Sustained donor engraftment and clinical remission after hematopoietic cell transplantation for vexas syndrome, but with observed toxicities: Results from a prospective phase II clinical Trial

Abstract

Abstract VEXAS is a recently described autoinflammatory disorder with bone marrow failure, caused by somatic mutations in UBA1. Disease is largely refractory to medical therapy leading to glucocorticoid dependence. Allogeneic hematopoietic cell transplant (HCT) can provide a curative option. Genomically confirmed VEXAS patients were enrolled (NCT05027945) if they had any of the following: 1) VEXAS inflammatory phenotype requiring ≥0.5 mg/kg/day of prednisone, 2) Significant cytopenia or myeloid neoplasm, 3) Steroid refractory disease. Primary objectives included sustained donor engraftment by day 100 and reversal of VEXAS clinical phenotype without steroid dependence by 1 year. Patients with 8/8 HLA matched donors (cohort 1; n=6) received fludarabine 40 mg/m2 on D-6 to -3 and busulfan (daily 3600-4200 uM-min, total 44-52 mg*h/L) on D-6 to -4. Patients with 7/8 HLA or haploidentical-related donors (cohort 2; n=1) received fludarabine 30 mg/m2 on D-6 to -2, cyclophosphamide 14.5 mg/kg D-6 -5, 200 cGy TBI on D-1, and busulfan on D-4, -3 (daily 3600 uM-min, total 30 mg*h/L). GVHD prophylaxis included post-transplant cyclophosphamide on D+3, +4, MMF to D+35 and sirolimus to D+180. Pre-HCT dose of glucocorticoids was reinitiated D+5 and tapered depending on symptoms in patients 4-7. Levels of 23 cytokines (compared to healthy controls [HC]; Luminex), peripheral blood (PB) flow cytometry for T-, B-, NK-, and dendritic cells, and monocytes, and error corrected sequencing for UBA1 mutation (minimum VAF of 0.5%) were performed pre-HCT, D+30, D+180 and yearly. Median age of 7 patients transplanted since March 2023 was 63 years. Disease burden was high with median of 5 (range 4-8) organ systems involved. Most had comorbidities with a median HCT-CI score of 4 (0-5). Patients were heavily pre-treated with a median of 4 prior steroid-sparing therapies (range 0-6) and all were steroid-dependent with median prednisone dose pre-HCT of 25mg/day (15-40mg/day). Steroid-sparing therapies (n=5) immediately prior to HCT included anti-IL-1 (n=2), anti-IL-6 (n=2) and JAKi (n=1). Median follow-up was 490 days. Six evaluable patients had sustained donor engraftment based on ANC and chimerism by D100; one patient who died on D+87 had full chimerism D+60. Median time of neutrophil engraftment was D+17 (13-25), and platelet engraftment was D+28 (13-63). Four patients had poor graft function requiring use of TPO agonists. By D+30, low-level UBA1 was detected in just 1 patient at VAF 0.6% and absent in the rest (n=6); all evaluable patients were UBA1 negative by D+180 (n=5), and 1 year (n=3). At 1 year, all 5 evaluable patients had sustained engraftment and remained in clinical remission. Myelotoxicity including mucositis and infections were observed in nearly all patients. Four patients experienced VEXAS flares around the time of engraftment requiring the use of high-dose steroids +/- tocilizumab, including, fever/altered mental status, labyrinthitis, small bowel inflammation/obstruction, and facial rash. CRP up-trended peri-engraftment in all patients (median 85.6mg/L [range 44.4-268]). Grade III/IV acute GVHD occurred in 28% (n=2) patients (lower GI n=2, liver n=1, skin n=2); one ultimately died from steroid refractory aGVHD and the other from infectious complications. Acute Grade I/II skin GVHD occurred in another 2 patients, managed with topical steroids. Two patients developed moderate chronic skin GVHD, both of whom have successfully tapered off steroids and remain on ruxolitinib and belumosudil. Cytokines elevated pre-HCT, including TNFa, IL6, IL18, and sCD25, were further increased at engraftment. Both patients with Grade III-IV acute GVHD had substantial increase in IL27, TNFRI, and REG3 at engraftment, and on D+30, and in CCL-2 after conditioning. Pre-HCT PB flow cytometry analyses showed increased CD57+CD4+ and CD57+CD8+ T cells, mature B cells, and CD57+ NK cells. At D+30, PD-1+CD4+ and PD-1+CD8+, and immature NK cell subfractions were increased compared to pre-HCT. Successful engraftment with eradication of UBA1 and resolution of VEXAS phenotype was achieved, but we observed significant systemic inflammation in early peri-HCT period requiring mitigating therapies. GVHD rates were high, resulting in mortality. Reduction in myelotoxicity, better control of early inflammation, and optimization of GVHD prophylaxis, have been implemented going forward. Comprehensive profiling of immunologic changes during HCT may inform future strategies.