A study from Vanderbilt University has uncovered subtle but significant differences in cellular organization between normal and quiescent Crohn’s disease tissues, challenging the assumption that intestinal tissue in disease remission completely returns to normal. The research, presented by Gaurav Rudravaram, a doctoral researcher at Vanderbilt University, employed advanced multiplexed immunofluorescence (MxIF) imaging and innovative computational methods to analyze cellular neighborhoods in the terminal ileum.[1]

“Histology is a powerful tool for studying Crohn’s disease at a cellular level,” Rudravaram noted during his talk. “While previous research has extensively characterized the active phase of Crohn’s disease, marked by high neutrophil infiltration, we wanted to investigate whether cellular neighborhoods truly return to normal during disease remission.”

The study analyzed tissue samples from nine individuals (four normal controls and five with quiescent Crohn’s disease) using MxIF imaging with 15 markers to classify cells into 13 cell subtypes. The researchers developed a sophisticated computational approach to analyze cellular neighborhoods, moving beyond traditional methods of simply counting nearby cells.

“In previous approaches, researchers would count the number of cell types in proximity to a cell of interest,” Rudravaram noted. “However, this method does not account for cells that might be technically ‘nearest’ but still relatively far away. We introduced a distance-weighting term that considers both the frequency of specific cell types and their proximity to the cell in question.”

The researchers processed over 100,000 nuclei in their dataset using a combination of nuclear segmentation, marker gating, and dimensionality reduction techniques. They employed the BIRCH clustering algorithm and developed a modified enrichment analysis formula that accounts for variations in tissue and nuclei numbers across samples.

“Our enrichment analysis wasn’t focused simply on whether a particular cell type occurred more often in a cluster,” Rudravaram clarified. “Instead, we wanted to understand if certain cell types appeared more or less frequently in particular disease states.”

The analysis revealed several patterns of cellular organization. In the 20-cluster analysis, monocytes were enriched in normal tissue, whereas leukocytes were enriched in two clusters in quiescent tissue. When the resolution was increased to 40 clusters, the differences were more pronounced. Normal tissues demonstrated enrichment of monocytes, macrophages, and leukocytes, whereas quiescent tissues showed unique patterns of enrichment of enteroendocrine cells and leukocytes. “We observed that the neighborhood weighting for cytotoxic T cells was much higher around leukocytes in the quiescent state than in the normal state,” Rudravaram explained. “Similar patterns were seen with CD4+ and CD8+ T cells.”

The team employed a leave-one-out analysis, removing one tissue sample at a time, to verify that the observed patterns were not driven by a single subject. The enrichment patterns were confirmed through this validation process. As Rudravaram noted, this approach could help improve our understanding of Crohn’s disease pathology and elucidate differences in the tissue microenvironment and cellular organization during different disease states. However, further studies with larger sample sizes are required to fully understand the clinical significance of these cellular neighborhood patterns and explore whether specific neighborhood configurations correlate with the likelihood of disease recurrence or response to therapy. Ultimately, this could lead to the identification of new therapeutic targets or biomarkers for monitoring disease status and predicting the risk of relapse.Gaurav Rudravaram, doctoral researcher at Vanderbilt University, USA.

[1] Gaurav Rudravaram, Identifying cellular neighborhood phenotypes differentiating normal and quiescent Crohn’s disease via MxIF. Presented at SPIE 2025 Digital and Computational Pathology conference, February 19, 2025; San Diego, CA.