Prediction model of poorly differentiated colorectal cancer (CRC) based on gut bacteria

Identification of predictors for lymph node metastasis in T2 colorectal cancer: retrospective cohort study from a high-volume hospital

BACKGROUND

Colorectal cancer (CRC) is the most prevalent malignant tumor of the digestive system globally, ranking third in incidence and second in mortality. In previous studies, the rate of lymph node metastasis (LNM) in T2 CRC ranged from 18.0 to 28.0%. We aim to identify T2 CRC patients without LNM and thereby mitigate the complications and potential impact on the quality of life associated with surgery.

METHODS

In this retrospective study, 787 cases with T2 CRC were selected. The preoperative and postoperative clinicopathological features were retrospectively studied. Univariate analysis and multivariate analysis were performed using binary logistic regression to determine the predictive factor for LNM. Odds ratio (OR) and 95% confidence interval (CI) were conducted.

RESULTS

184 (23.4%) patients were diagnosed with LNM, including 144 (78.3%) patients with N1stage and 40 (21.7%) patients with N2 stage. According to univariate analysis and multivariate analysis, poorly differentiated tumors (p = 0.003, OR = 4.405, 95%CI: 1.632-11.893), perineural invasion (p = 0.001, OR = 4.789, 95%CI: 1.958-11.716), and lymphovascular invasion (p = 0.001, OR = 2.779, 95%CI: 1.497-5.159) were independent risk factors of LNM, while male (p = 0.017, OR = 0.652, 95%CI: 0.459-0.926) and elevated preoperative PLR (p = 0.048, OR = 0.996, 95%CI: 0.993-1.000) seemed to be independent protective factors. Larger tumor size did not show significant association with LNM.

CONCLUSIONS

Approximately three-quarters of T2 CRC patients are likely to avoid unnecessary surgery. Female, poorly differentiated tumors, perineural invasion, and lymphovascular invasion are expected to be used as predictors of LNM in T2 CRC.

Candidate Biomarkers Associated With Circulating Tumor Cell Status in Metastatic Colorectal Cancer

BACKGROUND

Colorectal cancer (CRC) ranks as the third most prevalent cancer worldwide. Recent studies suggest the promising potential of microRNAs (miRNA) in predicting the status of circulating tumor cells (CTC), and their combined analyses could pave the way for significant advancements in assessing the risk of metastatic cancer. Here, we investigate the circulating miRNA signatures associated with CTC status in metastatic CRC (mCRC).

METHODS

The CTC status of mCRC patients was assessed using AdnaTest ColonCancer technology, which detects tumor cells using an immunomagnetic approach and characterizes them based on colon-specific surface markers. The miRNA profiles were analyzed using the Agilent miRNA microarray in 8 CTC-positive, 8 CTC-negative, and eight healthy individuals. The functional implications of dysregulated miRNAs and their interactions with target mRNAs, TFs, and lncRNAs were determined through a comprehensive in silico analysis. Candidate miRNAs that were differentially expressed in CTC-positive and CTC-negative groups, which have prior evidence for their role in CRC biology, were validated using qPCR.

RESULTS

We identified two groups of dysregulated miRNAs associated with CTC status and multiple candidate biomarkers in suggested miRNA regulatory networks. Three miRNAs (hsa-miR-199a-5p, hsa-miR-326, hsa-miR-500b-5p), which were downregulated in the CTC-positive group compared to the CTC-negative group, were confirmed by qPCR and prioritized as candidate predictors of CTC status in mCRC.

CONCLUSION

Our findings suggest biomarker candidates that can be used to predict CTC status in individuals with mCRC. This might also provide new insights into new translational medicine applications in the management of mCRC through miRNA-based CRC-associated CTC detection.

Intestinal microbiota as biomarkers for different colorectal lesions based on colorectal cancer screening participants in community

Introduction

The dysregulation of intestinal microbiota has been implicated in the pathogenesis of colorectal cancer (CRC). However, the utilization of intestinal microbiota for identify the lesions in different procedures in CRC screening populations remains limited.

Methods

A total of 529 high-risk individuals who underwent CRC screening were included, comprising 13 advanced adenomas (Aade), 5 CRC, 59 non-advanced adenomas (Nade), 129 colon polyps (Pol), 99 cases of colorectal inflammatory disease (Inf), and 224 normal controls (Nor). 16S rRNA gene sequencing was used to profile the intestinal microbiota communities. The Gut Microbiota Health Index (GMHI) and average variation degree (AVD) were employed to assess the health status of the different groups.

Results

Our findings revealed that the Nor group exhibited significantly higher GMHIs and the lowest AVD compared to the four Lesion groups. The model incorporating 13 bacterial genera demonstrated optimal efficacy in distinguishing CRC and Aade from Nor, with an area under the curve (AUC) of 0.81 and a 95% confidence interval (CI) of 0.72 to 0.89. Specifically, the 55 bacterial genera combination model exhibited superior performance in differentiating CRC from Nor (AUC 0.98; 95% CI, 0.96-1), the 25 bacterial genera combination showed superior performance in distinguishing Aade from Nor (AUC 0.95). Additionally, the 27 bacterial genera combination demonstrated superior efficacy in differentiating Nade from Nor (AUC 0.82). The 13 bacterial genera combination exhibited optimal performance in distinguishing Inf from Nor (AUC 0.71).

Discussion

Our study has identified specific microbial biomarkers that can differentiate between colorectal lesions and healthy individuals. The intestinal microbiota markers identified may serve as valuable tools in community-based CRC screening programs.

Distinct prebiotic effects of polysaccharide fractions from Polygonatum kingianum on gut microbiota

This study investigated the physicochemical properties, digestive stability, and in vitro fermentation behavior of Polygonatum kingianum polysaccharide (PKP) fractions (PKP60, PKP70, PKP80) obtained through graded ethanol precipitation. High-performance gel permeation chromatography revealed significant molecular weight differences among the fractions, while reverse-phase high-performance liquid chromatography indicated consistent monosaccharide types with variations in their proportions. Uronic acid analysis confirmed that all polysaccharide fractions met the criteria for neutral polysaccharides. Congo red staining confirmed the presence of a triple-helix structure in all PKP fractions. Comprehensive analysis demonstrated that these fractions remained stable during in vitro digestion, as evidenced by consistent molecular weights and total carbohydrate content, with no significant production of free monosaccharides or reducing sugars. All PKP fractions were fermented by gut microbiota, resulting in the production of short-chain fatty acids. Beta diversity and structural analyses of gut microbiota revealed distinct modulatory effects associated with each PKP fraction. The PKP fractions promoted probiotic growth, especially PKP70, which significantly enhanced Bifidobacterium proliferation, indicating strong prebiotic potential. These findings underscore the importance of isolation and purification methods in determining the functionality and gut microbiota-modulating effects of plant-derived polysaccharides, emphasizing the need for in-depth research that extends beyond merely evaluating their source.

Quantitative parameters of dual-layer spectral detector computed tomography for evaluating differentiation grade and lymphovascular and perineural invasion in colorectal adenocarcinoma

PURPOSE

To explore the predictive value of dual-layer spectral detector CT (SDCT) quantitative parameters for determining differentiation grade, lymphovascular invasion (LVI) and perineural invasion (PNI) in colorectal adenocarcinoma (CRAC) patients.

METHODS

A total of 106 eligible patients with CRAC were included in this study. Spectral parameters, including CT values at 40 and 100 keV, the effective atomic number (Zeff), the iodine concentration (IC), the slope of the spectral Hounsfield unit (HU) curve (λHU), and the normalized iodine concentration (NIC) in the arterial phase (AP) and venous phase (VP), were compared according to the differentiation grade and the status of LVI and PNI. The diagnostic accuracies of the quantitative parameters with statistical significance were determined via receiver operating characteristic (ROC) curves, and the area under the curve (AUC) was calculated.

RESULTS

There were 57 males and 49 females aged 43-86 (69 ± 10) years. The measured values of the spectral quantitative parameters of the CRAC were consistent within the observer (ICC range: 0.800-0.926). The 40 keV-AP, IC-AP, NIC-AP, 40 keV-VP, and IC-VP were significantly different among the different differentiation grades in the CRAC (P = 0.040, AUC = 0.673; P = 0.035, AUC = 0.684; P = 0.031, AUC = 0.639; P = 0.044, AUC = 0.663 and P = 0.035, AUC = 0.666, respectively). A statistically significant difference was observed in 40 keV-VP, 100 keV-VP, Zeff-VP, IC-VP, and λHU-VP between LVI-positive and LVI-negative patients (P = 0.003, AUC = 0.688; P = 0.015, AUC = 0.644; P = 0.001, AUC = 0.688; P = 0.001, AUC = 0.703 and P = 0.003, AUC = 0.677, respectively). There were no statistically significant differences in the values of the spectral parameters of the PNI state of patients with CRAC (P > 0.05).

CONCLUSION

The quantitative parameters of SDCT had good diagnostic efficacy in differentiating between different grades and statuses of LVI in patients with CRAC; however, SDCT did not have value for identifying the state of PNI.

Gut microbial subtypes and clinicopathological value for colorectal cancer

BACKGROUND

Gut bacteria are related to colorectal cancer (CRC) and its clinicopathologic characteristics.

OBJECTIVE

To develop gut bacterial subtypes and explore potential microbial targets for CRC.

METHODS

Stool samples from 914 volunteers (376 CRCs, 363 advanced adenomas, and 175 normal controls) were included for 16S rRNA sequencing. Unsupervised learning was used to generate gut microbial subtypes. Gut bacterial community composition and clustering effects were plotted. Differences of gut bacterial abundance were analyzed. Then, the association of CRC-associated bacteria with subtypes and the association of gut bacteria with clinical information were assessed. The CatBoost models based on gut differential bacteria were constructed to identify the diseases including CRC and advanced adenoma (AA).

RESULTS

Four gut microbial subtypes (A, B, C, D) were finally obtained via unsupervised learning. The characteristic bacteria of each subtype were Escherichia-Shigella in subtype A, Streptococcus in subtype B, Blautia in subtype C, and Bacteroides in subtype D. Clinical information (e.g., free fatty acids and total cholesterol) and CRC pathological information (e.g., tumor depth) varied among gut microbial subtypes. Bacilli, Lactobacillales, etc., were positively correlated with subtype B. Positive correlation of Blautia, Lachnospiraceae, etc., with subtype C and negative correlation of Coriobacteriia, Coriobacteriales, etc., with subtype D were found. Finally, the predictive ability of CatBoost models for CRC identification was improved based on gut microbial subtypes.

CONCLUSION

Gut microbial subtypes provide characteristic gut bacteria and are expected to contribute to the diagnosis of CRC.

Predictive modeling of colorectal cancer using exhaustive analysis of microbiome information layers available from public metagenomic data

This study aimed to compare the microbiome profiles of patients with colorectal cancer (CRC, n = 380) and colorectal adenomas (CRA, n = 110) against generally healthy participants (n = 2,461) from various studies. The overarching objective was to conduct a real-life experiment and develop a robust machine learning model applicable to the general population. A total of 2,951 stool samples underwent a comprehensive analysis using the in-house MetaBakery pipeline. This included various data matrices such as microbial taxonomy, functional genes, enzymatic reactions, metabolic pathways, and predicted metabolites. The study found no statistically significant difference in microbial diversity among individuals. However, distinct clusters were identified for healthy, CRC, and CRA groups through linear discriminant analysis (LDA). Machine learning analysis demonstrated consistent model performance, indicating the potential of microbiome layers (microbial taxa, functional genes, enzymatic reactions, and metabolic pathways) as prediagnostic indicators for CRC and CRA. Notable biomarkers on the taxonomy level and microbial functionality (gene families, enzymatic reactions, and metabolic pathways) associated with CRC were identified. The research presents promising avenues for practical clinical applications, with potential validation on external clinical datasets in future studies.

Circulating Bacterial DNA in Colorectal Cancer Patients: The Potential Role of Fusobacterium nucleatum

Intestinal dysbiosis is a major contributor to colorectal cancer (CRC) development, leading to bacterial translocation into the bloodstream. This study aimed to evaluate the presence of circulated bacterial DNA (cbDNA) in CRC patients (n = 75) and healthy individuals (n = 25). DNA extracted from peripheral blood was analyzed using PCR, with specific primers targeting 16S rRNA, Escherichia coli (E. coli), and Fusobacterium nucleatum (F. nucleatum). High 16S rRNA and E. coli detections were observed in all patients and controls. Only the detection of F. nucleatum was significantly higher in metastatic non-excised CRC, compared to controls (p < 0.001), non-metastatic excised CRC (p = 0.023), and metastatic excised CRC (p = 0.023). This effect was mainly attributed to the presence of the primary tumor (p = 0.006) but not the presence of distant metastases (p = 0.217). The association of cbDNA with other clinical parameters or co-morbidities was also evaluated, revealing a higher detection of E. coli in CRC patients with diabetes (p = 0.004). These results highlighted the importance of bacterial translocation in CRC patients and the potential role of F. nucleatum as an intratumoral oncomicrobe in CRC.

Targeting gut-microbiota for gastric cancer treatment: a systematic review

Background

Preclinical research has identified the mechanisms via which bacteria influence cancer treatment outcomes. Clinical studies have demonstrated the potential to modify the microbiome in cancer treatment. Herein, we systematically analyze how gut microorganisms interact with chemotherapy and immune checkpoint inhibitors, specifically focusing on how gut bacteria affect the pharmacokinetics and pharmacodynamics of cancer treatment.

Method

This study searched Web of Science, Scopus, and PubMed until August 2023. Studies were screened by their title and abstract using the Rayyan intelligent tool for systematic reviews. Quality assessment of studies was done using the JBI critical appraisal tool.

Result

Alterations in the gut microbiome are associated with gastric cancer and precancerous lesions. These alterations include reduced microbial alpha diversity, increased bacterial overgrowth, and decreased richness and evenness of gastric bacteria. Helicobacter pylori infection is associated with reduced richness and evenness of gastric bacteria, while eradication only partially restores microbial diversity. The gut microbiome also affects the response to cancer treatments, with higher abundances of Lactobacillus associated with better response to anti-PD-1/PD-L1 immunotherapy and more prolonged progression-free survival. Antibiotic-induced gut microbiota dysbiosis can reduce the anti-tumor efficacy of 5-Fluorouracil treatment, while probiotics did not significantly enhance it. A probiotic combination containing Bifidobacterium infantis, Lactobacillus acidophilus, Enterococcus faecalis, and Bacillus cereus can reduce inflammation, enhance immunity, and restore a healthier gut microbial balance in gastric cancer patients after partial gastrectomy.

Conclusion

Probiotics and targeted interventions to modulate the gut microbiome have shown promising results in cancer prevention and treatment efficacy.Systematic review registration: https://osf.io/6vcjp.

Colorectal cancer prognosis based on dietary pattern using synthetic minority oversampling technique with K-nearest neighbors approach

Generally, a person's life span depends on their food consumption because it may cause deadly diseases like colorectal cancer (CRC). In 2020, colorectal cancer accounted for one million fatalities globally, representing 10% of all cancer casualties. 76,679 males and 78,213 females over the age of 59 from ten states in the United States participated in this analysis. During follow-up, 1378 men and 981 women were diagnosed with colon cancer. This prospective cohort study used 231 food items and their variants as input features to identify CRC patients. Before labelling any foods as colorectal cancer-causing foods, it is ethical to analyse facts like how many grams of food should be consumed daily and how many times a week. This research examines five classification algorithms on real-time datasets: K-Nearest Neighbour (KNN), Decision Tree (DT), Random Forest (RF), Logistic Regression with Classifier Chain (LRCC), and Logistic Regression with Label Powerset (LRLC). Then, the SMOTE algorithm is applied to deal with and identify imbalances in the data. Our study shows that eating more than 10 g/d of low-fat butter in bread (RR 1.99, CI 0.91-4.39) and more than twice a week (RR 1.49, CI 0.93-2.38) increases CRC risk. Concerning beef, eating in excess of 74 g of beef steak daily (RR 0.88, CI 0.50-1.55) and having it more than once a week (RR 0.88, CI 0.62-1.23) decreases the risk of CRC, respectively. While eating beef and dairy products in a daily diet should be cautious about quantity. Consuming those items in moderation on a regular basis will protect us against CRC risk. Meanwhile, a high intake of poultry (RR 0.2, CI 0.05-0.81), fish (RR 0.82, CI 0.31-2.16), and pork (RR 0.67, CI 0.17-2.65) consumption negatively correlates to CRC hazards.

Gut microbes on the risk of advanced adenomas

BACKGROUND

More than 90% of colorectal cancer (CRC) arises from advanced adenomas (AA) and gut microbes are closely associated with the initiation and progression of both AA and CRC.

OBJECTIVE

To analyze the characteristic microbes in AA.

METHODS

Fecal samples were collected from 92 AA and 184 negative control (NC). Illumina HiSeq X sequencing platform was used for high-throughput sequencing of microbial populations. The sequencing results were annotated and compared with NCBI RefSeq database to find the microbial characteristics of AA. R-vegan package was used to analyze α diversity and β diversity. α diversity included box diagram, and β diversity included Principal Component Analysis (PCA), principal co-ordinates analysis (PCoA), and non-metric multidimensional scaling (NMDS). The AA risk prediction models were constructed based on six kinds of machine learning algorithms. In addition, unsupervised clustering methods were used to classify bacteria and viruses. Finally, the characteristics of bacteria and viruses in different subtypes were analyzed.

RESULTS

The abundance of Prevotella sp900557255, Alistipes putredinis, and Megamonas funiformis were higher in AA, while the abundance of Lilyvirus, Felixounavirus, and Drulisvirus were also higher in AA. The Catboost based model for predicting the risk of AA has the highest accuracy (bacteria test set: 87.27%; virus test set: 83.33%). In addition, 4 subtypes (B1V1, B1V2, B2V1, and B2V2) were distinguished based on the abundance of gut bacteria and enteroviruses (EVs). Escherichia coli D, Prevotella sp900557255, CAG-180 sp000432435, Phocaeicola plebeiuA, Teseptimavirus, Svunavirus, Felixounavirus, and Jiaodavirus are the characteristic bacteria and viruses of 4 subtypes. The results of Catboost model indicated that the accuracy of prediction improved after incorporating subtypes. The accuracy of discovery sets was 100%, 96.34%, 100%, and 98.46% in 4 subtypes, respectively.

CONCLUSION

Prevotella sp900557255 and Felixounavirus have high value in early warning of AA. As promising non-invasive biomarkers, gut microbes can become potential diagnostic targets for AA, and the accuracy of predicting AA can be improved by typing.

Relationship of Metabolic Dysfunction-Associated Steatohepatitis-Related Hepatocellular Carcinoma with Oral and Intestinal Microbiota: A Cross-Sectional Pilot Study

Background and Objectives: The incidence of metabolic dysfunction-associated steatohepatitis (MASH)-related hepatocellular carcinoma (HCC) is increasing worldwide, alongside the epidemic of obesity and metabolic syndrome. Based on preliminary reports regarding the potential association of HCC and periodontitis, this study aimed to analyze the involvement of periodontal bacteria as well as the oral and intestinal bacterial flora in MASH-related HCC (MASH-HCC). Materials and Methods: Forty-one patients with MASH and nineteen with MASH-HCC participated in the study, completing survey questionnaires, undergoing periodontal examinations, and providing samples of saliva, mouth-rinsed water, feces, and peripheral blood. The oral and fecal microbiome profiles were analyzed by 16S ribosomal RNA sequencing. Bayesian network analysis was used to analyze the causation between various factors, including MASH-HCC, examinations, and bacteria. Results: The genus Fusobacterium had a significantly higher occupancy rate (p = 0.002) in the intestinal microflora of the MASH-HCC group compared to the MASH group. However, Butyricicoccus (p = 0.022) and Roseburia (p < 0.05) had significantly lower occupancy rates. The Bayesian network analysis revealed the absence of periodontal pathogenic bacteria and enteric bacteria affecting HCC. However, HCC directly affected the periodontal bacterial species Porphyromonas gingivalis, Tannerella forsythia, Fusobacterium nucleatum, and Prevotella intermedia in the saliva, as well as the genera Lactobacillus, Roseburia, Fusobacterium, Prevotella, Clostridium, Ruminococcus, Trabulsiella, and SMB53 in the intestine. Furthermore, P. gingivalis in the oral cavity directly affected the genera Lactobacillus and Streptococcus in the intestine. Conclusions: MASH-HCC directly affects periodontal pathogenic and intestinal bacteria, and P. gingivalis may affect the intestinal bacteria associated with gastrointestinal cancer.

Machine learning for predicting colon cancer recurrence

INTRODUCTION

Colorectal cancer (CRC) is a global public health concern, ranking among the most commonly diagnosed malignancies worldwide. Despite advancements in treatment modalities, the specter of CRC recurrence remains a significant challenge, demanding innovative solutions for early detection and intervention. The integration of machine learning into oncology offers a promising avenue to address this issue, providing data-driven insights and personalized care.

METHODS

This retrospective study analyzed data from 396 patients who underwent surgical procedures for colon cancer (CC) between 2010 and 2021. Machine learning algorithms were employed to predict CC recurrence, with a focus on demographic, clinicopathological, and laboratory characteristics. A range of evaluation metrics, including AUC (Area Under the Receiver Operating Characteristic), accuracy, recall, precision, and F1 scores, assessed the performance of machine learning algorithms.

RESULTS

Significant risk factors for CC recurrence were identified, including sex, carcinoembryonic antigen (CEA) levels, tumor location, depth, lymphatic and venous invasion, and lymph node involvement. The CatBoost Classifier demonstrated exceptional performance, achieving an AUC of 0.92 and an accuracy of 88 % on the test dataset. Feature importance analysis highlighted the significance of CEA levels, albumin levels, N stage, weight, platelet count, height, neutrophil count, lymphocyte count, and gender in determining recurrence risk.

DISCUSSION

The integration of machine learning into healthcare, exemplified by this study's findings, offers a pathway to personalized patient risk stratification and enhanced clinical decision-making. Early identification of individuals at risk of CC recurrence holds the potential for more effective therapeutic interventions and improved patient outcomes.

CONCLUSION

Machine learning has the potential to revolutionize our approach to CC recurrence prediction, emphasizing the synergy between medical expertise and cutting-edge technology in the fight against cancer. This study represents a vital step toward precision medicine in CC management, showcasing the transformative power of data-driven insights in oncology.

Investigating the influence of the gut microbiome on cholelithiasis: unveiling insights through sequencing and predictive modeling

BACKGROUND

Cholelithiasis is one of the most common disorders of hepatobiliary system. Gut bacteria may be involved in the process of gallstone formation and are, therefore considered as potential targets for cholelithiasis prediction.

OBJECTIVE

To reveal the correlation between cholelithiasis and gut bacteria.

METHODS

Stool samples were collected from 100 cholelithiasis and 250 healthy individuals from Huzhou Central Hospital; The 16S rRNA of gut bacteria in the stool samples was sequenced using the third-generation Pacbio sequencing platform; Mothur v.1.21.1 was used to analyze the diversity of gut bacteria; Wilcoxon rank-sum test and linear discriminant analysis of effect sizes (LEfSe) were used to analyze differences in gut bacteria between patients suffering from cholelithiasis and healthy individuals; Chord diagram and Plot-related heat maps were used to analyze the correlation between cholelithiasis and gut bacteria; six machine algorithms were used to construct models to predict cholelithiasis.

RESULTS

There were differences in the abundance of gut bacteria between cholelithiasis and healthy individuals, but there were no differences in their community diversity. Increased abundance of Costridia, Escherichia flexneri, and Klebsiella pneumonae were found in cholelithiasis, while Bacteroidia, Phocaeicola, and Phocaeicola vulgatus were more abundant in healthy individuals. The top four bacteria that were most closely associated with cholelithiasis were Escherichia flexneri, Escherichia dysenteriae, Streptococcus salivarius, and Phocaeicola vulgatus. The cholelithiasis model based on CatBoost algorithm had the best prediction effect (sensitivity: 90.48%, specificity: 88.32%, and AUC: 0.962).

CONCLUSION

The identification of characteristic gut bacteria may provide new predictive targets for gallstone screening. As being screened by the predictive model, people at high risk of cholelithiasis can determine the need for further testing, thus enabling early warning of cholelithiasis.

T2-weighted imaging-based radiomic-clinical machine learning model for predicting the differentiation of colorectal adenocarcinoma

BACKGROUND

The study on predicting the differentiation grade of colorectal cancer (CRC) based on magnetic resonance imaging (MRI) has not been reported yet. Developing a non-invasive model to predict the differentiation grade of CRC is of great value.

AIM

To develop and validate machine learning-based models for predicting the differentiation grade of CRC based on T2-weighted images (T2WI).

METHODS

We retrospectively collected the preoperative imaging and clinical data of 315 patients with CRC who underwent surgery from March 2018 to July 2023. Patients were randomly assigned to a training cohort (n = 220) or a validation cohort (n = 95) at a 7:3 ratio. Lesions were delineated layer by layer on high-resolution T2WI. Least absolute shrinkage and selection operator regression was applied to screen for radiomic features. Radiomics and clinical models were constructed using the multilayer perceptron (MLP) algorithm. These radiomic features and clinically relevant variables (selected based on a significance level of P < 0.05 in the training set) were used to construct radiomics-clinical models. The performance of the three models (clinical, radiomic, and radiomic-clinical model) were evaluated using the area under the curve (AUC), calibration curve and decision curve analysis (DCA).

RESULTS

After feature selection, eight radiomic features were retained from the initial 1781 features to construct the radiomic model. Eight different classifiers, including logistic regression, support vector machine, k-nearest neighbours, random forest, extreme trees, extreme gradient boosting, light gradient boosting machine, and MLP, were used to construct the model, with MLP demonstrating the best diagnostic performance. The AUC of the radiomic-clinical model was 0.862 (95%CI: 0.796-0.927) in the training cohort and 0.761 (95%CI: 0.635-0.887) in the validation cohort. The AUC for the radiomic model was 0.796 (95%CI: 0.723-0.869) in the training cohort and 0.735 (95%CI: 0.604-0.866) in the validation cohort. The clinical model achieved an AUC of 0.751 (95%CI: 0.661-0.842) in the training cohort and 0.676 (95%CI: 0.525-0.827) in the validation cohort. All three models demonstrated good accuracy. In the training cohort, the AUC of the radiomic-clinical model was significantly greater than that of the clinical model (P = 0.005) and the radiomic model (P = 0.016). DCA confirmed the clinical practicality of incorporating radiomic features into the diagnostic process.

CONCLUSION

In this study, we successfully developed and validated a T2WI-based machine learning model as an auxiliary tool for the preoperative differentiation between well/moderately and poorly differentiated CRC. This novel approach may assist clinicians in personalizing treatment strategies for patients and improving treatment efficacy.

Commensal Fecal Microbiota Profiles Associated with Initial Stages of Intestinal Mucosa Damage: A Pilot Study

Progressive intestinal mucosal damage occurs over years prior to colorectal cancer (CRC) development. The endoscopic screening of polyps and histopathological examination are used clinically to determine the risk and progression of mucosal lesions. We analyzed fecal microbiota compositions using 16S rRNA gene-based metataxonomic analyses and the levels of short-chain fatty acids (SCFAs) using gas chromatography in volunteers undergoing colonoscopy and histopathological analyses to determine the microbiota shifts occurring at the early stages of intestinal mucosa alterations. The results were compared between diagnosis groups (nonpathological controls and polyps), between samples from individuals with hyperplastic polyps or conventional adenomas, and between grades of dysplasia in conventional adenomas. Some microbial taxa from the Bacillota and Euryarchaeota phyla were the most affected when comparing the diagnosis and histopathological groups. Deeper microbiota alterations were found in the conventional adenomas than in the hyperplastic polyps. The Ruminococcus torques group was enriched in both the hyperplastic polyps and conventional adenomas, whereas the family Eggerthellaceae was enriched only in the hyperplastic polyps. The abundance of Prevotellaceae, Oscillospiraceae, Methanobacteriaceae, Streptococcaceae, Christensenellaceae, Erysipelotrichaceae, and Clostridiaceae shifted in conventional adenomas depending on the grade of dysplasia, without affecting the major SCFAs. Our results suggest a reorganization of microbial consortia involved in gut fermentative processes.

From-Toilet-to-Freezer: A Review on Requirements for an Automatic Protocol to Collect and Store Human Fecal Samples for Research Purposes

The composition, viability and metabolic functionality of intestinal microbiota play an important role in human health and disease. Studies on intestinal microbiota are often based on fecal samples, because these can be sampled in a non-invasive way, although procedures for sampling, processing and storage vary. This review presents factors to consider when developing an automated protocol for sampling, processing and storing fecal samples: donor inclusion criteria, urine-feces separation in smart toilets, homogenization, aliquoting, usage or type of buffer to dissolve and store fecal material, temperature and time for processing and storage and quality control. The lack of standardization and low-throughput of state-of-the-art fecal collection procedures promote a more automated protocol. Based on this review, an automated protocol is proposed. Fecal samples should be collected and immediately processed under anaerobic conditions at either room temperature (RT) for a maximum of 4 h or at 4 °C for no more than 24 h. Upon homogenization, preferably in the absence of added solvent to allow addition of a buffer of choice at a later stage, aliquots obtained should be stored at either -20 °C for up to a few months or -80 °C for a longer period-up to 2 years. Protocols for quality control should characterize microbial composition and viability as well as metabolic functionality.