COX6B2, a biomarker for predicting the risk of recurrence of cholangiocarcinoma, and related products.

Abstract

This invention discloses a biomarker, COX6B2, and related products for predicting the risk of recurrence of cholangiocarcinoma. The biomarker COX6B2 has excellent predictive efficacy for the risk of recurrence of cholangiocarcinoma, with high accuracy, sensitivity, and specificity, and an AUC value as high as 0.87. It effectively solves the technical problem of the lack of reliable biomarkers to guide personalized treatment of cholangiocarcinoma patients in current clinical practice, and provides a key tool for the accurate assessment of the risk of recurrence of cholangiocarcinoma, which has important scientific significance and clinical application value.

Description

Technical Field

[0001] This invention belongs to the field of biomedical technology, specifically relating to a biomarker COX6B2 for predicting the risk of recurrence of cholangiocarcinoma and related products. Background Technology

[0002] Cholangiocarcinoma (CCA), a highly aggressive malignant tumor, is experiencing a continuous rise in global incidence. Although surgical resection remains the primary treatment, the high recurrence rate of 50%-70% after surgery severely restricts patients' survival prognosis and quality of life. Current clinical prognostic indicators, such as tumor size, lymph node metastasis, vascular invasion, and CA19-9 levels, have limited accuracy in predicting recurrence risk, highlighting the urgent need to discover more reliable biomarkers to guide personalized treatment strategies.

[0003] Cytochrome C oxidase subunit 6B2 (COX6B2), as the nuclear coding subunit of mitochondrial respiratory chain complex IV, plays a crucial role in oxidative phosphorylation (OXPHOS) and cellular energy metabolism. It is not only a core component of complex IV assembly and functional maintenance, participating in ATP generation and mitochondrial membrane potential stability through mediating the terminal steps of the electron transport chain, but also exhibits oncogenic properties through its regulatory role in metabolic reprogramming. Related studies have shown that COX6B2 is abnormally expressed in various malignant tumors such as lung adenocarcinoma and pancreatic ductal adenocarcinoma, but there are no studies or reports on its association with CCA recurrence. Summary of the Invention

[0004] In order to solve the above-mentioned technical problems existing in the field, the purpose of this invention is to provide a biomarker COX6B2 and related products for predicting the risk of recurrence of cholangiocarcinoma.

[0005] The present invention achieves the above-mentioned objectives by adopting the following technical solution:

[0006] A first aspect of the present invention provides a biomarker for predicting the risk of recurrence of cholangiocarcinoma.

[0007] Furthermore, the biomarker is COX6B2.

[0008] In this invention, COX6B2 comprises the COX6B2 gene and the COX6B2 protein. The COX6B2 gene is transcribed and translated into the COX6B2 protein product in the research subject (subject). The Gene ID of the COX6B2 gene is 125965, and its full name is cytochrome c oxidase subunit 6B2 [Homo sapiens (human)]. Detailed information about this gene can be obtained from the NCBI database (https: / / www.ncbi.nlm.nih.gov / gene / ) based on the aforementioned Gene ID.

[0009] In this invention, the biomarker refers to a molecular indicator with specific biological characteristics, biochemical features, or other properties, which can be used to determine the presence or absence of a specific disease or condition and / or the severity of a specific disease or condition. In a specific embodiment of this invention, the biomarker is COX6B2.

[0010] A second aspect of the invention provides the use of a reagent for detecting the expression level of COX6B2 in a sample in the preparation of a product for predicting the risk of recurrence of cholangiocarcinoma.

[0011] Furthermore, the reagent includes:

[0012] Reagents for detecting the protein expression level of COX6B2 in samples;

[0013] A reagent for detecting the level of COX6B2 DNA in a sample;

[0014] Reagents for detecting COX6B2 RNA levels in samples; or

[0015] A reagent for detecting the number of COX6B2-positive cells in a sample.

[0016] Furthermore, the reagents for detecting the protein expression level of COX6B2 in the sample include antibodies, aptamers, agglutinants, and / or small molecule compounds that specifically bind to the protein encoded by COX6B2.

[0017] Optionally, the reagents for detecting the COX6B2 DNA level in the sample include reagents for detecting the COX6B2 DNA expression level, DNA methylation level, DNA acetylation level, DNA phosphorylation level, DNA hydroxymethylation level and / or SNP site genotype in the sample.

[0018] Optionally, the reagents for detecting the RNA level of COX6B2 in the sample include reagents for detecting the expression levels of COX6B2 mRNA, lncRNA, and / or miRNA in the sample;

[0019] Optionally, the reagent for detecting the number of COX6B2-positive cells in the sample includes a reagent for detecting the number of COX6B2-positive cells by immunohistochemistry.

[0020] Furthermore, the reagents for detecting the COX6B2 DNA level in the sample include reagents for detecting the level using sequencing technology;

[0021] Optionally, the reagents for detecting the RNA level of COX6B2 in the sample include primers that specifically amplify COX6B2 and / or probes that specifically recognize COX6B2.

[0022] Furthermore, the samples include tissue samples, blood samples, serum samples, plasma samples, exosome samples, and / or cell samples.

[0023] In some implementations, the sequencing technology refers to any reagent capable of detecting the DNA expression level, DNA methylation level, DNA acetylation level, DNA phosphorylation level, DNA hydroxymethylation level, and / or SNP locus genotype of COX6B2 in a sample. Exemplarily, the sequencing technology includes, but is not limited to: Sanger sequencing technology, Illumina sequencing technology, Ion Torrent sequencing technology, PacBio sequencing technology, and Oxford Nanopore long-read sequencing technology.

[0024] In this invention, the sample refers to a composition obtained from or derived from the target subject, which contains cellular entities and / or other molecular entities to be characterized and / or identified, for example, based on physical, biochemical, chemical and / or physiological characteristics.

[0025] In some embodiments, the sample may be a tissue sample, blood, or other fluid sample of biological origin, such as a biopsy tissue sample or tissue culture or cells derived therefrom, obtained from a subject (the subject may be a human or a non-human mammal, preferably a human). The source of the tissue sample may be solid tissue, such as fresh, frozen, and / or preserved organ or tissue samples, biopsy tissue or aspirates; blood or any blood component; body fluids; cells from any stage of an individual's pregnancy or development; or plasma. The term "sample" includes biological samples that have been processed in any way after their acquisition, such as by reagent treatment, stabilization, or enrichment for certain components (such as proteins or polynucleotides), or embedding in a semi-solid or solid matrix for sectioning purposes. In this invention, there are no particular limitations on the sample, and applications related to the detection of COX6B2 in any sample for the purpose of predicting the risk of recurrence of cholangiocarcinoma in a subject fall within the scope of protection of this invention.

[0026] In some embodiments, the samples described in this invention include, but are not limited to: tissues, blood, tissue-derived cells, blood-derived cells, serum, plasma, lymph, synovial fluid, exosomes, cell extracts, feces, urine, saliva, sputum, synovial fluid, pleural effusion, peritoneal effusion, serous cavity effusion, lymph, cerebrospinal fluid, uterine cavity fluid, digestive juices, bile, pulmonary bronchial lavage fluid, organs, and any combination thereof. In a preferred embodiment, the sample is selected from the subject's tissue (e.g., subject-derived cholangiocarcinoma tissue).

[0027] In some embodiments, the subject refers to any animal, including both human and non-human animals. The term non-human animals includes all vertebrates, such as mammals, including non-human primates (especially higher primates), sheep, dogs, rodents (such as mice or rats), guinea pigs, goats, pigs, cats, rabbits, cattle, and any livestock or pets; as well as non-mammals, such as chickens, amphibians, reptiles, etc. In specific embodiments of the present invention, the subject is preferably a human.

[0028] In this invention, the primers refer to 7-50 nucleic acid sequences capable of forming base pairs complementary to the template strand and serving as a starting point for template strand replication. Primers are typically synthesized, but naturally occurring nucleic acids can also be used. The primer sequence does not necessarily need to be completely identical to the template sequence, as long as it is sufficiently complementary to hybridize with the template.

[0029] In this invention, the probe refers to a nucleic acid fragment, such as RNA or DNA, ranging from a few to hundreds of bases in length. This nucleic acid fragment can specifically bind to mRNA and can determine the presence of a specific mRNA through labeling. The probe can be prepared in the form of oligonucleotide probes, single-stranded DNA probes, double-stranded DNA probes, and RNA probes.

[0030] In this invention, the antibody refers to a specific immunoglobulin targeting an antigenic site. The antibody in this invention refers to an antibody that specifically binds to the COX6B2 protein described in this invention, and can be manufactured according to conventional methods in the art. The antibody can take the form of polyclonal or monoclonal antibodies, antibody fragments (such as Fab, Fab', F(ab')2, and Fv fragments), single-chain Fv (scFv) antibodies, multispecific antibodies (such as bispecific antibodies), monospecific antibodies, monovalent antibodies, chimeric antibodies, humanized antibodies, human antibodies, fusion proteins containing an antigen-binding site, and any other modified immunoglobulin molecule containing an antigen-binding site, provided that the antibody exhibits the desired biological binding activity. In a specific embodiment of this invention, the antibody is a rabbit-derived anti-COX6B2 monoclonal antibody (PA5-50213, ThermoFisher).

[0031] A third aspect of the present invention provides a product for predicting the risk of recurrence of cholangiocarcinoma.

[0032] Furthermore, the product contains reagents for detecting the expression level of COX6B2 in the sample.

[0033] Furthermore, the reagent is the reagent described in the second aspect of the present invention;

[0034] Optionally, the product includes a reagent kit, a chip, or a test strip.

[0035] In some implementations, the kits include, but are not limited to: ELISA kits, protein chip kits, RT-PCR kits, DNA chip kits, rapid detection kits, or MRM (multiple reaction monitoring) kits.

[0036] In some embodiments, the kit comprises primers or probes that specifically bind to COX6B2. In some embodiments, the detection chip comprises a solid support and probes specifically recognizing COX6B2 attached to the solid support.

[0037] In some embodiments, the kit may also contain one or more substances selected from the group consisting of: containers, instructions for use, positive controls, negative controls, buffers, auxiliaries, or solvents.

[0038] In some embodiments, the ELISA kit may further include elements necessary for performing the ELISA. The ELISA kit may contain an antibody specifically targeting a protein (the COX6B2 protein described in this invention). The antibody has high selectivity and affinity for the labeled protein, no cross-reactivity with other proteins, and may be a monoclonal antibody, polyclonal antibody, or recombinant antibody. Furthermore, the ELISA kit may contain an antibody specifically targeting a control protein. Additionally, the ELISA kit may further contain reagents capable of detecting the bound antibody, such as a labeled secondary antibody, a chromophore, an enzyme (e.g., conjugated to the antibody), a substrate thereof, or a substance capable of binding the antibody.

[0039] In some embodiments, the protein chip kit may include a chip carrier immobilized with probes (such as antibodies or aptamers) that specifically bind to the COX6B2 protein, and may also include blocking solution, washing buffer, labeling reagents (such as fluorescently labeled secondary antibodies) and substrates required for color development or detection, for high-throughput detection of the expression level of COX6B2 protein in samples.

[0040] In some embodiments, the RT-PCR kit may further include elements necessary for the reverse transcription polymerase chain reaction. The RT-PCR kit contains a pair of primers specifically targeting a gene encoding a marker protein. Each primer is a nucleotide having a nucleic acid sequence specifically targeting the gene. The RT-PCR kit may also include test tubes or suitable dishes, reaction buffers (at different pH values ​​and magnesium concentrations), deoxynucleotides (dNTPs), enzymes (e.g., Taq polymerase and reverse transcriptase), deoxyribonuclease inhibitors, ribonuclease inhibitors, DEPC-water, and sterile water.

[0041] In some implementations, the DNA chip kit may include a chip immobilized with a COX6B2 gene-specific probe, as well as hybridization buffer, washing solution, labeling reagent (such as biotin-labeled cDNA), signal detection reagent, etc., to detect the expression or mutation of the COX6B2 gene through nucleic acid hybridization reaction, thereby indirectly linking it to its protein function.

[0042] In some implementations, the rapid test kit may include antibodies or other binding molecules that specifically recognize COX6B2 protein, chromogenic reagents (such as colloidal gold, latex particles, etc.), sample processing solutions, reaction strips or cartridges, etc. It is simple and rapid to operate, and can intuitively display the expression level of COX6B2 protein in the sample, making it suitable for preliminary screening at the field or bedside.

[0043] In some implementations, the MRM (Multiple Reaction Monitoring) kit may include specific peptide standards for COX6B2 protein, isotope internal standards, protease digestion reagents, chromatographic columns, mobile phase additives, etc., and uses mass spectrometry multiple reaction monitoring technology to quantitatively detect COX6B2 protein in samples, with high specificity and sensitivity.

[0044] In some implementations, the chip (microarray chip) refers to a solid support containing linked nucleic acid or peptide probes. The array typically contains a variety of different nucleic acid or peptide probes attached to a substrate surface at different known locations. These arrays, also known as microarrays, can typically be produced using mechanosynthesis or photoguided synthesis, which combines photolithography and solid-phase synthesis methods. The array can contain a flat surface or can be nucleic acids or peptides on beads, gels, polymer surfaces, fibers such as optical fibers, glass, or any other suitable substrate. The array can be packaged in a way that allows for diagnostic or other manipulation of a fully functional device.

[0045] In some embodiments, the chip includes a gene chip and a protein chip; the gene chip includes a solid support and oligonucleotide probes ordered immobilized on the solid support, the oligonucleotide probes specifically corresponding to part or all of the sequence represented by COX6B2. The protein chip includes a solid support and specific antibodies or ligands for the COX6B2-encoded protein immobilized on the solid support. The solid support can be made of various commonly used materials in the chip field, including but not limited to: plastic products, microparticles, membrane carriers, etc.

[0046] In some implementations, the test strip may include a sample pad, a conjugate pad, a reaction membrane, an absorbent pad, and a base plate. The conjugate pad is immobilized with labeled anti-COX6B2 antibodies (such as colloidal gold-labeled rabbit monoclonal antibodies). The reaction membrane has a detection line (coated with anti-COX6B2 antibodies) and a control line (coated with anti-labeled antibodies). It may also be equipped with a sample processing solution. Through the specific binding reaction of antigen and antibody, the detection line and control line are made colored, allowing for a direct assessment of the expression level of COX6B2 protein in the sample.

[0047] A fourth aspect of the present invention provides a system or apparatus for predicting the risk of recurrence of cholangiocarcinoma.

[0048] Furthermore, the system or device includes a processor, an input module, and an output module;

[0049] The processor is used to perform logical operations on the input information using bioinformatics methods; the input module is used to input the COX6B2 expression level in the subject sample, and a computer-readable medium containing instructions that, when executed by the processor, execute an algorithm at the COX6B2 input expression level; the output module is used to output the subject's risk of bile duct cancer recurrence.

[0050] In this invention, the system or apparatus is a method for distinguishing different components, elements, parts, sections, or assemblies at different levels. However, if other terms can achieve the same purpose, they can be replaced by other expressions. Those skilled in the art will know that this invention can be implemented as a device, method, or computer program product. Therefore, the disclosure of this invention can be specifically implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. Furthermore, in some specific embodiments, this invention can also be implemented as a computer program product in one or more computer-readable media containing computer-readable program code.

[0051] The present invention also provides a computer-readable storage medium.

[0052] Furthermore, the computer-readable storage medium stores a computer program that, when executed by a processor, implements the system or apparatus of the present invention as described above.

[0053] The fifth aspect of the invention provides for any of the following applications:

[0054] (1) Application of reagents for detecting COX6B2 expression levels in samples in the preparation of systems or devices for predicting the risk of recurrence of cholangiocarcinoma;

[0055] (2) The application of the biomarker COX6B2 described in the first aspect of the present invention in combination with tumor markers CA19-9, CEA, vascular invasion indicators and tumor size indicators in the preparation of products for predicting the risk of recurrence of cholangiocarcinoma or in the construction of predictive models for predicting the risk of recurrence of cholangiocarcinoma.

[0056] In some implementations, the tumor marker CA19-9 refers to carbohydrate antigen 19-9, a tumor marker commonly used for the auxiliary diagnosis and recurrence monitoring of diseases such as cholangiocarcinoma (CCA).

[0057] In some embodiments, the tumor marker CEA refers to carcinoembryonic antigen, which may be elevated in various malignant tumors. In some embodiments of this application, it is used as one of the clinicopathological features associated with the risk of CCA recurrence to construct a composite risk stratification model.

[0058] In some embodiments, the vascular invasion index refers to the extent to which tumor cells invade surrounding blood vessels, and is an important pathological feature for assessing tumor invasiveness and prognosis. In some embodiments of this application, it is incorporated as a factor influencing the risk of CCA recurrence into a risk stratification model.

[0059] In some embodiments, the tumor size index refers to the size of the detected cholangiocarcinoma tumor and is a basic indicator describing the degree of tumor progression. In some embodiments of this application, it is used as one of the clinicopathological features for constructing a CCA recurrence risk prediction model.

[0060] The present invention also provides a method for predicting the risk of recurrence of cholangiocarcinoma.

[0061] Furthermore, the method includes detecting the expression level of COX6B2 in the test sample from the subject, and predicting the risk of bile duct cancer recurrence based on the detected COX6B2 expression level.

[0062] In this invention, any method that uses the biomarker COX6B2 described in this invention (or in combination with other indicators, for example, the biomarker COX6B2 in combination with the tumor marker CA19-9, the tumor marker CEA, the vascular invasion indicator, and the tumor size indicator) to predict the risk of bile duct cancer recurrence falls within the protection scope of this invention. It is not limited to the specific method of using the biomarker COX6B2. As long as the purpose of predicting the risk of bile duct cancer recurrence can be achieved or substantially achieved, it falls within the protection scope of this invention.

[0063] Compared with the prior art, the advantages and beneficial effects of the present invention are as follows:

[0064] This invention is the first to creatively discover that COX6B2 can serve as an effective biomarker for predicting the risk of bile duct cancer recurrence. Clinical validation in a training cohort (115 patients) and a validation cohort (48 patients) has demonstrated its excellent predictive efficacy for the risk of bile duct cancer recurrence, with high accuracy, sensitivity, and specificity, and an AUC value as high as 0.87. This invention effectively solves the current technical challenge of lacking reliable biomarkers to guide personalized treatment for bile duct cancer patients in clinical practice, providing a key tool for the accurate assessment of the risk of bile duct cancer recurrence and a basis for personalized treatment strategies for patients. It has significant scientific and clinical application value. Attached Figure Description

[0065] Figure 1 Bioinformatics analysis results: Figure A: Comparison of differentially expressed gene (DEG) distribution between the relapse and non-relapse groups in the TCGA-CHOL and GSE107943 datasets, showing the number of significantly upregulated DEGs in each dataset (27 in the TCGA-CHOL relapse group and 393 in the GSE107943 relapse group); Figure B: Kaplan-Meier curve of COX6B2 expression level and relapse-free survival (RFS) in the GSE107943 cohort, showing that high COX6B2 expression is significantly associated with increased relapse risk (log-rank test); Figure C: Volcano plot of differentially expressed genes between the relapse and non-relapse groups in the GSE107943 dataset, where COX6B2 shows significant upregulation (marked with red dots); Figure D: Volcano plot of differentially expressed genes between the relapse and non-relapse groups in the TCGA-CHOL dataset, where COX6B2 also shows an upregulation trend (marked with red dots).

[0066] Figure 2The expression and predictive efficacy of COX6B2 in CCA were validated. Figures AC and DE show the immunohistochemical staining results of COX6B2 in CCA tissue, indicating weak (A), medium (B), and strong (C) staining intensities (primarily cytoplasmic expression). Figures DE and E show the ROC curve (D) and bootstrap internal validation curve (E) for COX6B2 predicting CCA recurrence in the training cohort, with an AUC of 0.87 (95% CI: 0.80–0.93) and an optimal cutoff value of 0.73. Figures F and G show the ROC curve (F) and validation results (G) for COX6B2 predicting CCA recurrence in the independent validation cohort, with an AUC of 0.81 (95% CI: 0.69–0.93), confirming its cross-cohort reliability.

[0067] Figure 3 Based on COX6B2, a prognostic stratification and nomogram model were established. Figure AB shows the Kaplan-Meier curves of relapse-free survival (RFS, A) and overall survival (OS, B) for the high and low COX6B2 expression groups in the training cohort, demonstrating a significant difference between the two groups (log-rank P < 0.01). Figure CD shows the Kaplan-Meier curves of RFS (C) and OS (D) for the high and low COX6B2 expression groups in the validation cohort, presenting a prognostic stratification consistent with the training cohort (log-rank P < 0.01). Figure EF shows the nomogram model integrating COX6B2 with clinicopathological features (CA19-9, CEA, etc.), visually demonstrating the predictive weight of each factor on relapse risk (E represents the training cohort, F represents the validation cohort).

[0068] Figure 4 The predictive performance and clinical application of the composite risk stratification model are shown in Figures AB and CD. Figures AB and AB show the ROC curves of the composite model (COX6B2 + clinicopathological features) in the training cohort (A) and validation cohort (B), respectively. Both models have an AUC of 0.93, significantly better than COX6B2 alone. Figures CD and CD show the RFS Kaplan-Meier curves of the high-risk and low-risk groups defined by the composite model in the training cohort (C) and validation cohort (D), respectively. The differences between the two groups are significant (log-rank P < 0.01). Figure EF shows a bar chart comparing the 5-year recurrence rate (90.62%-95.59%) of the high-risk group and the recurrence-free rate (85.11%-87.50%) of the low-risk group in the training cohort (E) and validation cohort (F), visually demonstrating the model's risk stratification efficacy. Detailed Implementation

[0069] The present invention will be further illustrated below with reference to specific embodiments. These specific embodiments are for illustrative purposes only and should not be construed as limiting the invention. Those skilled in the art will understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the invention. The scope of the invention is defined by the claims and their equivalents. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains.

[0070] The reagents, raw materials, and experimental consumables used in this invention are readily available to those skilled in the art and, unless otherwise specified, can be obtained commercially. Experimental methods not specifying particular conditions in this invention are typically performed under conventional conditions or according to the manufacturer's recommendations. In particular, the following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention in any way. It should be noted that the experimental conditions and results described in the following examples are for illustrative purposes only and should not, and will not, limit the invention as described in the claims.

[0071] Example: Screening and validation of biomarkers for predicting the risk of cholangiocarcinoma recurrence

[0072] 1. Experimental Methods

[0073] (1) Bioinformatics Analysis

[0074] Transcriptomic analysis data with matched clinical annotations were obtained from two independent cholangiocarcinoma cohorts.

[0075] The TCGA-CHOL dataset includes RNA-seq profiles and recurrence status for 36 tumors, retrieved from the UCSC Xena platform (https: / / xena.ucsc.edu). Patients were stratified into a recurrence group (n=10) and a non-recurrence group (n=14).

[0076] The GSE107943 dataset was obtained from the Gene Expression Omnibus (https: / / www.ncbi.nlm.nih.gov / geo / ) and includes 21 relapsed cases and 9 non-relapsed cases selected using the same clinical criteria.

[0077] Differential expression analysis was performed using DESeq2 (v1.38.3), and significantly differentially expressed genes (DEGs) were identified based on a strict double threshold: absolute log2 fold change (|log2FC|>1) and adjusted p-value (false detection rate, FDR) <0.05. Survival-related genes were systematically identified using a three-step workflow: (1) batch log-rank test (Kaplan-Meier stratification by median expression) was performed on recurrent upregulated genes in GSE107943, (2) univariate Cox regression (proportional hazard model, p<0.05 threshold), and (3) cross-cohort validation was performed by intersecting significant survival genes in GSE107943 with recurrent upregulated DEGs in TCGA-CHOL.

[0078] (2) Patient cohort

[0079] This study included two independent cohorts: a training cohort (containing 115 patients who underwent radical surgery at the Second Hospital of Hebei Medical University between July 2010 and August 2024) and a validation cohort (containing 48 patients from Hengshui People's Hospital). Both cohorts met the standardized inclusion criteria. The institutional review committee approved the study protocol and obtained written informed consent from all participants. Tumor staging followed the 2017 UICC TNM classification (8th edition). Recurrence was diagnosed by enhanced CT, MRI, or DSA imaging, combined with elevated serum CA19-9 (>37 U / mL), and confirmed by histopathology where feasible. Disease-free survival (DFS) was calculated from surgery to first recurrence (intrahepatic / extrahepatic), and overall survival (OS) was defined as the time from surgery to death or last follow-up (censored for surviving patients). Subsequent protocols were implemented as previously described.

[0080] In the training cohort, the ratio of cholangiocarcinoma tissue samples from patients with recurrent cholangiocarcinoma to cholangiocarcinoma tissue samples from patients with non-recurrent cholangiocarcinoma was 72:43.

[0081] In the validation cohort, the ratio of cholangiocarcinoma tissue samples from patients with recurrent cholangiocarcinoma to cholangiocarcinoma tissue samples from patients with non-recurrent cholangiocarcinoma was 31:17.

[0082] (3) Immunohistochemical analysis

[0083] Immunohistochemical staining for COX6B2 expression was performed according to standard protocol. After antigen retrieval, tissue sections were incubated with rabbit-derived anti-COX6B2 monoclonal antibody (PA5-50213, ThermoFisher; 1:100 dilution), followed by incubation with horseradish peroxidase-conjugated secondary antibody (PV-9000, ZSGB-BIO). 3,3'-diaminobenzidine was used for staining, and hematoxylin was used for counterstaining. Staining assessment was performed according to established semi-quantitative criteria, evaluating staining intensity (0 = none; 1 = weak cytoplasmic staining; 2 = moderate membranous / cytoplasmic staining; 3 = strong diffuse staining) and cell distribution (0 = ≤10%; 1 = 11-25%; 2 = 26-50%; 3 = 51-75%; 4 = >75% positive cells). The Immune Response Score (IRS) is the product of the intensity and distribution scores, and the final classification is defined as: negative (IRS≤3), weak (4-6), moderate (7-9), or strong (>9). All assessments were performed by two independent pathologists unaware of the clinical outcomes. For cases with inconsistencies (IRS score difference >2 points), consensus review was conducted using a multi-observer microscopy system.

[0084] (4) Statistical analysis

[0085] Continuous variables were classified into two categories using a cohort-specific median threshold. Between-group comparisons (relapsed vs. non-relapsed) were performed using the χ² method. 2 Tests were performed (categorical variables) and the Mann-Whitney U test (continuous variables were tested for Shapiro-Wilk normality). The discriminative power of the Cox model was quantified using time-dependent receiver operating characteristic (ROC) analysis. For the training cohort, internal validation was performed using 1000 guided resampling samples to compute adjusted area under the curve (AUC) estimates and bias-adjusted 95% confidence intervals. The DeLong test was used for paired ROC curve comparisons. Survival endpoints (OS / RFS) were calculated from surgical intervention to event occurrence or last follow-up and analyzed using Kaplan-Meier estimates and log-rank tests. A phased Cox regression approach was performed: (1) univariate screening (P < 0.1 inclusion threshold), (2) multivariate analysis with backward elimination (P < 0.05 retained). All analyses were performed using SPSS 29.0 and R 4.4.2, as well as the survival, pROC, and survminer packages.

[0086] 2. Experimental Results

[0087] (1) Identification of COX6B2 as a potential biomarker for predicting recurrence of cholangiocarcinoma

[0088] Analysis of the TCGA-CHOL dataset showed 27 significantly upregulated differentially expressed genes in the recurrent cholangiocarcinoma group, while 393 significantly upregulated genes were found in the GSE107943 dataset. Figure 1 A). Batch log-rank test identified 206 genes significantly associated with recurrence, univariate Cox regression analysis identified 252 candidate genes associated with the recurrence time of cholangiocarcinoma, and cross-validation revealed 166 overlapping genes. Subsequent cross-validation with upregulated genes in the TCGA-CHOL recurrence group identified COX6B2 as a key overlapping gene.

[0089] In the GSE107943 queue, the log-rank test ( Figure 1 (B) The results showed that COX6B2 expression level was significantly associated with recurrence-free survival. Univariate Cox regression analysis also confirmed that COX6B2 overexpression was significantly associated with an increased risk of cholangiocarcinoma recurrence (HR = 1.351, 95% CI = 1.023–1.783, p = 0.034). These results indicate that for every unit increase in COX6B2 expression, the risk of cholangiocarcinoma recurrence increases by 35.1%.

[0090] In addition, the volcano map visually depicts GSE107943 ( Figure 1 C) and TCGA-CHOL Figure 1 D) Distribution patterns of differentially expressed genes between the recurrent and non-recurrent cholangiocarcinoma groups in the dataset. COX6B2 showed a consistent upregulation trend in both cohorts of the recurrent cholangiocarcinoma group.

[0091] Cross-database bioinformatics analysis revealed that COX6B2 exhibited an overexpression pattern in patients with recurrent cholangiocarcinoma. Its expression level was significantly negatively correlated with recurrence-free survival, suggesting it may serve as an independent molecular biomarker for postoperative recurrence of cholangiocarcinoma.

[0092] (2) Training and validation of COX6B2 in predicting recurrence in patients with cholangiocarcinoma

[0093] Immunohistochemical analysis showed that COX6B2 was mainly expressed in the cytoplasm of cholangiocarcinoma tissues, with staining intensities categorized as weak, medium, and strong. Figure 2 We established a training cohort consisting of 115 patients with a recurrence rate of 62.6% and a validation cohort consisting of 48 patients with a recurrence rate of 64.6%. The clinicopathological characteristics of the two groups were similar to validate their clinical predictive value. The median age of the training and validation cohorts was 61 years and 63.5 years, respectively (Table 1). There were no significant differences in recurrence rate and baseline characteristics between the two groups (P>0.05), thus minimizing the risk of selection bias.

[0094] Table 1. Clinicopathological characteristics of the clinical cohort

[0095]

[0096]

[0097] Cox proportional hazards regression analysis showed that COX6B2 had excellent relapse discrimination ability in the training cohort (AUC = 0.87, sensitivity 88%, specificity 71%, 95% CI: 0.80–0.93; optimal cutoff value = 0.73). Figure 2 DE. Bootstrap internal validation (1000 resampling) further enhanced its robustness (AUC = 0.859, sensitivity 80.7%, specificity 81.1%, 95% CI: 0.827–0.865). COX6B2 maintained consistent diagnostic performance in the independent validation cohort (AUC = 0.81, sensitivity 84%, specificity 71%, 95% CI: 0.69–0.93). Figure 2 The cross-cohort reliability was confirmed by FG. In summary, COX6B2 demonstrated excellent and stable recurrence predictive efficacy through multi-phase clinical validation. Its high AUC value and cross-cohort reproducibility highlight its translational potential as a non-invasive diagnostic biomarker for postoperative CCA recurrence.

[0098] (3) Development and verification of prognostic nomogram based on COX6B2

[0099] Semi-quantitative immunohistochemical analysis classified CCA tissues into four expression levels: negative, weak, intermediate, and strong. ROC curve analysis determined the optimal cutoff value to be 0.73 (Youden index: sensitivity 88%, specificity 71%), enabling the four-level classification to be divided into a clinically operable binary risk model (low-risk vs. high-risk). The validation cohort was stratified using the same Youden index threshold as the training cohort to ensure model reproducibility. The median follow-up time in the training cohort was 21.2 months (range: 4.2–55.8 months), and the median follow-up time in the validation cohort was 21 months (range: 5.2–36.2 months). Survival analysis showed that in the training cohort, the high-risk group had a significantly longer recurrence-free survival (RFS; log-rank test, P < 0.01). Figure 3 A) and overall survival (OS); log-rank test, P<0.01; Figure 3 B) Significantly shortened. Similar prognostic stratification (RFS) was also observed in the independent validation cohort. Figure 3 C; OS: Figure 3 D, log-rank test, P<0.01).

[0100] We utilized data from the two independent cohorts (training and validation cohorts) to construct a prognostic nomogram by integrating the novel biomarker COX6B2 with established pathological risk indicators (including CA19-9, CEA, vascular invasion, and tumor size) to enhance clinical application. Figure 3 Traditional risk factors contribute little to the predictive model, while COX6B2 has a significant predictive weight, highlighting its crucial role in prognostic assessment.

[0101] After integrating the variables initially screened by univariate analysis (P<0.1, Table 2, where CI is confidence interval; HR is hazard ratio; *, P<0.05), multivariate Cox regression analysis showed that CEA, CA19-9, and COX6B2 were independent predictors of prognosis in both cohorts. Among them, COX6B2 was the most robust predictor, with an HR of 4.94 in the training cohort (95% CI = 2.6–9.4, P<0.01) and an HR of 4.38 in the validation cohort (95% CI = 1.52–12.62, P<0.05).

[0102] Table 2. Univariate and multivariate Cox proportional hazards regression analysis of relapse.

[0103]

[0104]

[0105] This nomogram not only visually highlights the superior prognostic value of COX6B2 compared to traditional biomarkers, but also demonstrates cross-cohort robustness as confirmed by multivariate analysis. These results underscore the clinical translational potential of the model provided in this invention.

[0106] (4) The risk stratification model integrating COX6B2 and clinicopathological features enhances the prediction of CCA recurrence.

[0107] To overcome the limitations of traditional clinicopathological markers in assessing the risk of CCA recurrence, we constructed a comprehensive risk stratification model that combines COX6B2 with established clinicopathological features such as CA19-9, CEA, vascular invasion, and tumor size. This model demonstrated superior predictive performance compared to single markers. In the training cohort, the model achieved an AUC of 0.929 (…). Figure 4 A), reaching 0.928 in the verification queue. Figure 4B) Both are superior to the AUC of single biomarkers. Furthermore, the Delong test showed that the diagnostic accuracy of this combined model was significantly higher than that of COX6B2 alone in both the training and validation cohorts (P<0.05). Table 3 details the advantages of this combined approach and lists comprehensive performance metrics, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). These findings highlight the potential of this combined model as a valuable clinical tool, enabling more accurate risk stratification and supporting personalized treatment strategies for CCA patients.

[0108] Table 3 shows the model's performance in assessing relapse risk.

[0109]

[0110] This invention aims to develop a precise and clinically operable risk stratification tool to identify high-risk cholangiocarcinoma patients requiring postoperative intensive therapy, while preventing overtreatment in low-risk cholangiocarcinoma patients. The final model has been applied to both the training (n=115) and validation (n=48) groups, classifying patients into high-risk and low-risk groups. In the training cohort, 59.1% (68 / 115) of patients were classified as high-risk, and 40.9% (47 / 115) were classified as low-risk; in the validation cohort, 66.7% (32 / 48) of patients were classified as high-risk, and 33.3% (16 / 48) were classified as low-risk.

[0111] Analysis showed that the RFS differences between the high-risk and low-risk groups in both cohorts were statistically significant (P<0.01), demonstrating the reliability and consistency of the model. Figure 4 CD). The relapse rate was significantly higher in the high-risk subgroup, with 95.59% (65 / 68) of patients in the training cohort and 90.62% (29 / 32) of patients in the validation cohort relapsing within 5 years. Figure 4 These results highlight the urgent need for adjuvant chemotherapy in this population to prevent relapse. In contrast, in low-risk patients, 85.11% (40 / 47) of patients in the training cohort and 87.5% (14 / 16) of patients in the validation cohort remained relapse-free during the same period. Figure 4 The combined risk stratification model (EF) indicates that these individuals can safely forgo unnecessary treatment and its associated toxicities and costs. Furthermore, the combined model outperformed COX6B2 alone, improving NPV by 11%. The higher NPV significantly reduced the likelihood of misdiagnosing low-risk patients as high-risk patients, thereby optimizing clinical decision-making. In conclusion, the risk stratification model effectively distinguishes between patients requiring adjuvant chemotherapy and those not requiring it, thus optimizing treatment strategies and improving patient outcomes.

Claims

1. Application of reagents for detecting COX6B2 expression levels in samples in the preparation of products for predicting the risk of recurrence of cholangiocarcinoma.

2. The application according to claim 1, characterized in that, The reagents include: Reagents for detecting the protein expression level of COX6B2 in samples; A reagent for detecting the level of COX6B2 DNA in a sample; Reagents for detecting COX6B2 RNA levels in samples; or A reagent for detecting the number of COX6B2-positive cells in a sample.

3. The application according to claim 2, characterized in that, The reagents used to detect the protein expression level of COX6B2 in the sample include antibodies that specifically bind to the protein encoded by COX6B2.

4. The application according to claim 2, characterized in that, The reagents used to detect the COX6B2 DNA level in the sample include reagents for detecting the COX6B2 DNA expression level in the sample.

5. The application according to claim 2, characterized in that, The reagents used to detect the RNA level of COX6B2 in the sample include reagents for detecting the mRNA expression level of COX6B2 in the sample.

6. The application according to claim 2, characterized in that, The reagents for detecting the number of COX6B2-positive cells in the sample include reagents for detecting the number of COX6B2-positive cells by immunohistochemistry.

7. The application according to claim 4, characterized in that, The reagents used to detect the COX6B2 DNA level in the sample include reagents that detect the level using sequencing technology.

8. The application according to claim 5, characterized in that, The reagents used to detect the RNA level of COX6B2 in the sample include primers that specifically amplify COX6B2 and / or probes that specifically recognize COX6B2.

9. The application according to claim 1, characterized in that, The samples include tissue samples, blood samples, and / or exosome samples.

10. The application according to claim 9, characterized in that, The blood sample includes a serum sample or a plasma sample.

11. The application according to claim 1, characterized in that, The samples include cell samples.

12. A system or device for predicting the risk of recurrence of cholangiocarcinoma, characterized in that, The system or apparatus includes a processor, an input module, an output module, and a computer-readable medium storing computer-executable instructions; The processor is used to call instructions stored in the computer-readable medium to perform logical operations on the input information using bioinformatics methods; the input module is used to input the COX6B2 expression level in the subject sample; the instructions stored in the computer-readable medium are executed by the processor to perform the algorithm on the COX6B2 input expression level; and the output module is used to output the subject's risk of bile duct cancer recurrence.

13. Application of reagents for detecting COX6B2 expression levels in samples in the preparation of systems or devices for predicting the risk of recurrence of cholangiocarcinoma.

14. The use of the reagents for detecting the expression level of COX6B2 and the tumor marker CA19-9 as described in claim 1, the reagents for detecting the tumor marker CEA, the reagents for detecting vascular invasion indicators, and the reagents for detecting tumor size indicators in combination in the preparation of a product for predicting the risk of recurrence of cholangiocarcinoma or in the construction of a predictive model for predicting the risk of recurrence of cholangiocarcinoma.

Images