IDH-mutant myeloid neoplasms are associated with seronegative rheumatoid arthritis and innate immune activation

Abstract

High prevalence of IDH mutations in seronegative rheumatoid arthritis (RA) with myeloid neoplasm, elevated 2-hydroxyglutarate, dysregulated innate immunity, and proinflammatory microenvironment suggests causative association between IDH mutations and seronegative RA. Our findings merit investigation of IDH inhibitors as therapeutics for seronegative IDH-mutated RA.

Figure 1.

High prevalence of IDH mutation in myeloid neoplasm with RA. (A) In all, 9.5% of cases with MDS had AIRD compared with only 4.7% of cases with AML (P = .003). (B) Inflammatory arthritis is the most prevalent AIRD in the MN-AIRD cohort. Inflammatory arthritis includes RA plus peripheral spondyloarthropathy. (C) Volcano plot comparing the blood counts, bone marrow blast, cytogenetic changes, and somatic mutation profile of MN-AIRD vs MN without AIRD. (D) Enrichment of IDH mutations in MN-AIRD compared with MN without AIRD, especially MDS-AIRD compared with MDS without AIRD. (E) Unexpected enrichment of IDH mutation observed in male cases with MN-AIRD but not in female cases with MN-AIRD. (F) High burden of IDH mutation in MN with RA compared with MN with other subtype of AIRD. (G) Frequency of IDH mutation according to subtype of AIRD in MN. (H) Strikingly high frequency of IDH mutation in seronegative RA compared with seropositive RA. (I) Majority of seronegative RA harbors IDH1 mutation, in contrast to IDH2 mutation in other MN-AIRDs. (J) In an independent validation cohort, one-third of patients with IDH^mut CMML have AIRD. The χ² test was used to determine associations between categorical variables. CTD, connective tissue diseases.

Figure 2.

IDH-mutant clones are likely to be present prior to diagnosis of autoimmune disease. (A) Interval between AIRD and MN diagnosis was shorter in IDH-mutated MN-AIRD compared with IDH wild-type MN-AIRD particularly in (B) IDH1-mutated compared with IDH2-mutated MN-AIRD. (C) Locally estimated scatterplot smoothing (LOESS) analysis to calculate clonal expansion kinetics of IDH^mut and to evaluate if IDH clones were present at the time of AIRD diagnosis. Within each “window,” a weighted average was calculated, and the sliding window passes along the x-axis. The shaded area indicates the 95% confidence interval. Gray lines represent patients. (D) Aberrantly high ratio of proinflammatory classical to nonclassical monocytes in patients with MN-AIRD (n = 4) compared with patients with MN without AIRD (n = 5) and age-matched healthy controls (n = 3). (E) Proinflammatory cytokines secreted by innate immune cells including GM-CSF, IL-12, fractalkine, IL-15, and IL-1β were significantly higher in bone marrow plasma of MN-AIRD (n = 9) compared with MN without AIRD (n = 109). (F) Aberrantly high macrophage phagocytic activity, assessed by pHrodo Red Staphylococcus aureus Bioparticles uptake, in subjects with MN (n = 9) compared with healthy controls (n = 3). (G) Schema of IDH mutational status interrogation in various cell populations (CD14⁺ monocytes, CD19⁺ B cells, and CD3⁺ T cells) in 2 patients with IDH mutation. (H) All IDH^mut MN with high bone marrow plasma level of 2-HG had AIRD compared with only 25% of cases with low 2-HG level. (I) Aberrantly high in vitro macrophage activity in subject with IDH^mut MN (n = 5) compared with healthy control (n = 3). All bars indicate mean, and all error bars indicate SD. The Mann-Whitney test was used to detect statistically significant differences between cohorts. The χ² test was used to determine associations between categorical variables. FGF-2, fibroblast growth factor 2; TGFα, transforming growth factor α; GM-CSF, granulocyte-macrophage colony-stimulating factor; IFN-α2, interferon α2; IFNγ, interferon γ; VAF, variant allele frequency.