Transcriptomic-Based Alternative to Clinical Classification of Human Ulcers

Lower extremity wounds are differentiated based on their physical location and underlying etiology. Clinicians frequently use classification schemes to characterize wounds and their potential outcomes. A proof of concept study was performed using RNA-sequencing data from lower extremity ulcers. We identified three clusters of samples with unique transcriptomic profiles that better reflected the biological/cellular environment of the wound as compared to clinical observations of wound duration or wound state. We identified a cluster of samples showing enrichment of inflammatory and infection gene expression? Cluster 3 showed enrichment of inflammatory and infection response genes, wheras Clusters 1 and 2 represented unique groups of wounds with differentially enriched reparative processes and less inflammation. We further developed a classifier based on these clusters using a small subset of 20 genes, which could accurately classify unknown samples into these clusters. Our results suggest these clusters may represent clinically relevant groups, which may serve as a more precise diagnostic tool when transcriptome technologies are more rapidly accessible.