Background Mutations in UBA1 are associated with VEXAS (Vacuoles, E1 enzyme, X-linked, Autoinflammatory, Somatic) syndrome, an adult-onset inflammatory disorder (Beck DB et al. NEJM 2020). Approximately 40% of VEXAS patients are also diagnosed with myelodysplastic syndromes (MDS). However, the prevalence and characteristics of UBA1 mutations in MDS are unknown. We hypothesize that UBA1 mutations underlie a proportion of MDS cases with systemic inflammation. Methods Molecular data were reviewed from a representative diagnostic and treatment-naive MDS cohort (n= 3,328) ascertained through the International Working Group for MDS (Bernard E et al NEJM Evid 2022). We prioritized 375 cases for UBA1 genotyping based on: 1) male gender; and 2) no identified driver mutations or few mutations, enriched in DTA (DNMT3A, TET2, ASXL1) genes; or 3) unclassifiable (MDS-U or MDS/MPN-U) per 2016 WHO Classification. We identified UBA1 hotspot mutations (M41T/V/L) with digital droplet PCR. Resulting UBA1 status was integrated with cytogenetic and sequencing data from 121 genes involved in myeloid pathogenesis to identify patterns of co-mutation. Clinical associations of UBA1 mutations were evaluated and where available, clinical history was reviewed for inflammatory conditions. Results We identified 30 UBA1 M41T/V/L mutations (median number mutant droplets: 1301; range: 4-12166) in 8% of the cohort (n=28 of 375 patients). This included 15 patients with M41T (c.122T>C), 7 with M41V (c.121A>G), 4 with M41L (c.121A>C), and 2 patients with more than one mutation (1 with M41V/T, and 1 with M41L/V). In 79% of UBA1-mutant cases (22/28), the karyotype was normal. Loss of the Y chromosome was observed in 6 cases. Three cases had MDS-associated cytogenetic abnormalities: del(11q), del(13q) and focal del(20q). In 43% of UBA1-mutant cases (12/28), UBA1 mutations were the sole genetic abnormalities. For a further 43% we observed co-mutation with DTA genes (DNMT3A, TET2, ASXL1) with median variant allele fraction (VAF) of 0.21 (0.02-0.56). SF3B1 co-mutations with VAF >30% were observed in 3 cases. Most UBA1-mutant cases were classified per 2016 WHO as MDS-SLD/MLD (n=18), but two cases were MDS with excess blasts 1, and one case was CMML-1 with 9% blasts. The median age at diagnosis was 72 years (44-89). No significant difference was found when comparing clinical parameters (blast counts, complete blood counts, age) between UBA1 mutant and wildtype MDS MDS. Clinical history was readily available for 4 UBA1-mutant patients. Inflammatory symptoms were reported in 3 patients, localized in the lungs (n=2), skin (n=2) and cartilage (n=1). Two patients also had a diagnosis of inflammatory/rheumatological disease (vasculitis and relapsing polychondritis), two patients had thrombosis, one had a history of monoclonal gammopathy of undetermined significance. The 4th patient, without inflammatory symptoms, had a prior diagnosis of T-cell large granular lymphocytic leukemia. Follow-up for transformation to acute myeloid leukemia (AML) and overall survival was available for 310 and 339 patients, respectively. No UBA1-mutant patients transformed to AML (n=24; median follow-up 2.6 years). There was no difference in overall survival between UBA1-mutant (n=27) and wildtype (n=312) patients (log-rank test p = 0.7), however, when subsetting for patients with no mutations detected by targeted DNA sequencing (UBA1-mutant n=7; wildtype n=45), UBA1-mutant patients had worse overall survival (median 5.2 vs 7.3 years; log-rank test p = 0.03). ConclusionUBA1 mutation in isolation or in combination with other mutations implicated in myeloid pathogenesis account for a significant proportion (8%) of patients diagnosed and treated as MDS but without established disease classification. Ongoing work includes 1) clinical review of inflammatory symptoms for the remaining UBA1-mutant cases and 2) profiling a larger cross-section of MDS by a UBA1 targeted panel to characterize mutations beyond the M41 locus.
Read full abstract