fetching data ... 
Background: Vacuole, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome (VS) is a life-threatening autoinflammatory disease caused by acquired pathogenic UBA1 variants and is associated with high mortality and adverse events. The prognostic factors proposed to date are primarily based on retrospective studies and may be affected by selection bias favoring surviving cases as well as by ambiguities in the follow-up period. Identifying reliable prognostic factors for survival is essential for risk stratification, optimization of treatment, enabling early intervention, guiding clinical decisions, informing patients, and supporting drug development to improve overall outcomes.
Objectives: We aimed to identify baseline predictors of poor outcomes in VS though prospective study.
Methods: We conducted a nationwide prospective registry in Japan. All patients with suspected VS, defined by a previously published score of ≥3 [1], underwent next-generation sequencing to detect pathogenic UBA1 variants and to quantify baseline variant allele frequency (VAF). Disease activity (VEXASCAF; REF) [2], serum C-reactive protein (CRP) levels, prednisolone (PSL) dose, adverse events (AEs), mortality, and transfusion dependency were assessed every 3 months for 1 year.
To validate the association between poor outcomes and VAF of pathogenic UBA1 variants, we also analyzed our recently published retrospective dataset [3]. All patients provided written informed consent. The study was approved by the Yokohama City University Certified Institutional Review Board (F240500053), and the protocol was registered in the University Hospital Medical Information Network (UMIN000053426)
Results: We enrolled 60 patients with pathogenic UBA1 variants (all male; median age 74 years) and 45 male patients with VEXAS-like syndrome without UBA1 variants as controls. Over a median follow-up of 343 days, VEXASCAF and CRP decreased, whereas the PSL dose did not change significantly. Grade ≥3 AEs occurred in 50% of patients within 1 year. In multivariable Cox regression, higher baseline VAF and higher PSL dose independently predicted the primary poor outcome, defined as death or transfusion dependency. The association between higher VAF and poor outcomes was replicated in an independent cohort of 34 patients with VS. After adjustment for disease onset, the UBA1 p.Met41Leu variant was associated with better survival. Patients with VEXAS-like syndrome showed similarly poor outcomes.
Conclusions: Baseline pathogenic UBA1 VAF is a strong prognostic biomarker for poor outcomes in VS and may aid risk stratification.
| VEXAS | VEXAS-like | p | |
|---|---|---|---|
| Participants -no. | 60 | 45 | |
| Male sex -no. (% ) | 60 (100) | 45 (100) | 1.0 |
| Median age at time analysis (range ) | 74.0 (55-89, IQR 71.0-77.0) | 73.0 (43-88, IQR 67.0-81.0) | 0.35 |
| Observation period (days ) | 343 | 213 | 0.0033 |
| Pathogenic UBA1 variant -no. (% ) | 60 (100) | 0 (0) | |
| p. Met41Thr (c.122T > C) -no. (% ) | 22 (36.6) | 0 | |
| p. Met41Val (c.121A > G) -no. (% ) | 12 (20.0) | 0 | |
| p. Met41Leu (c.121A > C) -no. (% ) | 22 (36.6) | 0 | |
| Splice variant -no. (% ) | 3 (5.0) | 0 | |
| p. Ala478Ser -no. (% ) | 1 (1.7) | 0 | |
| VEXASCAF (range ) | 3 (0-8) | 3 (0-7) | 0.93 |
| VEXASCAF components | |||
| Fever -no. (% ) | 28 (48.3) | 29 (64.4) | 0.10 |
| Auricular chondritis -no. (% ) | 13 (22.0) | 6 (13.3) | 0.26 |
| Nose chondritis -no. (% ) | 5 (8.8) | 1 (2.2) | 0.16 |
| Respiratory symptoms -no. (% ) | 15 (25.9) | 14 (31.1) | 0.56 |
| Arthralgia -no. (% ) | 25 (43.1) | 20 (44.4) | 0.89 |
| Arthritis -no. (% ) | 12 (21.1) | 7 (15.6) | 0.48 |
| Thrombosis -no. (% ) | 5 (8.8) | 6 (13.3) | 0.46 |
| Dermatosis -no. (% ) | 36 (61.0) | 25 (55.6) | 0.58 |
| Eye symptoms -no. (% ) | 16 (27.6) | 8 (17.8) | 0.24 |
| Vasculitis -no. (% ) | 7 (13.2) | 8 (18.6) | 0.47 |
| Periorbital edema -no. (% ) | 4 (7.0) | 4 (8.9) | 0.73 |
| Hematological manifestation | |||
| Myelodysplastic syndrome -no. (% ) | 24 (45.3) | 23 (54.8) | 0.36 |
| Macrocytic anemia no. (% ) | 50 (83.3) | 22 (48.9) | 0.0002* |
| Variable at study entry | HR | 95% CI | p |
|---|---|---|---|
| VAF | 1.036 | 1.005-1.069 | 0.023 |
| Age | 1.072 | 0.994-1.155 | 0.070 |
| CRP mg/L | 1.003 | 0.994-1.013 | 0.470 |
| PSL mg/day | 1.042 | 1.003-1.082 | 0.033 |
REFERENCES: [1] Maeda, A., et al., Efficient detection of somatic UBA1 variants and clinical scoring system predicting patients with variants in VEXAS syndrome. Rheumatology (Oxford), 2023.
[2] Kirino, Y., et al., Low remission rates and high incidence of adverse events in a prospective VEXAS syndrome registry. Rheumatology (Oxford), 2024.
[3] Maeda, A., et al., High Glucocorticoid Dependency and Limited Therapeutic Response in Japanese Patients with VEXAS Syndrome: A Multicentre Retrospective Study. Mod Rheumatol, 2025.
Acknowledgments: NIL.
Disclosure of Interests: Yohei Kirino Novartis, Amgen, Novartis, Sobi, Amgen, Sysmex, Ayaka Maeda: None declared, Hideaki Nakajima: None declared.