Absence of NLRP3 somatic mutations and VEXAS‐related UBA1 mutations in a large cohort of patients with Schnitzler syndrome

To the Editor, Schnitzler's syndrome (SchS) is an extremely rare systemic autoinflammatory disease (SAID) characterized by a late onset of urticarial rash, recurrent fever, bone pain, arthralgia, elevated acutephase reactants, and a monoclonal gammopathy involving IgMκ chains (classical type) and rarely IgG (variant type). To date, about 300 cases have been reported worldwide.1 The diagnosis of SchS is currently based on a set of clinical criteria, that is, the Strasbourg diagnostic criteria.2 It is of prime importance to establish this diagnosis given the therapeutic efficacy of interleukin1 antagonists.3 No genetic predisposition has been formally identified so far. Two myeloid lineagerestricted somatic NLRP3 mutations were reported in 2015 in two SchS patients with the IgGκ variant type.4 Nevertheless, no NLRP3 mutation was identified in subsequent studies involving 32 SchS patients.5,6 More recently, somatic mutations in exon 3 of UBA1 were reported in a syndrome affecting men with adultonset SAID and named VEXAS (vacuoles, E1 enzyme, Xlinked, autoinflammatory, somatic).7 Notably, in this first VEXAS study, 20% of the patients presented multiple myeloma or monoclonal gammopathies of unknown significance (MGUS), thereby making UBA1 a potential candidate gene for SchS. These data prompted us to assess the contribution of NLRP3 and UBA1 somatic mutations, as well as of the main genes so far involved in SAID patients with cutaneous lesions in SchS pathophysiology. To this end, we screened for mutations of all these genes in the largest cohort of SchS patients constituted so far (40 unrelated cases) and at the origin of the validation of Strasbourg diagnostic criteria of SchS.2 We performed deeptargeted nextgeneration sequencing (NGS— mean sequencing depth of 943X) of 19 genes so far involved in SAIDs with cutaneous manifestations (NLRP3, NLRC4, MEFV, MVK, TNFRSF1A, IL1RN, IL36RN, LPIN2, PSTPIP1, TNFAIP3, IL36RN, CARD14, IL10, IL10RA, IL10RB, NOD2, PLCG2, LYN, and NLRP1), and Sanger sequencing of exon 3 of UBA1 (NM_003334) in all patients. None of the 40 SchS patients carried a somatic NLRP3 mutation. Notably, the two previously reported patients with a diagnosis of SchS and an NLRP3 mosaic mutation4 had a transient monoclonal IgG, an observation raising the possibility of a lateonset NLRP3AID diagnosis. Indeed, lateonset urticarial lesions can represent the main clinical manifestation in patients with NLRP3 mosaic mutations.8 As for UBA1, the gene involved in VEXAS patients among whom 20% have multiple myeloma or MGUS,7 no somatic mosaic mutation was identified in the third exon of UBA1 in our cohort of 40 patients with SchS. These results, therefore, do not support the hypothesis of UBA1 somatic mutations in SchS pathophysiology. The study of 18 other genes involved in a SAID phenotype associated with cutaneous manifestations identified no somatic mosaic mutations in the 40 SchS patients. We, however, identified 33 rare heterozygous germline variations (Table S1). According to the American College of Medical Genetics recommendations used to assess their pathogenicity, 24 variations were classified as likely benign and 9 as variations of unknown significance (VUS) (i.e., 4 in NLRP1, 2 in IL10RA, 1 in IL10RB, 1 in NOD2, and 1 in TNFRSF1A) (Table S1). Although 4 variations were found in NLRP1, it is important to underline that all of them have a CADD (Combined AnnotationDependent Depletion score, https://cadd.gs.washi ngton.edu/) score <15, a finding that argues against their pathogenicity. In conclusion, no somatic or germline pathogenic variations have been identified in NLRP3, UBA1, and 18 other SAID genes involved in a disease phenotype associated with cutaneous signs, in the large cohort of 40 SchS patients at the origin of the validation of Strasbourg diagnostic criteria of SchS.