Objective: Using a combination of clinical, radiographic, functional, and serum protein biomarker assessments, this study was aimed at defining the prevalence and clinical characteristics of interstitial lung disease (ILD) in a large cohort of patients with anti-Jo-1 antibodies.
Methods: A review of clinical records, pulmonary function test results, and findings on imaging studies determined the existence of ILD in anti-Jo-1 antibody-positive individuals whose data were accumulated in the University of Pittsburgh Myositis Database from 1982 to 2007. Multiplex enzyme-linked immunosorbent assays (ELISAs) for serum inflammation markers, cytokines, chemokines, and matrix metalloproteinases in different patient subgroups were performed to assess the serum proteins associated with anti-Jo-1 antibody-positive ILD.
Results: Among the 90 anti-Jo-1 antibody-positive individuals with sufficient clinical, radiographic, and/or pulmonary function data, 77 (86%) met the criteria for ILD. While computed tomography scans revealed a variety of patterns suggestive of underlying usual interstitial pneumonia (UIP) or nonspecific interstitial pneumonia, a review of the histopathologic abnormalities in a subset of patients undergoing open lung biopsy or transplantation or whose lung tissue was obtained at autopsy (n = 22) demonstrated a preponderance of UIP and diffuse alveolar damage. Analysis by multiplex ELISA yielded statistically significant associations between anti-Jo-1 antibody-positive ILD and elevated serum levels of C-reactive protein (CRP), CXCL9, and CXCL10, which distinguished this disease entity from idiopathic pulmonary fibrosis and anti-signal recognition particle antibody-positive myositis. Recursive partitioning further demonstrated that combinations of these and other serum protein biomarkers can distinguish these disease subgroups at high levels of sensitivity and specificity.
Conclusion: In this large cohort of anti-Jo-1 antibody-positive individuals, the incidence of ILD approached 90%. Multiplex ELISA demonstrated disease-specific associations between anti-Jo-1 antibody-positive ILD and serum levels of CRP as well as the interferon-gamma-inducible chemokines CXCL9 and CXCL10, highlighting the potential of this approach to define biologically active molecules contributing to the pathogenesis of myositis-associated ILD.