Abstract: FR-PO847
Self-Clustering of Tissue Gene Expression to Classify Patients with Lupus Nephritis
Session Information
- Glomerular Diseases: Immunology, Inflammation - I
November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Glomerular Diseases
- 1202 Glomerular Diseases: Immunology and Inflammation
Authors
- Almaani, Salem, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
- Yu, Lianbo, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
- Song, Huijuan, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
- Parikh, Samir V., The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
- Ayoub, Isabelle, The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
- Mejia-Vilet, Juan M., Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico, Mexico
- Malvar, Ana, Hospital Fernandez, Buenos Aires, Argentina
- Rovin, Brad H., The Ohio State University Wexner Medical Center, Columbus, Ohio, United States
Background
Histologic classification of kidney biopsy in lupus nephritis (LN), while used for treatment decisions, is not sufficiently robust to account LN's molecular heterogeneity that affects treatment response and outcomes. We tested whether unsupervised clustering of LN biopsies based on tissue gene expression was feasible to classify LN. We postulated that such a classification of LN would reflect disease pathobiology and would be more relevant to managing LN with drugs targeting specific pathogenic pathways.
Methods
Transcript levels of >500 genes involved in autoimmunity were measured using NanoString in microdissected glomeruli from 57 LN patient biopsies, and then used for unsupervised hierarchical clustering. For each gene, mRNA abundance was compared between each cluster group (CG) and the mean abundance of the other groups to determine genes that were differentially expressed. Differentially-expressed genes from each CG were used for pathway analysis. Demographic, clinical and histopathologic data were also compared between CGs using ANOVA and Fisher’s exact test as appropriate.
Results
Clustering resulted in 4 CGs. There were no significant differences in baseline creatinine, proteinuria, NIH activity or chronicity indices, or ISN/RPS class between CGs. Canonical pathway and upstream regulator analysis differentiated CGs (Table).
Conclusion
Transcript expression in the glomerular compartment of LN kidney biopsies identifies 4 subsets of patients. Inflammatory pathways expression appears to be highest in CG2 followed by CG4, and relatively suppressed in CG1 & 3. We suggest it may be feasible to tailor treatment to patients based on their CG classification of injury pathways that are differentially expressed.