Predicting pulmonary fibrosis disease course from past trends in pulmonary function

Abstract

Background: The clinical course of idiopathic pulmonary fibrosis (IPF) is characterized by progressive decline in lung function and eventual mortality. We sought to determine if future declines in pulmonary function, mortality, or both can be predicted from prior trends in pulmonary function tests (PFTs).

Methods: Data from 1981 to 2008 on 4,431 PFTs and mortality were analyzed from 734 subjects with IPF. The Kaplan-Meier method was used for mortality analyses. Mixed models were used to describe longitudinal pulmonary function dynamics, since PFTs were observed at varying time points from baseline.

Results: During the first year of follow-up, 135 subjects (73%) had stable FVC while 50 subjects (37%) showed a decline in FVC. During months 12 to 24 (1-2 years after diagnosis), a stable FVC occurred with the same frequency among both subjects whose FVC had declined during year 1 and whose FVC had remained stable (84.0% and 80.7%, respectively; P=.59). Among subjects alive at the end of year 1, those with a stable FVC were more likely to be alive at the end of year 2 than those whose FVC declined (hazard ratio [HR], 0.91 [95% CI, 0.87-0.94] and HR, 0.71 [95% CI, 0.62-0.78], respectively).

Conclusions: PFT decline predicts early mortality, but not future declines in physiology, regardless of time since diagnosis.

Figure 1.

Mixed models analysis of the trend in FVC (solid lines: mean FVC, dashed lines: 95% CI) for patients with a stable FVC the year prior (solid line) vs those who declined the year prior (tick marks). *P < .0001 for the difference in intercept between stable and declined; P < .0001 for the difference in trajectory. #P < .0001 for the difference in intercept between stable and declined for the difference in trajectory 0.78. PFT = pulmonary function test.