Background:

VEXAS syndrome is an adult-onset autoinflammatory disease caused by somatic UBA1 mutations. Up to 50% of VEXAS meet WHO criteria for MDS, but the diagnostic and prognostic value of MDS-related morphological and genetic features in this setting are unclear. The relevance of standard MDS prognostic models, such as IPSS-R and IPSS-M, also remains to be established in VEXAS-associated MDS.

Methods: Multicenter retrospective study of the FRENVEX network, including genetically confirmed VEXAS patients (pts) diagnosed between 2009 and 2024 (retrospectively for pts diagnosed before 2021) with available bone marrow assessment and molecular data (31-gene NGS panel). Patients were classified according to WHO 2022 as VEXAS with MDS (VEXAS-MDS) or without (w/o) MDS. VEXAS-MDS cases were matched 1:3 to MDS controls without UBA1 mutation based on sex, age, and IPSS-M. The 3 groups were compared.

Results:

Among 174 VEXAS pts (172 men; median age 74 years), 78 (45%) met WHO 2022 criteria for MDS (VEXAS-MDS). Anemia was more pronounced in both VEXAS w/o MDS (median Hb 9.8 g/dL) and VEXAS-MDS (9.3 g/dl) compared to MDS controls (n=234; 10.6 g/dl; both p<.001). Erythroblastopenia (≤10% marrow erythroblasts) was more frequent in VEXAS w/o MDS (20% pts) and VEXAS-MDS (17% pts) than in MDS (7% pts; p=0.01 and 0.02, respectively). Bone marrow hypercellularity and megakaryocytic hyperplasia (grade >3, Thiele et al., Haematologica 2005) were also more common in VEXAS w/o MDS (34% and 20% pts) and VEXAS-MDS (46% and 27% pts) than in MDS controls (23% and 13% pts; all p<0.05).

IPSS-R cytogenetics were favorable (very good or good) in 91% VEXAS w/o MDS and 95% of VEXAS-MDS pts, compared with 83% of MDS controls (p= NS).

Somatic MDS-related co-mutations (other than UBA1) were observed in 44 (46%) VEXAS w/o MDS, 39 (50%) VEXAS-MDS, and 180 (77%) MDS controls (p<0.01 for both). Mutations affecting 1, 2, 3, or ≥4 MDS-related genes occurred in 24 (25%), 16 (17%), 4 (4%), and 0 (0%) VEXAS w/o MDS; 20 (26%), 12 (15%), 5 (6%), and 2 (3%) VEXAS-MDS; and 93 (30%), 78 (25%), 28 (9.0%), and 19 (6%) MDS pts, respectively. Most frequent mutations involved epigenetic regulators, in 30 (31%) VEXAS w/o MDS, 27 (35%) VEXAS-MDS, and 127 (41%) MDS pts (p= NS), while spliceosome mutations were less frequent in VEXAS w/o MDS (8%) and VEXAS-MDS (14%) compared to MDS pts (37%, both p<.01).

In VEXAS patients, association with MDS was characterized by more frequent multilineage dysplasia (p<.01), higher marrow blast counts (p<.01), and lower CRP (p=0.02).

Median follow-up was 4.0 years (95% CI, 3.6–4.4). No AML transformation occurred in VEXAS w/o MDS and VEXAS-MDS patients. Median overall survival (OS) was 4.5 years (95% CI, 3.1–NR) in VEXAS w/o MDS and 5.5 years (95% CI, 4.8–8.0) in VEXAS-MDS (p=0.56), significantly shorter than in MDS controls, who, as stated before, were matched with VEXAS cases in particular for IPSS-M (median OS 10.7 years; 95% CI, 8.2–NR; p=0.003 with VEXAS w/o MDS; p=0.036 with VEXAS-MDS).

Discriminative performance for OS using Harrell's concordance index was limited in VEXAS w/o MDS (IPSS-R: 0.54; IPSS-M: 0.58) and VEXAS-MDS (0.59; 0.55), but higher in MDS controls (0.72 for both).

In VEXAS patients (with or w/o MDS), only high-risk IPSS-M (>0.5) was associated with inferior OS (median 2.2 years [95% CI, 0.94–NR]; HR 10.7, 95% CI 2.30–49.6; p=0.002). No significant impact on OS was observed for UBA1 genotype (HR 1.09 for M41V; p = 0.80) or for the co-mutation profile.

Conclusions

In this large VEXAS cohort, patients, regardless of MDS association, had distinct hematologic and genetic features compared with MDS controls, including hypercellular marrow, erythroblastopenia, more favorable cytogenetics, and a restricted mutational profile. To our knowledge, this is the first study demonstrating that, even after matching for IPSS-M, age, and sex, VEXAS patients (with or without MDS) have poorer OS than MDS without VEXAS.

Conventional MDS classifications and prognostic scores failed to stratify OS, highlighting the need for VEXAS-specific tools integrating both inflammatory and clonal parameters. These findings may have important therapeutic implications, particularly regarding the use of MDS-directed treatments frequently considered in VEXAS, such as azacitidine in particular by our group (Jachiet et al., Blood 2025), and more rarely, allogeneic SCT.

This content is only available as a PDF.
2025
Sign in via your Institution