Autoimmune and autoinflammatory diseases
Home · Research areas

VEXAS syndrome

VEXAS syndrome is an acquired autoinflammatory disease of adults, caused by a somatic variant of the UBA1 gene in bone marrow cells. It combines multisystem inflammatory manifestations, cytopenias and bone marrow abnormalities, and its management is shared between internal medicine and haematology. Arsène Mekinian is first author of the American College of Rheumatology guidance statement on its diagnosis and management.

VEXAS syndrome is an acquired autoinflammatory disease of adults caused by somatic variants of the UBA1 gene in haematopoietic cells.

First identified in 2020, it is an important model at the interface between systemic inflammation, clonal haematopoiesis and myeloid disease.

The disease combines, to varying degrees, multisystem inflammatory manifestations, cytopenias and bone marrow abnormalities. Its management therefore requires close collaboration between internal medicine and haematology.

Arsène Mekinian's work addresses the clinical and haematological characterisation of VEXAS, prognostic factors, treatment strategies and the development of tools allowing these strategies to be assessed prospectively.

What establishes his expertise in this field

What he coordinates or directs.

  • First author of the American College of Rheumatology text on the diagnosis and management of VEXAS syndrome, an international expert guidance statement published in Arthritis & Rheumatology 2026;78(3):509-522, DOI 10.1002/art.43287. He is one of the eight equally contributing co-first authors and one of the two corresponding authors, together with the last named author. [Ref. 1]
  • Cofounder and coordinator of FRENVEX, the French study group on VEXAS syndrome, since 2020. The record published by Sorbonne Université Formation Continue is the source that attributes this dual role to him.
  • Head of the constitutive site of the CeRéMAIA reference centre (French reference centre for autoinflammatory diseases and amyloidosis) at Saint-Antoine hospital, whose scope covers acquired inflammatory conditions including VEXAS syndrome. The record of the constitutive site published by the FAI²R network states that this site is coordinated by him; national coordination of the centre belongs to another site.
  • Principal investigator of the phase II trial GFM-AZA-SAID, NCT02985190, on azacitidine in the autoimmune and inflammatory manifestations associated with myelodysplastic syndromes that are dependent on or resistant to corticosteroid therapy, completed on 30 August 2022 according to ClinicalTrials.gov.

His position in the publications cited here. Of the 29 references cited at the end of this page, Arsène Mekinian is first author of 2, last author of 5, and corresponding author of 2. One of the two first positions is a co-first position declared by the article itself, which names eight equally contributing co-first authors. The positions of all 29 references were verified on the MEDLINE record. A thirtieth reference is cited at the end of the page without his signature, the prevalence study published in JAMA in 2023: it is there because the page uses the frequency figure it establishes, and it does not enter any of these totals.

In biomedical research, the first author is generally the person who carried out the work day to day, and the last author the one who designed and supervised the study. The literature describes this as an informal practice rather than a rule: international recommendations leave the order of authors to the authors themselves, and reading seniority from position becomes unreliable as the list of authors grows. [How to read author position](/en/method/author-position/)

An autoinflammatory disease of clonal origin

VEXAS is caused by the acquisition, during life, of a somatic variant of UBA1 within haematopoietic cells. The abnormality arises in bone marrow cells and is not inherited from the parents; it is not passed on to children. The name is an acronym for vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic.

UBA1 encodes the main E1 enzyme that initiates the ubiquitylation of proteins. The gene lies on the X chromosome, which is why the disease is described almost exclusively in men, but the variant is acquired, not transmitted: X-linked here describes the location of the gene, not a mode of inheritance. The guidance statement carries, with 100% agreement, the statement that VEXAS is not a heritable disease: no case of transmission to offspring had been documented at its date of writing. [Ref. 1]

It occurs in the great majority of cases in men over 50, and may combine fever and a biological inflammatory syndrome, skin manifestations, chondritis, ocular or pulmonary involvement and thrombotic events.

From the haematological standpoint, macrocytic anaemia and other cytopenias are frequent. Dysplastic bone marrow abnormalities may be observed, and some patients have an associated myelodysplastic syndrome.

Diagnosis rests on identifying a pathogenic somatic variant of UBA1, sought on a blood or bone marrow sample. Because the abnormality lies in the marrow cells, the disease is inflammatory and haematological from the outset.

How common it is, and in whom

The true frequency of the disease is not known, and the figures that circulate are estimates of different kinds, obtained by different routes. They are reported here as each source writes them.

The guidance statement of the American College of Rheumatology cites, from an American study of a single health system in which Arsène Mekinian did not take part, a prevalence of pathogenic UBA1 variants of 1 individual in 13,591, and derives from it, in its introduction, an order of magnitude close to one million people worldwide, while writing that complete worldwide data are lacking. That extrapolation rests on that one study. [Ref. 1]

The VEXAS page of the FAI²R network, in its September 2024 version written by Prof. Sophie Georgin-Lavialle, gives a different order of magnitude for France: about one man in 4,000 over the age of 50, about 300 patients counted in France at that date, at least 1,000 worldwide, 95% of them men, with a youngest reported patient aged 23. These are estimates by a national reference network, not a measured prevalence.

Two things follow for a reader. The disease is rare but probably under-diagnosed: the guidance statement writes that an undetermined number of patients remain undiagnosed or wrongly diagnosed. [Ref. 1] And forms in women exist, in particular where there is loss of one X chromosome, but they are poorly documented; the search for X monosomy has a particular value in women. [Ref. 1]

How the diagnosis is made

This is the part of the subject on which a clinician most often needs a precise answer, and the guidance statement devotes seven of its 28 statements to the search for the UBA1 variant. What follows summarises its positions; it does not replace reading the text itself, which is freely available. [Ref. 1]

When to suspect it. Persistently raised inflammatory markers, together with skin, ocular, pulmonary or cartilaginous manifestations, or with a cytopenia, should raise the diagnosis. The pre-test probability is higher in men, after 50, and when acute-phase proteins are high. Macrocytic anaemia is the most frequent haematological abnormality without being constant, and macrocytosis may be isolated and precede everything else. [Ref. 1]

Vacuoles are suggestive, not diagnostic. Vacuoles in marrow erythroid or myeloid precursors are also found in deficiencies, copper deficiency in particular, and after chemotherapy. Their absence does not justify abandoning the test when suspicion persists. The guidance statement cites, for the combination of marrow vacuoles and relevant clinical signs, a positive predictive value of 87.5% for a pathogenic UBA1 variant; genetic testing remains necessary for the diagnosis in every case. [Ref. 1]

Where the variant is. Most typical presentations are due to a missense or splice-site variant in exon 3 of UBA1, around methionine 41. That statement was voted with 100% agreement. Sanger sequencing of exon 3 also covers the Ser56 residue, whose variants give a more moderate inflammatory phenotype. A test targeted on exon 3 remains an option, provided one knows that it misses variants outside exon 3. [Ref. 1]

The technique sets the detection threshold. The guidance statement indicates that Sanger sequencing detects little below a variant allele burden of about 20%, whereas targeted next-generation sequencing, droplet digital PCR and high-resolution melting analysis go down to around 1%. [Ref. 1]

Ongoing treatment can distort the result. Treatment in progress, hypomethylating agents in particular, lowers the mutational burden and may falsify the test. This is the least consensual of the 28 statements, at 87% agreement, and one of the most consequential: the timing of the sample relative to treatment is part of the request. [Ref. 1]

If the blood test is negative and suspicion persists, the panel indicates checking that the whole gene was covered, then considering a marrow sample analysed with the method most sensitive to low allele burdens. The variant is myeloid and absent from lymphocytes. It has been detected in skin or nails, but possibly through blood contamination, since fibroblast culture does not find it. [Ref. 1]

One laboratory detail with practical weight. C-reactive protein is a more reliable marker of inflammation than the erythrocyte sedimentation rate in this disease, because the haematological abnormalities raise the second independently of inflammation. [Ref. 1]

Imaging does not make the diagnosis. A multicentre retrospective study of 125 18F-FDG-PET/CT scans in 66 patients with genetically confirmed VEXAS, of which he is 31st of the 32 named authors, found abnormal uptake most often in bone marrow (83%) and lymph nodes (53%), with pulmonary involvement in 38% and vascular involvement in 11%. The number of organs with abnormal uptake was higher in active disease than in remission (median 2 versus 1), but 90% of the scans performed during presumed clinical remission still showed abnormalities, bone marrow in 57%. Its authors conclude that the findings are frequent but non-specific, that diagnostic usefulness is limited, and that the value of the examination may lie in monitoring rather than in diagnosis. [Ref. 29]

From systemic inflammation to clonal haematopoiesis

VEXAS has profoundly changed the understanding of certain inflammatory diseases of adults.

It shows that severe systemic inflammation can result directly from an acquired somatic abnormality in a haematopoietic clone.

This relationship between myeloid clone and inflammation leads beyond the traditional separation between autoinflammatory disease and haematological disease.

It raises several questions: which determinants of the clone influence the inflammatory phenotype? Why do some patients develop an associated haematological disease? And should the priority be to control the inflammation or to target the clone that sustains it?

Two pieces of work in which he takes part show how the clone and the tissue lesion are connected. Paired sequencing of bone marrow and of skin biopsies in eight patients with VEXAS found the same loss-of-function UBA1 variant in both samples, which supports the view that the dermal infiltrate derives from the pathological myeloid clone; he is 19th of the 27 named authors. [Ref. 16] Bulk RNA sequencing of lesional skin in 20 patients, six of them with VEXAS, found transcriptomic profiles in VEXAS and in myelodysplasia cutis that are closely related and dominated by interferon signatures and by tumour necrosis factor and interleukin 1β pathways; he is 9th of the 22 named authors. That study rests on six VEXAS samples, and its authors present the therapeutic implication as a hypothesis, not as a result. [Ref. 22]

The anaemia of the disease has also been re-examined. A 2026 study combining analyses in patients with in vitro base editing of UBA1 in CD34+ cells, of which he is 18th of the 26 named authors, reports that mature red cells show no ubiquitylation defect and that marrow erythroblasts carry no UBA1 variant beyond the basophilic stage, edited erythroid precursors dying early in differentiation. Its authors propose that the anaemia of VEXAS should be regarded as a mosaic erythroblastopenia, whose severity depends on the size and quality of the UBA1 wild-type compartment. This is a mechanistic proposal drawn from patient samples and from a cell model; it has no therapeutic consequence established at this stage. [Ref. 26]

Characterising phenotypes and prognosis

The French VEXAS network made it possible to assemble multicentre cohorts quickly and to define the spectrum of the disease more precisely. In 2022, the description of 116 French patients referred to a multicentre registry between November 2020 and May 2021 established this clinical spectrum: skin lesions in 83% of patients, non-infectious fever 64%, weight loss 62%, pulmonary involvement 50%, ocular signs 39%, relapsing chondritis 36%, venous thrombosis 35%, lymphadenopathy 34%, arthralgia 27%. A haematological disease was present in 58 cases, half the series: 58 myelodysplastic syndromes and 12 monoclonal gammopathies of undetermined significance, every patient with a gammopathy also having a myelodysplastic syndrome. The UBA1 variants were distributed as p.M41T (45%), p.M41V (30%), p.M41L (18%) and splice-site variants (7%). Arsène Mekinian is 73rd of the 73 named authors of this series. [Ref. 2]

This work showed substantial heterogeneity between patients, both clinically and prognostically. After a median follow-up of three years, 18 patients had died, that is 15.5%, nine of them from infection. An unsupervised analysis separated three groups of patients, whose five-year survival probabilities were 84.2%, 50.5% and 89.6%, the lowest survival corresponding to the group with an underlying myelodysplastic syndrome. The p.Met41Leu variant was associated with a better prognosis. [Ref. 2]

The study of neurological manifestations, published in 2025, gives the order of magnitude of the registry at its date: of 291 patients included in the French VEXAS registry between November 2020 and March 2023, 17, that is 6%, had central or peripheral nervous system involvement. [Ref. 10] The numbers in the different analyses are successive numbers and are not cumulative.

The aim of this work is to develop a prognostic stratification capable of guiding the intensity and the nature of treatment strategies.

Organ involvement, one system at a time

Beyond the founding series, several studies describe one organ at a time. Each rests on small numbers and none is prospective; the figures are given here with their denominators, because in a rare disease the denominator is the information.

Lung. The 2023 pulmonary series, of which he is 4th of 43 named authors, analysed the chest computed tomography scans available in the French cohort, which that article describes as comprising 114 patients: 51 patients had a scan that could be reviewed, and 45 of them, that is 39% of the cohort, had pleuropulmonary abnormalities attributed to VEXAS after adjudication, with ground-glass opacities in 87% of them and pleural effusion in 53%. Those 45 patients were men in 95% of cases, median age 67.0 years at symptom onset; 44% reported dyspnoea and 40% cough; all had lung opacities, with consolidation in 49%, septal lines in 51% and reticulation in 38%. Most improved on prednisone, but usually above 20 mg a day. The authors note that clinical features, laboratory features and median survival did not differ from the rest of the cohort, which leads them to write that pleuropulmonary involvement may have been under-diagnosed elsewhere in the cohort. [Ref. 6] That article dates the French cohort at 114 patients over the period from November 2020 to May 2021, whereas the dermatological series gives 116 patients over exactly the same period. Both articles carry his name; both figures are reported here as each article writes them, and the discrepancy is not adjudicated. [Ref. 2, ref. 6]

Skin. A multicentre retrospective study of the cutaneous manifestations of 59 patients was published in 2023, of which he is 4th of the 34 named authors; the MEDLINE record of that article carries no abstract, and it is cited here by its title, its type and his author position, without any figure. [Ref. 18]

Cartilage. The comparison between idiopathic relapsing polychondritis and VEXAS-associated relapsing polychondritis, a French series of 95 patients of which he is 28th of the 28 named authors, separates the two situations clearly. Compared with the 40 patients with idiopathic disease, the 55 with VEXAS-associated disease were men in 96% of cases against 30%, older at diagnosis, 66 against 44 years, and had more fever (60% against 10%), skin lesions (82% against 20%), ocular involvement (57% against 28%), pulmonary infiltrates (46% against 0%) and cardiac involvement (11% against 0%), with a higher median C-reactive protein (64 mg/L against 10 mg/L). A myelodysplastic syndrome was present in 75% of the VEXAS group and in none of the other. Remission was obtained in 27% against 90%, and six patients (11%) died in the VEXAS group after a median follow-up of 37 months, against none in the other group after 92 months. [Ref. 17]

Heart. Non-ischaemic cardiac involvement was looked for in a French multicentre series of which he is 9th of the 11 named authors, run from 1 November 2020 to 30 June 2024: of 299 patients with VEXAS, 8 (2.7%) had a cardiac manifestation, and 5 (1.7%) a non-ischaemic one and were analysed, all men, median age 65 years, with a median interval of 2.3 years between the first symptoms of VEXAS and the cardiac event. Pericarditis, myocarditis, heart failure, complete atrioventricular block and cardiac AA amyloidosis were each observed in one or two patients. Four of the five died; only one of those deaths is not labelled non-cardiovascular in the article's table, a cardiac arrest at induction of anaesthesia. Five patients is a case series, not a frequency estimate. [Ref. 20]

Kidney. A clinicopathological description of kidney features in VEXAS was published in 2025, of which he is 34th of the 37 named authors; its MEDLINE record carries no abstract, and it is cited here without any figure. [Ref. 25]

Vessels. A case-control study published in 2026, of which he is 21st of the 22 named authors, compared polyarteritis nodosa associated with VEXAS or with chronic myelomonocytic leukaemia against primary polyarteritis nodosa, each case matched to three controls. Twenty-three secondary cases were included, 12 VEXAS and 11 leukaemia-associated. VEXAS-associated disease occurred later (73 years against 54), with skin involvement in 100%, orchitis in 67% and ocular involvement in 58%, was less likely to reach remission (adjusted odds ratio 0.15, 95% CI 0.02 to 1.00) and relapsed more (adjusted hazard ratio 3.24, p = 0.043); mortality was 50%, against 17% in primary disease. Twelve patients is a very small group, and the confidence interval on remission includes 1. [Ref. 28]

Nervous system. The 2025 series of neurological manifestations, 17 patients of 291, is described above. A published correction to that article appeared in 2025 and is listed with the references. [Ref. 10]

The haematological side, and the myelodysplastic syndrome question

Four statements of the guidance statement deal with the marrow. A bone marrow aspirate is indicated in a patient with VEXAS who has a cytopenia, to rule out an associated haematological disease, and the text recommends it before starting a disease-modifying treatment. It is accompanied by karyotype and by next-generation sequencing for coexisting somatic mutations; when the diagnosis was made by Sanger sequencing or by a test targeted on UBA1, the myeloid panel should still be performed. [Ref. 1]

Interpreting the marrow is the difficult point. Signs of dysplasia are frequent without the 2022 WHO and ICC criteria for myelodysplastic syndrome being met, and vacuolation alone is not enough to conclude dysplasia, since it belongs to the disease itself. The guidance statement reports that the literature places the frequency of associated myelodysplastic syndrome between one third and one half of patients at diagnosis, but that with a stricter definition — requiring transfusion-dependent anaemia and a platelet count below 100 G/L in addition to morphological or genetic abnormalities — the frequency falls to 18%. Depending on the definition used, the same patient does or does not have a myelodysplastic syndrome. [Ref. 1]

The text settles two frequent situations. Isolated vacuoles, with no other WHO-defined dysplastic feature, a normal karyotype and no other somatic mutation, should not lead to a diagnosis of myelodysplastic syndrome, but to closer surveillance. Conversely, a blast count of 5 to 20%, which is rare, or a cytogenetic abnormality that defines a myelodysplastic syndrome, do establish it. The clonal landscape is dominated by DNMT3A and TET2, the karyotype is normal in more than half of cases, and mutations in myeloid driver genes are present in close to 60% of patients, with or without a myelodysplastic syndrome. [Ref. 1]

A 2026 letter from the French group, of which he is 25th of the 26 named authors, addresses the characteristics and the prognostic significance of myelodysplasia-related features in VEXAS. Its MEDLINE record carries no abstract; it is cited here by its title, its type and his author position, without any figure. [Ref. 27]

The American College of Rheumatology guidance statement

Arsène Mekinian is first author of the international American College of Rheumatology guidance statement on the diagnosis and management of VEXAS syndrome, published online on 11 August 2025 and paginated in the March 2026 issue of Arthritis & Rheumatology (2026;78(3):509-522). The text has 58 named authors; he is the first of them, one of the eight co-first authors designated as having contributed equally, and one of the two corresponding authors together with the last named author. The full text is freely available in PubMed Central, and the ACR announced it in a press release. [Ref. 1]

The ACR classifies this text among its guidance statements, a format it distinguishes from its clinical practice guidelines. The distinction is not cosmetic: a guidance statement is the format the college uses when the evidence does not support a formal graded recommendation, and every statement in this one is conditional. The panel brought together 57 members from 18 countries and nine disciplines, without patient representatives. The stated method was a literature review from January 2020 to August 2024, a two-day workshop in Paris in May 2024, then a single round of anonymous electronic voting with a consensus threshold set at 80% agreement. The workshop was supported by the FAI²R network; the text carries the endorsement of EuroBloodNet and of the ISSAID society. [Ref. 1]

It carries 28 statements, in four domains: six on clinical and laboratory features, seven on the search for the UBA1 variant, four on the diagnosis of an associated myelodysplastic syndrome, 11 on outcome, prognosis and management. All crossed the threshold, with agreement rates from 87% to 100% for 52 or 53 voters depending on the statement. [Ref. 1]

On prognosis, the text reports a median survival from symptom onset of about ten years, thromboembolic disease in about half of patients, and 33 transplanted patients published at its date of writing, 27 of them alive. [Ref. 1]

On treatment, the statements set out the following positions. Corticosteroids are usually necessary, at moderate to high doses — the article cites the order of 10 to 25 mg of prednisone per day or more — with slow tapering to the lowest effective dose, and corticosteroid-free remission is unusual in VEXAS, unlike many rheumatological diseases. Treatments targeting inflammatory pathways, Janus kinase inhibitors and interleukin 6 inhibitors, are presented as more effective than methotrexate, azathioprine and B-cell-directed treatments; among Janus kinase inhibitors, ruxolitinib may be preferred, in part because it can be titrated over a wider dose range, the article indicating 10 mg twice a day to 20 mg twice a day. Interleukin 1 inhibitors may be considered, with the reservation of severe injection-site reactions reported with anakinra and of the small amount of data on canakinumab. Clone-reducing treatments such as azacitidine may be effective in some patients. Allogeneic haematopoietic stem cell transplantation may be curative in selected patients. [Ref. 1]

One statement excludes a treatment. B-cell-depleting treatments are said to be contraindicated: they are not effective in VEXAS and may worsen the B-cell lymphopenia associated with the disease. This is directly actionable for a practitioner receiving a patient already treated under another diagnosis. [Ref. 1]

A single statement, voted with 98% agreement, brings together three preventions to consider in every patient: prophylaxis of opportunistic infections, Pneumocystis and alphaherpesviruses in particular; prevention of thromboembolic disease, present in about half of patients, venous above all but also arterial, with thromboprophylaxis in at-risk situations such as hospitalisation, unless contraindicated; and limitation of the adverse effects of prolonged corticosteroid therapy, with attention to bone and to cardiovascular risk. Two practical points are added in the body of the text: a skin lesion that does not respond to treatment warrants considering a biopsy to rule out non-tuberculous mycobacterial infection before increasing immunosuppression, and any new or recurrent symptom calls for a broad differential diagnosis, since infections, thromboembolic disease and treatment complications are frequent and can imitate a flare. [Ref. 1]

What the guidance statement says it does not know

The site's approach is to write the negative results and the limits as carefully as the positive ones. The guidance statement does the same, and the following is its own account of what it could not settle. The final statement, voted with 100% agreement, is that clinical trials will be needed to determine which treatment is the most effective. [Ref. 1]

  • No randomised controlled trial had been conducted at the date of writing. Most published data are isolated observations or small series, and prospective cohorts are lacking.
  • No medicine has established efficacy beyond corticosteroids, in the authors' own terms. The hierarchy between classes rests on retrospective series and is therefore exposed to indication bias.
  • When to start treatment is not fixed, although the disease is a progressive clonal process of the marrow: the authors indicate that this timing may weigh as much as the choice of molecule.
  • The duration and choice of anticoagulation after a thromboembolic event are not defined.
  • The candidates for hypomethylating agents are not identified.
  • The timing of allogeneic transplantation and the selection of candidates are designated as the most difficult question.
  • At the date of the text there was neither a standardised definition of treatment response nor a validated activity score. This is one of the four reasons the authors give for judging the quality of the evidence low, together with selection bias, follow-up that is often too short, and imprecision in the diagnosis of associated myelodysplastic syndrome.
  • The true frequency of the disease is not known, as set out above.

Two further limits belong to the method itself and are stated in the text: no formal systematic review was carried out, the panel included no patient representatives, and voting took place in a single round. [Ref. 1]

Two complementary treatment approaches

The particular nature of VEXAS leads to two broad treatment strategies being considered. No medicine holds a marketing authorisation in this indication.

The first is to control the inflammation, in particular with corticosteroids and with various treatments targeting cytokine or intracellular signalling pathways.

The second aims to act on the haematopoietic clone, in particular with treatments used in haematology or, in selected patients, with allogeneic haematopoietic stem cell transplantation.

This distinction between inflammatory targeting and clonal targeting is one of the main current lines of therapeutic research. A 2024 review of which he is 3rd of the four named authors sets out the same division, and adds a third element that is easily forgotten: supportive care, that is red cell and platelet transfusion, erythropoiesis-stimulating agents, thromboprophylaxis and anti-infective prophylaxis. [Ref. 19]

Targeted therapies

The French and international cohorts have assessed several treatment classes, in particular JAK inhibitors, IL-6 inhibitors and IL-1 inhibitors. The series of 110 patients treated with targeted therapies, published in 2024 by the FRENVEX group, covered 194 lines of treatment and gave, at three months, an overall response rate of 24% on Janus kinase inhibitors, 32% on interleukin 6 inhibitors, 9% on interleukin 1 inhibitors and 0% on anti-TNF agents; at six months, the rates were 30% and 26% for the first two classes. He is 72nd of the 73 named authors. [Ref. 8]

He is also a co-author, 6th of 25 named authors, of the international retrospective analysis of 30 patients treated with various Janus kinase inhibitors, whose authors report, on ruxolitinib, clinical remissions and a reduction in corticosteroid therapy in most patients. [Ref. 4] In 2026, the international comparison of anakinra and canakinumab in 47 men, of whom 44 received anakinra and nine canakinumab, showed an overall response at three months of 22% on anakinra and 78% on canakinumab, with a median duration of treatment maintenance of 54 months for canakinumab at 300 mg per month against 1 month for anakinra; he is 36th of the 36 named authors. [Ref. 12]

These studies have made it possible to identify signals of efficacy that vary by class and by patient, but they rest mainly on observational data. They therefore do not yet allow an optimal treatment sequence to be established.

Reducing corticosteroid dependence and achieving durable control of inflammation are major objectives, given in particular the infectious risk associated with the disease and with its treatment, a risk that was highest on Janus kinase inhibitors in the study of serious infections from the French registry. [Ref. 7]

Targeting the haematopoietic clone

Azacitidine is another approach, studied particularly in patients with associated myeloid abnormalities.

The French series have reported, in some patients, inflammatory and haematological improvement, sometimes associated with a marked reduction in the UBA1 mutational burden. As early as 2022, a first series of 11 patients treated with azacitidine for VEXAS syndrome with myelodysplastic syndrome reported a clinical response in five of them, that is 46%, with response durations of 6, 8, 12, 21 and 27 months, two of these responses still ongoing at the date of the analysis; he is 17th of the 17 named authors. [Ref. 3] In 2025, the analysis of 88 patients treated with azacitidine, of whom 80% met the 2022 WHO criteria for myelodysplastic syndrome, showed an inflammatory response in 41% of patients at six months and 54% at twelve months, transfusion independence in 65% and platelet improvement in 77%, with a molecular response, defined as a fall of at least 25% in the variant allele frequency of UBA1, in 65% of patients; the variant allele frequency fell below 2% in 43% of cases, and infections were reported in 34% of patients. He is 53rd of the 54 named authors. [Ref. 11]

He is also first author and corresponding author of a prospective phase II trial of azacitidine in corticosteroid-dependent or refractory systemic autoimmune and inflammatory disorders and VEXAS syndrome associated with myelodysplastic syndrome or chronic myelomonocytic leukaemia, published in 2022 in Leukemia as a letter, with no indexed abstract: the trial is cited here by its title, its type and his author position, without any figure. [Ref. 5]

How long such a response can last is documented by a single published observation, of which he is 12th of the 13 named authors: one patient received azacitidine for 105 cycles over ten years, with complete remission of skin symptoms, withdrawal of corticosteroids, and a UBA1 clone that became undetectable with a 1% sensitivity assay. He then relapsed in December 2019 with an expanding UBA1 clone, did not respond to shortened azacitidine cycles or to ruxolitinib, and died of a pulmonary infection in December 2022. This is one patient. It shows what is possible and, in the same report, that the control obtained was not definitive. [Ref. 21]

These observations strengthen the hypothesis that controlling the clone can bring about control of the inflammation. They nevertheless require prospective confirmation and a better definition of the patients likely to benefit. The guidance statement writes explicitly that it is not yet known whether patients with an associated myelodysplastic syndrome are the best candidates or whether the benefit extends more widely. [Ref. 1]

Allogeneic haematopoietic stem cell transplantation

Transplantation is currently the approach capable of achieving eradication of the pathological clone, since the variant is restricted to the haematopoietic compartment, but its use is limited by its toxicity and by the characteristics of the patients concerned. The guidance statement reports 33 transplanted patients published at its date of writing, 27 of them alive, that is 81.8%, and eleven with post-transplant molecular data showing complete eradication of the UBA1 variant, most often on a follow-up of less than twelve months. [Ref. 1] Arsène Mekinian is a co-author, 24th of 33 named authors, of the European multicentre EBMT study on transplantation in VEXAS syndrome. [Ref. 9]

The central question is therefore the selection of candidates and the optimal timing of transplantation. The guidance statement names as candidates patients with progressive and severe bone marrow failure, transfusion-dependent in particular, or with refractory inflammation. [Ref. 1]

Age, comorbidities, the severity of the inflammatory disease, haematological involvement and previous treatments, including prolonged corticosteroid exposure, must all be brought into this assessment.

Its exact place in the treatment strategy remains to be defined prospectively.

Anaemia, transfusion and supportive care

Anaemia in VEXAS is usually non-regenerative, and red cell transfusion dependence has been reported in 23% to 83% of patients depending on the series. Outside transfusion, no treatment of this anaemia is validated.

A FRENVEX multicentre retrospective study published in 2025, of which he is 13th of the 13 named authors, assessed erythropoiesis-stimulating agents and luspatercept in 45 anaemic patients (haemoglobin below 10 g/dL) treated between 2020 and 2024 in eight centres, 82% of whom had an associated myelodysplastic syndrome and 67% of whom were transfusion-dependent. An erythroid haematological improvement was obtained in 14 of 45 patients, that is 31%, at week 16. Response was associated with a lower baseline endogenous erythropoietin level (median 33 IU/L in responders against 300 IU/L in non-responders, p = 0.01) and with non-classical UBA1 variants; among patients with an erythropoietin level above 200 IU/L, only 2 of 18 responded. Loss of response reached 19% at one year and 38% at two years in the results section of the article, whose discussion writes 36.3% at two years instead, without the internal gap being explained. Eight patients were switched to luspatercept after failure of an erythropoiesis-stimulating agent, four of whom responded at week 16: a number too small to support any conclusion about that drug, and its authors say so. [Ref. 23]

Two things are worth retaining. Response is a minority outcome, and a baseline erythropoietin level identifies in advance most of the patients who will not respond. Neither observation comes from a controlled trial.

Infection, thrombosis and prevention

Infection is the leading identified cause of death in the published series, and it is the point on which the available data are most directly actionable.

Infectious complications were the subject of a dedicated study from the French registry, published in 2024, which brought together 74 patients and 133 infectious episodes, of which 59% were pulmonary, with microbiological confirmation in 76% of cases. Risk was associated with age above 75 years, the p.Met41Val variant and arthralgia, and it was highest on Janus kinase inhibitors, with a hazard ratio of 3.84. Twenty-seven patients died during follow-up, 15 of them from an infection. He is 42nd of the 45 named authors. [Ref. 7]

A systematic review published in 2025, of which he is 4th of the 11 named authors, pooled 57 studies and 813 patients. Infections were frequent, severe in 40 to 60% of cases and fatal in 6 to 15%. Pulmonary infections were the most common, followed by cutaneous infections and bacteraemia. Opportunistic pathogens — Pneumocystis jirovecii, Legionella pneumophila, non-tuberculous mycobacteria, varicella zoster virus — were frequently reported even in patients not receiving immunosuppressive therapy, which its authors read as suggesting an intrinsic immune dysfunction of the disease itself. They propose considering co-trimoxazole prophylaxis (or another Pneumocystis prophylaxis, atovaquone or pentamidine) and valaciclovir in patients at high risk — those on immunosuppressive therapy, and those with lymphopenia, the p.Met41Val variant, or previous severe or recurrent infection — posaconazole in neutropenic patients taking azacitidine, and vaccination against pneumococcus, varicella zoster virus, influenza and SARS-CoV-2. The review's own conclusion is that these preventive strategies remain to be evaluated prospectively. [Ref. 24]

Thromboembolic disease concerns about half of patients, venous above all but also arterial, and the guidance statement includes its prevention in the single statement on prevention, while stating that the duration and choice of anticoagulation after an event are not defined. [Ref. 1]

Towards structured therapeutic trials

One of the historical limitations of research in VEXAS was the absence of standardised definitions of disease activity and of treatment response.

Recent international work in which Arsène Mekinian takes part has developed a consensus definition of flare as well as a disease-specific activity index, the VEXAS-DAI. He is 6th of 14 named authors of the consensus definition of flare, established by a modified Delphi method with an agreement threshold of 75% and published in 2026, which defines a flare as an active inflammatory manifestation requiring an increase in corticosteroid therapy and distinguishes three categories. [Ref. 13] He is 10th of 26 named authors of the VEXAS-DAI activity index, built by the same method on 158 test cases, which assesses twelve domains for a total score of 0 to 40. [Ref. 14]

These tools should make it possible to standardise the assessment of patients and the endpoints of future trials. Their prospective validation is still under way, which means that their performance has not yet been measured. [Ref. 13, ref. 14]

The phase II PAXIS trial is also assessing pacritinib against placebo in VEXAS syndrome (NCT06782373). He is 5th of 15 named authors of the protocol article, published in 2026: that article describes the design of the trial, it reports no results from it. [Ref. 15]

FRENVEX and multidisciplinary organisation

Care and research are developed in the department of internal medicine and clinical immunology at Saint-Antoine Hospital, AP-HP, Sorbonne Université, a constitutive site of CeRéMAIA, the reference centre for autoinflammatory diseases and inflammatory amyloidosis. The FAI²R network page for that site states that this constitutive site is coordinated by Prof. Arsène Mekinian; national coordination of the centre lies with another site. The same page lists the site's referring physicians in internal medicine, rheumatology, neurology, cardiology, haematology and intensive care, as well as the referring physicians in nephrology, dermatology, respiratory medicine, obstetrics and gynaecology and ophthalmology at the associated sites. It is this multidisciplinary organisation that the guidance statement holds as the framework for management, with referral to an expert centre where one exists. [Ref. 1]

The department takes part in the FRENVEX network, the French VEXAS study group. The acronym appears in the published literature under several forms — French VEXAS study group, French VEXAS registry, FRENVEX group — and two peer-reviewed articles, two only, gloss it. The 2025 azacitidine study writes "FRENVEX (French VEXAS study group)" in its abstract. [Ref. 11] The 2025 FRENVEX study of anaemia writes, in its full text, "eight centers of the French Vexas (FRENVEX) group". [Ref. 23] The second wording places the bracket immediately after "French Vexas", which points to an acronym built on FRENch and VEXas. That is as far as the public sources go: none spells the acronym out letter by letter and none gives an official form, so this page proposes none. Published inclusions begin in November 2020, within weeks of the first description of the disease. The full text of the 2024 study of serious infections states that it was conducted in the forty centres of the group; that figure holds for that study and at its date. [Ref. 7]

Why a national collection is necessary follows from the way these patients reach hospital. One patient enters internal medicine for unexplained fever, another rheumatology for chondritis, dermatology for a rash, haematology for macrocytic anaemia, nephrology for renal failure. No single door sees the whole picture, and no single centre gathers enough observations to describe the disease.

His Sorbonne Université Continuing Education profile describes Arsène Mekinian as co-founder and coordinator of FRENVEX and links him to the immuno-haematology and VEXAS group at Institut Imagine; the affiliation carried on the ACR guidance statement also mentions Institut Imagine and the reference centre. That profile is a university staff page, whose content is in practice supplied by the person concerned: it establishes that the statement exists and where it is published, not an act of foundation. None of the group's publications names a coordinator, and no consulted source gives a year of foundation.

Several things about the network are simply not public, and it is better to say so than to let the gap be filled with invention. At 14 September 2026, none of the following is published: the expansion of the acronym letter by letter, the date the group was constituted, its legal status, its sponsor, its governance, the regulatory formalities of the collection, its data access policy, an up-to-date number of centres, or a current number of patients. The German Multicenter, Interdisciplinary National VEXAS Registry (NCT06377462) is unrelated to FRENVEX and should not be attributed to it.

This network brings together internal medicine, haematology and various organ specialties, and provides the support for the French and international cohorts and collaborative work cited on this page. Requests for advice between practitioners go through the department secretariat.

What remains to be shown

Despite the rapid progress made since VEXAS was identified, the optimal treatment strategy is not established. The main uncertainties concern the sequence of treatments, the respective place of inflammatory targeting and clonal targeting, the indications for transplantation and the biomarkers allowing response to be predicted.

No medicine holds a marketing authorisation in this indication. No randomised controlled trial had been published in VEXAS syndrome at the date of the protocol article of the PAXIS trial, in February 2026; of that trial, only the protocol has been published. [Ref. 15]

Most of the available treatment data still rest on retrospective cohorts. The treatment data of which he is a co-author are retrospective and multicentre, without a comparator arm or randomisation. The hierarchy between treatment classes that emerges from them remains open to indication bias, since the choice of treatment depended on the physician and on the patient's condition. The response rates observed at three months in the series of 110 patients, between 24% and 32% for the two most effective classes, leave most patients without a complete or partial response at that point. [Ref. 8] The comparison between anakinra and canakinumab rests on nine patients treated with canakinumab, a number that does not allow a firm conclusion. [Ref. 12] For azacitidine, the series of 88 patients is retrospective, and its authors themselves conclude that larger prospective trials are needed. [Ref. 11] For anaemia, the erythroid response rate at week 16 was 31%, and the luspatercept arm rests on eight patients. [Ref. 23]

The place of allogeneic haematopoietic stem cell transplantation is not established. The timing of transplantation and the criteria for selecting candidates remain the most difficult question, because of age, comorbidities and prolonged corticosteroid exposure. [Ref. 1, ref. 9] The diagnosis of associated myelodysplastic syndrome depends on the definition used, which makes the proportion of patients concerned vary twofold. Until 2026 there was no validated activity score and no consensus definition of flare; the definition of flare and the VEXAS-DAI index have just been proposed and their prospective validation is under way. [Ref. 13, ref. 14]

The organ series are small. Non-ischaemic cardiac involvement rests on five patients, the VEXAS-associated polyarteritis nodosa comparison on twelve, the transcriptomic study of skin on six VEXAS samples, and the long-term azacitidine observation on one patient. [Ref. 20, ref. 28, ref. 22, ref. 21] The 18F-FDG-PET/CT study concludes against a diagnostic use of that examination. [Ref. 29] Four works cited here have neither an indexed abstract nor an exploited open-access full text, and no figure is given for them: the cutaneous series of 59 cases [Ref. 18], the kidney description [Ref. 25], the 2026 letter on myelodysplasia-related features [Ref. 27] and the 2022 phase II trial [Ref. 5]. The gap between 116 and 114 patients is not adjudicated. Two studies describe the same French cohort over the same inclusion window, November 2020 to May 2021, and do not give the same number. The 2022 dermatological series writes in its methods that 116 patients with VEXAS syndrome were referred to a French multicentre registry between those two dates. [Ref. 2] The 2023 pulmonary series writes in its methods that 114 patients were included in the French cohort of VEXAS syndrome between the same dates, and then reports 45 patients, that is 39%, a proportion consistent with a denominator of 114. [Ref. 6] The only difference available is therefore the verb each uses, "referred to a registry" on one side, "included in the cohort" on the other. Neither article comments on the gap, neither cites the other's figure, and nothing in the two texts allows one to say which is the exact denominator, or whether two referred patients were not included. Both carry his name. Both figures are reproduced here as each article writes them, and the gap is not adjudicated. [Ref. 2, ref. 6]

Finally, the true frequency of the disease is not known, the disease has been described almost exclusively in men over fifty, and forms in women remain poorly documented.

The priorities are now to develop prospective controlled trials, to validate the tools for activity and response, and to build an individualised treatment strategy integrating the inflammatory, haematological and molecular phenotype at the same time.

Sources and method

Every publication listed below was checked on 14 September 2026 on its full MEDLINE record, opened individually in the PubMed `?format=pubmed` display. Author positions are counted on the `FAU` fields alone, that is, on the named authors of the article. The `CN`, `FIR` and `IR` fields, which carry the name of a collaborative group and the list of its collaborators, never count as an author position; bibliographic aggregators that treat them as authors give wrong positions for this corpus.

Two spellings of the surname appear in MEDLINE, with and without an acute accent, sometimes for the same author in different journals. A search that assumes only one spelling misses part of the corpus and can wrongly conclude that he is not an author of a given article. On this page three works carry the accented spelling in the `FAU` field: the 2022 ruxolitinib series [Ref. 4], the 2025 azacitidine study [Ref. 11] and the 2024 cardiac series [Ref. 20]. The search used here was `VEXAS AND (Mekinian[All Fields] OR Mékinian[All Fields])`, which returned 43 records on 14 September 2026, of which 42 carry him as a named author; the references below are the subset cited on this page.

Every figure on this page comes from the MEDLINE record, from the open-access full text where one exists, or from the named institutional source. Where a record carries no abstract and no open-access full text, the work is cited by its title, its type and his author position, with no figure. Figures, tables and passages from articles under copyright are not reproduced.

References

Published work he has taken part in, on this subject.

General information only. This page does not replace individual medical advice. Updated: 19 September 2026.