Background/Purpose: Our group was the first to contend that STAT1 mediated genes were expressed in sarcoidosis. Since that time, several other groups have found similar results. In a direct extension of these data, JAK-inhibitors have shown benefit in treating sarcoidosis. There are now numerous JAK inhibitors approved to treat inflammatory and hematologic diseases, with varying specificity for JAK molecules. To design more robust clinical studies, we need to determine which JAK-inhibitors are most likely to work in sarcoidosis.

Methods: All publicly available gene expression microarray data from the Gene Ontology Omnibus (GEO) was queried. GEO was queried using the search terms “sarcoidosis” or “sarcoid”, and all human datasets which compared sarcoidosis to healthy controls were included. Microarray data were evaluated for differential expression using GEO2R, a limma based public access software. We analyzed the differential expression of the 4 JAK and 7 STAT genes in all sarcoidosis datasets to date. Significant differential expression was defined as a FC >1.5 or < -1.5, and an adjusted p-value or FDR p-value of < 0.05.

Microarray data were evaluated for differential expression using GEO2R, a limma based public access software available as part of the GEO database. For each dataset, sarcoidosis samples were compared to healthy controls. When sarcoid activity vs inactivity was reported (2 datasets), only active sarcoid samples were included. We analyzed the differential expression of the 4 JAK and 7 STAT genes in all sarcoidosis datasets to date. Significant differential expression was defined as a log2FC >0.585 or < -0.585 (corresponding to a FC >1.5 or < -1.5), and an adjusted p-value or FDR p-value of < 0.05.

Microarray data were evaluated for differential expression using GEO2R, a limma based public access software available as part of the GEO database. For each dataset, sarcoidosis samples were compared to healthy controls. When sarcoid activity vs inactivity was reported (2 datasets), only active sarcoid samples were included. Two analyses were planned. The first was an analysis of differential expression of the 4 JAK and 7 STAT genes. The second was a pathway-based analysis of all pathways known to signal through JAK-STAT. Pathways were evaluated using Gene Set Enrichment Analysis (GSEA).

Results: 12 datasets were included: 6 whole blood, 1 peripheral blood mononuclear cells, and 1 each of anterior orbit, lacrimal gland , bronchoalveolar lavage fluid, lung biopsy, and skin. STAT1 and JAK2 were the most commonly differentially expressed JAK-STAT genes. STAT1 was significantly differentially expressed in 85% of sarcoid datasets; the average FC was 2.6 compared to healthy controls. JAK2 was differentially expressed in 62% of datasets; the average FC was 1.7 compared to healthy controls.

Conclusion: This in-silico meta-analysis found that STAT1 and JAK2 are the most significantly expressed JAK-STAT genes in sarcoidosis tissues. We hypothesize that JAK-inhibitors targeting JAK2 are the most likely to be useful in sarcoidosis. Potential JAK inhibitor candidates for clinical trials include tofacitinib, ruxolitinib, baricitinib and peficitinib. We would not suggest trialing JAK1-specific inhibitors in sarcoidosis.

Figure 1: Average fold change of JAK-STAT genes across all sarcoidosis datasets.

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/stat1-and-jak2-are-the-most-appropriate-targets-of-jak-inhibitor-therapy-for-sarcoidosis-an-in-silico-meta-nalysis/