Real‐world outcomes of vacuoles, E1 enzyme, X‐linked, autoinflammatory, somatic syndrome at a tertiary referral centre

Vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic (VEXAS) syndrome is a recently described clonal myeloid disorder that presents with autoinflammatory manifestations. It arises from the acquisition of somatic mutations in UBA1, a gene that encodes the ubiquitin-like modifier activating enzyme E1 and lies on the X chromosome.1 This most commonly manifests as a treatment-refractory and often fatal autoinflammatory disease in older men, but women can rarely be affected.2, 3 Other systemic manifestations include dermatological involvement, colitis, pneumonitis and thrombosis.4 The haematological features of VEXAS include prominent vacuolation in the erythroid and myeloid precursors in the bone marrow, as well as a macrocytic anaemia and predisposition towards myelodysplastic syndromes (MDS) and monoclonal gammopathies.5 Identification of somatic, inactivating mutations in UBA1 is mandatory to establish the diagnosis of VEXAS. These most commonly occur at the p.Met41 codon, but novel mutations are now being described.6 Corticosteroids are effective in reducing autoinflammatory manifestations of VEXAS but result in substantial toxicity, particularly with long-term exposure.6, 7 First-line steroid-sparing treatments include anti-interleukin (IL)-1 agents, anti-IL6 agents (e.g. tocilizumab) and Janus kinase inhibitors for patients with primarily inflammatory manifestations.8, 9 For patients with concomitant MDS and primarily haematological manifestations, chemotherapy agents such as azacitidine are often used.10 Allogeneic haematopoietic cell transplantation (allo-HCT) is a potentially curative option but can be associated with significant treatment-related morbidity and mortality.11, 12 The optimal approach for allo-HCT in VEXAS is not yet known.13 The majority of data on outcomes in VEXAS syndrome come from large cohorts ascertained at the National Institutes of Health and by other large collaborative groups, and there remains a need to understand clinical presentations and treatment patterns in real-world settings. To address this, we retrospectively analysed a cohort of consecutive patients treated at a Canadian tertiary referral centre. Medical records were examined by two independent reviewers using a standardized form designed to capture demographic features, clinical characteristics, peripheral blood and bone marrow findings, treatment regimens, treatment-related complications and outcomes for VEXAS patients with a confirmed somatic UBA1 mutation diagnosed between 1 January 2021 and 5 October 2023. Some features of patients included in this cohort have been published in part previously.3, 5, 14 Sequencing was performed at the US National Institutes of Health (Sanger sequencing), the IWK Health Centre in Halifax, Nova Scotia (next-generation sequencing, NGS) or BluePrint Genetics (NGS). Treatments were counted if patients received them at any time point. Clinical responses after VEXAS treatment were categorized as per the FRENVEX criteria: Complete response (CR) required absence of VEXAS-related symptoms, C-reactive protein (CRP) ≤10 mg/L and prednisone ≤10 mg/day; partial response (PR) required absence of VEXAS-related symptoms, CRP reduction of 50% from pretreatment, and prednisone equivalent dose; and treatment failure (TF) was considered if there were persistent VEXAS-related clinical symptoms, persistent inflammatory biomarkers or an inability to reduce glucocorticoid therapy.15 Treatment sequencing was as per individual clinician decision; patients did not receive concurrent steroid-sparing therapies. Continuous data were presented as median (range), while categorical data were expressed as number (percentage). Statistical analysis was performed using GraphPad Prism version 10.0.0, GraphPad Software, Boston, Massachusetts, USA (www.graphpad.com). Local ethics approval was obtained (H23-02559) for this study. A total of 18 patients were identified, 17 males and one female patient with Turner syndrome (Table 1).3 The mean age of symptom onset was 67 years (range 49–81 years). The median age at diagnosis was 71 years (range 60–83 years). Symptom onset was prior to 2020 in 50% (9/18) patients, and the median time from symptom onset to diagnosis for patients with symptom onset after 2020 was 28 months (range 5–36 months). Of these patients, 33% (6/18) were deceased at last follow-up. The most common mutation types included p.Met41Thr (33%, 6/18), p.Met41Val (33%, 6/18) and the splicing variant c.118-1G>C (17%, 3/18). A UBA1 p.Asp585Glu variant was identified in a patient with typical clinical features of VEXAS. This is a rare mutation in the active adenylation domain of UBA1; it is absent from gnomAD version 4 and predicted to be likely pathogenic by Alpha Missense (0.995).16 A total of six patients (33%) had chondritis, although none had a previously established diagnosis of relapsing polychondritis. Cutaneous manifestations were the most prevalent clinical feature in this cohort, with 78% (14/18) of patients exhibiting skin lesions during the course of their disease. Neutrophilic dermatosis (Sweet syndrome) was diagnosed in 17% (3/18) of patients, tumid lupus erythematosus in one patient and erythema nodosum in one patient. This was followed by constitutional symptoms in 72% (13/18), pulmonary involvement in 56% (10/18) and venous thromboembolism in 50% (9/18). Ocular involvement was observed (28%, 5/18) and included giant cell arteritis (GCA), episcleritis and periorbital inflammation. Adult-onset Still's disease, rheumatoid arthritis and atypical GCA were diagnosed prior to VEXAS discovery in one patient each. Gastrointestinal involvement (11%, 2/18) included mesenteric panniculitis and recurrent colitis. Neurological involvement (17%, 3/18) included cranial nerve palsy, aseptic meningitis, acute sensorineural hearing loss, and strokes. A minority of the cohort had an established diagnosis of MDS (17%, 3/18) or monoclonal gammopathy of undetermined significance (17%, 3/18). The most common morphological finding on bone marrow biopsy was vacuolization in granulocytic precursors, seen in 78% (14/18) of patients. Significant dysplasia was noted in 67% (12/18) of patients and marrow hypercellularity in 61% (11/18) of patients. Most patients had a mild anaemia (median haemoglobin, 126 g/L) and elevated inflammatory markers (median CRP 67.1 mg/L, median ferritin 578 μg/L) at the time of symptom onset, with progressive anaemia (median haemoglobin, 100 g/L) and persistent elevation in inflammatory markers (CRP 34.1 mg/L, ferritin 597 μg/L) noted at the time of their VEXAS diagnosis. Figure 1 summarizes treatment responses and the consequent change in laboratory parameters to the three steroid-sparing agents used in this cohort. Amongst patients receiving azacitidine, we observed a CR rate of 60% (3/5) and a TF rate of 40% (2/5) with no PR seen. For patients receiving ruxolitinib, we observed a PR rate of 33% (1/3) and a TF rate of 66% (2/3) with no CR seen. Notably, one patient classified as a TF to ruxolitinib did have an improvement in clinical symptoms and a 25% reduction in steroid dose but had persistently elevated CRP. Of patients receiving tocilizumab, 14% (1/7) had a CR and 86% (6/7) had TF (Figure 1A). The median pretreatment steroid dose for azacitidine was 10 mg (range 10–30 mg) and that for posttreatment was 8 mg (range 5–30 mg). The median pretreatment steroid dose for ruxolitinib was 20 mg (range 17.5–30 mg) and that for posttreatment was 17.5 mg (range 15–30 mg). For tocilizumab, the median pretreatment steroid dose was 25 mg (range 10–40 mg) and that for posttreatment was 25 mg (range 5–40 mg) (Figure 1B,C). Patients receiving azacitidine had higher haemoglobin (123 g/L) and platelet counts (263 × 109/L) during treatment (Figure 1D–G). Treatment with ruxolitinib or tocilizumab was not associated with numerical improvements in median haemoglobin. A swimmer plot demonstrating treatment durations and outcomes is demonstrated in Figure 1H. Inflammatory biomarker trajectories are illustrated in Figure 1I. Three patients with VEXAS syndrome received allo-HCT. One patient had persistent and refractory autoinflammatory symptoms for over 20 years with pulmonary involvement requiring years of high-dose corticosteroids. He received a reduced intensity conditioning (RIC) allo-HCT from a haploidentical donor. He died of a pulmonary fungal infection at D+21 following allo-HCT. A second patient had clonal evolution with inv(2)(p23q21) in the marrow without morphological MDS and lung involvement. He received a RIC allo-HCT from a fully matched unrelated donor. He experienced late-onset Kaposi sarcoma, drug-resistant cytomegalovirus reactivation and steroid refractory gut and skin graft-versus-host disease and died at D+410 following transplant. A female patient with Turner syndrome had VEXAS and transfusion-dependent MDS and received a RIC allo-HCT from a fully matched sibling donor. She had no graft-versus-host disease or infections and remains alive at over 2 years post-transplant. In this study, we characterized the diagnostic features and clinical outcomes of patients with VEXAS syndrome in a real-world, tertiary referral setting. The median time from symptom onset to diagnosis for patients with symptom onset after the description of VEXAS in 2020 was 28 months, suggesting that there remain significant challenges with identifying and diagnosing VEXAS patients in a timely manner. We also identified that a considerable number of patients in our cohort have mutations that lie outside the p.Met41 codon. This is important to recognize for clinicians, as not all commercially available tests will sequence the entire UBA1 gene and may only target the p.Met41 hotspot. We also demonstrate that the clinical presentation of patients with VEXAS in a real-world setting is heterogeneous, and diagnosing these patients requires a high index of suspicion. In our cohort, we observed that patients who received azacitidine had both a reduction of steroid requirements and improvement in their haematological parameters. Most patients who received azacitidine did experience a clinical response. Ruxolitinib did induce clinical responses and reduce steroid requirements in some patients but was associated with a mild reduction in haemoglobin levels. We observed few clinical or haematological responses to tocilizumab. A small number of patients underwent allo-HCT, but this was associated with a high rate of non-relapse mortality due to profound infectious and organ-related complications, even with RIC conditioning approaches, although this is a limited sample size. However, one patient who received allo-HCT did respond well and remains alive at over 2 years follow-up. Future controlled, prospective studies will be required to definitively establish the standard of care in clinical management for VEXAS syndrome. LL and EJZ collected data, wrote the manuscript and generated figures. RS, AA, GS, KA, MH, LF, VC, AL, MY, LL, PY and TJN contributed to patient referrals. LYCC and RJS supervised the project. All authors approved the final manuscript. No funding received for this study. RJS reports honoraria from AbbVie, Jazz Pharmaceuticals, Kite/Gilead and Sobi; advisory board for AbbVie, Astellas, BMS and Novartis and research funding from Jazz Pharmaceuticals. LF reports honoraria from Novartis. This project was REB approved under certificate H23-02559. Consent requirement waived by REB. Clinical data cannot be made publicly available for privacy reasons, but the authors can be contacted to assess reasonable requests.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.