Predictors of diagnosis and survival in idiopathic pulmonary fibrosis and connective tissue disease-related usual interstitial pneumonia.

BACKGROUND: Although usual interstitial pneumonia (UIP) appears to portend better survival when associated with connective tissue disease (CTD-UIP), little is known about the presenting clinical, radiologic, and pathologic features that differentiate pathologically confirmed UIP with CTD from idiopathic pulmonary fibrosis (IPF). In patients with atypical radiologic and clinical features, what specific findings predict underlying IPF vs. CTD-UIP diagnosis and their respective long term survival?

METHODS: A large retrospective cohort analysis was done of consecutive patients seen from 1995 through 2010 with biopsy confirmed UIP completed or reviewed at our institution. CTD-UIP was defined by independent rheumatology consultation with exclusion of all other secondary causes of lung fibrosis. Primary clinical data was collected and compared for IPF and CTD-UIP along with logistic regression performed for predictors of disease likelihood and Cox proportional hazards analysis for predictors of survival.

RESULTS: Six hundred and twenty five patients were included in the study of which 89 had diagnosed CTD-UIP representing 7 disease entities. Survival was better among those with CTD-UIP except in UIP associated with rheumatoid arthritis, which had similar presenting features and survival to IPF. Predictors of underlying CTD included female gender, younger age, positive autoimmune serology, and inconsistent presenting radiologic findings. Only age and forced vital capacity corrected for a priori covariates were predictive of survival in CTD-UIP.

CONCLUSIONS: UIP pathology occurs frequently among patients with atypically presenting clinical and radiologic features, and may represent IPF or CTD-UIP with improved prognosis if underlying CTD is diagnosed. Presenting radiologic and pathologic features alone are not predictive of underlying secondary cause or survival between the two groups.