Therapeutic options in VEXAS syndrome: insights from a retrospective series

TO THE EDITOR:

VEXAS syndrome (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) has been recently described as the association of severe autoinflammatory manifestations and myeloid dysplasia.¹  This syndrome is due to a somatically acquired mutation affecting methionine 41 of the E1-ubiquitin ligase UBA1, leading to the expression of a catalytically impaired isoform that drives inflammation. As this gene is located on the X chromosome, only men appear to be affected by this syndrome. There is currently no published data about the efficacy of the different therapeutic options proposed to these patients; this prompted us to analyze the UBA1 gene in a cohort of patients from the Lyon University Hospital whose clinical phenotypes were compatible with the VEXAS syndrome in order to describe the response to therapy in these patients.

We identified 19 male patients with myeloid dysplasia and autoinflammatory disease such as relapsing polychondritis or Sweet syndrome. DNA extracted from blood (n = 9) or bone marrow (n = 10) samples was analyzed by polymerase chain reaction and high-resolution melting, followed by Sanger sequencing (primer sequences are presented in the supplemental Methods, available on the Blood Web site). Among these 19 patients, 9 (47%) had a mutation leading to the substitution of the methionine 41 of UBA1. Two additional patients had a newly described mutations of the splice motif at the junction of intron 2 and exon 3 (c.118-2A>C and c.119-1G>C). The analysis of the UBA1 messenger RNA found that in both patients, an alternative acceptor splice site located in exon 3 was used, resulting in a UBA1 protein lacking 4 amino acids, including methionine 41. We hypothesize that the translation is then initiated at the methionine 67, leading to a catalytically impaired shortened isoform (UBA1c), but in the absence of cells from these 2 patients, we were unable to confirm this at the protein level (supplemental Figure 1). Numerous vacuolized myeloid precursors were found on their bone marrow smears (supplemental Figure 2). Altogether, 11 of the 19 patients tested (58%) were confirmed to have VEXAS syndrome.

The clinical characteristics of the 11 VEXAS patients are described in Table 1 and supplemental Table 1, and individual data are presented in supplemental Table 2. The median age at disease onset was 66 years (range, 47-83 years). As expected, most patients had fever (91%), skin involvement (100%), and arthritis or arthralgia (100%). All patients had an elevated serum concentrations of C-reactive protein (median, 114 mg/L; range, 16.7-205), and most of them had an increased MCV (>100 fL for 7 patients [64%]). As compared with patients who had VEXAS, the 8 patients without UBA1 mutation had significantly less frequent arthralgia/arthritis and chondritis.

Six of them had a formal diagnosis of MDS according to the latest World Health Organization criteria, among whom 5 had refractory cytopenia without excess of blasts and one had a type 1 refractory anemia with excess of blasts. Interestingly, 73% of patients (8/11) had hypercellular bone marrow, and 6 of the 9 patients who underwent bone marrow biopsy presented fibrosis (grade I, n = 5; grade II, n = 1). After a median follow-up of 25.1 months (range, 12.9-95.1 months), 3 patients died (2 from a VEXAS-related complication and 1 from infection), and the overall survival of the patients with VEXAS syndrome at 5 years was 63% (supplemental Figure 3).

The median number of therapeutic lines was 3 (range, 0-6). Of note, 1 patient did not receive any specific treatment (except erythropoietin for MDS-related anemia) over a long period (87 months), suggesting that VEXAS syndrome can also present with more indolent evolution. All other patients received corticosteroids. In order to compare the effectiveness of the different therapeutic strategies in this retrospective study, we chose the time to the addition of a new steroid-sparing agent (“time to next treatment”) as the most objective marker. Data were censored at the time of last visit or death. We collected data to evaluate the response to immunosuppressive drugs such as corticosteroids (n = 10) and methotrexate (n = 3), cytokine-targeting agents such as anti-tumor necrosis factor α (anti-TNF-α) (adalimumab, n = 3) or anti-interleukin-6 (anti-IL-6) receptor (tocilizumab, n = 4), signaling inhibitors such as calcineurin inhibitor (cyclosporine, n = 3) and JAK inhibitors (ruxolitinib, n = 2; tofacitinib, n = 1), and the hypomethylating agent azacytidine (n = 4). Details of individual patient responses to each therapeutic line are presented in supplemental Table 4. As illustrated in Figure 1, most treatments were only transiently effective; the median time to next treatment was 3.4 months for adalimumab, 3.9 months for corticosteroids, 7.4 months for methotrexate, and 8 months for tocilizumab. The hypomethylating agent and signaling inhibitors seemed to achieve interesting results; the median time to next treatment was 12.7 months for cyclosporine, 21.9 months for azacytidine, and not reached for JAK inhibitors, even if the short duration of follow-up limits the evaluation of the latter. An illustration of the potential effectiveness of JAK inhibitors is provided by the dramatic regression of cutaneous lesions observed after the introduction of ruxolitinib as monotherapy following azacytidine in 1 patient, without any increase in corticosteroids, as shown in supplemental Figure 4. Finally, no improvement in cytopenia (especially anemia) was observed in any patient. For 7 patients evaluated with a bone marrow examination, we did not notice any change in the myelodysplastic features under treatment.

The recent recognition of VEXAS syndrome enables us to group a set of patients with variable autoinflammatory manifestations associated with cytopenias and dysplastic bone marrow in a common entity based on a unified pathophysiological process. As a consequence of the loss of methionine 41, the expression of a catalytically impaired isoform of UBA1 results in decreased ubiquitylation and endoplasmic reticulum stress in myeloid cells. Accordingly, these cells express higher levels of proinflammatory cytokines, which drive inflammation. Hence, different therapeutic strategies can be interesting according to which step of this cascade is targeted. Azacytidine is a US Food and Drug Administration–approved drug for high-risk MDS that might target myeloid cells harboring UBA1 mutation. Interestingly, azacytidine has been proposed as a therapeutic option in autoimmune disorders associated with MDS based on retrospective studies reporting a control of the inflammatory manifestations in two-thirds of the patients.²  The observation of some responses to this drug in the patients with VEXAS syndrome reported herein is in line with this previous observation. Among the strategies targeting the immune system, responses were observed with corticosteroids, but relapses or corticosteroid dependence limited their effectiveness over time. More interesting results were found with tocilizumab, cyclosporine, and JAK inhibitors. JAK inhibition was reported to have dramatic effects on the inflammation associated with primary myelofibrosis³  and is also an effective drug in graft-versus-host disease⁴  and several other immune-inflammatory diseases.^5,6  These interesting results warrant further evaluation and need to be interpreted with caution given the retrospective nature of this small cohort with short duration of follow-up.

We present herein an independent cohort beyond the initial description of the disease, confirming key clinical and laboratory features of VEXAS syndrome. The rapid identification of 11 VEXAS cases was based on the well-characterized phenotype described by Beck et al and suggests an unappreciated incidence of this syndrome among patients with myeloid dysplasia and inflammatory manifestations. We also describe 2 novel splice motif mutations leading to truncated UBA1 exon 3 with loss of methionine 41 as a new mechanism for VEXAS pathogenesis. Finally, we bring new insights about the therapeutic outcome of VEXAS patients in the absence of evidence-based recommendations. Randomized clinical trials evaluating different therapeutic strategies are awaited to improve the outcome of patients with VEXAS syndrome.