How can autoantibodies predict the long-term outcome of patients with interstitial lung disease? Results from a retrospective cohort study.

Abstract

This study aimed to investigate whether positive serum autoantibodies (AAbs) have any impact on survival and time evolution of radiological findings and pulmonary function indices in patients with interstitial lung disease (ILD). Ninety four patients with regular clinical, functional and high resolution computed tomography (HRCT) imaging follow-up for at least 12 consecutive months and complete testing for a panel of AAbs most commonly associated with ILD were enrolled in this retrospective two-center study. Eligible patients were divided into two groups based on the presence [ILD/AAb(+)] (n = 69) or absence [ILD/AAb(-)] (n = 25) of positive serum AAbs. All-cause mortality and longitudinal indicators of ILD progression such as a sustained decrease from baseline in absolute measurements of forced vital capacity (FVC) of ≥10% or single-breath diffusion capacity (DLCOSB) of ≥15% were the primary study endpoints. DLCOSB < 40% predicted on at least two consecutive measurements and progression of HRCT findings were our secondary endpoints. Kaplan-Meier (K-M) survival analysis and multivariate Cox proportional-hazards (PH) model were used to evaluate the prognostic significance of positive AAbs in the outcome of patients with ILD. ILD/AAb(+) patients were predominantly female (71% vs 32%), were significantly younger (54.8 ± 14.6 vs 66.8 ± 10.1 years), and had longer duration of follow-up (78.1 ± 53.1 vs 41.6 ± 26.7 months), compared with ILD/AAb(-) patients (p < .01 for each comparison). Baseline measurements of FVC (% pred.) and DLCOSB (% pred.) did not differ significantly between the two groups. At the end of follow-up, mortality rates and the percentage of patients with a sustained FVC decrease were lower in the ILD/AAb(+) group (p < .05 for each comparison). With the exception of DLCOSB < 40% pred., ILD/AAb(+) patients had a longer median time-to-event for each of the other studied outcomes (p < .01 for each K-M analysis). In addition, Cox PH models adjusted for age, smoking status, baseline pulmonary function tests and morphological pattern of ILD remained statistically significant in favor of the ILD/AAb(+) group (p < .05 for each comparison). AAb(+) patients with ILD seem to have a more favorable prognosis regarding all-cause mortality, long-term deterioration in lung function parameters and progression of HRCT findings than their AAb (-) counterparts.