Patient genetics shape the autoimmune response in the blistering skin disease pemphigus vulgaris

BACKGROUND AND AIM: Pemphigus vulgaris (PV) is known to have one of the strongest HLA associations among autoimmune diseases. DRB1*0402 and DQB1*0503 in particular are significantly overrepresented in PV patients in certain worldwide populations. Yet, there remain significant gaps in our understanding regarding the precise link between PV-associated HLA molecules, the specificity of the autoimmune response, and clinical expression. In this study we assessed correlations between factors including HLA genotype, ethnicity, autoantibody levels, and lesion distribution in a cohort of 293 patients. METHODS AND POPULATION: Participants were recruited from multiple outpatient dermatology clinic settings and patient support meetings in the USA. On intake, patients provided venous blood samples and answered questionnaires regarding their current disease activity. RESULTS: Eighty-one percent of patients typed as either DRB1*0402 or DQB1*0503 with a high prevalence of DRB1*0402 in patients of Ashkenazi Jewish or Caucasian (non-Jewish) descent (86% and 42%, respectively) and DQB1*0503 in patients of Southeast Asian descent (78%). Patients typing as HLA DRB1*0402 had higher levels of anti-desmoglein (Dsg)3 antibodies (204.6 +/- 340.5 IU/ml) than patients without DRB1*0402 (138.5 +/- 236.4 IU/ml) (p=0.03) and had mucosal only lesions more often than cutaneous only or mucocutaneous lesions. Patients typing as DQB1*0503 had higher levels of anti-Dsg1 antibodies (47.3 +/- 59.8 IU/ml) compared to other groups (27.8 +/- 43.7 IU/ml) (p=0.06) and higher rates of mucocutaneous disease than other lesion types. We also report an unexpected HLA association of DRB1*0804 in PV patients of African descent. Sixty-four percent of this population carried the DRB1*0804 allele, and presented with highly elevated levels of anti-Dsg3 (p=0.02). However, neither African heritage nor the presence of DRB1*0804 correlated with a predilection to any specific lesion morphology. Patients that carried neither DRB1*0402, nor DQB1*0503 or DRB1*0804 had the lowest levels of anti-Dsg3 antibodies (60.0 +/- 80.0 IU/ml) and the highest rate of solely cutaneous disease compared to carriers of these alleles. CONCLUSION: Our data illuminate the broader impact of genetic factors on disease development by showing that differences in HLA expression among patients and ethnicities play a large role in driving distinct patterns of antibody selection and disease phenotype in PV. These findings provide insights regarding clinical heterogeneity, and are relevant to developing improved, patient tailored management strategies.