Intestinal microbial markers for predicting the efficacy of drugs for inflammatory bowel disease and use thereof

Abstract

The application provides application of intestinal flora as a marker in preparation of a product for evaluating treatment effect of inflammatory bowel disease. The feasibility of using a microbial spectrum to predict the curative effect of drug treatment on patients with inflammatory bowel disease is achieved, and the achievement can accurately identify patients who need to optimize the treatment scheme in the early stage of the disease, thereby significantly improving the prognosis of the patients. Meanwhile, a construction method of an intestinal flora biomarker prediction model is provided, and through the combination of Faecalibacterium prausnitzii, Blautia massiliensis and Phascolarctobacterium faecium, the precise prediction of the invalid condition of drug treatment is realized, which highlights the potential of the new tool for early identification of patients who may need to optimize the treatment.

Description

Technical Field

[0001] This invention relates to the field of biotechnology, specifically to gut microbiota biomarkers for predicting the efficacy of drugs (e.g., 5-aminosalicylic acid, anti-tumor necrosis factor, ustekinumab) for inflammatory bowel disease and their applications. Background Technology

[0002] Inflammatory bowel disease (IBD), including ulcerative colitis (UC) and Crohn's disease (CD), is a chronic gastrointestinal disease characterized by recurrent inflammation, various complications, and unknown etiology. Ulcerative colitis originates in the colon and rectum, causing persistent or recurrent diarrhea, mucus and bloody stools, and can lead to complications such as toxic megacolon, gastrointestinal bleeding, and cancer. Crohn's disease can affect the entire digestive tract, including the colon, small intestine, stomach, and esophagus. Patients often experience abdominal pain and diarrhea, are prone to malnutrition, and may develop complications such as intestinal stricture, intestinal fistula, and intestinal obstruction, severely impacting their quality of life.

[0003] Current treatment options mainly include 5-aminosalicylic acid (5-ASA), glucocorticoids, immunosuppressants, and biologics. Mucosal healing (MH) is currently the treatment goal for IBD because it is associated with better disease prognosis.

[0004] 5-ASA is a first-line treatment for UC, with the fewest side effects and lowest cost among all IBD treatments. However, less than 50% of patients achieve hemorrhagic muscular dystrophy (MH) with 5-ASA. Identifying patients who fail 5-ASA treatment in the early stages of the disease allows for earlier escalation of treatment, enabling patients to reach treatment goals sooner and improving long-term prognosis. However, currently available clinical indicators are insufficient to predict 5-ASA treatment outcomes. Therefore, there is an urgent need for indicators that can be used to predict treatment efficacy, especially non-invasive testing methods.

[0005] Anti-tumor necrosis factor-α (TNF-α) agents are the earliest and most widely used biologics, capable of inducing and maintaining disease remission and MH. However, 20%–40% of IBD patients exhibit primary non-response to anti-TNF-α agents, and approximately 50% experience secondary non-response. Ustekinumab, an interleukin (IL)-12 / 23 inhibitor, is also a biologic, approved in China in 2020 for the treatment of inflammatory bowel disease. Recent studies suggest that gut microbiota has potential in predicting the efficacy of biologics in IBD patients, but reliable microbial biomarkers remain to be discovered.

[0006] The gut microbiome is central to the pathogenesis of IBD, and studies have shown that it interferes with pharmacokinetics and pharmacodynamics, affecting drug absorption, distribution, metabolism, excretion, and ultimately, therapeutic efficacy. Therefore, in-depth research to identify microbial biomarkers that can predict the efficacy of therapeutic drugs in IBD is of significant practical importance. Summary of the Invention

[0007] To address the lack of prediction of the treatment efficacy of inflammatory bowel disease in existing technologies, this invention provides an application of gut microbiota as a biomarker in the preparation of products that predict the efficacy of drugs for inflammatory bowel disease. The gut microbiota biomarker of this invention has high sensitivity and good specificity, and has profound application prospects for estimating the efficacy of drugs for inflammatory bowel disease.

[0008] In a first aspect, the present invention provides the use of gut microbial biomarkers in the preparation of products that predict the efficacy of drugs against inflammatory bowel disease.

[0009] Preferably, the intestinal microbial markers are selected from one or more of Faecalibacterium prausnitzii, Blautia massiliensis and Phascolarctobacterium faecium, especially a combination of three.

[0010] Preferably, the inflammatory bowel disease is selected from one or both of ulcerative colitis and Crohn's disease.

[0011] Preferably, the drug for treating inflammatory bowel disease is selected from one or more of 5-aminosalicylic acid, glucocorticoids, immunosuppressants, and biological agents.

[0012] More preferably, the drug for treating inflammatory bowel disease is selected from one or more of anti-tumor necrosis factor, interleukin (IL)-12 / 23 inhibitors, and 5-aminosalicylic acid.

[0013] In some embodiments of the present invention, the inflammatory bowel disease is ulcerative colitis, and the drug is 5-aminosalicylic acid.

[0014] In some embodiments of the present invention, the inflammatory bowel disease is ulcerative colitis, and the drug is an antitumor necrosis factor.

[0015] In some embodiments of the invention, the inflammatory bowel disease is ulcerative colitis or Crohn's disease, and the drug is an interleukin-12 / 23 inhibitor (e.g., ustekinumab).

[0016] The predictive criteria for the efficacy of drugs in treating inflammatory bowel disease are as follows: the abundance of gut microbiota markers in the patient's stool sample is measured. Compared with a normal reference, the lower the abundance of gut microbiota markers in the patient, the less effective the drug is in treating the patient's inflammatory bowel disease, and it is necessary to consider switching to other biologics and adjusting the treatment plan.

[0017] Preferably, the product can be a reagent, reagent kit, test strip, or instrument platform.

[0018] Specifically, the test sample for the product is feces.

[0019] A second aspect of the invention provides a kit for predicting the efficacy of a drug for inflammatory bowel disease, comprising a product for detecting gut microbiota markers as described in the first aspect (e.g., a reagent for detecting the gut microbiota marker 16S rRNA), wherein the drug and inflammatory bowel disease are as described in the first aspect of the invention.

[0020] Specifically, the kit also includes reagents for extracting gut microbial genomic DNA from samples.

[0021] A third aspect of the present invention provides a method for screening gut microbiota biomarkers, comprising the following steps:

[0022] S1. Obtain clinical information data and stool samples of the disease at baseline and during follow-up, and analyze them;

[0023] S2. DNA extraction and metagenomic sequencing screening were performed on the fecal samples.

[0024] S3. Preprocess and perform microbiome analysis on DNA sequence data of the disease at baseline and during follow-up to identify gut microbiota biomarkers.

[0025] Preferably, the disease is an inflammatory bowel disease, including one or both of ulcerative colitis and Crohn's disease.

[0026] Preferably, the DNA extraction method includes extracting intestinal microbial genomic DNA using the Omega MAG-BIND Soil DNA Kit (M5635-02), quantifying the DNA using a Qubit 4 fluorometer, and assessing the quality by agarose gel electrophoresis.

[0027] Preferably, the DNA sequence data and microbiome processing software includes one or more of Trimmomatic (v0.39), KNEADDATA (v0.12.0), Metaphlan (v4.0.6), HUMAnN (v3.7), Chocophlan, Uniref90 database, R (v4.2.0), and Huttenhower Lab Galaxy.

[0028] In a fourth aspect, the present invention provides a predictive model for the efficacy of a drug in treating inflammatory bowel disease, wherein the predictive model uses the gut microbiota markers described in the first aspect as variables.

[0029] In one embodiment of the present invention, the variables of the prediction model are a combination of F. prausnitzii, B. massiliensis and P. faecium.

[0030] Preferably, the prediction model is a random forest model.

[0031] A fifth aspect of the present invention provides a method for constructing a prediction model as described in the fourth aspect, comprising the following steps: inputting biomarker data from samples into a random forest (RF) binary classifier machine learning model, and training the model using a relative abundance table.

[0032] A sixth aspect of the present invention provides a system for predicting the efficacy of a drug for inflammatory bowel disease, comprising:

[0033] The data processing module is used to receive or input gut microbial abundance data in fecal samples from patients with inflammatory bowel disease, wherein the gut microbial flora includes the gut microbial biomarkers described in the first aspect of the present invention.

[0034] The judgment and output module is used to obtain and output a prediction result of whether drug treatment is effective for the patient with inflammatory bowel disease after the receiving or input is completed, through the prediction model described in the fourth aspect of the present invention.

[0035] In some embodiments of the present invention, the system further includes a data processing module for collecting gut microbial abundance data in samples.

[0036] A seventh aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, can perform the functions of the system described in the sixth aspect of the present invention.

[0037] In an eighth aspect of the invention, an electronic device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the functions of the system described in the sixth aspect of the invention.

[0038] In a ninth aspect of the present invention, a method for predicting the efficacy of a drug for inflammatory bowel disease is provided, comprising the following steps:

[0039] (1) Detect the abundance of gut microbiota in a fecal sample of a patient to be predicted, wherein the gut microbiota includes the gut microbiota markers described in the first aspect of the present invention;

[0040] (2) Input the data obtained in step (1) into the prediction model described in the fourth aspect of the present invention, and output the result of whether the drug treatment is effective for the patient.

[0041] Specifically, the method further includes the following steps: DNA extraction and metagenomic sequencing of fecal samples from the patient to be predicted.

[0042] Specifically, the patient is a mammal, such as a human.

[0043] The beneficial effects of this invention are:

[0044] 1. This invention provides the application of gut microbiota as biomarkers in the preparation and prediction of drug efficacy for inflammatory bowel disease (IBD), demonstrating the feasibility of using gut microbiota profiling to predict the efficacy of drug treatment for IBD patients. This achievement can accurately identify patients requiring escalated treatment regimens in the early stages of the disease, thereby significantly improving patient prognosis.

[0045] 2. This invention provides a method for constructing a gut microbiota prediction model, which, through the combination of F. prausnitzii, B. massiliensis and P. faecium, achieves accurate prediction of drug treatment ineffectiveness, highlighting its potential as a new tool for early identification of patients who may require escalation of treatment. Attached Figure Description

[0046] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings, wherein:

[0047] Figure 1 Key differences in the gut microbiome between the effective and ineffective groups at baseline. A: Stacked bar plots depict phylum-level differences in gut microbiome composition between the two groups; B: Proteobacteria in both groups; C: Alpha diversity of the microbiome; D: Beta diversity analysis at the species level using Aitchison distance (PCoA plot); E: Baseline differences in species between the two groups; F: Beta diversity analysis of functional profiles using Aitchison distance (PCoA plot); G: Baseline differences in pathways between the two groups; H: Spearman correlation between species and function.

[0048] Figure 2Longitudinal changes in species and pathways between the effective and ineffective groups. A: Significant difference in species richness between the two groups at follow-up; B: Relative abundance of the three bacteria continued to decrease in ineffective patients at baseline and during follow-up; C: Log2-fold change (FC) in *Proteus vulgaris* after treatment compared to baseline samples.

[0049] Figure 3 ROC curves of the test queue of the RF classifier built based on baseline microbial variables.

[0050] Figure 4 ROC curve of the IBDMDB queue based on the RF classifier.

[0051] Figure 5 Relative abundance of *Bacterium praechoicum*, *Blautia massiliensis*, and *Bacterium fecalis* in IBDMDB.

[0052] Figure 6 ROC curves of an RF classifier constructed to predict the efficacy of anti-tumor necrosis factor therapy in patients with ulcerative colitis based on gut microbiota markers.

[0053] Figure 7 ROC curves of an RF classifier constructed to predict the efficacy of ustekinumab treatment in patients with inflammatory bowel disease based on gut microbiota markers. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] The term "patient" encompasses mammals. Examples of mammals include, but are not limited to, any member of the class Mammalia: humans; non-human primates (such as chimpanzees and other ape and monkey species); livestock such as cattle, horses, sheep, goats, and pigs; domestic animals such as rabbits, dogs, and cats; and laboratory animals, including rodents such as rats, mice, and guinea pigs. In this respect, the mammal is human.

[0056] In this invention, the term "microorganism" refers to a large group of organisms including bacteria, viruses, fungi, as well as some small protozoa and microalgae.

[0057] Statistical analysis: The normality of the data was tested using the one-sample Kolmogorov-Smirnov test. Continuous variables were expressed as mean ± standard deviation, and categorical variables as frequency or percentage. Independent samples t-tests were used for between-group comparisons of normally distributed data, while the Mann-Whitney-Wilcoxon test was used for non-normally distributed data. Fisher's exact test or chi-square test was used for categorical variables. 2 Correlation tests were conducted to explore the correlations between continuous variables. All statistical tests were two-tailed tests with a significance level of 0.05.

[0058] Ethical considerations: The study was approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University (2017-P2-094-01) and informed consent was obtained from all participants.

[0059] Example 1: Patient Sample Collection and Preliminary Analysis

[0060] 1. Sample collection

[0061] In a prospective IBD registry at Beijing Friendship Hospital, all patients were diagnosed with ulcerative colitis (UC) according to the Lennard-Jones criteria (Lennard-Jones, JE, Classification of inflammatory bowel disease. Scand J Gastroenterol Suppl 1989, 170, 2-6; discussion 16-9). Follow-up assessments were conducted every 3 months. Detailed clinical information, including demographic details, history of inflammatory bowel disease, clinical presentation, medication history, laboratory results, and endoscopic findings, was collected at enrollment and at each visit.

[0062] Patients were included according to the following criteria: (1) aged 18 years or older; (2) diagnosed with active UC at the time of onset (Mayo Endoscopic Score [MES] ≥ 2); (3) received standardized 5-ASA treatment; (4) had good adherence to 5-ASA treatment; and (5) underwent colonoscopy 6-12 months after 5-ASA treatment to evaluate UC.

[0063] The exclusion criteria are as follows: (1) those with a history of tumors or gastrointestinal surgery; (2) those who have taken antibiotics within 3 months prior to stool sample collection; (3) those who have been diagnosed with infectious colitis by microbiological examination within 30 days prior to stool sample collection; and (4) those who have been exposed to corticosteroids, immunosuppressants, or received biological therapy.

[0064] A standardized 5-ASA treatment regimen is defined as a sufficient dose of 5-ASA for induction of remission (oral 5-ASA ≥3 g / day and / or topical treatment ≥1 g / day) or maintenance (oral 5-ASA ≥2 g / day and / or topical treatment 2-3 g / week).

[0065] Adherence to 5-ASA therapy: Good medication adherence was determined by a Morisky Medication Adherence Scale (MMAS) score of 6-8. Effective 5-ASA therapy was considered to have achieved mucosal healing (MH) (MES = 0 or 1) after 6-12 months of treatment. Patients who did not achieve MH were assigned to the ineffective group. (For UC patients, MH was defined as the absence of tissue fragility, bleeding, erosion, and ulceration in all visible areas of the intestinal mucosa.)

[0066] All stool samples collected at home by patients were collected using stool collection tubes provided by the researchers and transported to the laboratory within 24 hours. The stool samples were stored at -80°C prior to DNA extraction and metagenomic sequencing.

[0067] 2. Preliminary Analysis

[0068] A total of 51 eligible UC patients participated in the study. Among these patients, 26 reached MH during follow-up. A total of 75 stool samples were collected, including 51 at baseline and 24 after treatment (14 in the effective group and 10 in the ineffective group). The specific results of the baseline characteristics of the included patients are shown in Table 1. The results showed that there were no significant differences between the effective and ineffective 5-ASA groups in terms of age, sex, disease severity, inflammation severity, or inflammatory markers. This means that clinical characteristics cannot be used as a predictor of 5-ASA.

[0069] Table 1. Baseline characteristics of UC patients included in this study

[0070]

[0071]

[0072] Example 2 Screening markers

[0073] 1. Fecal DNA extraction and Shortgun metagenomic sequencing

[0074] In Example 1, fecal samples were used to extract gut microbial genomic DNA using the OMEGA Mag-Bind Soil DNA Kit (M5635-02) (Omega Bio-Tek, Norcross, GA, USA). Extracted samples were stored at -20°C for further evaluation. DNA was quantified using a Qubit 4 fluorometer (Invitrogen, USA), and quality was assessed by agarose gel electrophoresis. DNA libraries were prepared using the Illumina TruSeq Nano DNALT Library Preparation Kit (400bp insert size) (Illumina, USA) and sequenced on the Illumina Novaseq platform, generating 6GB of data per sample with 2×150bp paired end reads.

[0075] 2. Sequence data preprocessing and microbiome analysis

[0076] Raw sequencing data were processed using Trimmomatic (v0.39) to remove adapter sequences, perform quality control, and filter low-quality reads. Subsequently, KneadData (v0.12.0) was used to remove reads aligned to the human reference genome to ensure that subsequent analyses included only sequences of microbial origin. Fecal microbial taxonomy analysis was performed using MetaPhlAn (v4.0.6) for species annotation, calculating microbial abundance at each taxonomic level (phylum, class, order, family, genus, species).

[0077] HUMAnN (v3.7) uses the UniRef90 database, which combines DIAMOND with ChocoPhlAn and EC-screening, to quantitatively analyze the relative abundance of functional pathways in each fecal sample. Only species and functional characteristics prevalent in at least 10% of patient samples are retained. Subsequently, the raw reading counts are converted to relative abundance by normalizing the total readings for each sample to ensure data comparability.

[0078] Data analysis of the fecal microbiome was performed using R (v4.2.0). Alpha diversity was assessed using the Shannon index and species richness index. Beta diversity, reflecting changes in community structure and function among different samples, was visualized using principal coordinate analysis (PCoA) based on Aitchison distance. The impact of confounding factors on Aitchison distance was assessed using permuted multivariate ANOVA (n=999).

[0079] Multivariate analysis was performed using the linear model (Maaslin2) (v1.18.0) statistical framework implemented in the Huttenhower Lab Galaxy instance (http: / / huttenhower.sph.harvard.edu / galaxy / ) to identify differentially expressed microbiome and functional characteristics, with significance set at p<0.05.

[0080] 3. Results

[0081] a) Proteobacteria status in stool samples at baseline in patients in the effective and ineffective groups after 5-ASA treatment.

[0082] The analysis of Proteobacteria in stool samples from patients in the effective and ineffective groups after 5-ASA treatment at baseline was performed. Specific results are as follows: Figure 1 As shown. Patients in the 5-ASA-ineffective group showed increased abundance of Proteobacteria in samples taken at baseline (p = 0.035). Figure 1 A, B). No significant differences in α-diversity were observed between the two groups in the overall composition of the gut microbiome. Figure 1 C). A trend toward separation was observed in the β-diversity analysis (Aitchison distance), and the difference was statistically significant (p = 0.054). Figure 1 D).

[0083] Multivariate analysis using a linear model (Maaslin2) identified a map of species differences between the two groups. Figure 1 E). In baseline stool samples from patients who did not respond to 5-ASA treatment, *Faecalibacterium prausnitzii*, *Blautia massiliensis*, *Phascolarctobacterium faecium*, *Blautia SGB4815*, *Coprococcus comes*, and *Peptostreptococcus stomatis* were significantly reduced, while *Klebsiella pneumoniae*, *Eggerthella sinensis*, and GGB80090 SGB1690 were significantly increased.

[0084] There was no significant difference in β-diversity regarding the functional potential of the gut microbiota. Figure 1 F). Pathways crucial for short-chain fatty acid (SCFA) synthesis, such as the super-pathway of pyruvate fermentation to butyrate and Clostridium pyruvate-butanol fermentation, were significantly enriched in baseline fecal samples from patients in the effective group. Figure 1G). Notably, baseline gondolic acid biosynthesis was negatively correlated with several species enriched in the effective group, including F. prausnitzii, C. comes, Blautia SGB4815, and P. stomatis. Figure 1 H).

[0085] b) The presence of Proteobacteria in stool samples from patients in the effective and ineffective groups after 5-ASA treatment.

[0086] The presence of Proteobacteria in stool samples from patients in the effective and ineffective groups after 5-ASA treatment was analyzed. Specific results are as follows: Figure 2 As shown. After 5-ASA treatment, 12 bacteria, including F. prausnitzii, B. massiliensi, P. faecium, Ruminococcus sp AF 13_28, Blautia sp Marseille P3087, Anaerostipes hadrus, Lacrimispora celerecrescens, Bifidobacterium longum, Actinomyces massiliensis, Enterococcus SGB6173, Actinomyces SGB17163, and Lachnospiraceae bacterium, were reduced in the ineffective group; while 6 bacteria, including Escherichia coli, Enterococcus avium, Enterococcus faecalis, and GGB3746, were reduced. SGB5089, *Limosilactobacillus mucosa*, and *Lactaseibacillus paracasei* were enriched in the invalid group. Figure 2 A).

[0087] In patients who did not respond to 5-ASA treatment, the levels of F. prausnitzii, B. massiliensis, and P. faecium in fecal samples remained consistently reduced at baseline and during follow-up. Figure 2 B). In patients who did not respond to 5-ASA treatment, the abundance of F. prausnitzii after treatment was even lower than the baseline level ( Figure 2 C).

[0088] Example 3: Construction of a 5-ASA efficacy prediction model

[0089] 1. Construction of a Random Forest-Based Diagnostic Model

[0090] A random forest (RF) binary classifier machine learning model was used in Scikit-Learn (v.1.5.2) software. This algorithm demonstrated superior performance compared to other machine learning models used for microbiome data in previous studies, and the model was trained using a relative abundance table. Cross-validation was performed by iteratively (10 times) training the RF model with a training / test set split ratio of 70%–30%. The `predict_proba` function was used to estimate the probability of each sample in different classes. The optimal threshold was determined using the Youden indexing method on the training set. Model performance was evaluated using receiver operating characteristic (ROC) curves and area under the curve (AUC). Model predictive performance was measured using multiple metrics, including sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV).

[0091] Considering the continued decrease in the abundance of *F. prausnitzii*, *B. massiliensis*, and *P. faecium* in the ineffective group both at baseline and after treatment, the low baseline abundance of these three species was used to construct a predictive model for 5-ASA treatment failure. A robust random forest classification model was constructed, which predicted a mean AUC of 0.80 for 5-ASA treatment failure. Figure 3 ).

[0092] Example 4: Validation of the 5-ASA efficacy prediction model

[0093] To further test the general applicability of this model, it was analyzed in the Inflammatory Bowel Disease Multi-Omics Database (IBDMDB) ( https: / / ibdmdb.org / results The predictive model's results were externally validated. Using eligible patients from the IBDMDB cohort (n=15) as an independent dataset (Table 2), the predictive model was externally validated. The model predicted an AUC of 0.82 for 5-ASA treatment failure, with specificity, NPV, and PPV of 0.88, 0.70, and 0.80, respectively. Figure 4 ). Figure 5 F. prausnitzii, B. massiensis, and P. faecium were identified as specific bacterial markers, which were persistently reduced in ineffective patients at baseline and follow-up. Therefore, the combination of these three gut microbiota markers—F. prausnitzii, B. massiensis, and P. faecium—provided the best predictive power.

[0094] Table 2. Baseline characteristics of the validation queue

[0095]

[0096]

[0097] Example 5: Validation of the predictive model for the therapeutic effects of other biological agents

[0098] The public dataset PRJNA685168 was used to validate a previously constructed model for predicting the efficacy of biologic therapy in patients with inflammatory bowel disease based on gut microbiota markers. This dataset can be obtained from the PRISM metagenomics data available in the Sequence Reading Archive (SRA) (https: / / www.ncbi.nlm.nih.gov / Traces / study / ?acc=PRJNA685168&o=acc_s%3Aa). The dataset included patients with inflammatory bowel disease (10 patients) and ulcerative colitis (11 patients). These patients received anti-tumor necrosis factor and anti-interleukin 12 / 23 (ustekinumab) therapy from baseline. Clinical remission was assessed at 14 weeks, defined as an HBI or SCCAI score ≤2.

[0099] Baseline fecal metagenomic data of different subgroups of patients were analyzed, and ROC curves were plotted. Specific results are as follows: Figure 6-7 As shown in Table 3, the baseline characteristics of patients with ulcerative colitis are presented. Figure 6 The model was used to predict the response to antitumor necrosis factor therapy in 11 patients with ulcerative colitis. The results showed that the AUC was 0.93, the sensitivity was 0.67, the specificity was 1.00, the accuracy was 0.82, the PPV was 1.00, and the NPV was 0.71, indicating that the model has good predictive value for the response to antitumor necrosis factor therapy in patients with ulcerative colitis.

[0100] Table 3. Baseline characteristics of patients with ulcerative colitis

[0101]

[0102] Table 4 shows the baseline characteristics of patients with inflammatory bowel disease. Figure 7 The model predicted the response to ustekinumab treatment in 10 patients with inflammatory bowel disease (including ulcerative colitis or Crohn's disease). The results showed an AUC of 0.95, a sensitivity of 0.86, a specificity of 1.00, an accuracy of 0.90, a PPV of 1.00, and an NPV of 0.75, indicating that the model performed well in predicting the efficacy of ustekinumab treatment.

[0103] Table 4. Baseline characteristics of patients with inflammatory bowel disease

[0104]

[0105]

[0106] Note: One case of UC was included in the validation cohort, but the extent of the colonic lesion was not included in the statistics. The HBI index, or Harvey-Bradshaw Index, is a practical tool used to assess the severity (activity) of Crohn's disease in patients.

[0107] Based on the above validation results, this model provides strong support for optimizing individualized treatment regimens for patients with inflammatory bowel disease using anti-tumor necrosis factor and ustekinumab, and offers new insights for exploring related microbial biomarkers.

[0108] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various simple modifications can be made to the technical solution of the present invention, and these simple modifications all fall within the protection scope of the present invention.

[0109] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. The application of a gut microbiome marker in the preparation of products for predicting the efficacy of drugs against inflammatory bowel disease, characterized in that, The gut microbial markers mentioned are a combination of three species: Faecalibacterium prausnitzii, Blautia massiliensis, and Phascolarctobacterium faecium.

2. The application according to claim 1, characterized in that, The inflammatory bowel disease mentioned is selected from one or both of ulcerative colitis and Crohn's disease.

3. The application according to claim 1, characterized in that, The drug for treating inflammatory bowel disease is selected from one or more of the following: anti-tumor necrosis factor, interleukin (IL)-12 / 23 inhibitors, and 5-aminosalicylic acid.

4. The application according to claim 1, characterized in that, The inflammatory bowel disease mentioned above is ulcerative colitis, and the drug mentioned above is 5-aminosalicylic acid; Or the inflammatory bowel disease may be ulcerative colitis, and the drug may be an anti-tumor necrosis factor; Or the inflammatory bowel disease may be ulcerative colitis or Crohn's disease, and the drug may be an interleukin-12 / 23 inhibitor.

5. A predictive model for the efficacy of a drug in treating inflammatory bowel disease, characterized in that, The prediction model uses the gut microbial biomarkers described in claim 1 as variables.

6. The prediction model according to claim 5, characterized in that, The prediction model is a random forest model.

7. The prediction model according to claim 5, characterized in that, The variables in the prediction model are a combination of Faecalibacterium prausnitzii, Blautia massiliensis, and Phascolarctobacterium faecium.

8. A system for predicting the efficacy of drugs for inflammatory bowel disease, characterized in that, The system includes: The data processing module is used to receive or input gut microbiota abundance data in fecal samples from patients with inflammatory bowel disease, wherein the gut microbiota includes the gut microbiota markers as described in claim 1; The judgment and output module is used to obtain and output a prediction result of whether drug treatment is effective for the patient with inflammatory bowel disease after the receiving or input is completed, through the prediction model described in claim 5.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to perform the functions of the system of claim 8.

Images