Early radiographic and clinical features associated with bronchiectasis in children

Abstract

Bronchiectasis among children living in developing regions is associated with respiratory infections during early childhood, but specific risk factors that precede childhood bronchiectasis are not fully characterized. We hypothesized that severe respiratory syncytial viral (RSV) infection in infancy would increase the risk of bronchiectasis among Alaska Native children in rural Alaska. This was a follow-up cohort study of a 1993-1996 case-control study of RSV-hospitalized case patients and their controls. For each 5-8-year-old former case-patient and control subject, we reviewed medical records, interviewed parents, performed physical examinations and spirometry, collected sera, and analyzed all historical chest radiographs. Ten (11%) RSV cases and 10 (9%) controls had radiographic evidence of bronchiectasis. The mean age at radiographic diagnosis of bronchiectasis was 3.3 years (range, 1.2-6.1 years). Children were more likely to develop bronchiectasis if their chest radiographs, when they were < 2 years of age, showed lung parenchymal densities (RR = 3.9, P < 0.013), persistent parenchymal densities > 6 months' duration (RR = 3.0, P = 0.02), or infiltrates on multiple episodes (test for trend, P = 0.003). Radiographic features of hyperinflation and atelectasis among children < 2 years old were not associated with eventual bronchiectasis. A single severe infection with RSV alone did not predispose Alaska Native infants to bronchiectasis. Childhood bronchiectasis was associated with lung and hence airway injury, manifested on radiographs by parenchymal densities or "pneumonia" rather than by hyperinflation or atelectasis.