LUNG INVOLVEMENT in VEXAS SYNDROME

ABSTRACT

Background: Vacuoles, E1 enzyme, X-linked, autoinfammatory, somatic (VEXAS) syndrome is a recently identifed disorder caused by somatic mutations in the UBA1 gene of myeloid cells. Various manifestations of pulmonary involvement have been reported, but a detailed description of lung involvement and radiologic fndings is lacking. Objectives: To describe lung involvement in VEXAS syndrome. Methods: A retrospective cohort study was conducted of all patients iden-tifed at the Mayo Clinic with VEXAS syndrome since October 2020. Clinical records and chest high resolution computed tomography (HRCT) scans were reviewed. Results: Our cohort comprised 22 white men with a median age of 69 years (IQR 62-74, range 57-84). Hematologic disorders including multiple myeloma, myelodysplastic syndrome and pancytopenia were present in 10 patients (45%), rheumatologic diseases including granulomatosis with poly-angiitis, IgG4-related disease, polyarteritis nodosa, relapsing polychondritis, and rheumatoid arthritis were found in 10 patients (45%), and 4 patients had dermatologic presentations including Sweet syndrome, Schnitzer-like syndrome or drug rash with eosinophilia skin syndrome (DRESS). VEXAS syndrome-related features included fever (18, 82%), skin lesions (20, 91%), lung infiltrates (12, 55%), chondritis (10, 45%), venous thromboembolism (12, 55%), macrocytic anemia (21, 96%), and bone marrow vacuoles (21, 96%). Other manifestations observed were arthritis, scleritis, hoarseness and hearing loss. Median erythrocyte sedimentation rate (ESR) was 69 mm/1st hour (IQR 34.3-118.8) and median C-reactive protein (CRP) of 55.5 mg/dL (IQR 11.4-98.8). The somatic mutations affecting methionine-41 (p.Met41) in UBA1 gene were 11 (50%) p.Met41Thr, 7 (32%) p.Met41Val, 2 (9%) p.Met41Leu, and 2 (9%) in the splice site. All patients received glu-cocorticoids (GC) (median duration of treatment was 2.6 years);21 (96%) received conventional immunosuppressive agents (methotrexate, aza-thioprine, mycophenolate, leflunomide, cyclosporin, hydroxychloroquine, tofacitinib, ruxolitinib) and 9 (41%) received biologic agents (rituximab, tocilizumab, infliximab, etanercept, adalimumab, golimumab, abatacept). Respiratory symptoms included dyspnea and cough present in 21 (95%) and 12 (55%), respectively, and were documented prior to VEXAS diagnosis. Most of the patients were non-smokers (14, 64%) and obstructive sleep apnea (OSA) was present in 11 patients (50%). Seven patients (32%) used non-invasive ventilation, 6 used C-PAP, and 1 used Bi-PAP. Bronchoalveolar lavage (BAL) was available in 4 patients, and the findings were compatible with neutrophilic alveolitis in 3. Two patients had lung biopsies (2 transbronchial and 1 surgical) that showed ATTR amyloidosis and organizing pneumonia with lymphoid interstitial pneumonia, respectively. Pulmonary function tests were available in 9 (41%) patients and showed normal results in 5;3 patients had isolated reduction in DLCO and 1 with mild restriction. On chest HRCT, 16 patients (73%) had parenchymal changes including ground-glass opacities in 9, septal thickening in 4, and nodules in 3;pleural effusions were present in 3 patients, air-trapping in 3 patients and tracheomalacia in 1 patient. Follow-up chest HRCT was available for 8 patients (36%), the ground-glass opacities resolved in 5 patients, 3 patients manifested new or increased ground-glass opacities, and 1 patient had increased interlobular septal thickening. After 1 year of follow-up, 4 patients (17%) had died;3 due to pneumonia (2 COVID-19,1 bacterial) and 1 due to heart failure. VEXAS flares occurred in 18 patients (82%), the maximum number of relapses was 7, and they were mainly managed with GC and with changes in the immuno-suppressive regimen. Conclusion: Pulmonary involvement was documented by chest HRCT in most patients with VEXAS syndrome. Respiratory symptoms occurred in over one half of patients and about 20% had PFT abnormalities. The pulmonary manifestations of VEXAS are nonspecifc and characterized predominantly by infamma-tory parenchymal involvement.