CHICAGO — Somatic mutations in IDH1 and IDH2 genes appear to be linked with inflammation, regardless of whether patients also have a hematologic malignancy, and are further associated with rheumatologic conditions in large biobanks, according to research presented at American College of Rheumatology (ACR) 2025 Annual Meeting.

The research further suggested that targeted treatment with IDH (isocitrate dehydrogenase) inhibition could improve inflammatory manifestations in these patients with otherwise undiagnosed symptoms.

“We suggest that IDH1/2 mutations are driving a treatable premalignant inflammatory condition and that screening of somatic IDH1/2 mutations in patients with undiagnosed systemic inflammatory disorder could inform targeted treatment,” Flore Castellan, PhD, a postdoctoral fellow at NYU Grossman School of Medicine in New York City, told attendees.

“Somatic mutations are increasingly recognized as drivers of inflammation, particularly in the setting of clonal hematopoiesis (CH), the clonal expansion of hematopoietic stem and progenitor cells carrying somatic mutations,” Castellan said. The co-occurrence of CH and persistent cytopenia is called clonal cytopenia of unknown significance (CCUS).

“Besides the potential for transformation to neoplasm, the impact of mutations in CH can vary widely, from having no apparent clinical consequence to contributing to an increased risk of rheumatic and cardiovascular conditions to causing somatic monogenic diseases such as VEXAS syndrome,” she said. VEXAS syndrome involves somatic mutations in the UBA1 gene, usually only in myeloid origin cells.

Given the recent discovery of VEXAS syndrome and the role of somatic UBA1 mutations in it, Castellan and her colleagues hypothesized that other somatic variants in novel genes could be an underlying contributor to inflammation in undiagnosed conditions. They therefore conducted a novel gene discovery analysis in a cohort of 265 adult patients with undiagnosed systemic autoinflammatory disease.

All the patients had VEXAS-like severe systemic inflammation, but they were negative for pathogenic variants, including no UBA1 variant. The patients were predominantly men (73%) and mostly between ages 50-80. Most (72.8%) had no hematologic diagnoses while 15.5% had myelodysplastic syndromes, 7.5% had myeloid neoplasm, and 4.2% had lymphoid neoplasm.

First, the researchers looked for shared rare missense variants and identified shared somatic gain-of-function variants in KRAS and NRAS, which are already known to cause Ras-associated autoimmune lymphoproliferative disorder. The most shared missense variants in the cohort were at IDH1 Arg132 and IDH2 Arg140, which occurred in 5.7% of the patients with VEXAS-like symptoms. Further, four of these 15 patients did not have a hematologic malignancy.

The researchers then grouped all the rare mutations per gene and conducted a gene burden analysis to compare with 558 matched control individuals. The only variants that remained significantly more common in their cohort after correcting for false discovery rates were IDH1 and IDH2 (P < .01) and SRSF2 (P < .05)

IDH1/2 code for metabolic enzymes involved in the citric acid cycle, and they can contain hotspot mutations that are frequently found in cancer, especially in glioma and blood neoplasms, Castellan said. Normally, IDH converts isocitrate to alpha-ketoglutarate, but mutant IDH diverts that reaction and instead leads to the production of 2-hydroxyglutarate, an oncometabolite. Previous research has already linked these IDH1/2 hotspot mutations to inflammation in myeloid neoplasms, but Castellan and her colleagues wanted to find out whether inflammation could result from these mutations even in the absence of hematologic malignancies.

So they looked closely at the four patients with IDH1/2 mutations but without hematologic malignancies. All four had episodic symptoms, rheumatoid arthritis, skin inflammation, elevated acute phase reactants (APR), and steroid dependency, and three of the four had cytopenia.

“The findings suggest that IDH1 and IDH2 mutations can drive inflammation even in the absence of hematologic malignancy,” Castellan said.

They then corroborated these findings by looking for IDH1/2 mutations in a cohort of 1100 Mayo Clinic patients with CH or CCUS. They identified 16 patients with these mutations, half of whom had inflammatory symptoms. Among those eight patients, all but one had cytopenia, all but one had elevated APR, all but two had steroid dependency, and half had rheumatoid arthritis.

Validation of Association Between IDH1/2 Hotspot Mutations and Inflammatory Disorders

From here, the researchers sought to validate the association between IDH1/2 hotspot mutations and inflammatory disorders. They looked for these mutations in the UK Biobank, which had approximately 450,000 participants with whole exome sequencing data, and the All of Us Research Program, which includes approximately 250,000 participants with whole genome sequencing data. They identified 112 people in the UK Biobank and 60 people in the All of Us program who had IDH1/2 mutations and no history of hematologic malignancies.

The researchers found that IDH1/2 and SRSF2 mutations were significantly associated with higher odds of disease as well as an association with higher C-reactive protein (CRP) levels. The odds of association for mutations in these three genes in the UK Biobank were highest for those with polymyalgia rheumatica (P < .001) and giant cell arteritis (P < .001), but there were also significant associations between SRSF2 and cutaneous vasculitis (P < .001) and SRSF2 and giant cell arteritis/temporal arteritis (P < .01) and granulomatosis with polyangiitis (P < .01). In the All of Us cohort, IDH1/2 mutations were also significantly associated with giant cell arteritis (P < .001) and with polymyalgia rheumatica (P < .05).

Effect of IDH Inhibition

The next question, then, is whether inhibiting IDH can treat people with inflammatory symptoms. Castellan reported on findings from a phase 2 trial of 20 patients with IDH1 mutations and CCUS who took the IDH inhibitor ivosidenib over 5 years. In that trial, investigators saw a decrease in CRP levels, including a normalization of levels in eight of the 11 patients who had elevated levels at baseline. All of the enrolled patients with autoimmune disorders reported improvements in their symptoms.

Evidence of therapeutic effects in two patients over 2 years of treatment were seen in a resolution of both hematologic and inflammatory manifestations, including a decrease in 2-hydroxyglutarate and CRP levels and an increase in neutrophil count. The patients also had a decrease in IDH allele fraction in their blood, “suggesting that a mutant clone may even be contracting over time,” Castellan said.

Altogether, the results of these findings suggest value in testing patients with otherwise undiagnosed inflammatory conditions for an IDH1/2 mutation and, if one is present, treating them with an IDH inhibitor.

Francisca Sivera, MD, PhD, of the Hospital General Universitario Elda in Elda, Spain, said the presentation got her thinking about her cohort of patients in her clinic.

“Do I have patients that really have that mutation, and I’m mistreating and I’m misdiagnosing [them] just as normal inflammatory or normal autoimmune patients, but I could be offering them something else?” Sivera mused. “I’m wondering, how many of them are there out there, and how do I identify them?”

Since there’s no diagnostic test for polymyalgia rheumatica, it might be that she has patients who are clinically diagnosed with polymyalgia rheumatica but who do not actually have it.

The research did not have external funding. Castellan had no disclosures. Sivera reported receiving travel support from AbbVie, grants from Novartis and Lilly, and honoraria from Pfizer.

Tara Haelle is a science/health journalist based in Dallas.