LAUSR

Introduction We studied the distribution and in vitro pathogenicity of anti-DSG3 IgG subclasses during the course of pemphigus vulgaris (PV). Methods We longitudinally studied the distribution of anti-DSG3 IgG subclasses (before versus after treatment) in sera from PV patients, using an addressable-laser bead immunoassay (ALBIA). The in vitro pathogenicity of corresponding sera was tested using keratinocyte dissociation and immunofluorescence assays. Results Sixty-five sera were assessed at baseline (33 from patients treated with rituximab and 32 with corticosteroids). Sixty-three percent of these baseline sera contained 2 or more anti-DSG3 IgG subclasses versus 35.7% of sera from patients in complete remission (CR) and 75.0% of sera from patients with persistent disease activity after treatment. IgG4 was the most frequently detected anti-DSG3 IgG subclass, both in patients with disease activity and in those in CR. The presence of three or more anti-DSG3 IgG subclasses was predictive of relapse, in particular when it included IgG3, with a positive predictive value of 62.5% and a negative predictive value of 92%. While anti-DSG3 IgG4 Abs from sera collected before treatment were most often pathogenic, anti-DSG3 IgG4 from sera collected after treatment were pathogenic only after adjusting their titer to the one measured before treatment. The IgG3 fraction containing anti-DSG3 Abs also had an in vitro pathogenic effect. The disappearance of the pathogenic effect of some sera after removal of anti-DSG3 IgG3 suggested an additional effect of this IgG subclass. Conclusion The serum levels and number of anti-DSG3 IgG subclasses drive the pathogenic effect of pemphigus sera and may predict the occurrence of relapses.