Predicting lymph node metastasis in colorectal cancer: An analysis of influencing factors to develop a risk model

BACKGROUND

Colorectal cancer (CRC) is a significant global health issue, and lymph node metastasis (LNM) is a crucial prognostic factor. Accurate prediction of LNM is essential for developing individualized treatment strategies for patients with CRC. However, the prediction of LNM is challenging and depends on various factors such as tumor histology, clinicopathological features, and molecular characteristics. The most reliable method to detect LNM is the histopathological examination of surgically resected specimens; however, this method is invasive, time-consuming, and subject to sampling errors and interobserver variability.

AIM

To analyze influencing factors and develop and validate a risk prediction model for LNM in CRC based on a large patient queue.

METHODS

This study retrospectively analyzed 300 patients who underwent CRC surgery at two Peking University Shenzhen hospitals between January and December 2021. A deep learning approach was used to extract features potentially associated with LNM from primary tumor histological images while a logistic regression model was employed to predict LNM in CRC using machine-learning-derived features and clinicopathological variables as predictors.

RESULTS

The prediction model constructed for LNM in CRC was based on a logistic regression framework that incorporated machine learning-extracted features and clinicopathological variables. The model achieved high accuracy (0.86), sensitivity (0.81), specificity (0.87), positive predictive value (0.66), negative predictive value (0.94), area under the curve for the receiver operating characteristic (0.91), and a low Brier score (0.10). The model showed good agreement between the observed and predicted probabilities of LNM across a range of risk thresholds, indicating good calibration and clinical utility.

CONCLUSION

The present study successfully developed and validated a potent and effective risk-prediction model for LNM in patients with CRC. This model utilizes machine-learning-derived features extracted from primary tumor histology and clinicopathological variables, demonstrating superior performance and clinical applicability compared to existing models. The study provides new insights into the potential of deep learning to extract valuable information from tumor histology, in turn, improving the prediction of LNM in CRC and facilitate risk stratification and decision-making in clinical practice.

Oral oligofructose challenge reduces expression of glucose transport-1 and 5'-adenosine monophosphate-activated protein kinase in lamellar wall of Holstein heifer claw

The laminar tissue of bovine laminitis may undergo energy failure. The expression of glucose transport protein-1 (GLUT-1) and 5'-adenosine monophosphate-activated protein kinase (AMPK) affects the energy metabolism of digital laminar tissue. This study aimed to determine the expression of glucose uptake and AMPK in laminar wall corium of Holstein heifer claw by oral administration of oligofructose. A total of twelve clinically healthy Holstein heifers were selected and divided into two groups, including control (CON, n = 6) and experimental (OF, n = 6) groups. The heifers of OF group were given 17 g/kg BW oligofructose dissolved in water (20 mL/kg BW) and the heifers of CON group were given water only (20 mL/kg BW). The laminar tissues were collected after euthanasia. The amount of protein and transcript expression of AMPK and GLUT-1 were determined by western blot and quantitative reverse transcription-polymerase chain reaction (qRT-PCR), respectively. Expressions of phosphoenolpyruvate carboxy-kinase (PEPCK), receptor-c coactivator1-α (PGC-1α) and peroxisome proliferator-activated receptor-γ (PPAR-γ) were determined by qRT-PCR. The heifers of OF group showed no significant change in the expression and concentration of AMPK. The phosphor-(Thr172) AMPK and GLUT-1 were significantly decreased, while the gene contents of PPAR-γ and PGC-1α were significantly increased. The activation of AMPK and GLUT-1 in digital laminar tissues of heifers was inhibited, which may contribute to digital laminar tissue damage.

Characterization of the rumen microbiota and its relationship with residual feed intake in sheep

Feed efficiency is a highly important economic trait in sheep production and has a significant impact on the economic benefits of sheep farming. Microbial fermentation of the rumen has a vital role in the host's nutrition; the rumen microbiota might affect host feed efficiency. However, the relationship between the rumen microbiota and feed efficiency in sheep is unclear. In the present study, the microbiota of 195 Hu sheep was investigated and their residual feed intake (RFI), a commonly used measure of feed efficiency, was determined. From birth, all sheep were subjected to the same management practices. At slaughter, samples of liquid rumen contents were collected and subjected to amplicon sequencing for the 16S rDNA gene on the IonS5™XL platform. To identify the bacterial taxa differentially represented at the genus or higher taxonomy levels, we used linear discriminant analysis coupled with effect size and curve fitting. In the sheep rumen, the four most abundant phyla were Firmicutes, Bacteroidetes, Fibrobacteres, and Proteobacteria; and the dominant genera were unidentified Prevotellaceae, Fibrobacter, unidentified Lachnospiraceae, Saccharofermentans, and Succinivibrio. Pathway analysis of the 16S rDNA sequencing data from the rumen microbiota identified that carbohydrate metabolism was enriched. Using α-diversity analysis, we further identified that Observed species, ACE, Good's coverage, and Chao1 are more abundant (P < 0.01) in the low-RFI (L-RFI) group compared to the high-RFI (H-RFI) group. High-RFI sheep had a higher abundance of three bacterial taxa (Prevotellaceae, Negativicutes, and Selenomonadales), and one taxa was overrepresented in the L-RFI sheep (Succinivibrio), respectively. Furthermore, model fitting showed that Veillonellaceae, Sphaerochaeta, Negativibacillus, Saccharofermentans, and members of the Tenericutes, Kiritimatiellaeota, Deltaproteobacteria, and Campylobacterales were correlated with the sheep RFI classification and thus were indicative of a role in animal efficiency. Tax4Fun analysis revealed that metabolic pathways such as "energy metabolism," "metabolism of cofactors and vitamins," "poorly characterized," and "replication recombination and repair proteins" were enriched in the rumen from H-RFI sheep, and "genetic information processing" and "lipopolysaccharide biosynthesis" were overrepresented in L-RFI sheep rumen. In addition, six Kyoto Encyclopedia of Genes and Genomes orthology pathways were identified as different between H-RFI and L-RFI groups. In conclusion, the low RFI phenotype (efficient animals) consistently (or characteristically) exhibited a more abundant and diverse microbiome in sheep.

Transcriptome analysis of long noncoding RNAs ribonucleic acids from the livers of Hu sheep with different residual feed intake

Long noncoding RNAs (LncRNAs), as key regulators, have vital functions in various biological activities. However, in sheep, little has been reported concerning the genetic mechanism of LncRNA regulation of feed efficiency. In the present study, we explored the genome-wide expression of LncRNAs and transcripts of uncertain coding potential (TUCPs) in the livers of sheep with extreme residual feed intake (RFI) using RNA sequencing. We identified 1 523 TUCPs and 1 996 LncRNAs, among which 10 LncRNAs and 16 TUCPs were identified as being differentially expressed between the High-RFI and Low-RFI groups. Co-expression and co-localization methods were used to search for LncRNA and TUCP target genes, which identified 970/1 538 and 23/27 genes, respectively. Ontology and pathways analysis revealed that the LncRNAs/TUCPs that were highly expressed in the Low-RFI group are mostly concentrated in energy metabolism pathways. For example, LNC_000890 and TUCP_000582 might regulate liver tissue metabolic efficiency. The LncRNAs/TUCPs that were highly expressed in the High-RFI group are mostly enriched in immune function pathways. For example, TUCP_000832 might regulate animal health, thereby affecting feed efficiency. Subsequently, a co-expression network was established by applying the expression information of both the differentially expressed LncRNAs and TUCPs and their target mRNAs. The network indicated that differentially expressed genes targeted by the upregulated LncRNAs and TUCPs were mainly related to energy metabolism, while those genes targeted by the downregulated LncRNAs and TUCPs were mainly related to immune response. These results provide the basis for further study of LncRNA/TUCP-mediated regulation of feed efficiency.