Efficient detection of somatic UBA1 variants and clinical scoring system predicting patients with variants in VEXAS syndrome

To efficiently detect somatic UBA1 variants and establish a clinical scoring system predicting patients with pathogenic variants in VEXAS (vacuoles, E1 enzyme, X-linked, autoinflammatory, somatic) syndrome.Eighty-nine Japanese patients with clinically suspected VEXAS syndrome were recruited [81 males and 8 females; median onset age (IQR) 69.3 years (62.1-77.6)]. Peptide nucleic acid-clamping PCR (PNA-PCR), regular PCR targeting exon 3 clustering UBA1 variants, and subsequent Sanger sequencing were conducted for variant screening. Partitioning digital PCR (pdPCR) or targeted amplicon deep sequencing (TAS) was also performed to evaluate the variant allele frequency (VAF). We developed our clinical scoring system to predict UBA1 variant-positive and ‑negative patients and assessed the diagnostic value of our system using receiver operating characteristic (ROC) curve analysis.Forty patients with reported pathogenic UBA1 variants (40/89, 44.9%) were identified, including a case having a variant with VAF of 1.7%, using a highly sensitive method. Our clinical scoring system considering >50 years of age, cutaneous lesions, lung involvement, chondritis, and macrocytic anaemia efficiently predicted patients with UBA1 variants (the area under the curve for the scoring total was 0.908).Genetic screening with the combination of regular PCR and PNA-PCR detected somatic UBA1 variants with high sensitivity and specificity. Our scoring system could efficiently predict patients with UBA1 variants.