ABSTRACT
Hypomethylating agents (HMA) and allogeneic hematopoietic stem cell transplantation (alloHSCT) have both demonstrated remissions in VEXAS; however, comparative data is lacking. We conducted a multicenter, retrospective analysis of 66 patients diagnosed with VEXAS syndrome treated with HMA (n = 35) or alloHSCT (n = 31). Baseline characteristics such as genetics, co‐morbidities, and performance status were balanced between the groups, except older age in the HMA group. Median follow‐up from therapy initiation was 18 months (95% CI: 11–26), and 14 (21%) deaths were reported (alloHSCT n = 3; HMA n = 11). Among all evaluable patients within the alloHSCT cohort, all patients achieved molecular remission, and a substantial proportion of patients discontinued glucocorticoids (58%). In contrast, HMA therapy was associated with lower but meaningful rates of molecular remission (22%) and glucocorticoid discontinuation (6%). In a real‐world setting, HMA therapy was associated with a high discontinuation rate related to toxicity or lack of response. On multivariable analysis adjusted for age and Charlson Comorbidity Index, alloHSCT was associated with improved overall survival (HR = 0.20, 95% CI: 0.05–0.81; p = 0.024). This association remained consistent across multiple ancillary sensitivity analyses, including restriction to transplant‐eligible patients, patients aged ≤ 75 years, 1:1 matching, and propensity score‐based weighted analyses. Although limited by retrospective design, these findings suggest that alloHSCT remains an attractive and potentially curative strategy in selected patients with VEXAS. Prospective validation of these findings is warranted.
Keywords: allogeneic transplantation, azacitidine, HMA, molecular remission, myelodysplastic syndrome, UBA1, VEXAS
1. Introduction
VEXAS (Vacuoles, E1 enzyme, X‐linked, Autoinflammatory, Somatic) syndrome is a rare, acquired hemato‐inflammatory disorder caused by recurrent somatic mutations in the UBA1 gene [1, 2, 3]. Located on the X chromosome, this gene encodes the E1 ubiquitin‐activating enzyme that initiates cellular ubiquitylation [3]. Pathogenic variants predominantly cluster at codon methionine‐41 (p.Met41), leading to loss of the canonical cytoplasmic isoform (UBA1b) and the emergence of a catalytically impaired isoform (UBA1c) [3, 4, 5]. The downstream consequence is proteotoxic stress, ultimately resulting in myeloid‐driven systemic inflammation [3, 5].
Clinically, VEXAS syndrome primarily affects men over the age of 50, with an estimated prevalence of UBA1 variants in approximately 1 in 4000 in this demographic [2, 3, 4, 6]. The disease phenotype is characterized by an overlap of autoinflammatory manifestations, including neutrophilic dermatosis, chondritis, and pulmonary infiltrates, as well as progressive bone marrow failure [1, 3, 4]. Cytopenias are nearly universal, with macrocytic anemia and thrombocytopenia frequently progressing to transfusion dependency [3, 4]. Furthermore, there is an increased incidence of hematological malignancies, with 25%–55% of patients meeting criteria for myelodysplastic syndromes (MDS) in the literature, although such cases typically lack high‐risk molecular drivers commonly observed in primary myeloid neoplasms and exhibit low propensity for leukemic transformation [3, 7, 8].
The management of VEXAS syndrome remains challenging, in part due to the paucity of prospective randomized trials [3, 7]. Glucocorticoids are generally effective at controlling inflammatory manifestation early in the disease course; however, most patients become steroid‐dependent, often requiring high doses exceeding prednisone 15–20 mg/day, with substantial metabolic and infectious toxicity [2]. Targeted anti‐inflammatory agents, such as Janus kinase (JAK) inhibitors and interleukin‐6 (IL‐6), do not eradicate the underlying mutant clone but may provide transient palliative benefit in some cases [9, 10, 11]. Investigational focus has increasingly shifted toward disease‐modifying strategies targeting the UBA1‐mutant clone, including hypomethylating agents (HMA), particularly azacitidine (5‐azacitidine, AZA) [6, 12, 13] and allogeneic hematopoietic stem cell transplantation (alloHSCT) [14, 15, 16, 17, 18], both of which have demonstrated clinical and molecular remissions. However, these therapies have not been formally compared. Here, we report a multicenter experience comprising two distinct cohorts of patients with VEXAS syndrome treated with HMA based therapy or alloHSCT, accounting for variables such as age, genetic subtypes, and co‐morbidities.
2. Methods
This multicenter retrospective study was conducted after regulatory approval (Mayo Clinic IRB#23‐013439 with site‐specific approvals), consecutive patients aged ≥ 18 years with molecularly confirmed (UBA1‐mutated) VEXAS syndrome treated with at least one cycle of HMA or received alloHSCT between March 2019 and August 2025 across six participating institutions (Mayo Clinic = 44, New York University = 9, University of Kansas = 5, Emory University = 4, University of Michigan = 3, University of Washington = 1) were included. Eligibility criteria included glucocorticoid‐refractory inflammation (need for daily prednisone or equivalent dose ≥ 15 mg) and/or concurrent myeloid neoplasm with adequate performance status [Karnofsky performance status (KPS) > 70/Eastern Cooperative Oncology Group performance status (ECOG ≤ 2] (KPS 50%–70%, and ECOG PS 3 was allowed if it was directly attributable to VEXAS)]. Patients with uncontrolled infection, recent major cardiovascular events, or severe organ dysfunction were excluded (Supporting Information). UBA1 mutations (pathogenic/likely pathogenic) were detected through targeted sequencing [NGS, limit of detection 2% variant allele fraction (VAF)], and molecular response was assessed by quantitative droplet digital polymerase chain reaction testing (ddPCR, sensitivity of 0.5% VAF) in a subset of patients.
Outcomes included overall survival (OS), cumulative incidence of glucocorticoid discontinuation, and molecular remission (MR, undetectable UBA1 mutation by NGS, VAF < 2%). Glucocorticoid refractoriness was defined as inability to reduce prednisone or equivalent dose below 15 mg/day for over 2 months [7]. OS was defined as the time from initiation of therapy (HMA or alloHSCT) to death or last follow up. Kaplan–Meier methods were used to compute OS; Cox regression models assessed treatment influence on OS while adjusting for other clinical factors restricted to two covariates due to event count. To accommodate the retrospective nature of these analyses and potential inherent differences between treatment groups, four analytical approaches were employed: (1) analysis of all patients, (2) analysis of only patients ≤ 75 years of age at treatment, (3) analysis using patients treated with alloHSCT matched 1:1 to those treated with HMA based on age and the pre‐treatment Charlson Comorbidity Index (CCI) score, and (4) modeling matched patients while also weighting on inverse probability treatment weights (IPTW) based on propensity scores. Propensity scores were calculated based on factors associated with OS but not significantly different between treatment groups and not already matched on: Karnofsky performance status, abnormal cytogenetics, and recruiting site. Cumulative incidence of glucocorticoid discontinuation was analyzed with death as a competing event. Statistical analyses were performed in R (version 4.2.2).
3. Results
3.1. Patient Characteristics
Of 168 patients screened, 66 met inclusion criteria (Figure S1). All patients were male, with a median age of 69 years at treatment (range 49–82, Table 1). Median time from diagnosis to treatment was 11 months (range 0–40) and median number of prior lines of therapy for VEXAS syndrome was 3 (range 0–14). The primary indication for therapy with HMA or alloHSCT was glucocorticoid‐refractory inflammation (52%), followed by progressive bone marrow failure or myeloid neoplasm (26%), or both (23%). All patients had < 5% marrow blasts before treatment, 42% of patients fulfilled criteria for concurrent MDS, 22% with abnormal karyotype, and 40% harbored myeloid co‐mutations. The most common co‐mutations were DNMT3A (23%, including two cases of R882 variant) followed by TET2 (11%) and SF3B1 (7%). The most frequent variants in the UBA1 gene included p.Met41Thr (40%), p.Met41Val (27%), and p.Met41Leu (25%). The median Hematopoietic Cell Transplantation‐specific Comorbidity Index (HCT‐CI) was 4 (range 0–11), with median KPS of 80 (50–100) and ECOG of 1 (0–3) (Table 1).
TABLE 1.
Clinical characteristics, laboratory parameters, and treatment details of 66 patients with VEXAS syndrome stratified by treatment with alloHSCT or HMA.
| Variables | All patients N = 66 | AlloHSCT N = 31 | HMA N = 35 | p **** |
|---|---|---|---|---|
| Age at treatment in years; median (range) | 69 (49–82) | 65 (49–75) | 72 (57–82) | 0.001 |
| Age > 60 years; N (%) | 58 (88) | 27 (87) | 31 (88) | 1 |
| Male sex; N (%) | 66 (100) | 31 (100) | 35 (100) | 1 |
| UBA1 variants (Evaluable = 63) | N = 30 | N = 33 | ||
| p.Met41Leu; N (%) | 16 (25) | 7 (23) | 9 (27) | 0.95 |
| p.Met41Val; N (%) | 17 (27) | 9 (30) | 8 (24) | 0.82 |
| p.Met41Thr; N (%) | 25 (40) | 11 (37) | 14 (42) | 0.83 |
| Splice site mutation; N (%) | 3 (5) | 2 (7) | 1 (3) | 0.78 |
| Others; N (%) | 2 (3) | 1 (3) | 1 (3) | 1 |
| Pretreatment UBA1 variant allele fraction % (VAF); median (range) | N = 20 | N = 28 | 0.98 | |
| (Evaluable = 48) | 71 (13–95) | 71 (13–89) | 69 (25–95) | |
| Number of prior lines of therapy a ; median (range) | 3 (0–14) | 3 (1–11) | 3 (0–14) | 0.87 |
| Time from diagnosis to therapy in months b ; median (range) | 11 (0–40) | 13 (1–40) | 6 (0–31) | 0.092 |
| Indication for therapy: | ||||
| Glucocorticoid‐refractory inflammation; N (%) | 34 (52) | 18 (58) | 16 (46) | 0.6 |
| Progressive bone marrow failure/concurrent myeloid neoplasm; N (%) | 17 (26) | 7 (23) | 10 (28) | |
| Both; N (%) | 15 (23) | 6 (19) | 9 (26) | |
| Pretreatment prednisone dose in mg; median (range) | 22 (0–60) | 20 (0–40) | 25 (0–60) | 0.27 |
| Pretreatment leukocytes, × 109/L; median (range) | 5 (1–13) | 5 (1–12) | 6 (2–13) | 0.02 |
| Pretreatment hemoglobin, g/dL; median (range) | 9 (5–15) | 11 (5–15) | 9 (7–15) | 0.043 |
| Pretreatment platelets, × 109/L; median (range) | 125 (17–574) | 134 (31–275) | 119 (17–574) | 0.29 |
| Pretreatment bone marrow blasts < 5%; N (%) | 66 (100) | 31 (100) | 35 (100) | 1 |
| Accompanying myelodysplastic syndrome; N (%) | 28 (42) | 13 (42) | 15 (43) | 1 |
| Presence of myeloid co‐mutation; N (%) | N = 30 | N = 27 | ||
| (Evaluable = 57) | 23 (40) | 10 (33) | 13 (48) | 0.29 |
| Presence of abnormal karyotype; N (%) | N = 29 | N = 33 | ||
| (Evaluable = 62) | 14 (22) | 4 (14) | 10 (30) | 0.14 |
| (Evaluable = 57) c | N = 30 | N = 27 | ||
| DNMT3A co‐mutation; N (%) | 13 (23) | 5 (17) | 8 (30) | 0.35 |
| TET2 co‐mutation; N (%) | 6 (11) | 4 (13) | 2 (7) | 0.67 |
| SF3B1 co‐mutation; N (%) | 4 (7) | 3 (10) | 1 (4) | 0.61 |
| ASXL1 co‐mutation; N (%) | 2 (4) | 1 (3) | 1 (4) | 1 |
| CBL co‐mutation; N (%) | 1 (2) | 0 (0) | 1 (4) | 0.47 |
| DDX41 co‐mutation; N (%) | 1 (2) | 1 (3) | 0 (0) | 1 |
| EZH2 co‐mutation; N (%) | 1 (2) | 1 (3) | 0 (0) | 1 |
| KRAS co‐mutation; N (%) | 1 (2) | 0 (0) | 1 (4) | 0.47 |
| TP53 co‐mutation; N (%) | 1 (2) | 1 (3) | 0 (0) | 1 |
| ZRSR2 co‐mutation; N (%) | 1 (2) | 0 (0) | 1 (4) | 0.47 |
| HCT CI score; median (range) | N = 31 | N = 25 | 0.65 | |
| (Evaluable = 56) | 4 (0–11) | 4 (0–8) | 4 (0–11) | |
| Charlson CI score; median (range) | 4 (1–9) | 4 (2–8) | 4 (1–9) | 0.23 |
| Karnofsky performance status; median (range) | 80 (50–100) | 80 (70–100) | 80 (50–100) | 0.94 |
| Eastern Cooperative Oncology Group performance status (ECOG); median (range) | 1 (0–3) | 1 (0–2) | 1 (0–3) | 0.74 |
| HMA therapy details | ||||
| HMA type | ||||
| Azacitidine; N (%) | 27 (77) | |||
| Decitabine, N (%) | 8 (23) | |||
| Number of HMA cycles; median (range) | 5 (1–36) | |||
| Received ≥ 4 cycles; N (%) | 27 (77) | |||
| Received ≥ 6 cycles; N (%) | 17 (49) | |||
| Allogeneic hematopoietic stem cell transplant details | ||||
| Conditioning intensity | ||||
| Myeloablative; N (%) | 1 (3) | |||
| Reduced intensity/Non myeloablative; N (%) | 30 (97) | |||
| Conditioning regimens | ||||
| Fludarabine/Busulfan (reduced intensity dosing); N (%) | 9 (29) | |||
| Fludarabine/Thiotepa/Busulfan (reduced intensity dosing); N (%) | 5 (16) | |||
| Fludarabine/Melphalan; N (%) | 14 (41) | |||
| Fludarabine/cyclophosphamide/total body irradiation; N (%) | 2 (6) | |||
| Fludarabine/total body irradiation (myeloablative dose); N (%) | 1 (3) | |||
| Human leukocyte antigen (HLA) matching | ||||
| Match related; N (%) | 4 (13) | |||
| Match unrelated; N (%) | 21 (68) | |||
| Mismatch unrelated; N (%) | 2 (6) | |||
| Haplo‐identical; N (%) | 4 (13) | |||
| Stem cell source | ||||
| Peripheral blood graft; N (%) | 30 (97) | |||
| Bone marrow graft; N (%) | 1 (3) | |||
| Graft versus host disease prophylaxis | ||||
| Post‐transplant cyclophosphamide/tacrolimus/mycophenolate; N (%) | 27 (87) | |||
| Post‐transplant cyclophosphamide/tacrolimus/Abatacept; N (%) | 1 (3) | |||
| Tacrolimus/Methotrexate; N (%) | 2 (6) | |||
| Tacrolimus/Methotrexate/Abatacept; N (%) | 1 (3) | |||
| Lowest post‐treatment UBA1 VAF %; median (range) | N = 24 | N = 21 | < 0.001 | |
| (Evaluable = 45) | 0 (0–96) | 0 (0–0) | 7 (0–96) | |
| Deaths; N (%) | 14 (21) | 3 (10) | 11 (31) |
Note: The bold values highlight statisically significant p values (< 0.05).
Abbreviations: alloHSCT, allogeneic hematopoietic stem cell transplantation; BM, bone marrow; CCI, Charlson Comorbidity Index; ECOG, Eastern Cooperative Oncology Group; GVHD, graft‐versus‐host disease; HCT‐CI, Hematopoietic Cell Transplantation‐specific Comorbidity Index; HMA, hypomethylating agent; KPS, Karnofsky performance status; MDS, myelodysplastic syndrome; MMF, mycophenolate mofetil; PtCy, post‐transplant cyclophosphamide; Tac, tacrolimus; VAF, variant allele fraction.
p‐values comparing characteristics between patients treated with HMA vs. AlloHSCT. Continuous measures compared using two sample t‐tests or the nonparametric Wilcoxon rank sum test if data available for < 30 patients in a group. Categorical factors compared using chi‐square tests or Fisher exact tests (if number in each subset is < 5).
Prior lines of therapy included combinations of prednisolone, hydroxychloroquine, methotrexate, leflunomide, azathioprine, tocilizumab, rituximab, anakinra, ixekizumab, adalimumab, mycophenolate, colchicine, HMA, dapsone, tofacitinib, infliximab, baricitinib, and canakinumab.
Four patients were diagnosed with VEXAS syndrome after receiving HMA treatment for underlying hematological disorder and were excluded from the time from diagnosis to treatment analysis.
Six patients had > 2 concurrent non‐UBA1 mutations (alloHSCT‐4, HMA‐2).
3.2. Clinical Outcomes in the Overall Cohort
The median follow‐up for the entire cohort was 18 months (95% CI: 11–26), with 14 (21%) deaths reported (HMA‐11 and alloHSCT‐3). Median OS was not reached (95% CI: 31 months ‐ not reached; Figure 1A), with outcomes differing by treatment modality (Figures 2A–C and 3A–3B). At last follow up, 20 patients (30%) discontinued glucocorticoids with a median time to discontinuation of 7 months (range < 1–31); no patients progressed to acute myeloid leukemia. On univariable analysis, alloHSCT (HR 0.21, p = 0.018), KPS (HR 0.49, p = 0.003), ECOG (HR 3.05, p = 0.004), and abnormal cytogenetics (HR 4.98, p = 0.006) were associated with OS (Figure 3C).
FIGURE 1.
Overall Survival (OS) analyses in 66 patients with VEXAS syndrome treated with hypomethylating agents (HMA) or allogeneic hematopoietic stem cell transplantation (alloHSCT). (A) Kaplan–Meier curves for overall survival (OS) in all patients. At a median follow‐up of 18 months (95% CI: 11–26), median OS was not reached (NR, 95% CI: 31 months‐NR, range: 1–61 months); (B) Kaplan–Meier curves for overall survival (OS) in patients who underwent alloHSCT. At a median follow‐up of 20 months (95% CI: 11–30), median OS was not reached (NR, range: 1–61 months); and (C) Kaplan–Meier curves for overall survival (OS) in patients treated with HMA therapy. At a median follow‐up of 11 months (95% CI: 8–18), median OS was 29.6 months (NR, range: 1–60 months). [Color figure can be viewed at wileyonlinelibrary.com]
FIGURE 2.
Survival and treatment efficacy analyses in 66 patients with VEXAS syndrome stratified by treatment with hypomethylating agents (HMA) or allogeneic hematopoietic stem cell transplantation (alloHSCT). (A) Kaplan–Meier curves for overall survival (OS) in all patients. At a median follow‐up of 18 months (95% CI: 11–26), median OS was not reached (NR, range: 1–61 months) after alloHSCT and 29.6 months (range 1–60) after HMA (p = 0.0096); (B) 1:1 case matched analysis between alloHSCT and HMA patients based on age and CCI score at the time of treatment. Accounting for the matching in the Cox regression model, alloHSCT treated patients (n = 29) still had significantly better OS compared to HMA patients (n = 29) (log‐rank p = 0.006, HR = 0.19, 95% CI: 0.05–0.67); and (C) Cumulative incidence of glucocorticoid discontinuation, accounting for death as a competing risk. Compared to HMA, alloHSCT was associated with a higher incidence of glucocorticoid discontinuation and a lower incidence of death. The 12‐month cumulative incidence of glucocorticoid discontinuation rates was 34.1% versus 5.0% for patients treated with alloHSCT vs. HMA. [Color figure can be viewed at wileyonlinelibrary.com]
FIGURE 3.
Post‐diagnosis survival, 12 month landmark survival and prognostic factor analyses in patients with VEXAS syndrome stratified by treatment with hypomethylating agents (HMA) or allogeneic hematopoietic stem cell transplantation (alloHSCT). (A) Kaplan–Meier curves for post‐diagnosis overall survival (OS) in all patients. At a median follow‐up of 33 months (95% CI: 29.5–39.5 months), median OS from diagnosis was not reached in both treatment groups, and 24‐month OS rate was 100% in those treated with alloHSCT (deaths were after 2 years of diagnosis) versus 73% (95% CI: 58%–91%) in those treated with HMA; (B) Kaplan–Meier curves for 12‐month landmark OS analysis in 34 patients alive and in follow‐up at 12 months after start of treatment. At a median follow‐up of 17 months (95% CI: 13–29 months), median OS was not reached in those treated with alloHSCT versus 18.2 months in those treated with HMA (log‐rank p = 0.04); and (C) Forest plot illustrating the results of univariable Cox proportional hazards analysis for factors associated with overall survival. Hazard ratios (HRs) are presented with 95% confidence intervals (CIs). Variables with HR > 1 indicate increased risk of death. [Color figure can be viewed at wileyonlinelibrary.com]
3.3. Allogeneic HSCT Cohort
Thirty‐one patients (47%) underwent alloHSCT at a median age of 65 years (range: 49–75) and median time from diagnosis to treatment of 13 months (range: 1–40). The most frequent UBA1 variant was p.Met41Thr (37%), with a median pretreatment variant allele fraction of 71% (range 13%–89%). Median number of prior therapies was 3 (range 1–11), and glucocorticoid‐refractory inflammation remained the predominant indication for treatment (58%), followed by progressive bone marrow failure or myeloid neoplasm (23%), or both (19%). Concurrent MDS was present in 42% of patients, abnormal karyotype in 14%, and myeloid gene co‐mutations in 33% of patients. The median HCT‐CI was 4 (range 0–8), KPS was 80 (range 70–100), and ECOG was 1 (range 0–2) (Table 1). Donor sources included matched sibling (N = 4), matched unrelated (N = 21), mismatched unrelated (N = 2), and haploidentical (N = 4). Conditioning regimen was predominantly reduced intensity (fludarabine/melphalan N = 14, fludarabine/busulfan N = 9, and thiotepa/busulfan/fludarabine N = 5), while two patients received non‐myeloablative conditioning with fludarabine/cyclophosphamide/total body irradiation (TBI) and one patient received myeloablative conditioning with fludarabine/TBI. The graft‐versus‐host disease (GVHD) prophylaxis was post‐transplant cyclophosphamide/tacrolimus/mycophenolate mofetil in all patients except in two who received tacrolimus/methotrexate ± abatacept following a matched sibling donor. Stem cell graft source was T‐cell replete peripheral blood except for one patient who received a bone marrow graft (Table 1).
At last median follow‐up of 20 months (95% CI: 11–30), 3 (10%) deaths were recorded and median OS not reached (range 1–61 months, Figure 1). AlloHSCT was associated with high rates of glucocorticoid discontinuation (58%), with a median time to discontinuation of 7 months (range < 1–31). Molecular remission was achieved in all evaluable patients (24 of 24; 100%).
3.4. HMA Cohort
Thirty‐five patients (53%) received hypomethylating agent–based therapy, most commonly azacitidine (77%), with a median of 5 cycles (range 1–36). Twenty‐seven (77%) received ≥ 4 cycles, while 49% received ≥ 6 cycles (Table 1, Table S1). Patients were older (median age, 72 years) and had a shorter time from diagnosis to treatment (median 6 months). The most frequent UBA1 variant was p.Met41Thr (42%), followed by p.Met41Leu (27%) and p.Met41Val (24%), with a median pretreatment variant allele fraction of 69% (range 25%–95%, Table 1). Concurrent MDS was present in 43% of patients, abnormal karyotype in 30%, and myeloid co‐mutations in 48%. The median number of prior therapies was 3 (range 0–14), and glucocorticoid‐refractory inflammation was the primary treatment indication (46%), followed by progressive bone marrow failure or myeloid neoplasm (28%), or both (26%). Similar to the alloHSCT cohort, the median HCT‐CI was 4 (range 0–11, p = 0.65), KPS was 80 (range 50–100, p = 0.94) and ECOG PS was 1 (range 0–3, p = 0.74). Notably, 2 patients in this group had an ECOG PS 3 and KPS 50–70 attributed to VEXAS. At last follow‐up, 11 (31%) deaths were recorded, median OS was 29.6 months (95% CI: 18 months‐not reached; range: 1–60 months, Figure 1C). Treatment discontinuation occurred in 49% of patients, primarily due to infectious complications, cytopenias, or lack of response (Table S1). Rates of glucocorticoid discontinuation were low (6%), and molecular remission was achieved in 22% of evaluable patients.
3.5. Impact of Treatment on Survival
OS was significantly longer with alloHSCT versus (vs.) HMA (median not reached vs. 29.6 months; Figure 2A), even after adjusting for age and CCI (HR 0.20, 95% CI: 0.05–0.84, p = 0.029; Table 2). Results were consistent when OS was computed from time of diagnosis (Figure 3A), 12‐month landmark OS analysis (Figure 3B), and in comparison to other prognostically relevant variables (Figure 3C).
TABLE 2.
Model‐based results evaluating overall survival (OS) between patients with VEXAS syndrome treated with alloHSCT versus HMA, adjusted for Charlson Comorbidity Index (CCI). HR < 1 favors alloHSCT.
| Model | N | HR (95% CI) | p |
|---|---|---|---|
| All patients | 65 a | ||
| AlloHSCT | 0.20 (0.05–0.81) | 0.024 | |
| HMA (reference) | — | — | |
| Age | 0.97 (0.90–1.05) | 0.50 | |
| Charlson CI | 1.36 (0.99–1.87) | 0.058 | |
| All patients | 65 a | ||
| AlloHSCT | 0.24 (0.07–0.87) | 0.031 | |
| HMA (reference) | — | ||
| Charlson CI | 1.32 (0.96–1.80) | 0.084 | |
| Patients ≤ 75 years | 52 a | ||
| AlloHSCT | 0.20 (0.05–0.84) | 0.029 | |
| HMA (reference) | — | ||
| Charlson CI | 1.64 (1.07–2.53) | 0.025 | |
| Matched 1:1 (age + CCI) | 58 | ||
| AlloHSCT | 0.19 (0.05–0.67) | 0.010 | |
| HMA (reference) | — | ||
| Matched 1:1 + PS‐based IPTW | 54 b | ||
| AlloHSCT | 0.27 (0.08–0.91) | 0.035 | |
| HMA (reference) | — |
Abbreviations: CI, confidence interval; CCI, Charlson Comorbidity Index; HR, hazard ratio; IPTW, inverse probability treatment weighting; PS, propensity score.
CCI could not be reliably calculated in one patient.
Four patients excluded from IPTW analysis due to missing data.
To further evaluate the robustness of these results in this retrospective setting, we found that ancillary analyses had comparable results. Restricting the analysis to transplant‐eligible patients ≤ 75 years (n = 52), alloHSCT was associated with better OS than HMA (median not reached vs. 30.2 months; HR 0.20, 95% CI: 0.05–0.84, p = 0.029, Figure 4A, Table S2), including patients who only received AZA (median not reached vs. 31 months, HR 0.24, 95% CI 0.06–0.95, p = 0.042), only those who met alloHSCT eligibility criteria (median not reached vs. 30 months, HR 0.26, 95% CI 0.06–0.99, p = 0.037) and those who received ≥ 4 HMA treatment cycles (median not reached vs. 30.2 months, Figure 4B, HR 0.29, 95% CI: 0.07–1.13, p = 0.06). The OS difference attenuated to non‐significance among patients completing > 6 HMA cycles; however, this was an underpowered analysis with only 17 evaluable patients in the HMA group (HR 0.63, 95% CI: 0.10–3.84, p = 0.61, Figure 4C). In the 1:1 matching between alloHSCT and HMA patients based on age and CCI score at time of treatment (N = 58, 29 in each treatment group), alloHSCT was associated with improved OS (log‐rank p = 0.006, HR = 0.19, 95% CI: 0.05–0.67, Figure 2B). Results were consistent after incorporating IPTW into this matched model, based on Karnofsky PS, abnormal cytogenetics, and recruiting site. Here, alloHSCT patients still appeared to have significantly better OS compared to patients treated with HMA (HR = 0.27, 95% CI: 0.08–0.91; p = 0.035, Table 2).
FIGURE 4.
Sensitivity analyses in subgroups highlighting overall survival (OS) in patients with VEXAS syndrome stratified by treatment with hypomethylating agents (HMA) or allogeneic hematopoietic stem cell transplantation (alloHSCT). (A) Kaplan–Meier (KM) curves for OS Overall Survival (OS) in 53 patients aged ≤ 75 years. At a median follow‐up of 15 months (95% CI: 11–25), median OS was not reached (NR, range: 1–61 months) after alloHSCT and 30.2 months (range 1–60) after HMA (p = 0.020). KM curve includes all 53 patients ≤ 75 years; Cox model excludes one patient with unreliable CCI (n = 52); (B) OS in 58 patients treated either with ≥ 4 HMA treatment cycles or with alloHSCT. At a median follow‐up of 16 months (95% CI: 11–26), median OS was not reached (NR, range: 1–61 months) after alloHSCT and 30.2 months (range 4–60) after HMA (p = 0.06); (C) OS in 48 patients treated either with ≥ 6 HMA treatment cycles or with alloHSCT. At a median follow‐up of 16 months (95% CI: 11–26), median OS was not reached (NR, range: 1–61 months) after alloHSCT and not reached (range 6–60 months) after HMA (p = 0.61). [Color figure can be viewed at wileyonlinelibrary.com]
3.6. Patients Receiving HMA Prior to alloHSCT
Five alloHSCT patients received prior HMA (median 4 cycles, range 1–7) for concurrent MDS, none showing clinical response; 3 received HMA within 3 months pre‐transplant. The survival advantage of alloHSCT remained significant after excluding these patients (HR 0.2, 95% CI 0.07–0.9, p = 0.03).
3.7. Glucocorticoid Discontinuation and Molecular Remission
Competing risk regression models showed that alloHSCT was associated with a markedly higher cumulative incidence of glucocorticoid discontinuation (cause‐specific HR = 19.9, 95% CI: 2.09–190; p = 0.009), adjusted for age and CCI. The 12‐month cumulative incidence of glucocorticoid discontinuation was 34.1% vs. 5.0% for alloHSCT vs. HMA (Figure 2C). AlloHSCT was also associated with a higher MR rate (24/24 [100%] vs. 5/23 [22%], p < 0.01); assessment of MR was done after a median of 5 HMA cycles (range 2–13). Among MR achievers, 16 patients (alloHSCT = 14, HMA = 2) were confirmed to be in MR by ddPCR (limit of detection [LOD] 0.5%); no MR patient relapsed with re‐emergence of UBA1 clone at last follow‐up.
3.8. Treatment Toxicities
In the HMA group, 17 patients (49%) discontinued therapy after a median of 4 months (range 1–20) due to recurrent infections/sepsis (n = 8), severe prolonged cytopenias (n = 3), lack of response (n = 3), gastrointestinal adverse events (n = 1), patient preference (n = 1), or after achieving MR (n = 1) (43% due to toxicity, Table S1). In the alloHSCT group, grade 2 acute GVHD occurred in 4 (13%; skin = 2; gut = 2) patients, with no grade 3–4 GVHD, and 6 (19%) developed moderate to severe chronic GVHD (lung = 2; oral/gastrointestinal/cutaneous = 4). Causes of death in the HMA group included infections (n = 7; four bacteremia, one pneumonia, one fungal infection, one diverticulitis/decubitus ulcer), worsening cardiac disease (n = 2), and progressive inflammatory disease/bone marrow failure (n = 2). In the alloHSCT group, infections (n = 2; HHV6 infection and pneumonia) and pulmonary GVHD (N = 1) were the main causes of death. Notably, 8 of 11 (73%) HMA‐group deaths occurred after therapy discontinuation.
4. Discussion
Since its initial recognition in 2020, VEXAS syndrome has emerged as a prototypical disorder manifesting with systemic inflammatory disease driven by mutant hematopoietic stem and progenitor cells [1, 3, 5, 19]. Clinical management remains challenging, with glucocorticoids constituting initial therapy; however, responses are often transient and accompanied by substantial long‐term toxicity [20, 21]. Conventional steroid sparing agents such as methotrexate, mycophenolate mofetil, and hydroxychloroquine have limited efficacy [4, 9, 11].
Targeted cytokine directed therapies in VEXAS including JAK inhibitors and IL‐6/IL‐1 inhibitors have yielded variable results [10, 11, 22, 23]. Ruxolitinib showed a 6‐month clinical response rate of 87% in a multicenter study of 30 patients [10], contrasting with a 30% overall response rate in the FRENVEX cohort [23]. IL‐6 inhibitors achieved an overall response rate of 26% in the FRENVEX registry [23], and IL‐1 inhibitors showed limited efficacy (9% at 3 months) [23]. Importantly, these approaches do not meaningfully impact UBA1 clonal burden or alter the natural history of marrow failure in VEXAS [23].
The current multicenter study provides, to our knowledge, the first comparative analysis evaluating outcomes between alloHSCT and HMA in VEXAS syndrome. Principal observations are: (i) improved OS association with alloHSCT (HR = 0.20, p = 0.024); (ii) higher rate of molecular remission (100% vs. 22%); and (iii) higher rate of glucocorticoid discontinuation (58% vs. 6%). The 100% molecular remission rate and absence of UBA1 clone re‐emergence in the alloHSCT cohort highlight the curative potential of this modality, consistent with prior registry and institutional series [6, 7, 9, 11, 12, 13, 14, 15, 16, 17, 18, 24].
HMA demonstrated more modest efficacy. Molecular remission rates were low and treatment discontinuation due to toxicity was frequent—consistent with the FRENVEX azacitidine cohort, which reported an inflammatory response rate of 41% at 6 months, with 75% relapse among responders who discontinued azacitidine and a 34% discontinuation rate due to toxicity or treatment failure [6].
A finding meriting explicit discussion is the non‐significant OS difference in the ≥ 6 HMA cycle subgroup although numbers in this subgroup were limited (n = 17, HR 0.63, p = 0.61; Figure 4C). Tolerability of HMA is itself a clinically meaningful outcome. In our cohort, 43% of patients could not complete therapy due to toxicity, and 8 of 11 HMA‐group deaths followed therapy discontinuation. The intention‐to‐treat perspective most accurately reflects real‐world outcomes when counseling patients at diagnosis. Nonetheless, the ≥ 6‐cycle subgroup finding supports the hypothesis that patients who can tolerate and complete HMA may achieve meaningful disease control and should inform the design of future prospective studies evaluating HMA as a bridge to transplantation.
Exploratory analyses in our study identified alloHSCT, robust performance status and normal cytogenetics as factors associated with improved survival, whereas UBA1 variant subtype and co‐mutations were not independently predictive [3, 4, 25, 26] These findings suggest that alloHSCT may mitigate the adverse impact of traditional risk factors, supporting consideration of earlier transplant referral in appropriate candidates. Notably, despite the high prevalence of concomitant MDS, no cases of AML transformation were observed [3, 7].
Limitations include retrospective design, potential selection and immortal time bias, and a limited sample size. Multiple analytical strategies were employed to mitigate bias; however, residual confounding cannot be fully excluded—a limitation shared by the FRENVEX azacitidine cohort [6]. Variable HMA dosing and early treatment discontinuation may underestimate therapeutic benefit in selected patients. Nonetheless, these data offer a pragmatic view of real‐world management across two distinct therapeutic approaches and provide the evidential foundation for designing a future randomized clinical trial, a goal toward which this study establishes clinical equipoise.
5. Conclusions
AlloHSCT offers the potential for durable remission and cure in appropriately selected patients with VEXAS, whereas HMAs provide a less intensive, disease‐modifying option for those who are not immediate transplant candidates. Prospective studies are warranted to refine risk stratification, optimize treatment sequencing, and define the role of HMAs as a bridge to transplantation.
Author Contributions
A.A.M. designed the study. S.F., A.A., and S.M.G. did data collection and analysis. S.F. and A.A.M. wrote the manuscript. A.A.M., M.M.‐K.W., M.S., K.J.L., M.R.L., A.M., U.D., M.H., W.J.H., M.V.S., R.S.G., G.B., J.K., K.M., M.J.K., M.S., G.H., J.C., M.A.‐H., K.J.W., Y.K., A.A.‐K., K.H.B., N.G., A.N.S., J.P., S.C., M.R., H.M., T.B., N.P., J.G.‐L., J.P.M., E.F.K., A.L., T.K., D.D., J.F., M.M.P., H.B.A., D.B.B. contributed patients. S.M.G., T.L.L., C.F., M.M.P., H.B.A. participated in data analysis and interpretation. K.K.R. and H.O. participated in laboratory studies and their interpretation. All authors participated in writing, reviewed the final manuscript draft, and approved its content.
Funding
A.M. is supported by the Mayo Clinic Center for Clinical and Translational Science (CCaTS) Benefactor Career Development Award (Arnold and Kit Palmer Career Development Award in Cancer Research) and American Cancer Society (ACS) Clinician Scientist Development Grant (CSDG‐23‐1020347‐01‐IBCD).
Ethics Statement
The study was conducted under a central institutional review board (Mayo Clinic IRB: 23–013439) for a multicenter study in accordance with the Declaration of Helsinki.
Consent
This is waived due to minimal risk research and only de‐identified information is included in the manuscript.
Conflicts of Interest
Authors disclose no relevant conflicts of interest in relation to the manuscript. A.M. has received research funding to institution from BMS, Novartis, Sanofi, Gilead, Medexus, and Solu Therapeutics. S.C. has received Honorarium: Sanofi, Sobi, Ascentage Pharma, BMS, Pfizer, Legend Biotech, AstraZeneca; Institutional Research funding: C4 Therapeutics, CARSgen, Cynata Therapeutics, Johnson & Johnson, AstraZeneca, Incyte, Abbvie, Cullinan Therapeutics, National Marrow Donor Program. A.N.S. reports consulting fees (paid to institution) from Novartis Pharmaceuticals, grant funding (paid to institution through the Robert A. Winn Career Development Award) from Gilead Sciences, and research funding (paid to institution) from Kura Oncology, Daiichi‐Sankyo, and Chordia Therapeutics. M.R.L. has received research funding from Abbvie, Actinium, Amgen, Astellas, Pluristem, Sanofi, and is on the Speakers Bereau of Amgen and BeOne. U.D. has received funding from Genmab for a clinical trial, and received non‐pharma Speaker Honoraria from Oncology Brothers and Mashup Media. M.J.K. has received consulting fees paid to the institution from Amgen, Rigel, Sobi. D.B.B. has received consulting fees from Sobi, GSK, Novartis, and Alexion. M.A.‐H. advisory board for Incyte, Daiichi, Servier, Amgen, and Novartis. H.M. has performed consulting/advisory board duties for Geron, Ascentage, Abbvie, Syndax Therapeutics, Kite Gilead, Seres, Sobi, BMS, Incyte, Sanofi, Orum. The remaining authors declare no conflicts of interest.
Supporting information
Supporting Information: Describe the inclusion and exclusion criteria for the study.
Figure S1: The number of patients screened and included in the analysis.
Table S1: Treatment details of the hypomethylating agent therapy in 35 patients with VEXAS syndrome.
Table S2: Clinical characteristics and laboratory parameters of 53 patients aged ≤ 75 years at diagnosis with VEXAS syndrome stratified by treatment with allogeneic hematopoietic stem cell transplantation (alloHSCT) or hypomethylating therapy (HMA).
Acknowledgments
The authors acknowledge statistical support from the Mayo Clinic Center for Clinical and Translational Science (CCaTS). They also wish to express sincere gratitude to the patients and their families for their partnership, dedication, and tireless advocacy in the fight against this disease. Preliminary findings from this study were presented as an oral presentation at the 67th American Society of Hematology Annual Meeting, Orlando, Florida, December 2025.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Beck D. B., Ferrada M. A., Sikora K. A., et al., “Somatic Mutations in UBA1 and Severe Adult‐Onset Autoinflammatory Disease,” New England Journal of Medicine 383, no. 27 (2020): 2628–2638, 10.1056/NEJMoa2026834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Beck D. B., Bodian D. L., Shah V., et al., “Estimated Prevalence and Clinical Manifestations of UBA1 Variants Associated With VEXAS Syndrome in a Clinical,” Journal of the American Medical Association 329, no. 4 (2023): 318–324, 10.1001/jama.2022.24836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Beck D. B., Georgin‐Lavialle S., Kirino Y., Patel B. A., and Ferrari S., “VEXAS Syndrome,” Nature Reviews. Disease Primers 12, no. 1 (2026): 19, 10.1038/s41572-026-00695-w. [DOI] [PubMed] [Google Scholar]
- 4. Koster M. J., Lasho T. L., Olteanu H., et al., “VEXAS Syndrome: Clinical, Hematologic Features and a Practical Approach to Diagnosis and Management,” American Journal of Hematology 99, no. 2 (2024): 284–299, 10.1002/ajh.27156. [DOI] [PubMed] [Google Scholar]
- 5. Ferrada M. A., Savic S., Cardona D. O., et al., “Translation of Cytoplasmic UBA1 Contributes to VEXAS Syndrome Pathogenesis,” Blood 140, no. 13 (2022): 1496–1506, 10.1182/blood.2022016985. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jachiet V., Kosmider O., Beydon M., et al., “Efficacy and Safety of Azacitidine for VEXAS Syndrome: A Large‐Scale Retrospective Study From FRENVEX,” Blood 146, no. 12 (2025): 1450–1461, 10.1182/blood.2024028133. [DOI] [PubMed] [Google Scholar]
- 7. Mekinian A., Zhao L. P., Chevret S., et al., “A Phase II Prospective Trial of Azacitidine in Steroid‐Dependent or Refractory Systemic Autoimmune/Inflammatory Disorders and VEXAS Syndrome Associated With MDS and CMML,” Leukemia 36, no. 11 (2022): 2739–2742, 10.1038/s41375-022-01698-8. [DOI] [PubMed] [Google Scholar]
- 8. Sirenko M., Bernard E., Creignou M., et al., “Molecular and Clinical Presentation of UBA1‐Mutated Myelodysplastic Syndromes,” Blood 144, no. 11 (2024): 1221–1229, 10.1182/blood.2023023723. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Al‐Hakim A., Trikha R., Phyu Htut E. E., et al., “Treatment Outcomes in Patients With VEXAS Syndrome: A Retrospective Cohort Study,” Lancet Rheumatology 7, no. 7 (2025): e472–e484, 10.1016/S2665-9913(25)00034-7. [DOI] [PubMed] [Google Scholar]
- 10. Heiblig M., Ferrada M. A., Koster M. J., et al., “Ruxolitinib Is More Effective Than Other JAK Inhibitors to Treat VEXAS Syndrome: A Retrospective Multicenter Study,” Blood 140, no. 8 (2022): 927–931, 10.1182/blood.2022016642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Bourbon E., Heiblig M., Gerfaud Valentin M., et al., “Therapeutic Options in VEXAS Syndrome: Insights From a Retrospective Series,” Blood 137, no. 26 (2021): 3682–3684, 10.1182/blood.2020010177. [DOI] [PubMed] [Google Scholar]
- 12. Raaijmakers M., Hermans M., Aalbers A., et al., “Azacytidine Treatment for VEXAS Syndrome,” Hema 5, no. 12 (2021): e661, 10.1097/hs9.0000000000000661. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Hoff F. W., Trikha R., Groarke E. M., and Patel B. A., “Hypomethylating Agents in Vacuoles, E1 Enzyme, X‐Linked, Autoinflammatory, Somatic Syndrome (VEXAS): A Systematic Review,” British Journal of Haematology 208 (2026): 819–828, 10.1111/bjh.70302. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mangaonkar A. A., Langer K. J., Lasho T. L., et al., “Reduced Intensity Conditioning Allogeneic Hematopoietic Stem Cell Transplantation in VEXAS Syndrome: Data From a Prospective Series of Patients,” American Journal of Hematology 98, no. 2 (2023): E28–e31, 10.1002/ajh.26786. [DOI] [PubMed] [Google Scholar]
- 15. Gurnari C., Koster L., Baaij L., et al., “Allogeneic Hematopoietic Cell Transplantation for VEXAS Syndrome: Results of a Multicenter Study of the EBMT,” Blood Advances 8, no. 6 (2024): 1444–1448, 10.1182/bloodadvances.2023012478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Al‐Hakim A., Poulter J. A., Mahmoud D., et al., “Allogeneic Haematopoietic Stem Cell Transplantation for VEXAS Syndrome: UK Experience,” British Journal of Haematology 199, no. 5 (2022): 777–781, 10.1111/bjh.18488. [DOI] [PubMed] [Google Scholar]
- 17. Diarra A., Duployez N., Fournier E., et al., “Successful Allogeneic Hematopoietic Stem Cell Transplantation in Patients With VEXAS Syndrome: A 2‐Center Experience,” Blood Advances 6, no. 3 (2022): 998–1003, 10.1182/bloodadvances.2021004749. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Comont T., Heiblig M., Rivière E., et al., “Azacitidine for Patients With Vacuoles, E1 Enzyme, X‐Linked, Autoinflammatory, Somatic Syndrome (VEXAS) and Myelodysplastic Syndrome: Data From the French VEXAS Registry,” British Journal of Haematology 196, no. 4 (2022): 969–974, 10.1111/bjh.17893. [DOI] [PubMed] [Google Scholar]
- 19. Caggiano V., Sbalchiero J., Frassi M., et al., “Recurrent Fever and Association With Severe Organ Involvement, Mortality and Treatment Outcomes in VEXAS Syndrome: Data From the AIDA Network,” Frontiers in Immunology 17 (2026): 1753412, 10.3389/fimmu.2026.1753412. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Pietsch D., Kulasekararaj A., and Savic S., “Navigating Through Uncertainty‐Experience From the UK National VEXAS MDT,” British Journal of Haematology 208, no. 4 (2026): 1306–1313, 10.1111/bjh.70365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Beck D. B., Heiblig M., Savic S., et al., “PAXIS: A Randomized, Double‐Blind, Placebo‐Controlled, Dose‐Finding Phase 2 Study (Part 1) Followed by an Open‐Label Period (Part 2) to Assess the Efficacy and Safety of Pacritinib in Patients With VEXAS Syndrome,” Journal of Clinical Medicine 15, no. 4 (2026): 1426, 10.3390/jcm15041426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Vitale A., Caggiano V., Leone F., et al., “Efficacy and Safety Profile of Biotechnological Agents and Janus Kinase Inhibitors in VEXAS Syndrome: Data From the International AIDA Network VEXAS Registry,” Frontiers in Pharmacology 16 (2025): 1462254, 10.3389/fphar.2025.1462254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Hadjadj J., Nguyen Y., Mouloudj D., et al., “Efficacy and Safety of Targeted Therapies in VEXAS Syndrome: Retrospective Study From the FRENVEX,” Annals of the Rheumatic Diseases 83, no. 10 (2024): 1358–1367, 10.1136/ard-2024-225640. [DOI] [PubMed] [Google Scholar]
- 24. Zaimoku Y., Imi T., Hatada T., et al., “Prevalence and Outcome of VEXAS Syndrome in Unrelated Hematopoietic Cell Transplantation for Bone Marrow Failure,” Clinical and Experimental Medicine 25, no. 1 (2025): 300, 10.1007/s10238-025-01832-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Fernandez J. A., Kusne Y., Gutierrez‐Rodrigues F., et al., “DNMT3A Mutant Clonal Hematopoiesis in Vexas Syndrome Epigenetically Potentiates IL‐13 and IL‐15 Inflammatory Pathways and Is Associated With Increased Mortality,” Blood 144, no. Supplement 1 (2024): 2686–2686, 10.1182/blood-2024-204334.39724163 [DOI] [Google Scholar]
- 26. Kirino Y., Maeda A., Higashitani K., et al., “Peripheral Blood UBA1 Variant Burden Predicts Poor Outcomes in VEXAS Syndrome: A Nationwide Prospective Study,” Annals of the Rheumatic Diseases, ahead of print, April 1, 2026, 10.1016/j.ard.2026.03.003. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting Information: Describe the inclusion and exclusion criteria for the study.
Figure S1: The number of patients screened and included in the analysis.
Table S1: Treatment details of the hypomethylating agent therapy in 35 patients with VEXAS syndrome.
Table S2: Clinical characteristics and laboratory parameters of 53 patients aged ≤ 75 years at diagnosis with VEXAS syndrome stratified by treatment with allogeneic hematopoietic stem cell transplantation (alloHSCT) or hypomethylating therapy (HMA).
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.