LAUSR

OBJECTIVES To determine whether development of ANCA-associated vasculitis (AAV) shows a relationship to a prior infection and if prior infection affects disease characteristics and outcome. METHODS All incident cases of AAV diagnosed in a defined region of Sweden from 2000 through 2016 were identified. For each case, 10 individuals from the general population, matched for age, sex, and area of residence, were selected. Infections occurring in AAV patients and controls prior to the date of AAV diagnosis (index date for respective controls) were identified using an administrative database. Conditional logistic regression models were used to calculate odds ratios (OR) of developing AAV. Occurrence, clinical characteristics, and outcome of AAV were analysed with respect to prior infection. RESULTS Two-hundred-seventy patients with AAV (48% female) and 2687 controls were included. Prior to diagnosis/index date, 146 (54%) AAV patients had been diagnosed with infection vs 1282 (48%) controls, with OR for AAV 1.57 (95% CI 1.18-2.19) in those with infections of the upper respiratory tract and 1.68 (1.02-2.77) in those with pneumonia. Difference from controls was significant in patients with myeloperoxidase (MPO-) 1.99 (95% CI 1.25-3.1) but not in those with proteinase-3 (PR3)-ANCA 1.0 (0.61-1.52). Patients with prior infection showed higher disease activity at AAV diagnosis. No differences in disease characteristics, comorbidities, or outcome in those with and without prior infections were observed. CONCLUSIONS Respiratory tract infections are positively associated with development of MPO- but not PR3-ANCA-vasculitis. Prior infection is associated with higher disease activity at AAV diagnosis.