TL;DR
This study develops a novel clinical diagnostic model for Ankylosing Spondylitis (AS) by integrating immune cell infiltration data with clinical parameters. The model identifies monocyte and neutrophil counts as key independent diagnostic factors, offering a promising tool for earlier and more accurate AS screening, especially in patients with chronic back pain.
What Problem Does This Paper Address?
Ankylosing Spondylitis (AS) is a chronic inflammatory disease that significantly impacts patients' quality of life, yet its diagnosis remains challenging. The exact pathogenesis and local immune responses in AS are not fully understood, leading to diagnostic delays that can hinder timely treatment. Current diagnostic methods, like HLA-B27 testing, have limitations. This paper aims to address this by identifying key immune-related biomarkers and constructing a robust clinical diagnostic model to improve AS detection.
Methods
The researchers employed a multi-pronged approach:
- Microarray Analysis: They performed miRNA and mRNA microarrays on interspinous ligament tissue samples from AS patients and healthy controls to identify differentially expressed genes (DEGs).
- Bioinformatics: They constructed an miRNA-mRNA interaction network and identified "hub genes" (genes central to the network) using tools like FunRich and Cytoscape. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were conducted to understand the biological functions of these genes.
- Immune Cell Infiltration Analysis: CIBERSORT software was used to analyze immune cell composition in both microarray data (tissue) and publicly available GEO datasets (peripheral blood).
- Clinical Data Analysis: They collected clinical data and routine blood test results from a large cohort of AS patients and controls. Statistical analyses, including univariate and multivariate logistic regression, were used to identify independent risk factors for AS.
- Nomogram Construction: Based on the identified independent risk factors, a nomogram (a graphical tool for predicting outcomes) was constructed.
- Immunohistochemistry: This technique was used to validate the expression of a key identified gene, LYN, in AS and control tissue samples.
Datasets
- Interspinous Ligament Tissue Samples: Collected from three AS patients and three non-AS patients for miRNA and mRNA microarrays.
- Peripheral Blood Samples: Collected from 359 patients (197 AS, 162 non-AS) for clinical data and routine blood tests.
- GEO Database Datasets: GSE25101 and GSE73754 datasets of peripheral blood gene expression were used for validation of immune cell infiltration and LYN expression.
Key Insights and Ideas
- LYN as a Key Immune Hub Gene: The study identified LYN (a Src family tyrosine kinase) as a crucial immune-related hub gene in AS. Its expression was significantly correlated with dendritic cell activation and immune cell infiltration.
- Monocytes and Neutrophils as Independent Predictors: Beyond the well-known factors like age and gender, the counts of monocytes and neutrophils in peripheral blood emerged as significant, independent diagnostic factors for AS. This is a crucial finding for developing a practical diagnostic tool.
- Integrated Diagnostic Model: The combination of clinical data (age, gender) and routine blood counts (monocytes, neutrophils) proved effective in building a predictive model for AS, outperforming some existing diagnostic markers.
- Immune Cell Dysregulation in AS: The research reinforces the understanding that AS is an immune-related disease, highlighting the role of immune cell infiltration and dysregulation in its pathogenesis.
Key Results
- Differentially Expressed Genes: 406 differentially expressed mRNAs (DEmRNAs) and 225 differentially expressed miRNAs (DEmiRNAs) were identified from the interspinous ligament tissue microarray.
- miRNA-mRNA Network: A network was constructed, identifying LYN and LEFTY2 as overlapping hub genes.
- Immune Cell Correlation: LYN expression was significantly correlated with dendritic cell activation in AS patients.
- Nomogram Performance: The developed nomogram, incorporating gender, age, and counts of white blood cells, monocytes, and neutrophils, achieved an Area Under the Curve (AUC) of 0.855 and a C-index of 0.835, indicating good diagnostic accuracy.
- LYN Expression in AS: Immunohistochemistry confirmed significantly higher LYN expression in AS patient tissues compared to controls.
- GEO Dataset Validation: Analysis of GEO datasets confirmed significant differences in monocyte and neutrophil counts in AS patients and validated the correlation of LYN with these cells.
Limitations and Future Work
- Retrospective Study Design: The study is retrospective, which inherently has limitations.
- Limited Non-AS Group Data: The non-AS group included patients with thoracic spinal cord compression or lumbar spinal stenosis, and data on HLA-B27 testing was not available for this group, limiting certain analyses.
- Single-Center Study: The diagnostic model was developed and validated at a single center. Multi-center external validation with a larger sample size is crucial.
- Small Microarray Sample Size: While immunohistochemistry provided additional validation, the initial microarray sample size was limited.
- Mechanism of LYN-miRNA Link: The specific mechanisms by which has-miR-30b-5p and has-miR-212-3p are linked to LYN remain unknown and require further investigation.
The authors suggest that additional laboratory analyses and independent external data validation are necessary to confirm their findings and fully establish the diagnostic efficiency of the proposed model.
Summary updated