Decline in lung function does not predict future decline in lung function in cystic fibrosis patients.

Abstract

Despite the attention paid to minimizing lung function decline among cystic fibrosis (CF) patients, the effect of rate of decline on subsequent disease progression is poorly understood. We aimed to describe the rate of decline of FVC, FEV1 , and FEF25-75 and to test the hypothesis that rate of decline of each spirometric variable predicts subsequent rate of decline in that variable and each other variable. Data were from the Epidemiologic Study of CF, an observational study of North American CF patients from 1994 to 2005. For each year of age, patients' best percent predicted FEV1 and associated FVC and FEF25-75, were used to calculate 2-year slopes for each spirometric variable. Pearson correlations were calculated between reference slopes and follow-up slopes up to 8 years later and, for FEV1 , between reference slopes and level (not slope) of lung function up to 5 years later. Twenty six thousand, three hundred and ninety-three patients contributed 427,063 spirometries. Median 2-year slopes of all variables were negative for all ages >6 years and the magnitude varied with age, being greatest among 13-17 year olds, especially for FEF25-75 . There was no correlation (r < 0.10) between reference slopes and subsequent slopes 3-8 years later, either within or across variables. The correlation between 2-year FEV1 slopes and FEV1 level even 5 years later was moderate (0.37-0.49) across disease stage categories. Contrary to our hypothesis, rate of lung function decline did not predict future rate of decline either within or across spirometric variables. In contrast, FEV1 slope did have moderate predictive ability for subsequent FEV1 level. These findings are relevant for clinical care and for clinical trial design.