A scoring method for anti-B7H4 antibody drug conjugate therapy.

Abstract

The present invention relates to a method for predicting how a cancer patient will respond to antibody drug conjugate (ADC) therapy, comprising computing a predictive response score based on a single-cell ADC score for each cancer cell. For each cancer cell, a single-cell ADC score is computed based on the staining intensity of a dye in the membrane and / or cytoplasm of the cancer cell and the membrane and cytoplasm of adjacent cancer cells. The present invention also relates to predicting a cancer patient's response to ADC therapy by aggregating all single-cell ADC scores of a tissue sample using statistical operations, followed by treatment of the cancer with the relevant antibody drug conjugate.

Description

[Technical Field]

[0001] Cross-reference of related applications This application claims priority to U.S. Provisional Patent Application No. 63 / 077,607, filed on 12 September 2020. Each of the applications listed above is incorporated herein by reference as a whole for all purposes.

[0002] Integration by referencing electronically submitted documents The computer-readable nucleotide / amino acid sequence listing submitted concurrently with this specification is incorporated herein by reference in its entirety and is identified as follows: a single 14,351-byte ASCII (text) file named "B7H4-400-WO-PCT_seq-listing.txt" created on 2 September 2021.

[0003] The present invention relates to a method for computerizing a score indicating how cancer patients will respond to therapy using anti-B7H4 antibody drug conjugates. [Background technology]

[0004] Assessing a cancer patient's responsiveness to a given treatment is an essential step in determining a treatment regimen for that patient. Such assessments often involve histological analysis of tissue samples from cancer patients and include identifying and classifying cancers using standard grading schemes. Immunohistochemical (IHC) staining can be used to distinguish marker-positive cells expressing specific proteins from marker-negative cells that do not express proteins. IHC staining typically involves multiple dyes, including one or more dyes linked to protein-specific antibodies, and another dye that serves as a counterstain. A common counterstain is hematoxylin, which labels DNA and therefore stains the nucleus.

[0005] Using protein-specific staining or biomarkers, areas of cancer patient tissue that are likely to respond to a predetermined therapy can be identified. For example, biomarkers that stain epithelial cells can help identify suspicious tumor areas. Then, other protein-specific biomarkers are used to characterize cells within the cancerous tissue. Cells stained by a specific biomarker can be identified and quantified, and a score indicating the number of positively and negatively stained cells can then be visually estimated by the pathologist. This score can then be compared to scores calculated in the same way for other cancer patients. If the responses of these other patients to a given cancer treatment are known, the pathologist can predict how well the cancer patient will respond to a given treatment based on the comparison of the score calculated for the cancer patient with the scores of other patients. However, visual assessments by pathologists tend to be variable and subjective.

[0006] One promising cancer treatment involves an antibody-drug conjugate (ADC) containing a cytotoxic drug conjugated to an antibody, the antigen of which is expressed on the surface of cancer cells. The ADC binds to the antigen, selectively delivers the drug to cancer cells, and causes intracellular migration to accumulate the drug within those cancer cells and kill them. Computer-based methods are being explored to generate reproducible and objective scores indicating cancer patients' responses to treatments containing therapeutic B7-H4 antibody-drug conjugates. [Overview of the project] [Means for solving the problem]

[0007] A method for predicting how cancer patients will respond to therapies containing antibody-drug conjugates (ADCs) involves computer-processing a response score based on a single-cell ADC score for each cancer cell. An ADC comprises an ADC payload and an ADC antibody that targets a protein on each cancer cell. Tissue samples are immunohistochemically stained using a dye linked to a diagnostic antibody that binds to a protein on cancer cells in the tissue sample. Digital images of the tissue samples are acquired. Image analysis is performed on the digital images, and cancer cells are detected using a convolutional neural network. For each cancer cell, a single-cell ADC score is computer-processed based on the staining intensity of the dye in the membrane and / or cytoplasm of the cancer cell and / or in the membrane and cytoplasm of other cancer cells closer than a predetermined distance to that cancer cell. A response score predicting the cancer patient's response to ADC therapy is generated by aggregating all single-cell ADC scores from the tissue samples using statistical operations. Patients with a response score higher than a predetermined threshold are recommended for ADC-containing therapies.

[0008] In one embodiment, a method for generating a score indicating the viability of cancer patients treated with an antibody-drug conjugate (ADC) involves computer processing of single-cell ADC scores. Tissue samples from cancer patients are immunohistochemically stained using a dye linked to a diagnostic antibody. The ADC comprises an ADC payload and an ADC antibody that targets the B7-H4 protein on cancer cells. The diagnostic antibody binds to the B7-H4 protein on cancer cells in the tissue sample. A digital image of the tissue sample is obtained, and cancer cells in the digital image are detected using image analysis. For each cancer cell, a single-cell ADC score is computer processed based on the staining intensity of the dye in the membrane. The single-cell ADC score can also be optionally obtained based on the staining intensity of the dye in the cytoplasm of the cancer cell, as well as the staining intensity of the dye in the membrane and cytoplasm of other cancer cells closer than a predetermined distance to that cancer cell. The resulting quantitative continuous score (QCS) indicates the viability of the cancer patient and is generated by aggregating all single-cell ADC scores from the tissue sample using statistical manipulation. Aggregation of all single-cell ADC scores is performed by determining the mean, determining the median, or determining a displacement value with a predetermined percentage. In another embodiment, aggregation of all single-cell ADC scores includes a thresholding operation using a predetermined threshold. All cells with a single-cell ADC score greater than the predetermined threshold are labeled single-cell ADC positive. Aggregation is performed by determining the number of single-cell ADC positive cells divided by the total number of cancer cells.

[0009] In one embodiment, a method for predicting a cancer patient's response to an antibody-drug conjugate (ADC) includes detecting cancer cells and computer-processing a single-cell ADC score for each cancer cell. The ADC comprises an ADC payload and an ADC antibody that targets a protein on the cancer cell. The protein is B7-H4. A tissue sample is immunohistochemically stained using a dye linked to a diagnostic antibody. The diagnostic antibody binds to the protein on the cancer cells in the tissue sample. A digital image of the tissue sample is acquired, and cancer cells are detected in the digital image. For each cancer cell, a single-cell ADC score is computer-processed based on the staining intensity of the dye in the membrane. The single-cell ADC score may also be optionally based on the staining intensity of the dye in the cytoplasm of the cancer cell, as well as the staining intensity of the dye in the membrane and cytoplasm of other cancer cells closer than a predetermined distance to that cancer cell. The staining intensity of each membrane is computer-processed based on the average optical density of brown diaminobenzidine (DAB) signals in the membrane pixels, and the staining intensity of each cytoplasm is computer-processed based on the average optical density of brown DAB signals in the cytoplasmic pixels. The cancer patient's response to ADCs is predicted based on the aggregation of all single-cell ADC scores in the tissue sample using statistical manipulation.

[0010] In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 5 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 6 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 7 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 9 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 10 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 15 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 20 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 25 or higher.

[0011] In some embodiments, a patient scores positive if at least 50% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 50% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 60% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 60% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 70% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 70% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 80% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 80% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 90% of tumor cells have a membrane optical density of 8 or higher. In some embodiments, a patient scores positive if at least 95% of tumor cells have a membrane optical density of 5 or higher, 8 or higher, or 25 or higher. In some embodiments, a patient scores positive if at least 95% of tumor cells have a membrane optical density of 8 or higher.

[0012] In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 500 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1000 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1250 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1500 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1600 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1670 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1700 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1800 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 1900 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 2000 cells / mm2 or higher. In some embodiments, a patient scores positive if the density of tumor cells having at least 8 membrane optical densities is 3000 cells / mm2 or higher.

[0013] In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 90 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 91 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 92 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 93 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 94 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 95 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 96 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 97 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 98 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 99 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 99.5 or higher. In some embodiments, a patient is considered to have a positive score if their spatial proximity score is 99.8 or higher. In the above embodiment, the spatial proximity score is calculated as follows: Spatial proximity score = ([Number of tumor cells with OD>8] + [Number of tumor cells with OD<=8 that are 25um neighbors of at least one tumor cell with OD>8]) / [Total number of tumor cells].

[0014] In another embodiment, a method for identifying cancer patients who will exhibit a predefined response to an antibody-drug conjugate (ADC) includes generating a response score. The ADC comprises an ADC payload and an ADC antibody that targets a protein on cancer cells. Tissue samples from cancer patients are immunohistochemically stained using a dye linked to a diagnostic antibody that binds to a protein on cancer cells in the tissue sample. Digital images of the tissue samples are acquired, and cancer cells in the digital images are detected using a convolutional neural network. For each cancer cell, a single-cell ADC score is computer-processed based on the staining intensity of the dye in the membrane. The single-cell ADC score may also be optionally based on the staining intensity of the dye in the cytoplasm of the cancer cell, as well as the staining intensity of the dye in the membrane and cytoplasm of other cancer cells that are closer than a predetermined distance to the cancer cell for which the single-cell score is computer-processed. The QCS score is a response score generated by aggregating all single-cell ADC scores from the tissue sample using statistical manipulation. Cancer patients are identified as those who will exhibit a predefined response to the ADC based on whether their QCS score exceeds a threshold. Patients with a QCS score greater than the threshold are considered QCS positive, while all other patients are considered QCS negative. The predefined response is a reduction in mean tumor size. In another aspect, the difference between the QCS score and the threshold indicates the likelihood that a cancer patient will be accurately identified as having a predefined response to ADC. A small difference indicates a low likelihood of accurate identification of a cancer patient, while a large difference indicates a high likelihood.

[0015] In one embodiment, a method for treating a cancer patient with an antibody-drug conjugate (ADC) includes generating a QCS score in the form of a treatment score. The ADC comprises an ADC payload and an ADC antibody that targets a protein on cancer cells. The protein is B7-H4. A tissue sample from the cancer patient is immunohistochemically stained using a dye linked to a diagnostic antibody that binds to a protein on cancer cells in the tissue sample. A digital image of the tissue sample is acquired, and cancer cells are detected in the digital image. For each cancer cell, a single-cell ADC score is calculated based on the staining intensity of the dye in the membrane. The single-cell ADC score may also be optionally based on the staining intensity of the dye in the cytoplasm of the cancer cell, as well as the staining intensity of the dye in the membrane and cytoplasm of other cancer cells closer than a predetermined distance to that cancer cell. The treatment score is generated by aggregating all single-cell ADC scores from the tissue sample using statistical operations. The ADC-containing therapy is administered to the cancer patient if the treatment score exceeds a predetermined threshold.

[0016] In another embodiment, the method of treating cancer patients with an antibody-drug conjugate (ADC) is carried out by a clinician administering the therapy. The ADC comprises an ADC payload and an ADC antibody that targets a protein on cancer cells. The protein is B7-H4. Therapy containing the ADC is administered to cancer patients if the response score exceeds a predetermined threshold. The QCS response score was generated by aggregating single-cell ADC scores from cancer patient tissue samples using statistical manipulation. Each single-cell ADC score was computer-processed for each cancer cell based on the staining intensity of the dye in the membrane. Single-cell ADC scores can also be optionally obtained based on the staining intensity of the dye in the cytoplasm of each cancer cell, as well as the staining intensity of the dye in the membrane and cytoplasm of other cancer cells that are closer than a predetermined distance to each cancer cell for which the single-cell ADC score is computer-processed. Cancer cells were detected in digital images of cancer patient tissue samples. The tissue samples were immunohistochemically stained using a dye linked to a diagnostic antibody that binds to a protein on cancer cells in the tissue sample.

[0017] Other embodiments and advantages are described in the detailed description below. This summary is not intended to define the invention. The invention is defined by the claims.

[0018] The attached drawings illustrate embodiments of the present invention, where similar numbers indicate similar components. [Brief explanation of the drawing]

[0019] [Figure 1] The complete nucleotide sequence encoding human B7-H4 (SEQ ID NO: 1) is shown. [Figure 2] The nucleotide coding sequence for human B7-H4 (SEQ ID NO: 2) is shown. [Figure 3] The polypeptide sequence of human B7-H4 (SEQ ID NO: 3) is shown. [Figure 4] The amino acid sequence of the heavy chain HCDR1 of the anti-B7H4 antibody (SEQ ID NO: 4) is shown. [Figure 5] The amino acid sequence of the heavy chain HCDR2 of the anti-B7H4 antibody (SEQ ID NO: 5) is shown. [Figure 6] The amino acid sequence of the heavy chain HCDR3 of the anti-B7H4 antibody (SEQ ID NO: 6) is shown. [Figure 7] The amino acid sequence of the light chain LCDR1 of the anti-B7H4 antibody (SEQ ID NO: 7) is shown. [Figure 8] The amino acid sequence of the light chain LCDR2 of the anti-B7H4 antibody (SEQ ID NO: 8) is shown. [Figure 9] The amino acid sequence of the light chain LCDR3 of the anti-B7H4 antibody (SEQ ID NO: 9) is shown. [Figure 10] The amino acid sequence (SEQ ID NO: 10) of the heavy chain variable region (VH) of the anti-B7H4 antibody is shown. [Figure 11] The amino acid sequence of the light chain variable region (VL) of the anti-B7H4 antibody (SEQ ID NO: 11) is shown. [Figure 12] The amino acid sequence of the heavy chain of the anti-B7H4 antibody (SEQ ID NO: 12) is shown. [Figure 13] The amino acid sequence of the light chain of the anti-B7H4 antibody (SEQ ID NO: 13) is shown. [Figure 14]This is a flowchart illustrating the process by which an analysis system analyzes digital images of tissue from cancer patients and predicts how likely the cancer patients are to respond to therapies containing anti-B7H4 antibody drug conjugates. [Figure 15] Figure 14 shows a digital image illustrating the image analysis process in step 2. [Figure 16] This shows the image analysis process for detecting nuclear targets in cancer cells. [Figure 17] This illustrates an image analysis process for detecting membranes using nuclear targets. [Figure 18] This shows the image analysis process for detecting membrane targets of cancer cells. [Figure 19] This is a screenshot of the results of the image analysis process in an image analysis software environment. [Figure 20] This shows the script used for image analysis and the cell target measurements obtained using that script. [Figure 21] This shows quantitative results of staining intensity from image analysis using gray values ​​of membrane and cytoplasmic pixels. [Figure 22] The quantitative amounts of staining on the membrane and in the cytoplasm shown in Figure 21 are listed as examples. [Figure 23] This demonstrates the mechanism by which anti-B7H4 ADC therapy kills cancer cells. [Figure 24] Figure 22 shows the calculation of the single-cell ADC score for each of the three cells shown, based on cell separation that reflects the uptake of the ADC payload into adjacent cells. [Figure 25] The formula for calculating the single-cell score is shown below. [Figure 26] This shows the correlation between actual outcomes and manually calculated H-scores, enumerated by response classification, for 15 samples from the PDX breast test. [Figure 27] This graph compares the ratio score of manual work to the average H score. [Figure 28] This graph compares the ratio score of manual work to the median QCS score for optical density of B7H4 film staining. [Figure 29]This graph compares the intensity score of manual work to the average H score. [Figure 30] This graph compares the intensity score of manual work to the median QCS score for optical density of B7H4 film staining. [Figure 31] This shows the average H-score of PDX breast test samples within the three response classifications: PD, R, and SD. [Figure 32] This shows the median QCS scores of PDX breast test samples within the three response classifications: PD, R, and SD. [Figure 33] The box plot distribution of manual H-scores for samples related to response classification R, SD, and PD is shown. [Figure 34] The box plot distribution of QCS scores, as defined by the median film optical density of the sample for response classifications R, SD, and PD, is shown. [Figure 35] This shows the relationship between tumor growth percentage and QCS score (median membrane optical density of all tumor cells in the sample / 50% displacement). [Figure 36] The correlation line shows the relationship between tumor growth percentage and the added QCS score. [Figure 37] This paper describes how xenograft mouse models were used, along with image analysis and quantitative continuous scoring, to investigate the pharmacokinetics and pharmacodynamics of anti-B7H4 ADCs. [Figure 38] Figure 38A is a graph showing the cell ratio of epithelial cells stained with ADC-IgG across all epithelial cells, demonstrating the concentration dependence and dynamics of ADC binding to target cells. Figure 38B is a graph showing the percentage of gH2AX-positive epithelial cells, demonstrating the increase in DNA damage over time after ADC treatment and at different ADC concentrations. Figure 38C is a graph showing the increase in DNA damage over time after ADC treatment based on the percentage of cleaved caspase-3 (CC-3)-positive tumor area. Figure 38D is a graph showing the increase in tumor cell death over time after ADC treatment based on the decrease in cell density of all epithelial cells. [Figure 39]Figure 39A shows the median optical density of membrane staining of tumor tissues from 28 breast cancer patients cloned by transplantation into mice. Figure 39B shows the stained tissue of sample #21 from Figure 39A, where membrane staining with diagnostic antibody D11 was far less than staining with diagnostic antibodies A57.1 and Cal63. Figure 39C shows the stained tissue of sample #28 from Figure 39A, where membrane staining with diagnostic antibody A57.1 was significantly greater than staining with diagnostic antibodies D11 and Cal63. [Figure 40] Figure 40A is a digital image of sample #26 from Figure 39A, stained with the diagnostic antibody Cal63. Figure 40B is a digital image of sample #4 from Figure 39A, stained with the diagnostic antibody Cal63. [Figure 41] Figure 41A shows the distribution of cell-specific membrane staining for sample #26 before and after treatment with ADC without varying concentrations of payload. Figure 41B shows the distribution of cell-specific membrane staining for sample #4 before and after treatment with ADC E02-GL-SG3932. [Figure 42-1] Figure 42A shows the cell-specific distribution of bound IgG before and after ADC treatment for sample #26. Figure 42B shows the cell-specific distribution of bound IgG before and after ADC treatment for sample #4. [Figure 42-2] Figure 42C is a graph of the maximum membrane optical density of bound IgG at various dose levels of ADC E02-GL-SG3932. Figure 42D shows the percentage of cleaved caspase-3 (CC-3) positive tumors for sample #26 at various dose levels of ADC E02-GL-SG3932. [Figure 43] Figure 43A is a graph showing the membrane optical density of IgG for clone samples #26 and #4 at various time points after treatment with various doses of ADC. Figure 43B shows the median optical density of sample #26 at various time points after treatment with 7 mg / kg ADC E02-GL-SG3932. [Figure 44]Figure 44A shows the increase in DNA damage over time after ADC treatment for sample #26, based on the density of stained epithelial cells and the percentage of cleaved caspase-3 (CC-3) positive tumor cells at various time points after ADC treatment. Figure 44B shows the increase in DNA damage over time after ADC treatment for sample #4, based on the density of stained epithelial cells and the percentage of CC-3 positive tumor cells at various time points after ADC treatment. [Figure 45] Figure 45A shows the relationship between binary spatial proximity and film optical density at various time points after ADC treatment for sample #26. Figure 45B shows the relationship between binary spatial proximity and film optical density at various time points after ADC treatment for sample #4. [Figure 46] Figure 46A shows the relationship between continuous spatial proximity and film optical density at various time points after ADC treatment for sample #26. Figure 46B shows the relationship between continuous spatial proximity and film optical density at various time points after ADC treatment for sample #4. [Figure 47] Figure 47A shows the relationship between continuous spatial proximity and epithelial cell density at various time points after ADC treatment for sample #26. Figure 47B shows the relationship between continuous spatial proximity and epithelial cell density at various time points after ADC treatment for sample #4. [Modes for carrying out the invention]

[0020] The present invention provides a novel method for computerizing a score indicating how cancer patients will respond to therapy using a B7-H4 antibody drug conjugate. Another aspect of the present invention relates to a method for computerizing a score for cancer patients indicating the viability of cancer patients treated with an ADC. Another aspect of the present invention relates to a method for predicting a cancer patient's response to an ADC. Another aspect of the present invention relates to identifying cancer patients who will exhibit a predetermined response to an ADC. Yet another aspect of the present invention relates to a method for treating cancer patients by administering a therapy including an ADC if their treatment score exceeds a predetermined threshold.

[0021] I. Definition To facilitate understanding of this invention, several terms and phrases are defined below. The term "cancer" is used to have the same meaning as the term "tumor."

[0022] In this invention, "B7-H4" is synonymous with V-set domain-containing T cell activation inhibitor 1 encoded by the VTCN1 gene. B7-H4 is a transmembrane polypeptide of the B7 family of costimulatory proteins. B7-H4 is understood to be expressed on the surface of antigen-presenting cells for interaction with ligands of immune cells such as T lymphocytes, with CD28 being a potential ligand. The observation that B7-H4 is highly expressed on cells of various cancer types suggests that this molecule is a tumor-associated antigen. B7-H4 expression is not limited to specific cancer types, but rather serves as a target antigen for treating a wide range of cancer types. The expression of the B7-H4 protein can be detected using methods well known to those skilled in the art, such as immunohistochemistry (IHC).

[0023] The RNA, DNA, and amino acid sequences of B7-H4 are known to those skilled in the art and can be found in many databases, such as the National Center for Biotechnology Information (NCBI) and UniProt databases. Examples of these sequences found in UniProt are Q7Z7D3 (VTCN1_HUMAN) for human B7-H4 and Q7TSP5 (VTCN1_MOUSE) for mouse B7-H4. The nucleotide sequence encoding human B7-H4 may be Sequence ID No. 1 (Figure 1) or more preferably Sequence ID No. 2 (Figure 2). The polypeptide sequence of human B7-H4 is preferably Sequence ID No. 3 (Figure 3).

[0024] In the present invention, "anti-B7H4 antibody" refers to an antibody that specifically binds to B7-H4, preferably having the activity to bind to B7-H4, and thereby being internally transferred into B7-H4 expressing cells. In other words, "anti-B7H4 antibody" refers to an antibody that has the activity to bind to B7-H4 and then moves into B7-H4 expressing cells.

[0025] The anti-B7H4 antibody includes a heavy chain HCDR1 having the amino acid sequence of SEQ ID NO: 4 (Figure 4), a heavy chain HCDR2 having the amino acid sequence of SEQ ID NO: 5 (Figure 5), a heavy chain HCDR3 having the amino acid sequence of SEQ ID NO: 6 (Figure 6), a light chain LCDR1 having the amino acid sequence of SEQ ID NO: 7 (Figure 7), a light chain LCDR2 having the amino acid sequence of SEQ ID NO: 8 (Figure 8), and a light chain LCDR3 having the amino acid sequence of SEQ ID NO: 9 (Figure 9). The anti-B7H4 antibody includes a variable heavy chain containing the amino acid sequence of SEQ ID NO: 10 (Figure 10) and a variable light chain containing the amino acid sequence of SEQ ID NO: 11 (Figure 11). The anti-B7H4 antibody includes a heavy chain (e.g., including VH and a constant heavy chain) containing the amino acid sequence of SEQ ID NO: 12 (Figure 12) and a light chain (e.g., including VL and a constant light chain) containing the amino acid sequence of SEQ ID NO: 13 (Figure 13).

[0026] The terms “subject,” “individual,” and “patient” are used interchangeably herein to refer to mammalian subjects. In one embodiment, the “subject” is a human, a domesticated animal, a livestock animal, a sporting animal, and a zoo animal, such as a human, a non-human primate, a dog, a cat, a guinea pig, a rabbit, a rat, a mouse, a horse, a cow, etc. In one embodiment, the subject is a cynomolgus macaque (Macaca fascicularis). In a preferred embodiment, the subject is a human. In the method of the present invention, the patient may not have been previously diagnosed with cancer. Alternatively, the patient may have been previously diagnosed with cancer. The patient may also be a person who exhibits disease risk factors or is asymptomatic with respect to cancer. The patient may also have cancer or be at risk of developing cancer. Thus, in one embodiment, the method of the present invention may be used to confirm the presence of cancer in a patient. For example, the patient may have been previously diagnosed with cancer by alternative means. In one embodiment, the patient may have previously undergone cancer therapy.

[0027] In this invention, the term "QCS positive" refers to cancer that is likely to respond to anti-B7H4 ADC therapy. The term "QCS negative" refers to cancer that is unlikely to respond to anti-B7H4 ADC therapy. The acronym QCS stands for Quantitative Continuum Score. The results of the novel prediction method of this invention are generally referred to as a quantitative continuous score and may be an indicator of response score or predicted survival time.

[0028] II. Anti-B7H4 antibody drug conjugates. The anti-B7H4 antibody drug conjugate used in the present invention comprises an anti-B7H4 antibody or antigen-binding fragment conjugated to one or more cytotoxics selected from topoisomerase I inhibitors, tubricin derivatives, pyrrolobenzodiazepines, or combinations thereof. For example, the antibody or its antigen-binding fragment is conjugated to one or more cytotoxics selected from the group consisting of topoisomerase I inhibitors SG3932, SG4010, SG4057, or SG4052 (structures provided below); tubricin AZ1508, pyrrolobenzodiazepine SG3315, pyrrolobenzodiazepine SG3249, or combinations thereof.

[0029] The antibody or its antigen-binding fragment is preferably conjugated with a topoisomerase I inhibitor. A topoisomerase inhibitor is a compound that blocks the action of topoisomerase (topoisomerase I and II), which is a type of enzyme that controls changes in DNA structure by catalyzing the cleavage and recombination of the phosphate diester backbone of DNA strands during the normal cell cycle.

[0030] Typical examples of suitable topoisomerase I inhibitors are represented by the following compounds: [Formula 1] [ka]

[0031] The compounds shown above are denoted as A* and may be referred to herein as “drug units.” Compound A* is preferably provided with a linker for linking (preferably conjugating) to an antibody or antigen-binding fragment (which may be referred to as a “ligand unit”) as described herein. Preferably, the linker is conjugated (e.g.) to an amino residue, for example, an amino acid of an antibody or antigen-binding fragment as described herein, in a cleavable manner.

[0032] More specifically, examples of suitable topoisomerase I inhibitors are represented by the following compounds, indicated as "I": [Formula 2] [ka]

[0033] Compound I includes salts and solvates thereof, where R L This is a linker for linking to an antibody or its antigen-binding fragment (e.g., ligand unit) as described herein, wherein the linker "Ia" is preferably [Formula 3] [ka] Selected from, Q is, [Formula 4] [ka] And in the formula, Q X This is such that Q is an amino acid residue, a dipeptide residue, a tripeptide residue, or a tetrapeptide residue, and X in Equation 3 is [Formula 5] [ka] In the equation, a = 0 to 5, b1 = 0 to 16, b2 = 0 to 16, c1 = 0 or 1, c2 = 0 or 1, and d = 0 to 5, where at least b1 or b2 = 0 (i.e., at least one of b1 and b2 may not be 0), and at least c1 or c2 = 0 (i.e., at least one of c1 and c2 may not be 0). G in Equation 3 L This is a linker for linking to an antibody or its antigen-binding fragment as described herein. Alternatively, G in Formula 3 L The compound shown as "Ib": [Formula 6] [ka] And in the formula, R L1 and R L2is independently selected from H and methyl, or together with the carbon atom to which they are attached, forms a cyclopropylene or cyclobutylene group. In Formula 6, e is 0 or 1.

[0034] It will be understood by those skilled in the art that two or more of said agents (e.g., topoisomerase I inhibitors) can be conjugated to an antibody or an antigen-binding fragment thereof. For example, a conjugate of the present invention (e.g., an antibody-drug conjugate) has the general formula L-(D L ) p or a pharmaceutically acceptable salt or solvate thereof, where L is an antibody or an antigen-binding fragment thereof (e.g., a ligand unit) described herein, p is an integer from 1 to 20, and D L is a topoisomerase I inhibitor having a linker (e.g., a drug linker unit) represented by the following formula: [Formula 7]

Chemical formula

[0035] In Formula 7, R LL is a linker linked to an antibody or an antigen-binding fragment thereof (e.g., a ligand unit) described herein, and the linker is preferably selected from "Ia'" or "Ib'": [Formula 8]

Chemical formula

Chemical formula

[0036] Therefore, the antibody-drug conjugate described herein comprises a conjugate comprising an antibody or its antigen-binding fragment (e.g., ligand unit) covalently linked to at least one topoisomerase I inhibitor (e.g., a drug unit such as A* shown above). The inhibitor is preferably R L and / or R LL The antibody or its antigen-binding fragment is linked by a linker (e.g., a linker unit), such as the linker described above. In other words, the antibody-drug conjugate described herein comprises an antibody or its antigen-binding fragment (e.g., a ligand unit) having one or more topoisomerase I inhibitors linked preferably via a linker (e.g., a drug-linker unit). The antibody or its antigen-binding fragment (representing the ligand unit), as described more fully above, is a targeted drug that binds to a target moiety. More specifically, the ligand unit specifically binds, for example, to the B7-H4 protein on a target cell, thereby delivering the drug unit. Thus, the present invention also provides, for example, methods for the treatment of various cancers and other disorders (e.g., cancers / disorders associated with the presence of cancerous cells, preferably cells expressing B7-H4) with ADCs.

[0037] After intracellular translocation into cancer cells, the anti-B7H4 antibody-drug conjugate used in this invention is cleaved at the linker unit, releasing the drug units into the cancer cells. The anti-B7H4 antibody-drug conjugate used in this invention also has a bystander effect, where the anti-B7H4 antibody-drug conjugate is internally translocated into cancer cells expressing the target protein B7-H4, and the drug units subsequently exert an antitumor effect against adjacent cancer cells that do not express the target protein B7-H4.

[0038] III. Further priorities of preferred embodiments The following priorities may apply to all aspects of the invention described above, or to a single aspect. The priorities may be combined in any combination. Various definitions relating to certain terms in this section are provided under the heading "Further Chemical Definitions" below.

[0039] Certain characteristics of the topoisomerase I inhibitors described above are particularly preferred and are defined in more detail below. As an example, characteristic Q of the linker in Equation 4. X Preferred embodiments are outlined. In one embodiment, Q is an amino acid residue. The amino acid may be a natural amino acid or a non-natural amino acid. For example, Q may be selected from Phe, Lys, Val, Ala, Cit, Leu, Ile, Arg, and Trp, where Cit is citrulline. In one embodiment, Q comprises a dipeptide residue. The amino acids in the dipeptide may be any combination of natural and non-natural amino acids. In some embodiments, the dipeptide comprises a natural amino acid. If the linker is a cathepsin-unstable linker, the dipeptide is the site of action for cathepsin-mediated cleavage. In that case, the dipeptide is the cathepsin recognition site.

[0040] In one embodiment, Q is: NH -Phe-Lys- C=O , NH -Val-Ala- C=O , NH -Val-Lys- C=O , NH -Ala-Lys- C=O , NH -Val-Cit- C=O , NH -Phe-Cit- C=O , NH -Leu-Cit- C=O , NH -Ile-Cit- C=O , NH -Phe-Arg- C=O , NH -Trp-Cit- C=O , and NH -Gly-Val- C=O Selected from; In the formula, Cit is citrulline. Preferably, Q is: NH -Phe-Lys- C=O , NH -Val-Ala- C=O , NH -Val-Lys- C=O , NH -Ala-Lys- C=O , and NH -Val-Cit- C=O Selected from.

[0041] More preferably, Q is NH -Phe-Lys- C=O , NH -Val-Cit- C=O or NH -Val-Ala- C=O Selected from: Other suitable dipeptide combinations are NH -Gly-Gly- C=O , NH -Gly-Val-C=O , NH -Pro-Pro- C=O , and NH -Val-Glu- C=O Includes.

[0042] Other dipeptide combinations, including those described in Dubowchik et al., Bioconjugate Chemistry, 2002, 13, 855-869, which are incorporated herein by reference, may also be used.

[0043] In some embodiments, Q is a tripeptide residue. The amino acids in the tripeptide may be any combination of native and non-native amino acids. In some embodiments, the tripeptide contains native amino acids. If the linker is unstable to the cathepsin, the tripeptide is the site of action for cathepsin-mediated cleavage. In that case, the tripeptide is the recognition site for the cathepsin. Particularly interesting tripeptide linkers are: NH -Glu-Val-Ala- C=O , NH -Glu-Val-Cit- C=O , NH -αGlu-Val-Ala- C=O , and NH -αGlu-Val-Cit- C=O That is the case.

[0044] In some embodiments, Q is a tetrapeptide residue. The amino acids in the tetrapeptide may be any combination of native and non-native amino acids. In some embodiments, the tetrapeptide contains native amino acids. If the linker is unstable to the cathepsin, the tetrapeptide is the site of action for cathepsin-mediated cleavage. In that case, the tetrapeptide is the recognition site for the cathepsin. Particularly interesting tetrapeptide linkers are: NH -Gly-Gly-Phe-GlyC=O and NH -Gly-Phe-Gly-Gly C=O That is the case.

[0045] In some embodiments, the tetrapeptide is: NH -Gly-Gly-Phe-Gly C=O That is the case.

[0046] In the above expression of peptide residues, NH - represents the N-terminus of a residue, - C=O represents the C-terminus of a residue. The C-terminus is bound to the NH of compound A* shown in formula 1. Glu is a glutamic acid residue, i.e.: [Formula 10] [ka] It represents.

[0047] αGlu is the glutamic acid residue when bound via the α chain, i.e.: [Formula 11] [ka] It represents.

[0048] In one embodiment, the amino acid side chains are chemically protected, where appropriate. The side-chain protecting group may be any of the groups described above. The protected amino acid sequence is enzymatically cleavable. For example, a dipeptide sequence containing a Boc side-chain protected Lys residue is cathepsin-cleavable.

[0049] Protecting groups for amino acid side chains are well known in the art and are described in the Novabiochem Catalog and above.

[0050] G of the linker in Equation 3 L The following options may be selected: [ka] [Chemical formula or structure indication]

[0051] In the formula, Ar represents a C 5~6 arylene group, for example, phenylene, and X represents a C 1~4 alkyl.

[0052] In some embodiments, G L is selected from G L1-1 and G L1-2 In some of these embodiments, G L is G L1-1 G LL can be selected from the following. [Chemical formula or structure indication]

[0053] In the formula, Ar represents a C 5~6 arylene group, for example, phenylene, and X represents a C 1~4 alkyl. In some embodiments, G LL is selected from G LL1-1 and G LL1-2 In some of these embodiments, G LL is G LL1-1

[0054] X is preferably [Formula 12] [Chemical formula or structure indication] where a = 0 to 5, b1 = 0 to 16, b2 = 0 to 16, c = 0 or 1, d = 0 to 5, and where at least b1 or b2 = 0, and at least c1 or c2 = 0.

[0055] "a" may be 0, 1, 2, 3, 4 or 5. In some embodiments, a is 0 to 3. In some of these embodiments, a is 0 or 1. In further embodiments, a is 0.

[0056] "b1" may be 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16. In some embodiments, "b1" is 0 to 12. In some of these embodiments, "b1" is 0 to 8, and may be 0, 2, 3, 4, 5, or 8.

[0057] "b2" may be 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16. In some embodiments, "b2" is between 0 and 12. In some of these embodiments, "b2" is between 0 and 8, and may be 0, 2, 3, 4, 5, or 8. Preferably, only one of b1 and b2 may not be 0.

[0058] "c1" may be 0 or 1. "c2" may be 0 or 1. Preferably, only one of "c1" or "c2" may be non-zero.

[0059] "d" may be 0, 1, 2, 3, 4, or 5. In some embodiments, "d" is 0 to 3. In some of these embodiments, "d" is 1 or 2. In further embodiments, "d" is 2. In even further embodiments, "d" is 5.

[0060] In some embodiments of X, a is 0, b1 is 0, c1 is 1, c2 is 0, d is 2, and b2 may be from 0 to 8. In some of these embodiments, b2 is 0, 2, 3, 4, 5, or 8. In some embodiments of X, a is 1, b2 is 0, c1 is 0, c2 is 0, d is 0, and b1 may be from 0 to 8. In some of these embodiments, b1 is 0, 2, 3, 4, 5, or 8. In some embodiments of X, a is 0, b1 is 0, c1 is 0, c2 is 0, d is 1, and b2 may be from 0 to 8. In some of these embodiments, b2 is 0, 2, 3, 4, 5, or 8. In some embodiments of X, b1 is 0, b2 is 0, c1 is 0, c2 is 0, and one of a and d is 0. The other of a and d is from 1 to 5. In some of these embodiments, the other of a and d is 1. In some of the other of these embodiments, the other of a and d is 5. In some embodiments of X, a is 1, b2 is 0, c1 is 0, c2 is 1, d is 2, and b1 may be from 0 to 8. In some of these embodiments, b2 is 0, 2, 3, 4, 5, or 8.

[0061] In some embodiments, R L is Ib of Formula 6. In some embodiments, R LL is the group Ib' of Formula 9. R L1 and R L2 may be independently selected from H and methyl, or together with the carbon atom to which they are attached, form a cyclopropylene or cyclobutylene group.

[0062] In some embodiments, R L1 and R L2 are both H. In some embodiments, R L1 is H and R L2 is methyl. In some embodiments, R L1 and R L2 are both methyl.

[0063] In some embodiments, R L1 and R L2 These, together with the carbon atoms to which they are bonded, form a cyclopropylene group. In some embodiments, R L1 and R L2 These, together with the carbon atoms to which they are bonded, form a cyclobutylene group.

[0064] In group Ib of formula 6, in some embodiments, e is 0. In other embodiments, e is 1, and the nitro group may be at any available position on the ring. In some of these embodiments, it is in the ortho position. In others of these embodiments, it is in the para position.

[0065] In some embodiments in which the compounds described herein are provided in a form rich in a single enantiomer or one enantiomer, the form rich in one enantiomer has an enantiomer ratio greater than 60:40, greater than 70:30; greater than 80:20 or greater than 90:10. In further embodiments, the enantiomer ratio is greater than 95:5, greater than 97:3 or greater than 99:1.

[0066] In some embodiments, R in Equation 2 L The following options are available: [ka] [ka]

[0067] In some embodiments, R LL This is the R mentioned above. L It is a group derived from another group.

[0068] Following the priorities outlined above, certain preferred topoisomerase I linker formulas (e.g., those of drug linker units) are described herein. In some embodiments, compound I of formula 2 is compound I P : [Formula 13] [ka] Also, its salts and solvates. LP This is a linker for linking to an antibody or its antigen-binding fragment as described herein, the linker being selected from group Ia or Ib. Group Ia is of formula: [Formula 14] [ka] It is expressed by, in the formula, Q P teeth, [ka] (In the formula, Q XP Q P (However, it is such that it is an amino acid residue, a dipeptide residue, or a tripeptide residue); X in formula 14 P teeth, [Formula 15] [ka] In the equation, aP = 0 to 5, bP = 0 to 16, cP = 0 or 1, and dP = 0 to 5. G in Equation 14 L This is a linker for linking to an antibody or its antigen-binding fragment (e.g., ligand unit). aP may be 0, 1, 2, 3, 4, or 5. In some embodiments, aP is 0 to 3. In some of these embodiments, aP is 0 or 1. In further embodiments, aP is 0. bP may be 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16. In some embodiments, b is 0 to 12. In some of these embodiments, bP is 0 to 8, and may be 0, 2, 4, or 8. cP may be 0 or 1. dP may be 0, 1, 2, 3, 4, or 5. In some embodiments, dP is 0 to 3. In some of these embodiments, dP is 1 or 2. In further embodiments, dP is 2.

[0069] Base Ib is given by the formula: [Formula 16] [ka] Represented by R L1 and R L2 The carbon atom is independently selected from H and methyl, or together with the carbon atom to which they are bonded, to form a cyclopropylene or cyclobutylene group. In formula 16, "e" is either 0 or 1.

[0070] X in Equation 14 P In some embodiments, aP is 0, cP is 1, dP is 2, and bP may be 0 to 8. In some of these embodiments, bP is 0, 4, or 8.

[0071] Q in Equation 4 relating to Compound I X The priority is Q in Equation 14. XP This can be applied to the above G relating to compound I of formula 2. L , R L1 , R L2 The priority regarding and e is compound I of formula 13. P It can be applied to this.

[0072] In some embodiments, the antibody-drug conjugate L-(D L ) p The conjugate is L-(D LP ) p or a pharmaceutically acceptable salt or solvate thereof, where L is an antibody or its antigen-binding fragment (e.g., ligand unit), and D LP Compound III P : [Formula 17] [ka] It is a topoisomerase I inhibitor (e.g., a drug linker unit), R LLP This is a linker linked to an antibody or its antigen-binding fragment (e.g., a ligand unit), the linker being selected from group Ia' or Ib'. Group Ia' is given by formula: [Formula 18] [ka] It is expressed by, in the formula, Q P and X P As defined above, and G LL This is a linker attached to an antibody or its antigen-binding fragment (e.g., ligand unit). The group Ib' is given by formula: [Formula 19] [ka] It is expressed by, in the formula, R L1 and R L2 is as defined above; and p is an integer between 1 and 20.

[0073] In some embodiments, compound I of formula 2 is compound I P2 : [Formula 20] [ka] Also, its salts and solvates. LP2 This is a linker for linking to an antibody or its antigen-binding fragment, and the linker is selected from group Ia or Ib. Group Ia is given by formula: [Formula 21] [ka] It is represented by and Q is: [ka] And in the formula, Q X This is a case where Q is an amino acid residue, a dipeptide residue, a tripeptide residue, or a tetrapeptide residue. P2 teeth: [ka] In the equation, aP2 = 0 to 5, b1P2 = 0 to 16, b2P2 = 0 to 16, cP2 = 0 or 1, and dP2 = 0 to 5, where at least b1P2 or b2P2 = 0 (i.e., there may be cases where only one of b1 or b2 is not 0). G in Equation 21 L This is a linker for attaching to an antibody or its antigen-binding fragment (e.g., a ligand unit).

[0074] aP2 may be 0, 1, 2, 3, 4, or 5. In some embodiments, aP2 is 0 to 3. In some of these embodiments, aP2 is 0 or 1. In further embodiments, aP2 is 0. b1P2 may be 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16. In some embodiments, b1P2 is 0 to 12. In some of these embodiments, b1P2 is 0 to 8 and may be 0, 2, 3, 4, 5, or 8. b2P2 may be 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or 16. In some embodiments, b2P2 is 0 to 12. In some of these embodiments, b2P2 is 0 to 8 and may be 0, 2, 3, 4, 5, or 8. Preferably, only one of b1P2 and b2P2 may be non-zero. cP2 may be 0 or 1. dP2 may be 0, 1, 2, 3, 4, or 5. In some embodiments, dP2 is 0 to 3. In some of these embodiments, dP2 is 1 or 2. In further embodiments, dP2 is 2. In further embodiments, dP2 is 5.

[0075] Base Ib is given by the formula: [Formula 22] [ka] It is expressed by, in the formula, R L1 and R L2 These atoms, independently selected from H and methyl, or in combination with the carbon atom to which they are bonded, form a cyclopropylene or cyclobutylene group; and e is 0 or 1.

[0076] X in Equation 21 P2 In some embodiments, aP2 is 0, b1P2 is 0, cP2 is 1, and dP2 is 2, and b2P2 may be 0 to 8. In some of these embodiments, b2P2 is 0, 2, 3, 4, 5, or 8. P2 In some embodiments, aP2 is 1, b2P2 is 0, cP2 is 0 and dP2 is 0 and b1P2 is 0 to 8. In some of these embodiments, b1P2 is 0, 2, 3, 4, 5 or 8. P2 In some embodiments, aP2 is 0, b1P2 is 0, cP2 is 0 and dP2 is 1 and b2P2 is 0 to 8. In some of these embodiments, b2P2 is 0, 2, 3, 4, 5 or 8. P2 In some embodiments, b1P2 is 0, b2P2 is 0, cP2 is 0, and one of aP2 and dP2 is 0. The other of aP2 and d is 1 to 5. In some of these embodiments, the other of aP2 and d is 1. In other embodiments of these embodiments, the other of aP2 and dP2 is 5.

[0077] The above Q regarding compound I in equation 2 X Priorities regarding this are the base Ia of Equation 21. P2 Q in X This can be applied to the above G relating to compound I of formula 2. L , R L1 , R L2 The priority regarding and e is compound Ia of formula 21. P2 It can be applied to this.

[0078] In some embodiments, the conjugate L-(D L ) p The conjugate is compound L-(D LP2 ) p or a pharmaceutically acceptable salt or solvate thereof, where L is an antibody or its antigen-binding fragment (e.g., ligand unit), and D LP2Compound III P2 : [Formula 23] [ka] It is a topoisomerase I inhibitor (e.g., a drug linker unit), and R LLP2 This is a linker linked to an antibody or its antigen-binding fragment (e.g., a ligand unit), the linker being selected from group Ia' or Ib'. Group Ia' is given by formula: [Formula 24] [ka] It is expressed by the formula, where Q and X P2 As defined above, and G LL This is a linker attached to an antibody or its antigen-binding fragment.

[0079] Base Ib' is given by the formula: [Formula 25] [ka] It is expressed by, in the formula, R L1 and R L2 is as defined above; and p is an integer between 1 and 20.

[0080] Particularly preferred topoisomerase I inhibitors include those having the following formula: [Formula 26] [ka] [Formula 27] [ka] [Formula 28] [ka] [Formula 29] [ka] and / or [Formula 30] [ka]

[0081] Inhibitor SG3932 is particularly preferred. Therefore, in a preferred embodiment, the antibody or its antigen-binding fragment is conjugated to a topoisomerase I inhibitor having the following formula for SG3932: [Formula 26] [ka]

[0082] To avoid misunderstanding, the number "8" in formula 26 specifies that the structure within the square brackets is repeated eight times. Therefore, another expression for SG3932 is: [Formula 31] [ka] That is the case.

[0083] Another way to express SG4010 is, [Formula 32] [ka] That is the case.

[0084] Another way to express SG4057 is, [Formula 33] [ka] That is the case.

[0085] Another way to express SG4052 is, [Formula 34] [ka] That is the case.

[0086] The antibodies or antigen-binding fragments thereof described in this specification can be conjugated to one or more of the topoisomerase I inhibitors.

[0087] IV. Further Chemical Definitions. The following definitions specifically pertain to the description of the topoisomerase I inhibitors above and may more particularly belong to the section entitled "Additional Priorities of Preferred Embodiments".

[0088] C 5~6 Arylene: The term "C 5~6 arylene" as used herein refers to a divalent moiety obtained by removing two hydrogen atoms from the aromatic ring atoms of an aromatic compound.

[0089] In this context, a prefix (e.g., C 5~6 ) indicates the number of ring atoms or the range of the number of ring atoms, regardless of whether they are carbon atoms or heteroatoms. As in the case of "carboarylene group", all the ring atoms may be carbon atoms, in which case the group is phenylene (C6). Alternatively, the ring atoms may contain one or more heteroatoms, as in the case of "heteroarylene group". Examples of heteroarylene groups include, but are not limited to, those derived from the following: N1: Pyrrole (azole) (C5), pyridine (azine) (C6); O1: Furan (oxole) (C5); S1: Thiophene (thiol) (C5); N1O1: Oxazole (C5), isoxazole (C5), isoxazine (C6); N2O1: Oxadiazole (furazan) (C5); N3O1: Oxatriazole (C5); N1S1: Thiazole (C5), isothiazole (C5); N2: Imidazole (1,3-diazole) (C5), pyrazole (1,2-diazole) (C5), pyridazine (1,2-diazine) (C6), pyrimidine (1,3-diazine) (C6) (e.g., cytosine, thymine, uracil), pyrazine (1,4-diazine) (C6); and N3: Triazole (C5), Triazine (C6).

[0090] C 1~4 Alkyl: The term "C 1~4 alkyl", as used herein, relates to a monovalent moiety obtained by removing one hydrogen atom from one carbon atom of a hydrocarbon compound having 1 to 4 carbon atoms, which may be aliphatic or alicyclic, and may be saturated or unsaturated (e.g., partially unsaturated, fully unsaturated). The term "C 1~n alkyl", as used herein, relates to a monovalent moiety obtained by removing one hydrogen atom from one carbon atom of a hydrocarbon compound having 1 to n carbon atoms, which may be aliphatic or alicyclic, and may be saturated or unsaturated (e.g., partially unsaturated, fully unsaturated). Thus, the term "alkyl" includes subclasses such as alkenyl, alkynyl, cycloalkyl, etc., discussed below.

[0091] Examples of saturated alkyl groups include, but are not limited to, methyl (C1), ethyl (C2), propyl (C3), and butyl (C4). Examples of saturated straight-chain alkyl groups include, but are not limited to, methyl (C1), ethyl (C2), n-propyl (C3), and n-butyl (C4). Examples of saturated branched-chain alkyl groups include isopropyl (C3), isobutyl (C4), sec-butyl (C4), and tert-butyl (C4).

[0092] C 2~4 Alkenyl: The term "C 2~4 alkenyl", as used herein, relates to an alkyl group having one or more carbon-carbon double bonds. Examples of unsaturated alkenyl groups include, but are not limited to, ethenyl (vinyl, -CH=CH2), 1-propenyl (-CH=CH-CH3), 2-propenyl (allyl, -CH-CH=CH2), isopropenyl (methylvinyl, -C(CH3)=CH2), and butenyl (C4).

[0093] C 2~4 Alkynyl: The term "C2~4 As used herein, "alkynyl" refers to an alkyl group having one or more carbon-carbon triple bonds. Examples of unsaturated alkynyl groups include, but are not limited to, ethynyl (-C≡CH) and 2-propynyl (propargyl, -CH2-C≡CH).

[0094] C 3~4 Cycloalkyl: Term "C 3~4 As used herein, "cycloalkyl" refers to an alkyl group that is a monovalent moiety obtained by removing one hydrogen atom from one alicyclic ring atom of a cyclic hydrocarbon (carbocyclic) compound, having 3 to 7 carbon atoms including 3 to 7 ring atoms, and is also a cycloyl group. Examples of cycloalkyl groups include, but are not limited to, saturated monocyclic hydrocarbon compounds such as cyclopropane (C3) and cyclobutane (C4); and unsaturated monocyclic hydrocarbon compounds such as cyclopropene (C3) and cyclobutene (C4).

[0095] Connection indicator: In equation 35 below, superscript indicator C(=O) and NH This indicates the group to which atoms are bonded. [Formula 35] [ka]

[0096] For example, the NH group is shown as being bonded to a carbonyl group (not part of the illustrated portion), and the carbonyl group is shown as being bonded to an NH group (not part of the illustrated portion).

[0097] Synthesis of V. topoisomerase I inhibitors. Compound I of formula 2 (wherein R L (is the base Ia of equation 3) is equation 36: [Formula 36] [ka] (In the formula, R L*From the compound (where is -QH), formula 37: [Formula 37] [ka] It can be synthesized by linking the compound or its activated form. Such a reaction can be carried out under amide coupling conditions. The compound of formula 36 is derived from formula 38: [Formula 38] [ka] (In the formula, R L*prot Q-Prot N (In the formula, Prot N The compound of formula 38 can be synthesized by deprotecting the compound of (which is an amine protecting group). The compound of formula 39 can be synthesized using the Friedlander reaction: [Formula 39] [ka] It can be synthesized by coupling the compound with compound A3. Compound A3 is (S)-4-ethyl-4-hydroxy-7,8-dihydro-1H-pyrano[3,4-f]indridine-3,6,10(4H)-trione. The compound of formula 39 is of formula 40: [Formula 40] [ka] It can be synthesized by removing the trifluoroacetamide protecting group from the compound. The compound of formula 40 is coupled with compound I7:R L*prot It can be synthesized by -OH. Compound I7 is N-(4-amino-8-oxo-5,6,7,8-tetrahydronaphthalene-1-yl)-2,2,2-trifluoroacetamide. Compound I of formula 2 (wherein R L (where is group Ia of formula 3 or group Ib of formula 6) is derived from compound I11 to compound R LIt can be synthesized by coupling -OH or its activated form. Compound I11 is (S)-4-amino-9-ethyl-9-hydroxy-1,2,3,9,12,15-hexahydro-10H,13H-benzo[d]pyrano[3’,4’:6,7]indolizino[1,2-b]quinoline-10,13-dione.

[0098] Amine protecting groups are well known to those skilled in the art. In particular, reference is made to the disclosure of suitable protecting groups in Greene’s Protecting Groups in Organic Synthesis, Fourth Edition, John Wiley & Sons, 2007 (ISBN 978-0-471-69754-1), pages 696 - 871.

[0099] VI Anti-B7H4 antibody. The term "antibody" encompasses monoclonal antibodies and their fragments (e.g., those exhibiting the desired biological activity). In a preferred embodiment, the antibodies described herein are monoclonal antibodies. In a more preferred embodiment, the antibody is a fully human monoclonal antibody. In one embodiment, the methods of the invention may utilize polyclonal antibodies.

[0100] In particular, antibodies are proteins that contain at least one or two heavy (H) chain variable regions (abbreviated herein as VHCs) and at least one or two light (L) chain variable regions (abbreviated herein as VLCs). The VHC and VLC regions can be further subdivided into highly variable regions called “complementarity-determining regions” (“CDRs”), and are interspersed with more conserved regions called “framework regions” (FRs). The extent of the framework regions and CDRs is strictly defined (see Kabat, EA, et al. Sequences of Proteins of Immunological Interest, Fifth Edition, USD Department of Health and Human Services, NIH Publication No. 91-3242, 1991, and Chothia, C. et al. J.MoI. Biol. 196:901-917, 1987).

[0101] Preferably, each VHC and VLC consists of three CDRs and four FRs, arranged from the amino terminus to the carboxyl terminus in the following order: FR1, CDR1, FR2, DR2, FR3, CDR3, FR4. The VHC or VLC chain of the antibody may further include all or part of the heavy or light chain constant region. In one embodiment, the antibody is a tetramer of two immunoglobulin heavy chains and two immunoglobulin light chains, where the immunoglobulin heavy and light chains are interconnected, for example, by disulfide bonds. The heavy chain constant region includes three domains, CH1, CH2, and CH3. The light chain constant region consists of one domain, CL. The variable regions of the heavy and light chains contain binding domains that interact with the antigen.

[0102] The term “antibody” includes intact immunoglobulins of type IgA, IgG, IgE, IgD, and IgM (and their subtypes), the light chains of which may be kappa or lambda. As used herein, the term “antibody” also refers to an antibody that binds to one of the aforementioned markers, for example, a portion of a molecule in which one or more immunoglobulin chains are not of full length but still bind to a marker. Examples of binding sites encompassed by the term antibody include (i) a monovalent fragment consisting of a Fab fragment, VLC, VHC, CL, and CH1 domains; (ii) a bivalent fragment containing an F(ab')2 fragment and two Fab fragments linked by disulfide crosslinks in the hinge region; (iii) an Fc fragment consisting of VHC and CH1 domains; (iv) an Fv fragment consisting of the VLC and VHC domains of a single arm of the antibody; (v) a dAb fragment consisting of a VHC domain (Ward et al., Nature 341:544-546, 1989); and (vi) an isolated complementarity-determining region (CDR) with a sufficient framework for binding, such as the antigen-binding portion of the variable region. The antigen-binding regions of the light chain variable region and the heavy chain variable region, for example, the two domains of an Fv fragment, VLC and VHC, can be linked by a synthetic linker using a recombinant method, which allows them to be produced as a single protein chain (where the VLC and VHC regions pair to form a monovalent molecule) (known as single-chain Fv (scFv); see, e.g., Bird et al. (1988) Science lAl-ATi-Alβ; and Huston et al. (1988) Proc. Natl. Acad. ScL USA 85:5879-5883). Such single-chain antibodies are also included in the term antibody. These may be obtained using conventional techniques known to those skilled in the art, and these portions are screened for utility in the same manner as intact antibodies.

[0103] In one embodiment, the antibody or antigen-binding fragment is one or more selected from mouse antibodies, humanized antibodies, chimeric antibodies, monoclonal antibodies, polyclonal antibodies, recombinant antibodies, multispecific antibodies, or a combination thereof.

[0104] In one embodiment, the antigen-binding fragment is one or more selected from Fv fragment, Fab fragment, F(ab')2 fragment, Fab' fragment, dsFv fragment, scFv fragment, sc(Fv)2 fragment, or a combination thereof.

[0105] In preferred embodiments, the antibody or its antigen-binding fragment is a monoclonal antibody (mAb). In one embodiment, the antibody or its antigen-binding fragment (e.g., mAb) of the present invention is scFV.

[0106] In one embodiment, the antibody or its antigen-binding fragment can bind to B7-H4 molecules across species; for example, the antibody or fragment can bind to mouse B7-H4, rat B7-H4, rabbit, human B7-H4, and / or cynomolgus monkey B7-H4. In one embodiment, the antibody or fragment can bind to human B7-H4 and cynomolgus monkey B7-H4. In one embodiment, the antibody or antigen-binding fragment can also bind to mouse B7-H4. In one embodiment, the antibody or its antigen-binding fragment can specifically bind to B7-H4, for example, human B7-H4 and cynomolgus monkey B7-H4, but not specifically to human B7-H1, B7-H2, and / or B7-H3.

[0107] In one embodiment, the antibody or its antigen-binding fragment may include a heavy chain constant region or a fragment thereof, in addition to VH and VL. In one embodiment, the heavy chain constant region is a human heavy chain constant region, for example, a human IgG constant region, for example, a human IgG1 constant region. In one embodiment (preferably when the antibody or its antigen-binding fragment is conjugated with a drug such as a cytotoxic drug), a cysteine ​​residue is inserted between amino acids S239 and V240 of the CH2 region of IgG1. This cysteine ​​is referred to as the "239 insertion" or "239i".

[0108] VII. Preparation of anti-B7H4 antibody. The antibodies described herein can be obtained using conventional methods known to those skilled in the art, and their usefulness can be confirmed by conventional binding assays. For example, a simple binding assay involves incubating cells expressing an antigen together with the antibody. If the antibody is tagged with a fluorophore, the binding of the antibody to the antigen can be detected by FACS analysis.

[0109] The antibodies described herein can be produced in a variety of animals, including mice, rats, rabbits, sheep, monkeys, or horses. Antibodies can be produced after immunization with individual or multiple capsular polysaccharides. Blood isolated from these animals contains polyclonal antibodies (multiple antibodies that bind to the same antigen). Antigens can also be injected into chickens to produce polyclonal antibodies in egg yolk. To obtain monoclonal antibodies specific to a single epitope of an antigen, antibody-secreting lymphocytes are isolated from animals and immortalized by fusing them with cancer cell lines. The fused cells are called hybridomas and will grow continuously in culture, secreting antibodies. Single hybridoma cells are isolated by dilution cloning to produce cell clones, all of which produce the same antibody. These antibodies are called monoclonal antibodies. Methods for generating monoclonal antibodies are conventional techniques known to those skilled in the art (see, for example, *Making and Using Antibodies: A Practical Handbook*, GC Howard, CRC Books, 2006, ISBN 0849335280). Polyclonal and monoclonal antibodies are often purified using protein A / G or antigen affinity chromatography.

[0110] The antibodies or antigen-binding fragments described herein may be prepared as monoclonal anti-B7H4 antibodies, which can be prepared using the hybridoma method (e.g., the method described by Kohler and Milstein, Nature 256:495 (1975)). Using the hybridoma method, mice, hamsters, or other suitable host animals are immunized as described above, inducing lymphocyte production of antibodies that will specifically bind to the immune antigen. Lymphocytes can also be immunized in vitro. After immunization, lymphocytes can be isolated and fused with a suitable myeloma cell line, for example using polyethylene glycol, to form hybridoma cells, from which unfused lymphocytes and myeloma cells can be selectively removed. Next, hybridomas that produce monoclonal antibodies specifically directed against selected antigens, as determined by immunoprecipitation, immunoblotting, or in vitro binding assays (e.g., radioimmunoassay (RIA) or enzyme-linked immunosorbent assay (ELISA)), can be grown in vitro using standard methods (Goding, Monoclonal Antibodies: Principles and Practice, Academic Press, 1986) or in vivo as ascites tumors in animals. The monoclonal antibodies can then be purified from the culture medium or ascites using known methods.

[0111] Alternatively, antibodies or their antigen-binding fragments (e.g., as monoclonal antibodies) can also be prepared using recombinant DNA methods as described in U.S. Patent No. 4,816,567. Polynucleotides encoding the monoclonal antibody are isolated from mature B cells or hybridoma cells by RT-PCR or other methods using oligonucleotide primers that specifically amplify the genes encoding the antibody's heavy and light chains, and their sequences are determined using conventional procedures.

[0112] Next, isolated polynucleotides encoding the heavy and light chains are cloned into a suitable expression vector, and when this is transfected into host cells such as Escherichia coli (E. coli) cells, monkey COS cells, Chinese hamster ovary (CHO) cells, or myeloma cells, which normally do not produce immunoglobulin proteins, monoclonal antibodies are produced by the host cells. Alternatively, recombinant monoclonal antibodies or their antigen-binding fragments of the desired species can be isolated from phage display libraries expressing the desired species' CDR, as described in McCafferty et al., Nature 348:552-554 (1990); Clackson et al., Nature 352:624-628 (1991); and Marks et al., J.Mol.Biol.222:581-597 (1991).

[0113] Alternative antibodies can be produced by further modifying the polynucleotide encoding the antibody or its antigen-binding fragment of the present invention in several different ways using recombinant DNA technology. In some embodiments, for example, the constant domains of the light and heavy chains of a mouse monoclonal antibody can be (1) replaced with those regions of a human antibody, for example, to produce a chimeric antibody, or (2) replaced with a non-immunoglobulin polypeptide to produce a fusion antibody. In some embodiments, the constant region is truncated or removed to produce the desired antibody fragment of the monoclonal antibody. The specificity, affinity, etc., of the monoclonal antibody can be optimized using site-directed mutagenesis or high-density mutagenesis of the variable region.

[0114] In one embodiment, the antibody or its antigen-binding fragment is a human antibody or its antigen-binding fragment. Human antibodies can be prepared directly using various methods known in the art. Immortalized human B lymphocytes can be prepared, either immunized in vitro or isolated from immunized individuals that produce antibodies against a target antigen. See, for example, Cole et al., Monoclonal Antibodies and Cancer Therapy, Alan R. Liss, p. 77 (1985); Boemer et al., J. Immunol. 147(1): 86-95 (1991); and U.S. Patent No. 5,750,373.

[0115] In one embodiment, an antibody or its antigen-binding fragment can be selected from a phage library that expresses human antibodies, for example, as described in Vaughan et al., Nat. Biotech. 14:309-314 (1996); Sheets et al., Proc. Natl. Acad. Sci. USA, 95:6157-6162 (1998); Hoogenboom and Winter, J. Mol. Biol. 227:381 (1991); and Marks et al., J. Mol. Biol. 222:581 (1991). Methods for the preparation and use of antibody phage libraries are also described in U.S. Patent Nos. 5,969,108, 6,172,197, 5,885,793, 6,521,404; 6,544,731; 6,555,313; 6,582,915; 6,593,081; 6,300,064; 6,653,068; 6,706,484; and 7,264,963; as well as in Rothe et al., J. Molec. Biol. 376:1182-1200 (2008) (each of which is incorporated in whole by reference).

[0116] Affinity maturation strategies and chain shuffling strategies are known in this technology and can be used to produce high-affinity human antibodies or their antigen-binding fragments. See Marks et al., BioTechnology 10:779-783 (1992) (the entire work is incorporated by reference).

[0117] In one embodiment, the antibody or its antigen-binding fragment (e.g., a monoclonal antibody) may be a humanized antibody. Methods for manipulating, humanizing, or resurfacing non-human or human antibodies may also be used, and these are well known in the art. Humanized antibodies, resurfacing antibodies, or similarly manipulated antibodies may have one or more amino acid residues derived from a non-human source, such as, but not limited to, mouse, rat, rabbit, non-human primate, or other mammal. These non-human amino acid residues are often substituted by residues called “implant” residues, which are usually obtained from “implant” variable domains, constant domains, or other domains of known human sequences. Such implanted sequences can be used to reduce immunogenicity or to reduce, enhance, or modify binding, affinity, on-rate, off-rate, binding activity, specificity, half-life, or any other desirable properties known in the art. Preferably, CDR residues may be directly and most substantially involved in the effect on B7-H4 binding. Therefore, while it is preferable to maintain some or all of the non-human or human CDR sequence, the non-human sequences in the variable and constant regions can be replaced with human or other amino acids.

[0118] Antibodies can also be, at the discretion of the user, humanized, resurfacing, manipulated, or human antibodies can be manipulated to retain high affinity for the antigen B7-H4 and other advantageous biological properties. To achieve this goal, humanized (or human) or manipulated anti-B7H4 antibodies and resurfacing antibodies can be, at the discretion of the user, prepared by analytical processes of the parent sequence and various conceptual humanized and manipulated products using three-dimensional models of the parent sequence, manipulated sequence, and humanized sequence. Three-dimensional immunoglobulin models are commonly available and well known to those skilled in the art. Computer programs are available that illustrate and display the expected three-dimensional conformational structure of selected candidate immunoglobulin sequences. By examining these representations, it is possible to analyze the expected role of residues in the function of the candidate immunoglobulin sequence, i.e., analyze residues that affect the ability of the candidate immunoglobulin to bind to its antigen, such as B7-H4. In this way, FW residues can be selected and combined from consensus sequences and transposition sequences, thereby achieving desired antibody properties, such as increased affinity for the target antigen.

[0119] Humanization, resurfacing, or manipulation of the anti-B7H4 antibody or its antigen-binding fragment of the present invention may be carried out by any known method, for example, but not limited to, Jones et al., Nature 321:522 (1986); Riechmann et al., Nature 332:323 (1988); Verhoeyen et al., Science 239:1534 (1988); Sims et al., J.Immunol.151:2296 (1993); Chothia and Lesk, J.Mol.Biol.196:901 (1987); Carter et al., Proc.Natl.Acad.Sci.USA 89:4285 (1992); Presta et al. al., J. Immunol. 151:2623 (1993); US Patent No. 5,639,641, US Patent No. 5,723,323; , 824,514; 5,817,483; 5,814,476; 5,763,192; 5,723,323; 5,766 ,886 specification; 5,714,352 specification; 6,204,023 specification; 6,180,370 specification; 5,693,762 specification; 5,530,101 No. 5,585,089; No. 5,225,539; No. 4,816,567; No. 7,557,189; No. 7,538,195 ;and specification No. 7,342,110; International application specification No. PCT / US98 / 16280; PCT / US96 / 18978; PCT / US91 / 09630; PCT / US91 / 05939; PCT / US94 / 01234; PCT / GB89 / 01334; PCT / GB91 / 01134; PCT / GB92 / 01755; International publication brochure No. 90 / 14443; brochure No. 90 / 14424; brochure No. 90 / 14430; and European Patent No. 229246 (each of these, including the references cited therein, is incorporated herein in whole by reference).

[0120] Anti-B7H4 humanized antibodies and their antigen-binding fragments can also be produced in transgenic mice containing a human immunoglobulin locus capable of producing a complete repertoire of human antibodies without producing endogenous immunoglobulins during immunization. This approach is described in U.S. Patents No. 5,545,807; No. 5,545,806; No. 5,569,825; No. 5,625,126; No. 5,633,425; and No. 5,661,016.

[0121] In one embodiment, fragments of an antibody (e.g., an anti-B7H4 antibody) are provided (e.g., antibody fragments). Various methods are known for generating antibody fragments. Conventionally, these fragments are obtained by proteolytic digestion of intact antibodies, as described, for example, by Morimoto et al., J. Biochem. Biophys. Meth. 24:107-117 (1993) and Brennan et al., Science 229:81 (1985). In one embodiment, anti-B7H4 antibody fragments are generated by recombination. Fab, Fv, and scFv antibody fragments can all be expressed in Escherichia coli (E. coli) or other host cells and secreted therefrom, thereby enabling the generation of these fragments in large quantities. Such anti-B7-H4 antibody fragments can also be isolated from antibody phage libraries as discussed above. Anti-B7-H4 antibody fragments may also be linear antibodies as described in U.S. Patent No. 5,641,870. Other methods for generating antibody fragments will be obvious to those skilled in the art.

[0122] According to the present invention, the method can be adapted for the production of single-chain antibodies specific to B7-H4. See, for example, U.S. Patent No. 4,946,778. In addition, the method can be adapted for the construction of Fab expression libraries that enable the rapid and effective identification of monoclonal Fab fragments, or their derivatives, fragments, analogs, or homologs having desired specificity for B7-H4. See, for example, Huse et al., Science 246:1275-1281 (1989). Antibody fragments may be produced by methods known in the art, including, but not limited to, F(ab')2 fragments produced by pepsin digestion of antibody molecules; Fab fragments produced by reduction of disulfide crosslinks of F(ab')2 fragments; Fab fragments produced by treating antibody molecules with papain and a reducing agent; or Fv fragments.

[0123] In one embodiment, an antibody or its antigen-binding fragment described herein may be modified to increase its serum half-life. This can be achieved, for example, by incorporating a salvage receptor-binding epitope into the antibody or antibody fragment by mutation in an appropriate region of the antibody or antibody fragment, or by incorporating the epitope into a peptide tag and subsequently fusing it to the antibody or antibody fragment at either terminal or central (e.g., by DNA or peptide synthesis), or by YTE mutation. Other methods for increasing the serum half-life of an antibody or its antigen-binding fragment, such as conjugation with heterologous molecules like PEG, are known in the art.

[0124] Modified antibodies or their antigen-binding fragments as provided herein may include any type of variable region resulting in association between the antibody or polypeptide and B7-H4. In this regard, the variable region may include or be derived from any type of mammal capable of initiating a humoral response to a desired antigen and inducing immunoglobulin production. Thus, the variable region of an anti-B7H4 antibody or its antigen-binding fragment may be of human, mouse, non-human primate (e.g., cynomolgus monkey, macaque, etc.), or wolf (lupine) origin. In one embodiment, both the variable and constant regions of the modified antibody or its antigen-binding fragment are human. In one embodiment, the variable region of a compatible antibody (usually derived from a non-human source) may be manipulated or specifically modified to improve the binding properties of the molecule or reduce its immunogenicity. In this regard, variable regions useful in the present invention may be humanized by including an imported amino acid sequence or otherwise modified.

[0125] In one embodiment, both the heavy and light chain variable domains of an antibody or its antigen-binding fragment are modified by at least partial substitution of one or more CDRs and / or by partial substitution and resequencing of a framework region. The CDRs may originate from the same class or even the same subclass antibody from which the framework region originates, but it is assumed that the CDRs originate from a different class of antibody, and in certain embodiments from a different species of antibody. It is not necessary to replace all CDRs with the complete CDR of the donor variable region in order to transfer the antigen-binding ability of one variable domain to another. Rather, it is sufficient to transfer only the residues necessary to maintain the activity of the antigen-binding site. Considering the descriptions in U.S. Patents No. 5,585,089, No. 5,693,761, and No. 5,693,762, it is well within the capabilities of those skilled in the art to perform the usual experiments to obtain functional antibodies with reduced immunogenicity.

[0126] Despite the modification of the variable region, those skilled in the art will understand that the modified antibodies or antigen-binding fragments described herein will include antibodies (e.g., full-length antibodies or antigen-binding fragments thereof) in which one or more portions of the constant region domain are deleted or otherwise modified to produce desired biochemical properties, such as increased tumor localization or reduced serum half-life, compared to antibodies of substantially the same immunogenicity that include the natural or unmodified constant region. In one embodiment, the constant region of the modified antibody will include the human constant region. Modifications to the constant region that are compatible with the present invention include the addition, deletion, or substitution of one or more amino acids in one or more domains. That is, the modified antibodies disclosed herein may include modifications or alterations to one or more of the three heavy chain constant domains (CH1, CH2, or CH3) and / or the light chain constant domain (CL). In one embodiment, a modified constant region in which one or more domains are partially or completely deleted is intended. In one embodiment, the modified antibody will include a domain deletion construct or variant (ΔCH2 construct) in which the entire CH2 domain is removed. In one embodiment, the removed constant region domain can be replaced with a short amino acid spacer (e.g., 10 residues) that provides a certain degree of molecular flexibility typically granted by the missing constant region.

[0127] In addition to these structures, constant regions are known in this field to mediate several effector functions. For example, antibodies bind to cells via their Fc region, and the Fc receptor site on the antibody Fc region binds to Fc receptors (FcRs) on the cell. Several Fc receptors exist that are specific to different antibody classes, including IgG (gamma receptor), IgE (eta receptor), IgA (alpha receptor), and IgM (mu receptor). When antibodies bind to Fc receptors on the cell surface, several important and diverse biological reactions are triggered, including phagocytosis and destruction of antibody-coated particles, clearance of immune complexes, lysis of antibody-coated target cells by killer cells (known as antibody-dependent cell-mediated cytotoxicity, or ADCC), release of inflammatory mediators, placental cross-transfer, and regulation of immunoglobulin production.

[0128] In one embodiment, the antibody or its antigen-binding fragment results in a modified effector function that consequently affects the biological profile of the administered antibody or its antigen-binding fragment. For example, deletion or inactivation of the constant region domain (by point mutation or other means) may reduce Fc receptor binding of the modified antibody in circulation. In other cases, modification of the constant region may, in accordance with the present invention, regulate complement binding and therefore reduce the serum half-life and nonspecific association of the conjugated cytotoxin. Further modifications of the constant region may be used to remove disulfide bonds or oligosaccharide moieties that enable enhanced localization by increasing antigen specificity or antibody mobility. Similarly, modifications to the constant region according to the present invention can be readily fabricated using well-known biochemical or molecular engineering techniques that are well within the scope of the art.

[0129] In one embodiment, the antibody or its antigen-binding fragment does not have one or more effector functions. For example, in one embodiment, the antibody or its antigen-binding fragment does not have antibody-dependent cell-mediated cytotoxicity (ADCC) activity and / or complement-dependent cell-mediated cytotoxicity (CDC) activity. In one embodiment, the antibody or its antigen-binding fragment does not bind to Fc receptors and / or complement factors. In one embodiment, the antibody or its antigen-binding fragment does not have effector functions.

[0130] In one embodiment, the antibody or its antigen-binding fragment may be manipulated so that the CH3 domain of each modified antibody or fragment directly fuses to the hinge region. In other constructs, a peptide spacer may be inserted between the hinge region and the modified CH2 and / or CH3 domains. For example, a compatible construct can be expressed in which the CH2 domain is deleted and the remaining CH3 domain (modified or unmodified) is bound to the hinge region by a 5-20 amino acid spacer. By adding such a spacer, it is possible to ensure, for example, that the regulatory elements of the constant domain remain free and accessible, or that the hinge region remains flexible. In some cases, the amino acid spacer is found to be immunogenic and induce an undesirable immune response to the construct. In one embodiment, in order to maintain the desired biochemical quality of the modified antibody, any spacers added to the construct may be relatively unimmunogenic or even omitted entirely.

[0131] In addition to deletion of the entire constant region domain, the antibodies or antigen-binding fragments provided herein can be modified by partial deletion or substitution of several or even a single amino acid in the constant region. For example, a mutation in a single amino acid within a selected range of the CH2 domain may be sufficient to substantially reduce Fc binding, thereby increasing tumor localization. Similarly, one or more constant region domains that control effector function (e.g., complement C1Q binding) can be completely or partially deleted. Such partial deletions of the constant region can improve selected properties (e.g., serum half-life) of the antibody or antigen-binding fragment while preserving other desirable functions associated with the constant region domain in question. Furthermore, the constant regions of antibodies and their antigen-binding fragments can be modified by mutations or substitutions of one or more amino acids that enhance the profile of the resulting construct. In this regard, it is possible to interfere with the activity provided by a conserved binding site (e.g., Fc binding) while substantially maintaining the arrangement and immunogenicity profile of the modified antibody or antigen-binding fragment. In one embodiment, the addition of one or more amino acids to the constant region may be present to enhance desirable properties such as a decrease or increase in effector function, or to provide binding of more cytotoxins or carbohydrates. In one embodiment, it may be desirable to insert or replicate a specific sequence derived from a selected constant region domain.

[0132] The antibodies and antibody-binding fragments described herein further encompass variants and equivalents that are substantially homologous to the antibodies or antigen-binding fragments specifically described herein (e.g., mouse, chimeric, humanized, or human antibodies, or their antigen-binding fragments). These may, for example, include conservative substitution mutations, i.e., substitutions of one or more amino acids with similar amino acids. For example, a conservative substitution refers to the substitution of an amino acid with another amino acid within the same general class, such as substituting one acidic amino acid with another acidic amino acid, one basic amino acid with another basic amino acid, or one neutral amino acid with another neutral amino acid. The intended effects of conservative amino acid substitutions are well known in the art.

[0133] In one embodiment, an antibody or its antigen-binding fragment can be further modified to include additional chemical portions that are not normally part of the protein. These derivatized portions can improve the protein's solubility, biological half-life, or absorption. These portions can also reduce or eliminate any desirable side effects of the protein. An overview of these portions can be found in Remington's Pharmaceutical Sciences, 22nd ed., Lloyd V. Allen, Jr. (2012).

[0134] VIII. Use of anti-B7H4 antibody drug conjugates Preferably, proliferative disorders can be treated using conjugates. The term “proliferative disorder” relates to the unwanted or uncontrolled proliferation of excess or abnormal cells, such proliferation being undesirable, e.g., tumorigenic or hyperplastic proliferation, in vitro or in vivo. The term “proliferative disorder” may be alternatively referred to as “cancer.”

[0135] A suitable proliferative disorder (e.g., cancer) would preferably be characterized by the presence of cancer cells expressing B7-H4.

[0136] Examples of proliferative states include, but are not limited to, benign, premalignant, and malignant cell proliferations, such as neoplasms and tumors (e.g., histiocytoma, glioma, astrocytoma, osteoma), cancers (e.g., lung cancer, small cell lung cancer, gastrointestinal cancer, intestinal cancer, colorectal cancer, breast cancer, ovarian cancer, prostate cancer, testicular cancer, liver cancer, kidney cancer, bladder cancer, pancreatic cancer, brain cancer, sarcoma, osteosarcoma, Kaposi's sarcoma, melanoma), leukemia, psoriasis, bone diseases, fibroproliferative disorders (e.g., those of connective tissue), and atherosclerosis. Other cancers covered include, but are not limited to, hematological malignancies such as leukemia, and lymphomas such as non-Hodgkin lymphoma, as well as subtypes such as DLBCL, marginal zone, mantle, and follicular, Hodgkin lymphoma, AML, and other cancers of B or T cell origin. It can treat any type of cell, including, but not limited to, the lungs, gastrointestinal tract (e.g., intestines, large intestine), milk (breast), ovaries, prostate, liver (liver), kidneys (kidneys), bladder, pancreas, brain, and skin.

[0137] In some embodiments, the antibody-drug conjugate described herein is used to treat cancer selected from the list, including breast cancer, ovarian cancer, endometrial cancer, cholangiocarcinoma, lung cancer, pancreatic cancer, and gastric cancer. In some embodiments, the antibody-drug conjugate described herein is used to treat cancer selected from the list, including breast cancer, ovarian cancer, endometrial cancer, and cholangiocarcinoma. In some embodiments, the breast cancer is hormone receptor-positive (HR+) breast cancer. In some embodiments, the breast cancer is human epidermal growth factor receptor 2-positive (HER2+) breast cancer. In other embodiments, the breast cancer is trinegative breast cancer (TNBC). In some embodiments, the lung cancer is non-small cell lung cancer (NSCLC). In some embodiments, the NSCLC is squamous cell carcinoma. In other embodiments, the NSCLC is adenocarcinoma.

[0138] The antibody-drug conjugate of the present invention can delay, suppress, and destroy cancer cell proliferation. These actions can relieve symptoms caused by cancer, improve the quality of life (QOL) of cancer patients, maintain the life of cancer patients, and thereby achieve therapeutic effects. Even if cancer cells are not destroyed, suppressing or controlling their proliferation can lead to long-term survival and improved QOL for cancer patients.

[0139] The anti-H7B4 ADC of the present invention is expected to exert therapeutic effects when applied to patients as a systemic therapy, and also when applied locally to cancerous tissue.

[0140] Anti-H7B4 ADC can be administered as a pharmaceutical composition together with one or more pharmaceutically compatible components. The pharmaceutically compatible components can be appropriately selected from pharmaceutical additives commonly used in the art, taking into consideration the dosage and concentration of anti-H7B4 ADC. For example, the anti-H7B4 ADC used in the present invention can be administered as a pharmaceutical composition containing a buffer such as a histidine buffer, an excipient such as sucrose, and a surfactant such as polysorbate 80. The pharmaceutical composition containing anti-H7B4 ADC used in the present invention can be used as an injectable, aqueous injectable, or lyophilized injectable, or lyophilized injectable.

[0141] If the pharmaceutical composition containing the anti-H7B4 ADC used in the present invention is an aqueous injection, it may be diluted with a suitable diluent and subsequently administered by intravenous injection. Examples of diluents may include dextrose solution and physiological saline.

[0142] If the pharmaceutical composition containing the anti-H7B4 antibody drug conjugate used in the present invention is a lyophilized injectable preparation, it may be dissolved in injection-grade water, subsequently diluted to the required volume with a suitable diluent, and then administered as an intravenous injection. Examples of diluents include dextrose solution and physiological saline.

[0143] Examples of possible routes of administration for administering the pharmaceutical compositions described herein include intravenous, intradermal, subcutaneous, intramuscular, and intraperitoneal routes.

[0144] IX. A method for predicting patient response to anti-B7H4 antibody drug conjugates. Figure 14 is a flowchart of steps 1-6 of Method 7, in which an analysis system analyzes digital images of tissue from cancer patients to predict how likely a cancer patient is to respond to therapy containing an anti-B7H4 antibody drug conjugate (ADC). In one embodiment, the method predicts the response to ADC in patients with cancer selected from the group consisting of breast cancer, ovarian cancer, endometrial cancer, cholangiocarcinoma, lung cancer, pancreatic cancer, and gastric cancer. In one embodiment, the method predicts the response to ADC in patients with cancer selected from the group consisting of breast cancer, ovarian cancer, endometrial cancer, and cholangiocarcinoma. In one embodiment, the method predicts the response to ADC in patients with breast cancer. In one embodiment, the method predicts the response to ADC in patients with ovarian cancer. In one embodiment, the method predicts the response to ADC in patients with endometrial cancer.

[0145] In the first step 1, high-resolution digital images are obtained from tissue sections from cancer patients that have been stained using one or more biomarkers or stains.

[0146] To predict the effectiveness of ADC therapy, diagnostic biomarkers are used that have conjugated dyes that target the same protein targeted by the ADC therapy. In one embodiment, the anti-B7H4 ADC therapy to be scored is an anti-B7H4 antibody conjugated to a topoisomerase I inhibitor. In one embodiment, the anti-B7H4 ADC is [ka] It contains an anti-B7H4 antibody conjugated to a topoisomerase I inhibitor selected from the following.

[0147] In one embodiment, the anti-B7H4 ADC has the following structure: [ka] It contains an anti-B7H4 antibody conjugated to the topoisomerase inhibitor SG3932, represented by [the specified agent].

[0148] Therefore, in this embodiment, the diagnostic biomarker also targets the B7-H4 protein.

[0149] In step 2, a pre-trained convolutional neural network processes digital images of cancer patient tissue stained with a diagnostic antibody linked to a dye such as 3,3'-diaminobenzidine (DAB). The staining intensity of the dye in the cancer cell membrane is determined based on the average staining intensity of the dye across all pixels associated with the corresponding segmented membrane target. Furthermore, the staining intensity of the dye in a single pixel is computer-processed based on the red, green, and blue components of the pixel. The result of the image analysis processing is two back-image layers for each pixel in the digital image, representing the likelihood that the pixel belongs to the cell nucleus and the likelihood that the pixel belongs to the cell membrane.

[0150] In another embodiment of step 2, two pre-trained convolutional networks process a digital image of the tissue. The result of processing by the first network is a back image layer representing the likelihood that each pixel in the digital image belongs to a cell nucleus. The result of processing by the second network is a back image layer representing the likelihood that each pixel in the digital image belongs to a cell membrane.

[0151] In step 3, individual cancer cells are detected based on heuristic image analysis of the nucleus and membrane in the back layer. Cancer cell targets are generated, including cell membrane targets and, optionally, cell cytoplasmic targets.

[0152] In step 4, a single-cell ADC score is determined for each cancer cell. The single-cell ADC score is based on (1) the amount of DAB in the cell membrane and (2) the amount of ADC payload uptake. The amount of DAB is determined by the staining intensity of each membrane based on the average optical density of the brown diaminobenzidine (DAB) signal in the membrane pixels. The amount of ADC payload uptake is estimated based on the amount of DAB in the cell membrane and, optionally, in the cell cytoplasm, and, optionally, the amount of DAB in the membrane and cytoplasm of adjacent cells relative to the cell being scored. The amount of DAB in the cell cytoplasm is determined by the staining intensity of each cytoplasm, which is computer-processed based on the average optical density of the brown DAB signal in the cytoplasmic pixels. The amount of DAB in adjacent cells is determined for those cancer cells within a predetermined distance of the cell being scored.

[0153] In step 5, patient scores are computer-processed with respect to digital images of tissue based on statistical operations on the single-cell ADC score of all cancer cells in the digital images. Patient scores indicate how cancer patients will respond to therapies including anti-B7H4 ADCs. The parameters of the single-cell ADC score in step 4 and the type of statistical operation in step 5 are optimized using a training cohort of patients with a known response to ADC therapy. The goal of optimization is a low p-value in the Kaplan-Meier analysis of the score-positive group compared to score-negative patients in the training cohort.

[0154] In step 6, therapy including anti-B7H4 ADC is recommended for score-positive patients if the score is greater than a predetermined threshold.

[0155] X. Examples of prediction and scoring methods. A. Image analysis of stained tissue. The method shown in Figure 14 is described here in relation to specific images of stained cancerous tissue.

[0156] In step 1, the tissue sample is immunohistochemically stained using a dye linked to a diagnostic antibody that binds to relevant proteins on cancer cells in the tissue sample. Figure 15 (upper left image) is a digital image 8 of a portion of the stained tissue obtained in step 1. Image 8 shows tissue from a cancer patient immunohistochemically stained with an anti-B7H4 diagnostic antibody linked to a dye. In this example, the diagnostic antibody is an anti-B7H4 generating clone (represented as B7H4 IHC tool Ab) reformatted as a mouse anti-human B7H4 IgG1 clone. IgG1 indicates the isotype of the anti-B7H4 antibody. The staining protocol was developed using the Dako automated staining platform. The anti-B7H4 antibody binds to the membrane protein B7-H4 so that 3,3'-diaminobenzidine (DAB) staining indicates the location of protein B7-H4 in the tissue sample.

[0157] In step 2, image analysis is performed on digital image 8 to generate a back image layer of cancer cell nuclei and membranes using a convolutional neural network. Image analysis is used to detect cancer cells and their components, such as the nucleus, membrane, and cytoplasm. Figure 15 shows the image analysis process in step 2. For each pixel of digital image 8, the convolutional neural network generates a back layer (grayscale image) indicating the likelihood that each pixel belongs to either the cell nucleus (upper right image in Figure 15) or the membrane (lower left image in Figure 15). High probability is shown in black, and low probability is shown in white.

[0158] In one embodiment, the convolutional neural network includes a series of convolutional layers from an input image 8 directed to a bottleneck layer having a very small spatial size (1 to 16 pixels), and a series of inverse convolutional layers directed to a backward layer having the same size as the input image 8. This network structure is called a U-Net. The weight training of the convolutional neural network is performed by generating manual annotation layers for nuclei and membranes in multiple training images, and then adjusting the network weights using an optimization algorithm so that the generated backward layer most closely resembles the manually generated annotation layer.

[0159] In another embodiment, nucleus and membrane annotation layers are automatically generated across multiple training images and manually corrected. Epithelial regions and nuclear centers are manually annotated as regions and points, respectively. For each training image, membrane segmentation is automatically generated by applying a region growth-like algorithm (e.g., watershed segmentation) that is factored by the annotated nuclear centers and limited by the extent of the annotated epithelial regions. Taking the training images into account, nuclear segmentation is automatically generated by applying a blob detection algorithm (e.g., the Maximum Stable Extreme Region MSER algorithm) and selecting only the detected blobs containing nuclear centers annotated as nuclei. The automatically generated membrane and nuclear segmentation is visually reviewed and manually corrected if necessary. The correction process includes one of the following methods: rejecting inaccurately segmented membranes or nuclei, explicitly acknowledging accurately annotated membranes or nuclei, or refining the shape of the membranes or nuclei. For each image with annotated membranes or nuclei, an annotation layer is generated. In one embodiment, each pixel in the annotation layer is assigned "1" if it belongs to an annotated object (membrane or nucleus); otherwise, it is assigned "0". In another embodiment, the pixels in the annotation layer represent the distance to the nearest annotated object. The network weights are adjusted by an optimization algorithm so that the generated back layers are most similar to the automatically generated membrane and nucleus annotation layers.

[0160] Figure 16 shows step 3, in which individual cancer cell targets, each containing the cell membrane and cytoplasm, are detected. The heuristic image analysis process uses watershed segmentation to segment the cell nuclei using a back layer of nuclei generated by a convolutional neural network. Segmentation generates nuclear targets. Each nuclear target is assigned a unique identifier (UID). Individually identified nuclei are shown as dense targets in Figure 16 (bottom right image). Detected nuclei are also displayed as overlays in the input image 8 (top left image) and in the back layers relating to the nuclei (top right image) and membrane (bottom left image).

[0161] In one embodiment, watershed segmentation includes binarization of the layers behind nuclei having a predetermined first size threshold. All single concatenated pixels exceeding the first size threshold are considered to belong to the nucleus. Nuclei with an area smaller than 16 μm² are rejected. A UID is assigned to each nucleus. In a subsequent step, the nuclei grow toward the smaller nuclei's rear portions, where the pixels behind the added nuclei must be greater than a second predetermined threshold.

[0162] Figure 17 shows a further segmentation step that uses nuclear markers to detect and improve membrane segmentation. The region growth algorithm uses the detected nuclear markers as seeds to grow up to the membrane ridges in the rear layer of the membrane (approximating cell boundaries). The detected membrane (bottom right image) is shown as an overlay in the rear layer with respect to input image 8 (top left image) and nuclei (top right image) and membrane (bottom left image).

[0163] Figure 18 illustrates the detection and segmentation of membrane targets, which are segmented by growing the region of cell boundary pixels detected outside the potential membrane layer and behind a predetermined membrane layer threshold. Darker boundary regions indicate membrane targets. Each membrane target is assigned the same UID as its associated nuclear target.

[0164] The space between the membrane and the nucleus is assigned to the cytoplasm using the nuclear UID. For each membrane (see image in the upper left of Figure 18) and cytoplasm (see image in the upper left of Figure 18), the average optical density of DAB staining, along with the UID, is exported to a file on the hard drive. For each cell (defined if it includes the nucleus, cytoplasm, and membrane), the centroid position (x,y) of the cell within the glass slide is also exported. The file may reside on a hard disk, solid-state disk, or part of dedicated RAM in a computer system.

[0165] Figure 19 shows the results of image analysis in an image analysis software environment. Figure 19 (top left image) shows the separation of nuclear and membrane targets as overlays on a digital image of stained tissue. Figure 19 (bottom left image) shows the separation of nuclear targets as overlays on the optical density display of the digital image. Higher optical density pixels have a large amount of DAB, and higher optical density pixels have a small amount of DAB. The DAB optical density of each image pixel is computer-processed from the red-green-blue display of the image pixel by a red-green-blue space transformation such that the brown DAB component becomes an independent color, and by taking the logarithm of the brown component. Figure 19 (top right image) and Figure 20 (top panel) show the image analysis script used to generate the segmented images within the Definiens Developer XD platform. Figure 19 (bottom right image) and Figure 20 (bottom panel) show the exported measurements for all cell membrane and cytoplasmic targets in Image 8.

[0166] B. Calculation of predictive ADC score. A single-cell ADC score is computer-processed for each cancer cell in digital image 8 based on the optical density of DAB staining in membrane targets and, optionally, in cytoplasmic targets. The single-cell ADC score is also optionally based on the staining intensity of DAB dye in membrane targets and cytoplasmic targets of adjacent cancer cells that are closer than a predetermined distance to the cancer cell for which the single-cell ADC score is being computer-processed. The aggregation of single-cell ADC scores provides a predictor of the cancer patient's response to anti-B7H4 ADC therapy.

[0167] Figure 21 shows exemplary quantitative results of the optical density of staining from the image analysis of steps 2-3 in the schematic diagram, using gray values ​​for membrane and cytoplasmic pixels. Exemplary compartments into the cell nucleus, cell membrane, and cell cytoplasm in Figure 21 are obtained using the heuristic image analysis steps shown in Figures 15-20. In Figure 21, light gray values ​​are associated with high DAB optical density and therefore with a large amount of protein targeted by the diagnostic antibody. Dark gray values ​​are associated with low DAB optical density. Brighter pixels correspond to higher DAB optical density.

[0168] Figure 22 lists exemplary quantitative amounts of membrane and cytoplasmic staining in the images of Figure 21, and is partially reproduced in Figure 22. The optical densities of brown DAB signals from the membranes of the first, second, and third cells are 0.949, 0.369, and 0.498, respectively. The optical densities of brown DAB signals from the cytoplasm of the first, second, and third cells are 0.796, 0.533, and 0.369, respectively. In the schematic image of Figure 22, the first cancer cell 9 expresses a large amount of target protein B7-H4 and is very likely to be killed by the ADC payload entering the cell linked to the ADC antibody (Result 1 in Figure 23). The second cancer cell 10 and the third cancer cell 11 do not express sufficient amounts of target protein B7-H4 to be directly killed by anti-B7H4 ADC. However, due to the proximity of the second cancer cell 10 to the first cancer cell 9, the toxic payload released from the first cancer cell will also kill the second cancer cell 10 (Result 3 in Figure 23). The third cancer cell 11 remains active and may be the starting point for drug resistance mechanisms, ultimately leading to death in the patient.

[0169] Figure 23 illustrates the mechanism by which anti-B7H4 ADC therapy kills cancer cells. In the first step, the ADC antibody binds to the target protein B7-H4, inhibiting the innate function of the target protein that can lead to cell death. In the second step, the payload (e.g., a type I topoisomerase inhibitor) is internalized into the cell, killing the cell through the toxicity of the payload. This payload uptake depends on the amount of target protein on the membrane, as well as the difference in the amount of target protein on the membrane and in the cytoplasm. After uptake, the payload may be released from the cell into the surrounding tissue. In the third step, the payload may enter nearby cells and kill them as well. The spatial distribution of the payload in the tissue is achieved by passive diffusion.

[0170] Conventional IHC scoring reflects both the effect of target protein inhibition due to ADC binding and the effect of the cytotoxic payload that enters cancer cells along with the ADC antibody. Therefore, conventional scoring for ADC therapy does not reflect the importance of the presence of the target protein in the cytoplasm and the effect of the cytotoxic payload that diffuses into the tissue after being released from the first dead cancer cell. In contrast, the novel predictive ADC score measures the effect of the release of the cytotoxic payload on adjacent cancer cells.

[0171] In step 4 of the method in Figure 14, a single-cell ADC score is determined for each cancer cell. Figure 24 shows the calculation of the single-cell ADC score for each of the three cells shown in Figure 22 based on cell separation that reflects the uptake of the ADC payload into adjacent cells, incorporating exponential weighting coefficients. The single-cell score can be calculated based on the formula shown in Figure 25 or the formula enumerated in claim 4. The optical densities enumerated in Figure 22 with respect to the DAB signal from the cell membrane (0.949, 0.369, and 0.498) and cytoplasm (0.796, 0.533, and 0.369) are inputs to the calculation shown in Figure 24. Cells 9-11, 9, and 9, 9, and 11, respectively, have single-cell scores of 0.145, 0.012, and 0.064.

[0172] Therefore, the single-cell ADC score incorporates a measurement of the amount of target protein on the cell membrane using DAB optical density and, optionally, an estimation of the amount of ADC payload uptake. As shown in Figure 23, the uptake of the ADC payload with respect to a first cell depends on both the amount of dye in its membrane and cytoplasm, as well as the amount of dye in the membrane and cytoplasm of a second cell located in the vicinity of the first cell. More specifically, the vicinity may be a disk with a predetermined radius surrounding the first cancer cell. In one embodiment, the single-cell ADC score with respect to the first cancer cell is determined by a distance-weighted sum of several powers of the DAB optical density of the membrane and cytoplasmic target, where the center of the relevant cancer cell is closer to the center of the first cancer cell than a predetermined distance. In one embodiment, the predetermined distance is 50 μm, as used in the calculation in Figure 24. In another embodiment, the distance is 20 μm. In yet another embodiment, the distance weighting involves computer processing an exponential function of adjusted negative Euclidean distances in total from the center of the first cancer cell to the center of the other cancer cell. In another embodiment, the exponents in the sum are limited to 0, 1, and 2 (constant, linear term, square).

[0173] Figure 25 shows one embodiment of the formula for calculating the single-cell ADC score. The function a in the formula kl The distance from cell j to cell i is |r j -r i | Depends on ODM j This is the DAB optical density of the cell membrane, and ODC. j is the DAB optical density of the cytoplasm of cell j. The constants A_ij, r_norm, and d are the same for all types of cancer. However, the thresholds for the score used to determine whether a patient is eligible for ADC therapy are not the same for different types of cancer.

[0174] In step 5 of a method for demonstrating how a particular cancer patient will respond to ADC therapy, a response score is computer-processed with respect to a digital image 8 of tissue from the cancer patient based on staining for the target protein B7-H4, such that the score indicates how the cancer patient will respond to anti-B7H4 ADC therapy. The response score is generated by aggregating all single-cell ADC scores from the tissue sample using a statistical operation. Thus, the score is computer-processed based on a statistical value of the single-cell ADC score for all cancer cells detected in the digital image. In one embodiment, the statistical value is a default displacement value of the ADC payload uptake estimate for all cancer cells in image 8.

[0175] In another embodiment, the formula for calculating the single-cell ADC score considers only the optical density of the stained membrane of each cancer cell and does not consider cytoplasmic staining or whether other cancer cells are located within a predetermined distance of the cancer cell from which the single-cell ADC score is being calculated.

[0176] In step 6 of the method, anti-B7H4 ADC therapy is recommended for cancer patients if the response score is greater than a predetermined threshold. The predetermined threshold in step 6 and the displacement value in step 5 are determined by optimizing the prevalence of positive predictors, negative predictors, and positive recommendations using a cohort of patients with known single-cell ADC scores and therapy response parameters.

[0177] In another example of computer processing of response scores in step 5, a digital image of a tissue sample from a cancer patient is acquired and stained with a diagnostic antibody (dAB). Image analysis is performed on the digital image to identify cancer cells (c i ) 0≦i<N It detects cell c. i The optical density in the segmented film is

number

number

[0178] The term p() represents the positive function for each cell, indicating its potential to respond to anti-B7H4 ADC therapy. For a given cell, the positive function is positive with respect to the aggregate of dAB positivity associated with its neighboring cells. i and the associated positive value p(c i )teeth, [Formula 42]

number

number

number

number

[0179] In one example, the function h() is:

number

number

number

number

number

number

[0180] In Equation 42, ∨ is the aggregation operator for the set of adjacent cancer cells. The aggregation operation is a single numerical value representing the dAB positivity associated with the adjacent cells. In one example, the aggregation operation is the weighted arithmetic mean of dAB positivity. [Formula 45]

number

number

number

number

number

[0181] Positive function p(c i One specific example of this is: [Formula 48]

number

[0182] C. Verification of the prediction method. The accuracy of the novel predictive ADC scores generated according to the method shown in Figure 14 was validated based on patient-derived xenograft (PDX) studies of 15 patients with fluctuating B7-H4 expression levels as indicated by manual "H-scores". Briefly, tumor tissue fragments were subcutaneously transplanted into 6-8 week old female thymus-deficient nude mice. When the tumors reached an appropriate tumor volume range (typically 150-300 mm), the animals were randomized to treatment and control groups, and administration with B7H4-SG3932 ADC was initiated. Tumor-bearing mice were administered a single dose of B7H4-SG3932 ADC via intravenous injection. Animals were observed daily, and tumor dimensions and body weight were measured and recorded twice weekly. The tumor volume was measured using a digital caliper, and the tumor volume was calculated using the following formula: Tumor volume = [Length (mm) × Width (mm)² × 0.5²] (wherein length and width are the longest and shortest diameters of the tumor, respectively).

[0183] Figure 26 shows the correlation between actual outcomes and manually drawn H scores, enumerated by response classification, for 15 patients in the PDX breast study. Response classifications are PD for progressive disease, SD for stable disease, and R for response. The tissues analyzed in Figure 26 were from mice carrying tumor cells from 15 human cancer patients.

[0184] Figures 27-28 show the correlation of ratio scores between manually assessed H scores and QCS scores. Figure 27 compares the manually assessed ratio scores to the mean H score. Figure 28 compares the manually assessed ratio scores to the median QCS score (q50) for optical density of B7H4 film staining. A strong correlation exists between the QCS score (q50) and the manually assessed H score with respect to the ratio scores.

[0185] Figures 29-30 show the correlation between the intensity score results between manually assessed H scores and QCS scores. Figure 29 compares the manually assessed intensity score to the mean H score. Figure 30 compares the manually assessed intensity score to the median QCS score for optical density of B7H4 film staining. A weak correlation exists between the QCS score and the manually assessed H score with respect to intensity scores.

[0186] Figures 31-32 show the correlation between each patient's actual response classification as a function of mean H score and as a function of the patient's actual response classification as a function of median QCS score (q50). Figure 31 shows the patient's mean H score for each of the three response classifications: PD (progressive disease), R (response), and SD (stable disease). Figure 32 shows the patient's median QCS score (q50) for each of the three response classifications: PD, R, and SD. While mean H scores did not optimally differentiate patients in the PD and R response classifications, lower QCS scores tended to be associated with progressive disease, and higher QCS scores tended to be associated with a response to anti-H7B4 ADC.

[0187] Figures 33-34 show that the difference in baseline B7-H4 levels between the PD and R response classifications is significant, using both QCS and H scores. Figure 33 shows that the p-value of the t-test between the PD (progressive disease) and R (response) groups in the patient sample was 0.019 using the H score. Figure 34 shows that the p-value of the t-test between the PD and R groups in the patient sample was 0.011 using the QCS score.

[0188] Figure 35 shows the relationship between tumor growth percentage and QCS score (median membrane optical density of all tumor cells in the sample / 50% displacement). X-axis values ​​greater than 0 indicate that the tumor size increased during observation, while values ​​less than 0 indicate that the tumor size decreased during observation.

[0189] The QCS q50 score for each patient sample is computer-processed by determining the median membrane optical density of all tumor cells in B7H4-stained digital tissue images obtained from the patient sample.

[0190] Figure 36 shows the correlation between QCS score and tumor growth. A negative correlation was found between baseline B7-H4 optical density staining level, assessed using the QCS score, and post-treatment tumor size change. In the graph in Figure 36, the Spearman rho is -0.65, and p is equal to 0.011. Therefore, higher optical density of B7-H4 staining was associated with greater tumor size reduction.

[0191] D. Cases showing higher QCS scores based on diagnostic antibody staining correlate with a reduction in tumor cell density after ADC treatment. An example demonstrating the application of quantitative continuous scoring (QCS) included the use of scores and associated image analysis to assess the pharmacokinetics (PK) and pharmacodynamics (PD) of anti-H7B4 ADC E02-GL-SG3932. Figure 37 shows how the PK and PD studies were conducted. Using a xenograft mouse model, we tested how ADC reduced tumor growth in human breast cancer tissue. Human tumor cells from 28 breast cancer patients were transplanted into mice and cloned. The growth of transplanted tumor cells was tracked over time using antibodies against ADC E02-GL-SG3932, associated image analysis, and stained diagnostic antibodies.

[0192] Figures 38A–D are graphs summarizing the PK and PD results. Tumor samples were collected at different time points after treatment with a single dose of ADC and subsequently stained by immunohistochemistry. Using a deep learning-based QCS algorithm, ADC was quantified, and custom-made HALO solutions determined the extent of gH2AX and cleaved caspase 3 in tumor cells.

[0193] Figure 38A shows the binding and uptake of ADC in epithelial cells based on the cell ratio of epithelial cells by membrane staining for ADC-IgG (human immunoglobulin G1) against all epithelial cells. Cells were counted as stained only if they showed staining exceeding the maximum value observed in the control. Figure 38A shows that ADC uptake increased over time up to 96 hours after ADC treatment, and increased more with higher doses of ADC.

[0194] Figure 38B shows the time-dependent increase in DNA damage (gH2AX) after ADC treatment. DNA damage is represented by the percentage of epithelial cells that tested positive in the lesion using the gH2AX assay. The extent of γH2AX spread was determined by analyzing the slides using the commercially available image analysis system Halo (Indica Labs, Albuquerque, USA). Figure 38B shows that DNA damage peaked approximately 168 hours after ADC treatment.

[0195] Figure 38C shows the increase in DNA damage over time after ADC treatment, based on the percentage of cleaved caspase-3-positive tumor cells. The glass slides were analyzed using the HALO image analysis system.

[0196] Figure 38D shows the increase in tumor cell death over time after ADC treatment, based on a reduction in the cell density of all epithelial cells. Treatment with E02-GL-SG3932 resulted in a decrease in cell number and cell death compared to treatment with isotype ADCs.

[0197] Before image analysis was performed on the stained tissues, diagnostic antibodies closely associated with the ADC E02-GL-SG3932 antibody were first selected. Next, the diagnostic antibodies were used for staining. Initially, using a xenograft model, it was confirmed that membrane staining of tumor cells by dyes linked to the three diagnostic antibodies D11, A57.1, and Cal63 was highly correlated among the three. Figure 39A shows the median optical density of membrane staining of tumor tissues from 28 breast cancer patients cloned by transplantation into mice. The high correlation of staining in 28 distinct tumor samples obtained from the three diagnostic antibodies demonstrates that the degree of tumor progression was consistently measured by the diagnostic antibodies. In only two of the 28 tumor samples, one of the diagnostic antibodies yielded significantly different staining levels than the other two.

[0198] Figure 39B shows that for tumor clone sample #21, membrane staining with diagnostic antibody D11 was far less than staining with diagnostic antibodies A57.1 and Cal63. Figure 39C shows that for tumor sample #28, membrane staining with diagnostic antibody A57.1 was significantly more than staining with diagnostic antibodies D11 and Cal63. Therefore, diagnostic antibody Cal63 provided the most consistent membrane staining of the tumor samples and was used for PK and PD testing.

[0199] Figures 40A-B show digital images of clonal samples #26 and #4 stained with the diagnostic antibody Cal63 linked to a dye. Using image analysis and the QCS method, a single-cell score based on membrane staining was determined for each tumor cell in the digital image. Next, the distribution of similarly stained cells was determined. The pharmacokinetics (PK) and pharmacodynamics (PD) of ADC E02-GL-SG3932 were examined using the QCS score.

[0200] Figures 41A and 41B show the distribution of membrane optical density of samples stained with the anti-B7H4 diagnostic antibody Cal63 after treatment with various doses of E02-GL-SG3932 or ADCs without a cytotoxic payload. The graphs in Figures 41A and 41B show that ADC treatment increased target expression levels and was independent of ADC concentration at the levels tested.

[0201] Figure 41A shows the distribution of cell-specific membrane staining (membrane optical density) for clone sample #26 before and after treatment with ADCs containing various payload concentrations. The upper graph represents time zero, and the lower graph represents time 168 hours after treatment. The isotype control shows nearly the same staining distribution as the untreated cells at time zero. At time 168 hours after treatment, the number of cells with higher levels of membrane staining increased. The increased staining was relatively independent of the ADC concentration. The increased staining after ADC treatment demonstrates that ADCs induced high target levels of B7H4 antibody in the epithelial cell membrane.

[0202] Figure 41B shows the distribution of cell-specific membrane staining for clone sample #4 before and after treatment with ADC E02-GL-SG3932. The upper graph is at zero, and the lower graph is at 168 hours post-treatment. The upper graph shows that the median optical density of membrane staining for clone sample #4 at zero is approximately 60% of that for clone sample #26 in Figure 41A. At 168 hours post-treatment, the number of cells with higher levels of membrane staining increased. Starting from this low extent of membrane staining in tumor cells, ADC treatment induced high target levels.

[0203] Figures 42A-B show the cell-specific distribution of bound IgG before and after ADC treatment for clonal samples #26 and #4. The upper graphs in Figures 42A-B show that sample #26 has a higher baseline distribution of bound IgG than administered sample #4. The lower graphs in Figures 42A-B show the distribution of bound IgG after ADC treatment, demonstrating that sample #26, with its higher baseline target level, resulted in higher ADC binding compared to sample #4. Therefore, ADC binding to target cells is dependent on target expression.

[0204] Figure 42C is a graph of the maximum membrane optical density of bound IgG at various dose levels of E02-GL-SG3932. Figure 42C shows that the mean distribution of membrane optical density of bound IgG increased with increasing dose levels of ADCs without a payload.

[0205] Figure 42D shows the percentage of cleaved caspase-3-positive tumors in clone sample #26 at dose levels of ADC E02-GL-SG3932 ranging from 1 to 7 mg / kg. Figure 42D shows that at a dose level of 7 mg / kg, the percentage of positive tumor values ​​is higher for clone #26 than for clone #4.

[0206] Figures 43A and 43B are graphs showing the effect of target expression on the pharmacokinetics (PK) and pharmacodynamics (PD) of E02-GL-SG3932 at various dosages.

[0207] Figure 43A shows the membrane binding kinetics of E02-GL-SG3932. Figure 43A records the median membrane optical density of IgG for clones #26 and #4 at various time points after treatment with ADCs without varying doses. The median membrane optical density increased over time since treatment, reaching a maximum value approximately 96 hours after treatment.

[0208] Figure 43B shows the dynamics of ADC presence in the membrane relative to the cytosol. The graph in Figure 43B shows the median optical density of clone #26 at various time points after treatment with 7 mg / kg E02-GL-SG3932. Optical density is an indicator of target expression. The graph in Figure 43B shows optical density in the membrane and cytosol. Approximately 24 hours after treatment, target expression in the membrane increases at a faster rate than target expression in the cytosol. Therefore, ADC binding is dependent on both dose and target expression.

[0209] Figures 44A and 44B are graphs showing the effects of E02-GL-SG3932 on targeted expression of PK and PD at various dosages. Figures 44A and 44B also show that ADCs are effective against cell lines with low targeted expression. Figure 44A is for clone #26, and Figure 44B is for clone #4.

[0210] The upper graphs in Figures 44A and 44B show the density of stained epithelial cells at various time points after treatment with ADCs without a payload. The upper graph in Figure 44A shows that epithelial cell density decreased by 51.2% up to 288 hours after treatment of clone #26 with ADCs without a 7 mg / kg payload compared to treatment with an isotype control. The upper graph in Figure 44B shows that epithelial cell density decreased by 44.7% up to 288 hours after treatment of clone #4 with ADCs without a 7 mg / kg payload compared to treatment with an isotype control.

[0211] The lower graphs in Figures 44A and 44B show the increase in DNA damage over time after ADC treatment, based on the percentage of cleaved caspase-3 (CC-3)-positive tumor cells. Slide glass image analysis was performed using a custom-made HALO solution. The lower graph in Figure 44A shows a significant increase in the CC-3-positive area of ​​damaged tumor cells of clone #26 96 hours after treatment. The area of ​​CC-3-positive tissue is wider for ADCs without a higher dose payload. The lower graph in Figure 44B shows that the CC-3-positive area of ​​damaged tumor cells of clone #4 began to increase only moderately around 48 hours after treatment, even for ADCs without a higher dose payload of 7 mg / kg. Furthermore, the increase in CC-3-positive tissue was almost the same as the increase for isotype ADCs.

[0212] Figures 45A-B, 46A-B, and 47A-B show the effect of spatial proximity of stained epithelial cells on the pharmacokinetics (PK) of E02-GL-SG3932. The spatial proximity of stained tumor cells allows the toxic payload of the ADC, which binds to cancer cells expressing the target protein, to also kill nearby cells that do not express the target protein. Therefore, in areas where stained epithelial cells are in close proximity, the rate of cell death is higher, and consequently, the expression level of the target protein is higher.

[0213] Figures 45A and 45B show the relationship between binary spatial proximity (SPS at 25 μm) and membrane optical density at various time points after up to 312 hours of ADC treatment. For clone sample #26, Figure 45A shows a large linear correlation between binary SPS and median membrane optical density as the relationship changes over time. The correlation coefficient r is equal to 0.914 for the scatter plot of Figure 45A. Figure 45B shows an even larger linear correlation between binary SPS and membrane optical density as the relationship changes over time for clone #4. The correlation coefficient r is equal to 0.925 for the scatter plot of Figure 45B.

[0214] Figures 46A-B show the relationship between continuous spatial proximity (continuous SPS at 25 μm) and membrane optical density at various time points after ADC treatment. For clone sample #26, Figure 46A shows that there is an insufficient linear correlation between continuous SPS and median membrane optical density as the relationship changes over time. Figure 46A actually shows that for clone #26, which has a high staining membrane optical density approximately 168 hours after treatment, the optical density increases as continuous spatial proximity decreases. This divides the scatter plot into two distinct groups. Thus, for clone #26 with higher levels of target expression, target expression increased over time after treatment in tumor areas with lower continuous spatial proximity.

[0215] However, for clone #4, which has lower levels of target expression, Figure 46B does not show the emergence of a second distinct group in the scatter plot after a longer period following treatment. Instead, a moderate linear correlation remains in the association between continuous SPS and membrane optical density, as the association changes over time. The correlation coefficient r is equal to 0.617 for the scatter plot in Figure 46B.

[0216] Figures 47A and 47B show the relationship between continuous spatial proximity (continuous SPS at 25 μm) and epithelial cell density at various time points after up to 312 hours of ADC treatment. For clone sample #26, Figure 47A shows a moderate linear correlation between continuous SPS and epithelial density as the relationship changes over time. The correlation coefficient r is equal to 0.798 for the scatter plot of Figure 47A. Figure 47B shows a less favorable linear correlation between continuous SPS and epithelial cell density as the relationship changes over time for clone #4. The correlation coefficient r is equal to 0.686 for the scatter plot of Figure 47B. Figures 47A and 47B show that, over time after ADC treatment, a higher rate of cell death (and therefore lower cell density) is observed in areas of closer continuous spatial proximity among stained tumor cells. The higher rate of observed cell death in regions where stained epithelial cells are more nearby and therefore where target protein expression is higher supports the proposed third effect shown in Figure 23, in which non-expressing cancer cells near cells targeted by ADCs are also killed by the toxic payload of the ADCs. Furthermore, this supports the inclusion of exponential weighting coefficients in the formula for the single-cell ADC score shown in Figures 24-25, which explains cell separation that reflects the uptake of the ADC payload into adjacent cells.

[0217] While the present invention is described in relation to certain embodiments for teaching purposes, the invention is not limited thereto. Various modifications, adaptations, and combinations of features of the described embodiments can be practiced without departing from the scope of the invention as described in the claims.

Claims

1. A method for generating a response score to predict a cancer patient's response to an antibody-drug conjugate (ADC) comprising an ADC payload and an ADC antibody that targets a protein on cancer cells, wherein the protein is B7-H4. The method involves immunohistochemically staining a tissue sample using a dye linked to a diagnostic antibody, wherein the diagnostic antibody binds to B7-H4 on the cancer cells in the tissue sample; To obtain a digital image of the aforementioned tissue sample; To detect cancer cells in the aforementioned digital image; For each cancer cell, a single-cell ADC score is calculated by a computer system, wherein the single-cell ADC score is a score calculated for the cancer cell based on the staining intensity of the dye on the membrane of the cancer cell, the staining intensity of the dye in the cytoplasm of the cancer cell, the staining intensity of the dye on the membrane of other cancer cells, and the staining intensity of the dye in the cytoplasm of the other cancer cells, the other cancer cells being cancer cells closer than a predetermined distance to the cancer cell; and The method involves generating a response score by a computer system by aggregating all single-cell ADC scores from the tissue sample using statistical manipulation, wherein the response score indicates the likelihood that the cancer patient will respond to the ADC. Methods that include...

2. The method according to claim 1, wherein the detection of cancer cells includes detecting, with respect to each cancer cell, pixels belonging to the membrane and pixels belonging to the cytoplasm.

3. The method according to claim 1 or 2, wherein the dye is 3,3'-diaminobenzidine (DAB), and / or the staining intensity of each membrane is calculated based on the average optical density of brown diaminobenzidine (DAB) signals in the pixels of the membrane, and the staining intensity of each cytoplasm is calculated based on the average optical density of brown DAB signals in the pixels of the cytoplasm.

4. The method according to any one of claims 1 to 3, wherein the single-cell ADC score for the first cancer cell is calculated as a distance-weighted sum of several powers of the staining intensity of the dye in the membrane of the other cancer cell and the staining intensity of the dye in the cytoplasm of the other cancer cell, and the other cancer cell is a cancer cell having a center closer to the center of the first cancer cell than a predetermined distance.

5. The single-cell ADC score for cell i is, |r ｊ -r ｉ |<d{a ２０ (|r ｊ -r ｉ |)×ODM ｊ ２ +a １１ (|r ｊ -r ｉ |)×ODM ｊ ×ODC ｊ +a ０２ (|r ｊ -r ｉ |)×ODC ｊ ２ +a ００ (|r ｊ -r ｉ |)} is calculated as the sum of all cells j, |r ｊ -r ｉ | represents the distance between cell j and cell i, and a ｋｌ (|r ｊ -r ｉ |) represents the distance between each cell i and each cell j |r ｊ -r ｉ It is a function that depends on | and ODM ｊ However, this is the optical density of the dye in the membrane of cell j, and ODC. ｊ The method according to any one of claims 1 to 3, wherein the optical density is the optical density of the dye in the cytoplasm of cell j.

6. function a ｋｌ However, the default coefficient A ｏｏ A １ｏ A ｏ１ A １１ A ２０ A ｏ２ The relationship is expressed by: a ｋｌ (|r ｊ -r ｉ |) = A ｋｌ ×exp(-|r) ｊ -r ｉ | / r ｎｏｒｍ ) The distance between cell j and cell i in the given cell |r ｊ -r ｉ | depends on and / or | r ｊ -r ｉ The method according to claim 5, wherein | is the distance between the center of cell j and the center of cell i.

7. The aforementioned predetermined coefficient A ｏｏ A １ｏ A ｏ１ A １１ A ２０ A ｏ２ d and r ｎｏｒｍ The method according to claim 6, wherein the response of a cohort of trained patients is determined by optimizing the correlation between the therapeutic response and the response score.

8. The method according to any one of claims 1 to 7, wherein the aggregation of all single-cell ADC scores is obtained from the group consisting of determining the mean, determining the median, and determining a displacement value having a predetermined percentage.

9. The method according to any one of claims 1 to 8, wherein the ADC antibody comprises HCDR1, HCDR2, HCDR3, LCDR1, LCDR2, and LCDR3, or functional variants thereof, each comprising the amino acid sequences of SEQ ID NO: 4, SEQ ID NO: 5, SEQ ID NO: 6, SEQ ID NO: 7, SEQ ID NO: 8, and SEQ ID NO: 9, respectively.

10. The ADC antibody comprises a VH chain and a VL chain containing the amino acid sequences of SEQ ID NO: 10 and SEQ ID NO: 11, respectively, or functional variants thereof; and / or The ADC antibody, A heavy chain containing the amino acid sequence of SEQ ID NO: 12; and Light chain containing the amino acid sequence of SEQ ID NO: 13 The method according to claim 9, including the method described in claim 9.

11. The method according to any one of claims 1 to 10, wherein a patient having a response score higher than a predetermined threshold is recommended for a therapy including the ADC.

12. The ADC antibody is a humanized IgG1 monoclonal antibody, and / or The method according to any one of claims 1 to 11, wherein the ADC payload is a cytotoxic, and optionally the ADC payload is a topoisomerase I inhibitor.

13. The method according to any one of claims 1 to 12, wherein the cancer patient has a cancer selected from the group consisting of breast cancer, ovarian cancer, endometrial cancer, cholangiocarcinoma, lung cancer, pancreatic cancer, and gastric cancer.

14. The ADC is an anti-B7H4 antibody conjugated to a drug-linker, and the drug-linker has the following formula: 【Chemistry 1】 Represented by R ＬＬ The method according to any one of claims 1 to 13, wherein the linker is linked to the anti-B7H4 antibody.

15. The ADC is an anti-B7H4 antibody conjugated with a topoisomerase inhibitor, and the topoisomerase inhibitor is 【Chemistry 2】 The method according to any one of claims 1 to 14, selected from the following.

16. The ADC is an anti-B7H4 antibody conjugated with a topoisomerase inhibitor, and the topoisomerase inhibitor is 【Transformation 3】 The method according to any one of claims 1 to 15.

17. A method for identifying cancer patients for treatment with an anti-B7H4 antibody-drug conjugate (ADC) comprising an ADC payload and an ADC antibody that targets a protein on cancer cells, Perform the method described in any one of claims 1 to 16; and The computer system identifies cancer patients who are likely to benefit from the administration of the ADC if their response score exceeds a threshold. Methods that include...

18. The method according to claim 17, wherein the possible benefit is a reduction in average tumor size.

19. A pharmaceutical composition for use in a method of treating cancer in a patient, comprising an antibody-drug conjugate (ADC) comprising an ADC payload and an ADC antibody that targets a protein on cancer cells, wherein the protein is B7-H4, and the method is Perform the method described in any one of claims 1 to 17; and If the response score exceeds a predetermined threshold, the pharmaceutical composition is administered to the patient. The pharmaceutical composition comprising the above.

Images