpoint inhibitor based on the fold change of tumor-infiltrating lymphocyte density; and, if it is determined that the cancer patient will respond to the immune checkpoint inhibitor based on the treatment response prediction results, administering the immune checkpoint inhibitor to the patient.

[0105] The invention will be described in more detail through the following examples. However, these examples are for illustrative purposes only, and the scope of the invention is not limited to these examples. [Examples]

[0106] Example 1. Development of an AI-powered pathology slide image analyzer.

[0107] Computing systems 10 and 200 can support an AI-based pathology slide image analyzer. The AI-based pathology slide image analyzer was used to AI-based analyze the treatment response to immune checkpoint inhibitors in solid cancers. The AI-based pathology slide image analyzer can perform tissue segmentation into at least one cancer area or cancer stroma. The AI-based pathology slide image analyzer can acquire information regarding cell detection in pathological images. Based on the acquired information, the AI-based pathology slide image analyzer can estimate the spatial location of TILs within the tumor microenvironment (TME).

[0108] The AI-based pathology slide image analyzer was trained and validated using the datasets shown in Tables 1 and 2 below.

[0109] [Table 1]

[0110] [Table 2]

[0111] Example 2. Analyte and Analytical Method

[0112] 2-1. Subject of Analysis

[0113] Using an AI-based pathology slide image analyzer, tumor samples from the VOLTAGE study, a multicenter phase 1 / 2 study evaluating the efficacy of chemoradiotherapy (CRT) followed by five doses of nivolumab and surgery in patients with locally advanced rectal cancer (LARC), were analyzed.

[0114] Specifically, a total of 86 H&E-stained whole slide images (WSIs) collected from all patients enrolled in the VOLTAGE study, both before chemoradiotherapy (CRT) and after chemoradiotherapy (CRT) / before nivolumab administration, were analyzed. This analysis quantified tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment (TME) using the AI-based pathology slide image analyzer of Example 1, based on 16,443 WSIs, including pathologist annotations.

[0115] 2-2. Analysis method

[0116] Tumor cells were collected from patients with locally advanced rectal cancer (LARC) before chemoradiotherapy (CRT) (B1), after chemoradiotherapy (CRT) / before nivolumab administration (B2), after three doses of nivolumab (B3), and after five doses of nivolumab during surgery. The tumor cells were analyzed using the AI-based pathology slide image analyzer of Example 1, and the analysis results are schematically shown in Figure 5.

[0117] Figure 5 is a schematic diagram illustrating the process of collecting tumor cells from patients with locally advanced rectal cancer before chemoradiotherapy (CRT) (B1), after chemoradiotherapy (CRT) / before nivolumab administration (B2), after three doses of nivolumab (B3), and after five doses of nivolumab, and analyzing these cells during surgery using an AI-based pathology slide image analyzer.

[0118] The AI-based pathology slide image analysis used a total of 86 whole H&E slide images (WSI) collected from all patients enrolled in the VOLTAGE study at the pre-chemoradiotherapy (CRT) (B1) and post-chemoradiotherapy (CRT) / pre-nivolumab administration (B2) time points.

[0119] Example 3. Preclinical Outcome Analysis of Chemoradiotherapy (CRT)

[0120] 3-1. Correlation between pathological complete response (pCR) rate and the proportion of PD-L1-positive tumors

[0121] The correlation between the pre-chemoradiotherapy (CRT) pathological complete response (pCR) rate and the proportion of PD-L1-positive tumors in the subjects analyzed in Example 2 was examined.

[0122] Specifically, we analyzed the correlation between PD-L1, which shows enhanced immune activity in both tumor cells and immune cells, and the pathological complete response (pCR) rate. To compare the degree of PD-L1 expression within tumors, we defined a tumor proportion score (TPS) and calculated it for patients with microsatellite stability (MMS) and high-microsatellite instability (MSI-H). The TPS is expressed by the following formula (1).

[0123] Number 1 TPS = Number of PD-L1-positive tumor cells / Total number of living tumor cells × 100

[0124] Table 3 shows the results of an analysis of the correlation between the pathological complete response (pCR) rate and the proportion of PD-L1-positive tumors in MMS patients and MSI-H patients.

[0125] [Table 3]

[0126] As shown in Table 3, when the proportion of PD-L1-positive tumors was 1% or greater (TPS, tumor proportion score) ≥ 1%, the pCR rate in MMS patients was confirmed to be 75% (6 / 8). On the other hand, when the proportion of PD-L1-positive tumors was less than 1%, the pCR rate in MMS patients was 17% (5 / 30), and the pCR rate in MSI-H patients was confirmed to be 60% (3 / 5).

[0127] This means that in MMS patients with a TPS of 1% or higher before chemoradiotherapy (CRT), drug sensitivity to immune checkpoint inhibitors and the prognosis for anticancer treatment are favorable.

[0128] 3-2. Correlation between pathological complete response (pCR) rate and CD8 / eTreg ratio

[0129] In the subjects analyzed in Example 2, the correlation between the pathological complete response (pCR) rate before chemoradiotherapy (CRT) and the CD8 / eTreg ratio in tumor-infiltrating lymphocytes (TILs) was analyzed.

[0130] Specifically, Table 4 shows the results of an analysis of the correlation between pathological complete response rate (pCR rate) and the CD8+ T cell / effector regulatory T cell ratio (CD8 / eTreg ratio) in patients with microsatellite stability (MMS) and patients with high microsatellite instability (MSI-H).

[0131] [Table 4]

[0132] As shown in Table 4, when the CD8 / eTreg ratio was 2.5 or higher, the pCR rate was 78% (7 / 9) in MMS patients and 67% (2 / 3) in MSI-H patients. On the other hand, when the CD8 / eTreg ratio was less than 2.5, the pCR rate was 13% (2 / 15) in MMS patients and 50% (1 / 2) in MSI-H patients.

[0133] This means that MMS patients with a CD8 / eTreg ratio of 2.5 or higher before chemoradiotherapy (CRT) have good drug sensitivity to immune checkpoint inhibitors and a favorable prognosis for anticancer treatment.

[0134] 3-3. Correlation between pathological complete response (pCR) rate, the proportion of PD-L1-positive tumors, and the CD8 / eTreg ratio in MMS patients.

[0135] In the subjects analyzed in Example 2, the correlation between the pathological complete response (pCR) rate before chemoradiotherapy (CRT) and the CD8+ T cell / effector regulatory T cell ratio (CD8 / eTreg ratio) in tumor-infiltrating lymphocytes (TILs) was analyzed.

[0136] Specifically, Table 5 shows the results of analyzing the pathological complete response (pCR) rates of tumor cell percentage score (TPS) in Example 3-1 and the CD8 / eTreg ratio in Example 3-2 in MMS patients.

[0137] [Table 5]

[0138] As shown in Table 5, when the proportion of PD-L1-positive tumors was 1% or more and the CD8 / eTreg ratio was 2.5 or higher, the pCR rate was 100% (5 / 5). On the other hand, when the proportion of PD-L1-positive tumors was less than 1% and the CD8 / eTreg ratio was 2.5 or higher, the pCR rate was 50% (2 / 4). When the proportion of PD-L1-positive tumors was 1% or more and the CD8 / eTreg ratio was less than 2.5, the pCR rate was less than 75% (6 / 8). When the proportion of PD-L1-positive tumors was less than 1% and the CD8 / eTreg ratio was less than 2.5, the pCR rate was 19% (3 / 16).

[0139] This means that if the percentage of PD-L1-positive tumors before chemoradiotherapy (CRT) is 1% or higher and the CD8 / eTreg ratio is 2.5 or higher, then MMS patients have good drug sensitivity to immune checkpoint inhibitors and a good prognosis for anticancer treatment.

[0140] 3-4. Correlation between pathological complete response (pCR) rate and changes in tumor-infiltrating lymphocyte (TIL) density within the tumor microenvironment (TME).

[0141] In the subjects analyzed in Example 2, the correlation between the pathological complete response (pCR) rate before chemoradiotherapy (CRT) and the change in tumor-infiltrating lymphocyte (TIL) density within the tumor microenvironment (TME) was analyzed.

[0142] Specifically, the tumor-infiltrating lymphocyte (TIL) density within the tumor microenvironment (TME) was measured before chemoradiotherapy (pre-CRT, B1) and after chemoradiotherapy / pre-nivolumab administration (post-CRT / pre-nivolumab, B2). The density changes were calculated, and the values ​​are shown in Table 6.

[0143] [Table 6]

[0144] As shown in Table 6, we confirmed that when the tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment (TME) after chemoradiotherapy / pre-nivolumab administration (post-CRT / pre-nivolumab, B2) was 1.8 times or more compared to the TIL density in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1), the pCR rate was 75% (6 / 8) in MMS patients and 33.3% (1 / 3) in MSI-H patients. On the other hand, when the tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment after chemoradiotherapy and before nivolumab administration (post-CRT / pre-nivolumab, B2) was less than 1.8 times that of the tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1), the pCR rate was 16.7% (5 / 30) in MMS patients and 100% (1 / 1) in MSI-H patients.

[0145] This means that if the tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment after chemoradiotherapy / before nivolumab administration (post-CRT / pre-nivolumab, B2) is 1.8 times or more compared to the tumor-infiltrating lymphocyte (TIL) density in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1), then drug sensitivity to immune checkpoint inhibitors and the prognosis for anticancer treatment are favorable in MMS patients.

[0146] 3-5. Pathological complete response (pCR) rate and univariate analysis of each biomarker

[0147] Table 7 summarizes the pathological complete response (pCR) rates by biomarker measured in Examples 3-1 to 3-4. Specifically, the pCR rate was calculated separately for MSI status, TMB status, TPS, CPS, tumor-infiltrating lymphocyte (TIL) density within the tumor microenvironment (TME), and changes in TIL density, and these results are shown in Table 7. TPS is defined by Equation 1 above, and CPS (combined positive score) is defined by Equation 2 below.

[0148] Math 2 CPS = Number of PD-L1-positive (tumor cells + lymphocytes + macrophages) / Total number of living tumor cells × 100

[0149] [Table 7]

[0150] As shown in Table 7, the pathological complete response (pCR) rates were 28.9% (11 / 38) and 60% (3 / 5) for MSS patients and MSI-H patients, respectively; 33.3% (7 / 21) and 60% (3 / 5) for patients with a TMB of less than 10 and 10 or greater, respectively; and 23.5% (8 / 34) and 66.7% (6 / 9) for patients with a TPS of less than 1% and 1% or greater, respectively.

[0151] Furthermore, the pathological complete response (pCR) rates were 27.8% (10 / 36) and 57.1% (4 / 7) when the CPS was less than 1 and when it was 1 or greater, respectively. The intratumoral tumor-infiltrating lymphocyte (iTIL) count before chemoradiotherapy (pre-CRT, B1) was 37.7 / mm³. 2 If less than 37.7 / mm 2In the cases described above, the percentages were 25% (6 / 24) and 42.1% (8 / 19), respectively. In MSS patients, the percentages were 16.7% (5 / 30) and 75.0% (6 / 8), respectively, when the density of tumor-infiltrating lymphocytes (tTIL on B2 / B1) in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1) and before chemoradiotherapy / nivolumab administration (post-CRT / pre-nivolumab, B2) was less than 1.8x and when it was 1.8x or greater. In MSS and MSI-H patients, the density of tumor-infiltrating lymphocytes (tTIL on B2 / B1) in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1) and before chemoradiotherapy / nivolumab administration (post-CRT / pre-nivolumab, B2) was 16.7% (5 / 30) and 75.0% (6 / 8), respectively. We confirmed that the percentages for B2 / B1) being less than 1.8x and 1.8x or greater were 19.4% (6 / 31) and 63.6% (7 / 11), respectively.

[0152] Example 4. Analysis of the rate of change (fold change) in tumor-infiltrating lymphocyte (TIL) density

[0153] 4-1. Analysis of the percentage change (fold change) in tumor-infiltrating lymphocyte (iTIL) density before and after chemoradiotherapy / nivolumab administration.

[0154] Using the AI-powered pathology slide image analyzer of Example 1, the fold change in tumor-infiltrating lymphocyte (iTIL) density within the tumor was analyzed before chemoradiotherapy and after chemoradiotherapy / nivolumab administration.

[0155] Specifically, iTIL density and the rate of change in iTIL density (fold change) were measured for all samples (N=43), samples showing pathological complete response (pCR) (n=14), and samples not showing pathological complete response (pCR) (n=29), and the results are shown in Figure 6.

[0156] Figure 6 shows graphs of intra-tumoral tumal-infiltrating lymphocyte (iTIL) density and fold change in density measured using an AI-based pathology slide image analyzer before chemoradiotherapy and after chemoradiotherapy / nivolumab administration. It can be seen that the fold change in intra-tumoral tumal-infiltrating lymphocyte (iTIL) density within the tumor tends to increase both before and after chemoradiotherapy / nivolumab administration, regardless of whether a pathological complete response (pCR) is achieved. However, the increase is greater in cases of pathological complete response.

[0157] 4-2. Analysis of the percentage change (fold change) in tumor-infiltrating lymphocyte (sTIL) density in the matrix before and after chemoradiotherapy / nivolumab administration.

[0158] The percentage change in tumor-infiltrating lymphocyte (sTIL) density within the matrix (fold change) was analyzed using the AI-powered pathology slide image analyzer of Example 1 before and after chemoradiotherapy / nivolumab administration.

[0159] Specifically, sTIL density and the rate of change in sTIL density (fold change) were measured for all samples (N=43), samples showing pathological complete response (pCR) (n=14), and samples not showing pathological complete response (pCR) (n=29), and the results are shown in Figure 7.

[0160] Figure 7 shows graphs of the density and fold change of tumor-infiltrating lymphocytes (sTILs) in the matrix, measured using an AI-based pathology slide image analyzer before and after chemoradiotherapy / nivolumab administration. This confirms that, unlike tumor-infiltrating lymphocytes within the tumor, tumor-infiltrating lymphocytes in the matrix do not show a significant increase even in samples that achieved pathological complete response (pCR), and rather show a decrease in samples that did not achieve pathological complete response.

[0161] Example 5. Correlation between pathological complete response (pCR) rate, tTIL density change rate, TPS, and CPS in MSS-LARC patients.

[0162] In MSS-LARC patients, we analyzed the correlation between the pathological complete response (pCR) rate, the rate of change in tumor-infiltrating lymphocyte (TIL) density within the tumor microenvironment (TME), and TPS and CPS.

[0163] Table 8 shows the pathological complete response rate based on the percentage change in tumor-infiltrating lymphocyte density in the tumor microenvironment before chemoradiotherapy (pre-CRT, B1) and after chemoradiotherapy / pre-nivolumab administration (post-CRT / pre-nivolumab, B2) in MSS-LARC patients, as well as the pathological complete response rate based on the MSI status before chemoradiotherapy (pre-CRT, B1), TPS value, and CPS value.

[0164] [Table 8]

[0165] Table 9, on the other hand, shows the pathological complete response rate measured considering the tumor-infiltrating lymphocyte density in the tumor microenvironment before chemoradiotherapy and the rate of change in tumor-infiltrating lymphocyte density in the tumor microenvironment after chemoradiotherapy / before nivolumab administration, as well as the TPS value before chemoradiotherapy, for MSS-LARC patients.

[0166] [Table 9]

[0167] As shown in Tables 8 and 9, among patients whose rate of change in tumor-infiltrating lymphocyte density within the tumor microenvironment increased by 1.8 times or more, the pathological complete response rate was 100% for patients with a TPS of 1% or higher, and 67% for patients with a TPS of less than 1%. On the other hand, among patients whose rate of change in tumor-infiltrating lymphocyte density within the tumor microenvironment was less than 1.8 times, the pathological complete response rate was 50% for patients with a TPS of 1% or higher, and 8% for patients with a TPS of less than 1%.

[0168] This means that among patients whose tumor-infiltrating lymphocyte density in the tumor microenvironment increased by 1.8 times or more between before chemoradiotherapy and after chemoradiotherapy / nivolumab administration, those with a TPS of 1% or higher had better drug sensitivity to immune checkpoint inhibitors and a better prognosis for anticancer treatment.

[0169] Example 6. Quantification and Statistical Analysis

[0170] The F1 score and IoU index were used to evaluate the performance of cell and tissue models, respectively. Agreement or Cohen's kappa value was used to assess the difference between pathologist interpretations and AI analyzer interpretations. The chi-squared test or Fisher's exact test was used for group comparisons of categorical variables. The Student's t-test or the nonparametric Mann-Whitney U test was used for group comparisons of continuations variables.

[0171] All statistical analyses were performed using Python (version 3.7) and R software (version 4.2.3) (R Foundation for Statistical Computing, Vienna, Austria).

Claims

1. A method for predicting the therapeutic response to an immune checkpoint inhibitor (ICI), A machine learning model is used to determine the density of tumor-infiltrating lymphocytes (TILs) in the pathological image at the first time point. Using the aforementioned machine learning model, the steps include determining the density of tumor-infiltrating lymphocytes in the pathological image at a second time point, A step of determining the fold change rate of tumor-infiltrating lymphocytes using the density of tumor-infiltrating lymphocytes in the pathological image at the first time point and the density of tumor-infiltrating lymphocytes in the pathological image at the second time point, and A method comprising the step of predicting the therapeutic response of a cancer patient to an immune checkpoint inhibitor based on the rate of change in the density of tumor-infiltrating lymphocytes.

2. The process further includes the step of detecting cancerous regions and cancerous matrix regions in pathological images using the aforementioned machine learning model, The method according to claim 1, wherein the density of tumor-infiltrating lymphocytes is determined based on the area of ​​the tumor microenvironment including the cancerous region and the cancer matrix region and the number of tumor-infiltrating lymphocytes contained in the tumor microenvironment.

3. The pathological images at the first time point include a whole slide image (WSI) of the cancer patient taken before chemoradiotherapy (CRT). The method according to claim 1, wherein the pathological image at the second time point includes a WSI of a cancer patient obtained after chemoradiotherapy and before administration of the immune checkpoint inhibitor.

4. The method according to claim 1, wherein the tumor-infiltrating lymphocytes include at least one of intratumoral tumor-infiltrating lymphocytes (iTILs) or stromal tumor-infiltrating lymphocytes (sTILs).

5. The method according to claim 1, wherein the immune checkpoint inhibitor comprises nivolumab.

6. The method according to claim 1, wherein the step of predicting the therapeutic response of the cancer patient to an immune checkpoint inhibitor includes determining that the cancer patient will respond to the immune checkpoint inhibitor if the rate of change in tumor-infiltrating lymphocyte density is 1.8 or higher.

7. A step of deriving the percentage of PD-L1-positive tumor cells among all tumor cells from the pathological image at the first time point, and The method according to claim 1, further comprising the step of predicting a therapeutic response to an immune checkpoint inhibitor using the proportion of PD-L1-positive tumor cells.

8. The method according to claim 7, wherein the proportion of PD-L1-positive tumor cells is calculated as a TPS (tumor proportion score) or CPS (combined positive score) value.

9. The method according to claim 7, wherein the step of predicting a therapeutic response to an immune checkpoint inhibitor using the percentage of PD-L1-positive tumor cells includes determining that the tumor will respond to an immune checkpoint inhibitor if the percentage of PD-L1-positive tumor cells among all tumor cells is greater than 1%.

10. The method according to claim 1, wherein the cancer is a solid tumor.

11. The method according to claim 10, wherein the cancer is microsatellite stable (MSS) or microsatellite unstable (MSI-High).

12. The method according to claim 10, wherein the solid tumor is one or more selected from the group consisting of lung cancer, skin cancer, stomach cancer, gastrointestinal cancer, intestinal cancer, colorectal cancer, colon cancer, rectal cancer, pancreatic cancer, liver cancer, thyroid cancer, uterine cancer, cervical cancer, ovarian cancer, testicular cancer, prostate cancer, breast cancer, and oral cancer.

13. A computer program stored on a computer-readable recording medium for performing the method according to claim 1 on a computer.

14. As a computing system, At least one memory, and It includes at least one processor connected to the memory and configured to execute at least one computer-readable program contained in the memory, The aforementioned at least one processor is Using a machine learning model, we determined the density of tumor-infiltrating lymphocytes (TILs) in the pathological image at the first time point. Using the aforementioned machine learning model, the density of tumor-infiltrating lymphocytes in the pathological image at the second time point is determined. The density change rate (fold change) of tumor-infiltrating lymphocytes is determined using the density of tumor-infiltrating lymphocytes in the pathological image at the first time point and the density of tumor-infiltrating lymphocytes in the pathological image at the second time point. A computing system that predicts the therapeutic response of cancer patients to immune checkpoint inhibitors based on the rate of change in the density of tumor-infiltrating lymphocytes.

15. The pathological images at the first time point include a whole slide image (WSI) of the cancer patient taken before chemoradiotherapy (CRT). The computing system according to claim 14, wherein the pathological images at the second time point include WSI of the cancer patient obtained after chemoradiotherapy and before administration of the immune checkpoint inhibitor.

16. The computing system according to claim 14, wherein at least one processor determines that the cancer patient will respond to an immune checkpoint inhibitor if the rate of change in the density of tumor-infiltrating lymphocytes is 1.8 or higher.

17. A machine learning model is used to determine the density of tumor-infiltrating lymphocytes (TILs) in the pathological image at the first time point. Using the aforementioned machine learning model, the steps include determining the density of tumor-infiltrating lymphocytes in the pathological image at a second time point, A step of determining the fold change rate of tumor-infiltrating lymphocytes using the density of tumor-infiltrating lymphocytes in the pathological image at the first time point and the density of tumor-infiltrating lymphocytes in the pathological image at the second time point. A step of predicting the therapeutic response of cancer patients to immune checkpoint inhibitors based on the rate of change in the density of tumor-infiltrating lymphocytes, and A method for treating cancer, comprising the step of administering an immune checkpoint inhibitor to a cancer patient if it is determined, based on the treatment response prediction results, that the cancer patient will respond to an immune checkpoint inhibitor.

Images