Predictive marker for epithelial cancer

Abstract

The present invention relates to a method of predicting the clinical outcome of a subject diagnosed with epithelial cancer, such as oral squamous cell carcinoma, and biomarkers for use in the method.

Description

Technical Field

[0001] This invention belongs to the field of biomedicine. More specifically, this invention relates to biomarkers for predicting clinical outcomes in patients diagnosed with epithelial cancer (e.g., squamous cell carcinoma, particularly oral squamous cell carcinoma). Background of the Invention

[0003] Epithelial tissue (epithelium) is composed of closely packed epithelial cells attached to an underlying cellless basement membrane. Epithelial cells can be squamous, cuboidal, or columnar, and they can form simple epithelium that is only one cell thick or stratified epithelium with multiple cell layers. Epithelium covers external body surfaces (e.g., skin, alveoli, and digestive tract), lines internal cavities (e.g., peritoneal cavity), and forms a key component of many internal organs (e.g., kidneys, bladder, breasts and other glands, uterus). Epithelium acts as a selective barrier between external and deep tissue compartments in an organ-specific manner. In addition to protective and structural functions, epithelial cells typically have secretory and / or absorptive functions. Most surface epithelial cells are constantly renewed and exposed to environmental carcinogens, increasing their risk of accumulating carcinogenic mutations over time. Epithelial carcinoma (carcinoma) is the most common form of solid malignancy in humans, accounting for 80-90% of all cancer cases. Squamous cell carcinoma is a common type of epithelial carcinoma that originates from squamous epithelial cells, such as those found on the skin surface and in the oral cavity. Cubic and columnar epithelial cells can also undergo malignant transformation, forming other common epithelial cancers (such as adenocarcinomas of the lung, colon, kidney, and breast).

[0004] Oral squamous cell carcinoma (OSCC) is the most common form of head and neck cancer and a major global health problem. According to the latest global report, an estimated 377,713 new cases and 177,757 deaths occurred in 2020. Among the subtypes of OSCC, squamous cell carcinoma of the mobile tongue (OTSCC) is the most common and the most prevalent type of cancer in the oral cavity. Notably, OTSCC shows an increasing incidence in younger individuals and has the worst prognosis among all OSCC cases, further highlighting its significant clinical importance. Even in early-stage OTSCC (cT1-cT2 stages), nearly 20% of patients eventually face recurrence and cancer-related death. Currently, there are no effective methods to identify these high-risk patients. Therefore, new predictive biomarkers are urgently needed to guide appropriate treatment strategies for these patients. Summary of the Invention

[0005] The present invention relates to a method for predicting clinical outcomes in a subject diagnosed with epithelial carcinoma (as described in independent claim 1), and a kit for said method (as described in independent claim 14).

[0006] Some embodiments of the invention are set forth in the dependent claims. Other embodiments and aspects can be found in the detailed description of the invention, the accompanying drawings, and the examples. 。

[0007] Brief description of the attached figures

[0008] The accompanying drawings, which are included to further understand the invention and form part of this specification, illustrate embodiments of the invention and, together with the description, help to explain the principles of the invention. In the drawings: Figure 1 This is a schematic flowchart that summarizes the research scheme that led to the present invention.

[0009] Figure 2 The expression of PROX1, KI-67, and E-cadherin in representative OSCC tumors is illustrated. 'ac' shows the IMC expression data for E-cadherin, KI-67, and PROX1, respectively. The dashed boxes in 'ac' indicate the cropped regions magnified in 'df'. Cell segmentation masks for Prox1-positive cells are overlaid in 'df'. Arrows in 'ef' indicate KI-67 and PROX1 double-positive cells.

[0010] Figure 3 LEC proliferation: survival analysis and biomarker detection using ROC. ac shows Kaplan-Meier curves for recurrence-free survival (a), disease-specific survival (b), and overall survival (c) in two groups of patients divided according to the median LEC proliferation (3.4%). Patients with a total LEC count less than 15 were excluded. The lower group represents patients with LEC proliferation values ​​below the median. The percentages shown in ac represent the mean survival probability over a 3-year follow-up period. df shows the time-dependent receiver operating characteristic (ROC) estimates from the 3-year recurrence-free survival (d), disease-specific survival (e), and overall survival (f) data. The area under the ROC curve (AUC) value is shown in the figure.

[0011] Figure 4PROX1 and KI-67: Relapse-free survival analysis and ROC assessment. ab shows Kaplan-Meier curves for relapse-free survival in the low and high patient groups, divided by the median LEC count (a) and total proliferation value (b). The LEC count represents the total number of PROX1-positive cells in the sample. The total proliferation value represents the percentage of KI-67-positive cells in the sample. cd shows time-dependent receiver operating characteristic (ROC) estimates of 3-year relapse survival data based on LEC count (PROX1) and total proliferation (KI-67), respectively. The area under the ROC curve (AUC) values ​​are shown in the figure.

[0012] Figure 5A Representative IMC images of podoplanin (Pdpn), Prox1, and Ki-67 expression in OSCC tumors of patients with AC are shown. Scale bar: 100 µm.

[0013] Figure 5B Showing Figure 5A Magnified areas of the images (images i-iii). Dashed lines highlight representative Ki-67 positive LECs. Scale bar: 33 µm.

[0014] Figure 6 Representative images of Pdpn, Prox1, and Ki-67 expression in lymphatic vessels near the OSCC tumor in patient DF are shown. Dashed lines indicate the signal contours of Pdpn (column 2) or Prox1 (column 4). Scale bar: 20 µm per row.

[0015] Figure 7 A scatter plot of platypophysin and Prox1 expression in a random sample of 70,000 segmented cells is shown, divided by cell type. The vertical and horizontal dashed lines represent the positive cutoff values ​​for platypophysin and Prox1, respectively, calculated as the mean in LEC - 1SD.

[0016] Figure 8 Kaplan-Meier survival curves for LEC proliferation (LECp%) using flatfoot protein and Ki-67 (left column) or Prox1 and Ki-67 (right column) are shown. The cutoff value for Ki-67 positivity is 2.

[0017] Figure 9A The chart shown is a patient cohort described in Example 1, divided into different groups based on their comprehensive risk scores as described in Example 3.

[0018] Figure 9B Showing Figure 9A Kaplan-Meier survival curves for the patient group.

[0019] Figure 9C The Kaplan-Meier survival curves are shown for patient groups with a comprehensive risk score of less than or greater than 2.

[0020] Figure 10 LEC proliferation: Survival analysis based on lymph node status (pN). Kaplan-Meier plots ab show recurrence-free survival in patients with: (a) one or more metastatic regional lymph nodes (pN > 0); (b) no cancer cells in regional lymph nodes (pN = 0). Patients were divided into low and high LEC proliferation groups based on the median (3.4%). Patients with a total LEC count less than 15 were excluded. The low group represents patients with LEC proliferation values ​​below the median.

[0021] definition

[0022] Before describing the invention, it should be understood that this disclosure is not strictly limited to any particular composition, reagent, antibody, device, scheme, or method described herein, and these may vary. It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention, as the scope of the invention is limited only by the appended claims.

[0023] It should also be noted that, unless otherwise defined, all technical and scientific terms used herein have the meanings commonly understood by one of ordinary skill in the art to which this invention pertains.

[0024] Furthermore, it should be noted that certain features of this disclosure described in the context of individual embodiments for clarity may also be provided in a single embodiment. Conversely, multiple features of this disclosure described in a single embodiment for brevity may also be provided individually or in any suitable sub-combination. Moreover, even if not repeated, any feature, detail, or embodiment disclosed in the context of the methods provided herein also applies to the kits provided herein, and vice versa.

[0025] In this text, the meaning of singular nouns includes the meaning of plural nouns, and thus (unless otherwise stated) singular terms can also have the meaning of their plural forms. In other words, "a," "an," or "the" can refer to one or more.

[0026] In phrases like “X and / or Y”, the term “and / or” should be understood as “X and Y” or “X or Y”, and should be regarded as providing explicit support for either or both of these meanings.

[0027] The term "carcinoma" as used in this article refers to cancer that forms in epithelial tissue, and is therefore used interchangeably with the term "epithelial cancer." Different types of carcinoma include squamous cell carcinoma, adenocarcinoma, transitional cell carcinoma, and basal cell carcinoma.

[0028] Epithelial tissue is distributed throughout the body, forming channels that cover and line all the body's surface and internal structures. Most cancers affecting the skin, breast, kidneys, liver, lungs, pancreas, gastrointestinal tract, prostate, mouth, pharynx, and larynx are carcinomas. In fact, carcinomas account for 80% to 90% of all cancer cases.

[0029] As used in this article, "oral squamous cell carcinoma (OSCC)" refers to a malignant tumor that can occur in any part of the oral cavity, including the cheek, tongue, base of the tongue, gingiva, floor of the mouth, and hard palate. Among the sites of OSCC occurrence, oral-lingual squamous cell carcinoma (OTSCC) is the most common type of cancer in the oral cavity.

[0030] As used herein, the term "prediction" refers to determining the likelihood (good or bad) of a patient's specific clinical outcome after surgical resection of the primary tumor. The prediction method of this invention can be used clinically to make treatment decisions by selecting the most appropriate treatment option (including no treatment) for any given patient. The prediction method of this invention is a useful tool for predicting whether a patient is likely to benefit from a certain treatment option (e.g., chemotherapy and / or radiotherapy).

[0031] As used in this article, the term "positive clinical outcome" refers to any improvement in a patient's condition, including commonly used measurements in the field, such as prolonged recurrence-free survival (RFS), prolonged disease-specific survival (DSS), and prolonged overall survival (OS). An increased likelihood of a positive clinical outcome corresponds to a decreased likelihood of cancer recurrence. Conversely, a decreased likelihood of a positive clinical outcome corresponds to an increased likelihood of cancer recurrence. In this article, the terms "positive clinical outcome" and "good prognosis" are used interchangeably.

[0032] In other words, a "good clinical outcome" means that the patient's survival time will be longer than the median (or mean) observed in the general population of subjects with the same disease. Conversely, a "negative clinical outcome" means that the patient's survival time will be shorter than the median (or mean) observed in the general population of subjects with the same disease. In this article, the terms "negative clinical outcome" and "poor prognosis" are used interchangeably.

[0033] As used in this article, the term “recurrence” refers to the recurrence of cancer, whether it is a local recurrence (e.g., at the location where the cancer was before treatment) or a distant recurrence (e.g., metastasis).

[0034] The term “recurrence-free survival (RFS)” as used in this article refers to the length of time from the date of curative surgery to the time of cancer recurrence or death from any cause.

[0035] The term “disease-specific survival (DSS)” as used in this article refers to the length of time from the date of curative surgery to the time of cancer recurrence or death from a non-target cancer cause.

[0036] The term “overall survival (OS)” as used in this article refers to the length of time from the date of curative surgery to death from any cause.

[0037] As used herein, the term "risk classification" refers to the level of risk or predicted outcome at which a subject will experience a particular clinical outcome. Subjects may be grouped into risk groups based on the predictive methods of this invention. A "risk group" refers to a group of subjects or individuals with similar risk levels for a particular clinical outcome.

[0038] As used herein, the term "preset reference value" refers to a threshold or cutoff value determined based on comparative measurements between cancer patients with good clinical outcomes and those with poor clinical outcomes. For example, a preset reference value could refer to an LEC proliferation value below which a subject is likely to have a good clinical outcome and / or above which a subject is unlikely to have a good clinical outcome. In some cases, a preset reference value could also refer to the expression level of a selected biomarker below which a subject is likely to have a good clinical outcome and / or above which a subject is unlikely to have a good clinical outcome, and vice versa. Statistical methods for determining appropriate reference values ​​will be readily apparent to those skilled in the art.

[0039] As used in this article, the term "LEC proliferation value" refers to the percentage of proliferating lymphoendothelial cells (LECs) among all lymphoendothelial cells in the region of interest (ROI).

[0040] Typically, a threshold needs to be determined based on the function of the detection and the benefit / risk balance (false positive clinical consequences versus false negative clinical consequences) to obtain optimal sensitivity and specificity. Optimal sensitivity and specificity (and threshold) are often determined using receiver operating characteristic (ROC) curves based on experimental data. For example, after determining the LEC proliferation value and / or the expression level of the selected biomarker in the reference group, an algorithmic analysis can be used to statistically process the determined proliferation value and / or expression level in the test sample to obtain classification criteria meaningful for sample categorization (e.g., "high-risk" and "low-risk" groups).

[0041] The term “expression level” used in this article refers to the relative amount of a relevant biomarker in a cell or tissue sample.

[0042] The terms “subject,” “patient,” and “individual” used in this article are used interchangeably unless otherwise stated, and refer to mammals, particularly humans.

[0043] As used herein, the term "sample" refers to a tissue sample obtained from a subject. Specific examples of tissue samples include, but are not limited to, solid tissue samples, such as fresh, frozen, and / or preserved organ or tissue samples, biopsy specimens, and sections or smears prepared from any of the foregoing sources. Preferably, the tissue sample is taken from a resected tumor and includes the tumor center (or core) and the immediate surrounding tissue ("invasion border"). Typically, obtaining a biological sample to be analyzed from a subject is not part of the methods of this invention.

[0044] The term "sample" also includes samples that have been acquired and processed or manipulated in any suitable manner, including but not limited to washing, reagent treatment, fixation (e.g., formalin fixation), freezing, or embedding in a semi-solid or solid matrix for sectioning purposes. For the purposes of this disclosure, a "section" of a tissue sample refers to a single portion or sheet of a tissue sample, such as a thin slice of tissue or cells cut from a tissue sample.

[0045] As used in this article, the term "invasive border" refers to the region on each side of the boundary between malignant cells and normal host tissue. This term is used interchangeably with "invasive margin."

[0046] The terms “biomarker” and “marker” used in this article are used interchangeably and refer to naturally occurring molecules that are objective, quantifiable indicators of a specific characteristic of their origin.

[0047] For example, a biomarker used to predict clinical outcomes in a subject diagnosed with epithelial carcinoma refers to a molecule that differs in a biological sample taken from a subject with a specific cancer (e.g., OSCC) compared to a similar sample taken from a control subject (e.g., a subject without the cancer, i.e., an apparently healthy subject, or a subject with the cancer but with a different possible course or recovery). Therefore, the biomarker provides information about the likely outcome or recovery from the cancer (e.g., OSCC) and is quantitatively or qualitatively associated with the prognosis of the cancer (e.g., OSCC).

[0048] Therefore, the term "lymphoendothelial cell marker" refers to markers that identify a certain type of cell as a lymphoendothelial cell.

[0049] A biomarker that is "specific" to a target is one that exists only in the target and is almost non-targeted. For example, lymphoendothelial cell-specific biomarkers are biomarkers that are almost non-existent in cells other than lymphoendothelial cells.

[0050] The term “nuclear biomarker” as used in this article refers to a biomarker located in the cell nucleus.

[0051] As used herein, the term "binding buddy" broadly refers to any molecule capable of specifically binding to its target antigen (e.g., a biomarker). Non-limiting examples of binding buddies include, but are not limited to, antibodies, antibody mimics, oligonucleotides, and peptide aptamers.

[0052] As used in this article, the term "specific binding" refers to a binding where a molecule binds to a specific polypeptide or an epitope on a specific polypeptide, without substantially binding to any other polypeptide or polypeptide epitope.

[0053] As used herein, the term "antibody" generally refers to an immunoglobulin structure comprising two heavy chains and two light chains linked together by disulfide bonds, including monoclonal and polyclonal antibodies. Antibodies can exist as intact immunoglobulins or as a variety of well-characterized antigen-binding fragments or single-chain variants thereof, all of which are included in the term "antibody" herein. Non-limiting examples of said antigen-binding fragments include Fab fragments, F(ab)2 fragments, Fab' fragments, F(ab')2 fragments, Fd fragments, Fd' fragments, Fv fragments, and scFv. These fragments and variants can be prepared using recombinant DNA technology or by enzymatic or chemical separation methods known in the art for intact immunoglobulins.

[0054] As used herein, the term "treatment" refers to the administration of drugs (e.g., chemotherapy drugs) or other treatment modalities (e.g., radiotherapy) to patients in need to achieve objectives including improving, reducing, suppressing, or curing related epithelial cancer (e.g., OSCC). The dosage and regimen of such administration can be readily determined by a person of ordinary skill in the clinical field of cancer treatment based on various variables.

[0055] The term "effective amount" as used in this article refers to the amount of a drug (such as a chemotherapy drug) that at least mitigates the harmful effects of the associated epithelial cancer (such as OSCC). Invention Details

[0057] Based on evidence of differential proliferation of lymphoendothelial cells at the invasive border of OSCC tumors, this invention provides methods and means for predicting clinical outcomes of epithelial carcinomas (e.g., OSCC). The predictive biomarkers and related information provided by this invention enable physicians to make more informed treatment decisions and tailor treatment plans to the individual needs of patients, thereby maximizing treatment effectiveness while minimizing patient exposure to unnecessary treatments that not only provide no significant benefit but often pose serious risks due to adverse reactions.

[0058] In one aspect, the present invention provides a method for predicting clinical outcomes in a subject diagnosed with epithelial carcinoma, particularly after surgical resection of said cancer. This prediction is essentially based on the proportion of proliferating lymphoendothelial cells at the invasive boundary of the cancerous tissue. If this proportion increases compared to a preset reference value, the likelihood of a good clinical outcome decreases. Conversely, if the proportion does not increase or is below the preset reference value, the likelihood of a poor clinical outcome decreases.

[0059] In one embodiment, the present invention provides a method for predicting clinical outcomes in subjects diagnosed with OSCC, particularly after surgical resection of said OSCC. This prediction is essentially based on the proportion of proliferating lymphoendothelial cells in the invasive border of the OSCC. If this proportion increases compared to a preset reference value, the likelihood of a good clinical outcome decreases. Conversely, if the proportion does not increase or is below the preset reference value, the likelihood of a poor clinical outcome decreases.

[0060] The proportion of proliferating lymphoendothelial cells in a region of interest (e.g., the invasive boundary of an OSCC) can be determined by dividing the number of proliferating lymphoendothelial cells by the total number of lymphoendothelial cells.

[0061] Lymphoendothelial cells are identifiable, and their presence and quantity in a given tissue sample can be determined in a variety of ways. Typically (but not limited to this), one or more biomarkers of lymphoendothelial cells can be used. For example, PROX1 (Prospero homeobox 1) is a lymphoendothelial cell-specific protein biomarker that is expressed in the nuclei of both quiescent and proliferating lymphoendothelial cells.

[0062] Therefore, in some embodiments, the number of lymphoendothelial cells in the invasion boundary of a tissue sample containing cancer cells is determined by detecting and counting the number of cells expressing PROX1. This can be achieved by contacting a tissue sample (preferably a histological section) with a binding chaperone (e.g., an antibody) capable of selectively interacting with the PROX1 protein present in the tissue sample, and detecting any binding reaction between the PROX1 protein and the binding chaperone, thereby enabling the counting of lymphoendothelial cells (i.e., PROX1-positive cells) in the invasion boundary.

[0063] In addition, lymphoendothelial cells can be detected by other biomarkers of these cells. These biomarkers include, but are not limited to, CD31, Lyve-1, and foot protein, which can be used in any combination with PROX1, although not all and every combination is listed herein. CD31, Lyve-1, and foot protein are not nuclear biomarkers.

[0064] One issue associated with platypoprotein is that its expression is not limited to lymphoendothelial cells. As demonstrated in the experimental section, although most lymphoendothelial cells express platypoprotein, it primarily stains cancer cells overall. Therefore, platypoprotein alone cannot distinguish lymphoendothelial cells from other cells near lymphatic vessels. Thus, platypoprotein should only be considered as an alternative to PROX1 if it can be ensured that platypoprotein-positive cancer cells or other non-endothelial cells are not misclassified as lymphoendothelial cells.

[0065] In some implementations, lymphoendothelial cells in a tissue sample or a specific region thereof are detected by contacting the tissue sample (preferably a histological tissue section) with a binding partner (e.g., an antibody) capable of specifically interacting with a nuclear biomarker (e.g., PROX1) of lymphoendothelial cells, and detecting the interaction between the nuclear biomarker and the binding partner (if any).

[0066] The number of proliferating lymphoendothelial cells in a tissue sample can be determined, for example, by detecting and counting the number of proliferating cells expressing PROX1 or another nuclear biomarker of lymphoendothelial cells, and optionally one or more other lymphoendothelial cell biomarkers (e.g., CD31, Lyve-1, and / or platypophysin). Cell proliferation can be measured by various methods, as is well known to those skilled in the art.

[0067] In some embodiments, one or more proliferation markers may be used, including but not limited to KI-67, a nuclear protein known to be associated with cell proliferation. KI-67 is present in all active phases of the cell cycle but absent in quiescent cells. Since expression of the KI-67 protein is essential for the progression of the cell division cycle, it is an excellent marker of the proliferative state of a given cell population. Therefore, in some embodiments, KI-67 is used as a proliferation marker for lymphoendothelial cells, either alone or in combination with one or more other proliferation markers.

[0068] In some embodiments, proliferating cells in a tissue sample or a specific region thereof are detected by contacting the tissue sample (preferably a histological tissue section) with a binding chaperone (e.g., an antibody) capable of selectively interacting with the KI-67 protein expressed by proliferating cells present in the tissue sample, and detecting the interaction between KI-67 and the binding chaperone (if any).

[0069] Based on the above, the number of proliferating lymphoendothelial cells present in a tissue sample can be determined based on the co-expression of nuclear markers (e.g., PROX1) and proliferation markers (e.g., KI-67) of lymphoendothelial cells. In some embodiments, the co-expression can be determined by contacting a tissue sample (preferably a histological tissue section) with a first binding partner specific to PROX1 and a second binding partner specific to KI-67, and determining whether binding reactions between i) PROX1 and the first binding partner and ii) KI-67 and the second binding partner occur in the same cell. The binding reactions result in the formation of a first complex (i.e., a complex of the first binding partner and PROX1) and a second complex (i.e., a complex of the second binding partner and KI-67), respectively. In embodiments involving histological tissue sections, the first and second binding partners may be contacted simultaneously or sequentially with the same tissue section. Furthermore, different tissue sections, preferably sequential tissue sections, may be used.

[0070] Using nuclear biomarkers (e.g., PROX1) and nuclear proliferation biomarkers (e.g., KI-67) of lymphoendothelial cells ensures the correct identification of proliferating lymphoendothelial cells. For example, if cytoplasmic biomarkers of lymphoendothelial cells are used instead of nuclear biomarkers, it is difficult to ensure that, in all cases, signals derived from proliferation biomarkers and signals derived from lymphoendothelial cell biomarkers originate from the same cell.

[0071] The clinical outcomes of the method of this invention can be expressed, for example, as relapse-free survival (RFS), disease-specific survival (DSS), and overall survival (OS).

[0072] Indeed, as demonstrated in the examples, an increase in the number of proliferating lymphoendothelial cells at the invasive boundary of OSCC indicates a reduced likelihood of a favorable outcome as determined by a decrease in RFS, DSS, and OS. For example, when the median percentage of LEC proliferation, 3.4, was used as the cutoff value for stratifying each patient into low or high LEC proliferation values, the three-year RFS rate was 96% for patients with low LEC proliferation values ​​and 64% for patients with high LEC proliferation values. The formula used to calculate the LEC proliferation value is as follows: LEC proliferation value (%) = n(KI67+PROX1 positive cells) / n(PROX1 positive cells).

[0073] Using the same cutoff value, the three-year DSS rate was 100% for OSCC patients with low LEC proliferation values, while it was 73% for patients with high LEC proliferation values. Similarly, the three-year OS rate was 86% for patients with low LEC proliferation values, compared to 63% for patients with high LEC proliferation values.

[0074] In some implementations, the use of additional biomarkers can provide higher predictive value than using PROX1 and KI-67 alone. Therefore, detecting one or more additional biomarkers in a sample improves the percentage of truly good and truly bad predictions and reduces the percentage of false good or false bad predictions. Thus, the method of the present invention may include measuring more than one additional biomarker, including one or more biomarkers listed in Table 2. In such a method, PROX1 may be replaced or supplemented by the use of one or more other lymphoendothelial cell-specific potential nuclear biomarkers, and / or KI-67 may be replaced or supplemented by the use of one or more other proliferation markers.

[0075] In some embodiments, the method of the present invention may include detecting one or more biomarkers selected from those in Table 2 in lymphoendothelial cells, preferably identified based on their PROX1 expression or based on another nuclear biomarker specific to lymphoendothelial cells.

[0076] Typically, determining the expression level of a biomarker at the protein level involves contacting a sample obtained from a subject requiring this determination with a binding chaperone (e.g., an antibody) that specifically recognizes the target polypeptide under conditions of specific interaction between the binding chaperone and the biomarker, and detecting said interaction (if any); wherein the presence or extent of said interaction is correlated with the presence or expression level of the biomarker in the sample. Besides antibodies and their fragments, other binding chaperones suitable for determining biomarker expression at the protein level include, but are not limited to, oligonucleotides or peptide aptamers.

[0077] Immunohistochemistry (IHC) is a preferred method for determining the proportion of proliferating lymphoendothelial cells in tissue samples and / or for measuring the expression levels of selected biomarkers in cells. IHC specifically provides a method for in situ detection of targets in samples or tissue specimens. 。 IHC maintains the overall cellular integrity of the sample, thus enabling the detection of the presence and location of target molecules.

[0078] Typically, immunohistochemistry (IHC) includes the following steps: i) fixing tissue samples with formalin; ii) embedding the samples in paraffin; iii) staining sample sections; iv) incubating the sections with a binding chaperone specific to the selected biomarker and labeled with a detectable marker; v) rinsing the sections; and vi) detecting the biomarker-chaperone complex using an appropriate technique based on the detectable marker used. Multiple binding chaperones targeting different selected biomarkers can be used simultaneously, each labeled with a different detectable marker. Counterstaining, such as hematoxylin-eosin, DAPI, and Hoechst, can also be used if necessary.

[0079] For IHC, binding chaperones can be directly or indirectly labeled, for example, via secondary antibody-based labeling, thereby enabling the detection of the target protein (i.e., the selected biomarker). Exemplary labels include radioisotopes (e.g., 3 H, 14 C 32 P, 35 S or 125 I) Non-radioactive heavy metal isotopes (such as those listed in Table 2), fluorescent dyes (such as fluorescein, rhodamine, phycoerythrin, fluorescein), chromophores (such as rhodopsin), chemiluminescent agents (such as luminol, imidazole), enzymes (such as horseradish peroxidase, alkaline phosphatase, β-lactamase), ligands, bioluminescent proteins (such as fluorescein, luciferase), haptens (such as biotin), particles (such as gold), and combinations thereof.

[0080] In some embodiments, the resulting stained samples are imaged individually using a system for observing detectable markers and acquiring images (e.g., digital images of the stain). Image acquisition methods are well known to those skilled in the art. For example, after sample staining, any optical or non-optical imaging device can be used to detect the dye or biomarker marker, such as upright or inverted optical microscopes, scanning confocal microscopes, mass spectrometry flow cytometry devices, cameras, scanning or tunneling electron microscopes, scanning probe microscopes, and imaging infrared detectors. In some examples, images can be captured digitally. The acquired images can then be used to quantitatively or semi-quantitatively determine the amount of the target marker in the sample, the absolute number of target marker-positive cells, or the surface area of ​​target marker-positive cells.

[0081] There are also various automated sample handling, scanning, and analysis systems suitable for IHC in this field.

[0082] Besides IHC, other techniques can be used to determine the proportion of proliferating lymphoendothelial cells in tissue samples and the expression levels of selected biomarkers. For example, single-cell RNA sequencing can be performed on tumor-derived cell suspensions to identify cells expressing both PROX1 mRNA and KI-67 mRNA (MKI67). Furthermore, spatial transcriptomics, including single-cell RNA sequencing-based methods and fluorescence in situ hybridization (FISH), can be used. Spatial methods are preferred because they can identify tumor boundaries and direct the analysis to cells located within those boundaries.

[0083] In some implementations, this method may also include analysis of known prognostic factors to improve the prognostic capability of the method. These factors include, but are not limited to, one or more of the following: lymph node involvement (pN), tumor size (pT), depth of invasion, tumor growth pattern (budding), number of tumor-infiltrating lymphocytes, and T-cell activation markers (e.g., granzyme B). A comprehensive risk score can be calculated by integrating multiple prognostic factors with the percentage of LEC proliferation. Factors significantly and independently associated with high recurrence-free survival (hazard ratio <1) decrease the comprehensive risk score, while factors significantly and independently associated with low recurrence-free survival (hazard ratio >1) increase the comprehensive risk score. Patients with intermediate to high total comprehensive risk scores have an increased risk of cancer recurrence. Combining multiple prognostic factors with LEC proliferation assessment can provide more accurate predictions of cancer progression.

[0084] In the implementation plan, subjects with epithelial carcinoma whose clinical outcomes were to be predicted had already experienced lymph node metastasis. When patients with pathologically confirmed sentinel lymph node metastasis (N1, N2, or N3) were analyzed as a separate cohort, patients with a high number of proliferative LECs had worse outcomes than those with a low number of proliferative LECs. Figure 10(a). These findings have significant clinical value because they demonstrate that patients with local lymph node metastasis (which is itself a poor prognostic factor) can still be categorized into good and bad prognostic subgroups based on the number of proliferative LECs.

[0085] In the implementation plan, subjects with epithelial carcinoma whose clinical outcomes were to be predicted had no lymph node metastasis. When those patients without pathologically confirmed sentinel lymph node metastasis (N0) were analyzed as a separate cohort, patients with high numbers of proliferative LECs had worse outcomes than those with low numbers of proliferative LECs. Figure 10 (b) These analyses further demonstrate that the predictive value of proliferative LECs is independent of the patient's lymph node status. Furthermore, they have significant clinical value because they show that patients without any local metastases in the lymph nodes (and therefore typically representing early, low-risk cases) can be categorized into good and bad prognostic groups based on the number of proliferative LECs.

[0086] In some specific applications, the method of the present invention for predicting clinical outcomes in subjects with epithelial cancer (e.g., OSCC) may further include therapeutic intervention. Once an individual is identified as having a reduced likelihood of a good clinical outcome, he / she may be treated with any appropriate treatment known to those skilled in the art, including but not limited to chemotherapy, radiotherapy, immunotherapy, and / or targeted therapy.

[0087] This invention not only benefits individual patient care but also facilitates better screening and stratification of patients in clinical trials. For example, OSCC patients or other epithelial cancer patients classified as high-risk can be recruited into clinical studies with the aim of developing highly effective new personalized treatment tools, such as new medical procedures or drugs.

[0088] In one aspect, the present invention also provides a kit for predicting clinical outcomes in subjects diagnosed with epithelial cancer (e.g., OSCC).

[0089] In the implementation scheme, the kit includes reagents capable of indicating the presence of lymphoendothelial cell-specific nuclear biomarkers and reagents capable of indicating cell proliferation.

[0090] In the implementation scheme, the kit includes reagents capable of indicating the presence of PROX1 and / or reagents capable of indicating the presence of KI-67.

[0091] In the implementation scheme, the kit includes a first binding partner capable of specifically binding to PROX1 (as a reagent capable of indicating the presence of PROX1) and / or a second binding partner capable of specifically binding to KI-67 (as a reagent capable of indicating the presence of KI-67).

[0092] In any embodiment, the kit may further include one or more additional reagents capable of indicating the presence of the markers listed in Table 2, and / or include one or more binding chaperones capable of specifically binding to the protein markers listed in Table 2.

[0093] In any implementation, one or more binding partners can be independently, directly or indirectly, detectably tagged.

[0094] In some embodiments, the kit includes one or more binding partners capable of specifically binding to or otherwise detecting one or more biomarkers of the present invention associated with a possible outcome of the epithelial cancer (e.g., OSCC). In some embodiments, the kit may also include at least one detection reagent or detection device capable of indicating the binding of the one or more binding partners to the one or more biomarkers, or capable of indicating the presence or level of the one or more biomarkers in a sample obtained from a subject whose possible outcome of epithelial cancer (e.g., OSCC) is to be determined or predicted.

[0095] In some implementations, the kit also includes positive and / or negative control samples or preset reference values ​​that can be detected or used for comparison with patient samples.

[0096] It should be understood that the contents of the kit may vary depending on the detection technology used, which will be apparent to those skilled in the art.

[0097] Although the above description and experimental sections focus on OSCC, it is understood that this method is also applicable to other cancers. Non-limiting examples of these other cancers include gastrointestinal cancer, pancreatic cancer, lung cancer, breast cancer, uterine cancer, prostate cancer, kidney cancer, bladder cancer, or skin cancer.

[0098] It will be apparent to those skilled in the art that, with advancements in technology, the basic idea of ​​this invention can be implemented in various ways. Therefore, this invention and its embodiments are not limited to the examples described above, but can be varied within the scope of the claims.

[0099] Experimental Section

[0100] Example 1. Identification of prognostic factors for OSCC

[0101] Samples and preparation

[0102] Whole tumor sections (n ​​= 95, Table 1) of early-stage (cT1-T2) oral squamous cell carcinoma fixed in formalin and embedded in paraffin (FFPE) were used as the study samples. Appropriate hematoxylin and eosin (H&E) staining was used to identify the region of interest (ROI) for each tumor sample. The selection criteria for ROI (1 = most important, 3 = least important) are as follows: 1. Complete organization 2. The marginal area between the invasive tumor and the surrounding stroma. 3. Areas rich in lymphocytes.

[0103] In other words, the ROI is the high-quality, invasive edge of a tumor that is infiltrated by immune cells, containing both malignant cancer cells and normal cells.

[0104] Table 1. Patient cohort; clinical and pathological characteristics

[0105] 1 Median (interquartile range); n (%) IMC staining failed, the sample was missing or there was no tumor residue in the section.

[0106] HUS = Helsinki University Hospital; KYS = Kuopio University Hospital; OYS = Oulu University Hospital; TAYS = Tampere University Hospital; TYKS = Turku University Hospital

[0107] Imaging Mass Cytometry (IMC) Staining Protocol

[0108] Perform immunostaining on the samples according to the following protocol: 1. Dewaxing and hydration; 2. Thermally induced epitope retrieval in citrate buffer (pH 6.0) (Dako target retrieval) (Aptum 2100 retrieval system); 3. Cut a circular region (diameter = 6 mm) around the predetermined ROI from the tissue section and circle the region with a hydrophobic IHC pen; 4. Block with PBS (phosphate-buffered saline) containing 5% BSA (bovine serum albumin) at room temperature for 45 minutes. 5. Prepare a pooled mixture of 25 commercially available metal-coupled antibodies or self-made conjugated antibodies (dilution range 1:50 – 1:400) in PBS containing 0.5% BSA (see Table 2); 6. Apply the antibody mixture to each excised tumor slice and incubate overnight at +4°C; 7. Dilute Cell-ID intercalation agents Ir (Ir191 and Ir193) 1:200 in PBS and incubate with the slides at room temperature for 30 minutes (for cell nucleus detection); 8. Air dry the sample at room temperature and store it.

[0109] Table 2. Antibody group used for IMC

[0110] Cancer cells

[0111] Immunosuppressive markers

[0112] Extracellular matrix / cytoskeleton

[0113] leukocyte

[0114] lymph

[0115] Cell proliferation

[0116] It is important to note that virtually all the markers used can stain several cell types and can be categorized into more than one class (the simple classification used for the current group generation is shown here).

[0117] IMC imaging

[0118] Imaging of immunostained samples was performed using the Hyperion™ imaging system and CyTOF software.

[0119] Using information from the corresponding H&E slices, the pre-defined ROI for each sample is identified based on the optical panoramic image generated during the Hyperion preparation step. IMC data (27 channels) from a 1mm x 1mm region (one ROI / sample) is acquired and the resulting data is stored in a .mcd file.

[0120] Data preparation and cell segmentation

[0121] Data preparation and cell segmentation steps were performed according to the IMC segmentation pipeline developed by Vito Riccardo Tomaso Zanotelli and Bernd Bodenmiller (ImcSegmentationPipeline: A pixel-classification based multiplexed image segmentation pipeline). Zenodo [Online] 2022).

[0122] The Ilastik pixel classification training set is generated using a stack of images randomly cropped from each sample.

[0123] Image channels of nuclear markers (Ir193, Ir193), E-cadherin, flatfoot protein, vimentin, CD31, TIM-3, Prox1, CD8a, granzyme B, CD3, IDO1, CD206, and CD45 were used for manual pixel labeling in the training set. Three regions were labeled: nucleus, cytoplasm, and background.

[0124] The final pixel probabilities generated by the trained Ilastik pixel classification algorithm are exported as RGB images and used to create cell segmentation masks in CellProfiler.

[0125] Single-cell and image features are measured in CellProfiler and used to generate spatial single-cell datasets in R for downstream analysis.

[0126] Data Analysis

[0127] Downstream analysis was conducted following the IMC analysis workflow of Windhager et al. An end-to-end workflow for multiplexed image processing and analysis. 2020, bioRxiv.

[0128] Prox1-positive cells (named LEC, n = 4773) were sorted from single-cell data for individual analysis. Successful sorting was confirmed by visualizing the sorted cells on IMC images.

[0129] Batch effect correction was performed using the nearest neighbor (MNN) method. Principal component analysis (PCA) was then performed using the corrected expression data.

[0130] Unsupervised clustering using the PhenoGraph clustering algorithm (Levine et al. Data- Driven Phenotypic Dissection of AML Reveals Progenitor-like Cells that Correlate with prognosis. 2015, Cell, 162(1), 184–197).

[0131] A unique cluster of LECs characterized by high KI-67 expression (named LECk) was identified, and the LEC proliferation value (%) for each sample was calculated by dividing the number of LECk cells by the total number of LECs in the sample. The proliferation value was calculated for each sample containing at least 15 LECs.

[0132] Patients were divided into two groups based on their LEC proliferation values. The low and high groups were defined using the median (3.4%) cutoff value. The survival distributions of the two patient groups were compared using the log-rank test, and the Kaplan-Meier method of Heagerty, Lumley, and Pepe was used. Time-dependent ROC curves for censored survival data and a diagnostic marker . 2000, Biometrics, 56(2), 337–344.4), estimated time-dependent ROC curves based on censored survival data.

[0133] result

[0134] Kaplan-Meier estimation was used to analyze recurrence-free survival (RFS), disease-specific survival (DSS), and overall survival (OS) in the study cohort. Figure 3 (A-3C). Patients were divided into two groups based on their LEC proliferation (LECp) values. The low-value group included patients with values ​​below the median, while the high-value group included patients with values ​​above the median. The three-year RFS probability was 64% in the high-value group, compared to 96% in the low-value group. Similar trends were observed in DSS and OS probabilities. The DSS was 73% in the low-value group and 100% in the high-value group; the OS was 63% in the low-value group and 86% in the high-value group. When patients with sentinel lymph node positivity (pN+) status were analyzed separately, the three-year RFS was 0% in the high-LECp group and 100% in the low-LECp group. Figure 10 (a) Similarly, in patients with sentinel lymph node negative (pN0) status, the three-year RFS was 75% in the high LECp group and 95% in the low LECp group. Figure 10 b).

[0135] LEC proliferation was further evaluated as a potential biomarker throughout the patient cohort using time-dependent receiver operating characteristic (ROC) analysis. Figure 3 (D-3F). During the three-year follow-up period, LEC proliferation values ​​showed an AUC of 0.863 for predicting RFS, 0.984 for DSS, and 0.684 for OS.

[0136] Prox1 and Ki-67 were also assessed separately by calculating the total number of Prox1-positive cells (LEC count) and the total percentage of Ki-67-positive cells (total proliferation). Kaplan-Meier analysis was performed to categorize patients into high and low groups based on the median values ​​of the two markers. Figure 4 A and 4B). Neither biomarker showed a significant difference in survival (RFS, DSS, or OS) when used alone. ROC analysis of three-year RFS ( Figure 4 The results (C and 4D) showed that the AUC value for LEC count was 0.618 and the AUC value for total proliferation was 0.572, further indicating that neither biomarker alone had prognostic value in the context of this study.

[0137] Some of the results are summarized in Table 3 below. In Table 3, LEC proliferation refers to cells expressing both PROX1 and KI-67, while LEC count refers to cells expressing only PROX1.

[0138] Table 3

[0139] 1. Kaplan-Meier % (95% CI), 2. Log-rank test

[0140] Example 2. Prognostic value of flatfoot protein

[0141] The potential prognostic value of flatfoot protein expression in OSCC tissue samples described in Example 1 was evaluated.

[0142] Based on the segmented IMC data described in Example 1, all LECs and tumor cells were identified using training data obtained through manual sorting and random forest classification. The identification results were visually verified for tumor sections. Cutoff expression values ​​were calculated for the identified LECs as mean - 1 SD (standard deviation), PROX1 (0.9), and PDPN (2.1). The cutoff value for KI-67 positivity (2) was determined by visual inspection. PROX1+, PDPN+, and KI-67+ cells were then sorted from the IMC data using these cutoff values.

[0143] In addition to visual inspection, expression data from a random sample of 70,000 cells were analyzed to assess the expression profiles of PROX1 and PDPN in LEC and tumor cells. The proportion of KI-67 positivity in PROX1+ and PDPN+ cells was calculated for each sample to compare the prognostic value of the combination of the two markers.

[0144] The results showed that flatfoot protein was highly expressed in OSCC tumor cells, while Prox1 maintained its specificity for LEC, even in areas of high tumor enrichment. Figure 5A and 5B These results indicate that platypophytes cannot distinguish lymphatic vessels (LECs) near OSCC tumor cells. Furthermore, while some lymphatic vessels can be visualized using platypophytes, they cannot distinguish individual endothelial cell nuclei from the nuclei of other cells closely associated with lymphatic vessels, such as lymphocytes and fibroblasts. This leads to an overestimation of LEC proliferation because Ki-67 is only expressed in the nucleus. Figure 6 (See Table 4). In summary, these data indicate that most flatfoot protein 1+Ki67+ cells are not LECs.

[0145] Most LECs express both Prox1 and flatfoot protein, but the vast majority of flatfoot protein+ cells are tumor cells or other non-LEC cell types. Figure 7 Ki-67 positivity in platypophysin+ cells was higher than the median but did not significantly predict overall survival or relapse-free survival. Conversely, Ki-67 positivity in Prox1+ cells was higher than the median and significantly associated with decreased overall survival and relapse-free survival. Figure 8 (Table 4).

[0146]

[0147] Example 3. Comprehensive Risk Score

[0148] Using the patient cohort and methods described in Example 1, various cell populations were analyzed based on IMC data to assess their prognostic value, both individually and in combination.

[0149] Using random forest classification and manual sorting training data, predefined cell populations were identified, including LECs (Prox1+), T cells (CD3+, CD45+), CD3-leukocytes (CD3-, CD45+), and tumor cells (E-cadherin+). These cell populations were further analyzed using the marker-based clustering method described in Example 1. Combined with spatial information, the analysis revealed several cell phenotypes with significant independent prognostic value: LECp (Prox1+, Ki-67+), Tc granzyme B+ (CD3+, CD45+, CD8a+, granzyme B+), and Tc TILs (CD3+, CD45+, CD8a+, distance from the nearest tumor <0 µm). For each subpopulation, its number in the sample was calculated as a percentage of the corresponding parental cell type. Values ​​above the median were considered "high," and values ​​below or equal to the median were considered "low."

[0150] The multivariate hazard ratio (HR) for each clinical parameter and cell phenotype was calculated using a Cox proportional hazards model. The multivariate model included known prognostic parameters: pN, pT, and age. Factors with significant independent prognostic value were selected for comprehensive risk analysis (Table 5). Factors with an HR less than 0 decreased the comprehensive risk score by 1 point, while factors with an HR greater than 0 increased the comprehensive risk score by 1 point.

[0151] Finally, patients were grouped according to their total risk score. Figure 9A Patients with high-risk scores (2 and 3) showed significantly worse recurrence-free survival and overall survival compared to those with low-risk scores (-1, 0, and 1). Figure 9B The estimated mean recurrence-free survival was 52% for patients with high-risk scores and 92% for patients with low-risk scores. Figure 9C These results suggest that combining LEC proliferation measurements with other prognostic parameters has prognostic potential.

[0152] Table 5

[0153] 1 HR makes adjustments based on age, PT (practice time), and PN (professional knowledge and skills).

Claims

1. A method of predicting the clinical outcome of a subject diagnosed with an epithelial cancer, comprising: i) determining the proportion of proliferative lymphatic endothelial cells in the invasive front of a cancer tissue sample obtained from the subject, wherein the lymphatic endothelial cells are identified by a nuclear biomarker specific for lymphatic endothelial cells; ii) comparing the proportion determined in step i) to a pre-set reference value; and ii) concluding that the likelihood of the subject obtaining a good clinical outcome is decreased when the proportion determined in step i) is higher than the pre-set reference value; or concluding that the likelihood of the subject obtaining a good clinical outcome is increased when the proportion determined in step i) is lower than the pre-set reference value.

2. The method of claim 1, wherein the proportion of proliferative lymphatic endothelial cells is calculated by dividing the number of proliferative lymphatic endothelial cells by the total number of lymphatic endothelial cells in the area of interest in the invasive front of the cancer tissue sample.

3. The method of claim 1 or 2, wherein the nuclear biomarker is Prospero Homeobox Protein 1 (PROX1).

4. The method of any one of claims 1-3, wherein the proliferative cells are identified based on the proliferation marker KI-67.

5. The method of any one of claims 1-4, wherein the proliferative lymphatic endothelial cells are identified based on the co-expression of the proliferation marker KI-67 and the lymphatic endothelial cell marker Prospero Homeobox Protein 1 (PROX1).

6. The method of any one of claims 1-3, wherein the tissue sample is a histological tissue sample.

7. The method of claim 6, wherein the co-expression of KI-67 and PROX1 is detected by contacting the histological tissue sample with a detectably labeled anti-KI-67 antibody and a detectably labeled anti-PROX1 antibody.

8. The method of any one of claims 1-5, wherein the expression of KI-67 and PROX1 is detected by single cell RNA sequencing or spatial transcriptomics.

9. The method of any one of claims 1-8, further comprising determining the expression level of one or more biomarkers listed in Table 2 and comparing the determined expression level to a respective pre-set reference value.

10. The method of claim 9, wherein the expression level is determined by a technique selected from the group consisting of immunohistochemistry, single cell RNA sequencing, and spatial transcriptomics.

11. The method of any one of claims 1-10, wherein the epithelial cancer is squamous cell carcinoma, adenocarcinoma, transitional cell carcinoma, or basal cell carcinoma.

12. The method of claim 11, wherein the squamous cell carcinoma is oral squamous cell carcinoma (OSCC), preferably selected from the group consisting of tongue OSCC, base of tongue OSCC, buccal OSCC, gingival OSCC, floor of mouth OSCC, and hard palate OSCC.

13. The method of any one of claims 1-12, wherein the clinical outcome is represented by recurrence-free survival, disease-specific survival, or overall survival. ​ 14. A kit for use in the method of any one of claims 1-13, comprising a reagent capable of indicating the presence of a nuclear biomarker specific for lymphatic endothelial cells and a reagent capable of indicating cell proliferation.

15. The kit of claim 14, comprising at least one of a reagent capable of indicating the presence of PROXl and a reagent capable of indicating the presence of KI-67.

16. The kit of claim 15, comprising at least one of a first binding partner capable of specifically binding to PROXl and a second binding partner capable of specifically binding to KI-67.

17. The kit of any one of claims 14-16, further comprising one or more additional reagents capable of indicating the presence of a marker listed in Table 2, and / or one or more binding partners capable of specifically binding to a protein marker listed in Table 2.

18. The kit of claim 16 or 17, wherein one or more of the binding partners are detectably labeled, directly or indirectly, independently of one another.

Images