Daily Home Spirometry: An Effective Tool for Detecting Progression in Idiopathic Pulmonary Fibrosis

Abstract

Rationale: Recent clinical trial successes have created an urgent need for earlier and more sensitive endpoints of disease progression in idiopathic pulmonary fibrosis (IPF). Domiciliary spirometry permits more frequent measurement of FVC than does hospital-based assessment, which therefore affords the opportunity for a more granular insight into changes in IPF progression.

Objectives: To determine the feasibility and reliability of measuring daily FVC in individuals with IPF.

Methods: Subjects with IPF were given handheld spirometers and instruction on how to self-administer spirometry. Subjects recorded daily FEV₁ and FVC for up to 490 days. Clinical assessment and hospital-based spirometry was undertaken at 6 and 12 months, and outcome data were collected for 3 years.

Measurements and main results: Daily spirometry was recorded by 50 subjects for a median period of 279 days (range, 13-490 d). There were 18 deaths during the active study period. Home spirometry showed excellent correlation with hospital-obtained readings. The rate of decline in FVC was highly predictive of outcome and subsequent mortality when measured at 3 months (hazard ratio [HR], 1.040; 95% confidence interval [CI], 1.021-1.062; P ≤ 0.001), 6 months (HR, 1.024; 95% CI, 1.014-1.033; P < 0.001), and 12 months (HR, 1.012; 95% CI, 1.007-1.016; P = 0.001).

Conclusions: Measurement of daily home spirometry in patients with IPF is highly clinically informative and is feasible to perform for most of these patients. The relationship between mortality and rate of change of FVC at 3 months suggests that daily FVC may be of value as a primary endpoint in short proof-of-concept IPF studies.

Keywords: biomarker; clinical trials; interstitial lung disease; personalized medicine.

Figure 1.

Comparison of hospital- and home-based spirometry readings. Bland–Altman plots comparing hospital- and home-based readings of (A) FVC and (B) FEV₁. Hospital-based readings were those obtained immediately before enrollment of subjects into the study. The value for home-based readings was taken as the mean of all the daily readings recorded by subjects during the first 7 days of the study.

Figure 2.

Individual examples of disease behavior. Daily FVC measurements for subjects with (A) inexorably progressive disease, (B) rapidly progressive disease, and (C) an acute exacerbation. Each point represents a single FVC measurement. The subject in A died of respiratory failure at 725 days. The subject in B died at Day 202, and the subject in C, despite losing 20% of FVC in a 3-week period, survived until Day 952.

Figure 3.

Rate of decline in FVC. Linear regression lines for each study subject demonstrate the rate of annual decline in FVC relative to baseline. Linear regression was performed using all readings available between baseline and Day 365 without imputation.

Figure 4.

Relationship between 3- and 12-month rate of FVC change and subsequent survival. Kaplan–Meier plots demonstrate the effect of rate of change in FVC on subsequent survival at (A) 3 months and (B) 12 months. At 12 months, subjects were dichotomized into those with >10% rate of decline in FVC (red line) (n = 28) or <10% FVC rate of change (blue line) (n = 22). At 3 months, subjects were dichotomized into those with >5% rate of change in FVC (red line) (n = 19) or <5% rate of change in FVC (blue line) (n = 31). Rate of change was calculated by linear regression analysis of all points between baseline and 3 and 12 months, respectively. Rate of change is reported relative to baseline values, which were calculated by taking the mean of all the daily readings recorded by subjects during the first 7 days of the study.

Figure 5.

Three-month rate of change predicts disease progression at 1 year. Disease progression at 1 year was defined as death or >10% relative decline in hospital-measured FVC at 12 months. (A) Scatter plot with box and whiskers demonstrating the relationship between rate of change in FVC at 3 months as measured by daily home spirometry and subsequent disease progression at 12 months. (B) Receiver-operating curve analysis assessing predictive value of 3-month FVC rate of change in determining subsequent disease progression. A 5% rate of change in 3-month FVC had a 62.1% sensitivity (i.e., two-thirds of those with disease progression at 1 yr could be identified at 3 months on the basis of having a 5% rate of FVC decline) and a specificity of 90.5% (i.e., <1 in 10 subjects with a 5% rate of decline in FVC at 3 months subsequently failed to fulfill the criteria for disease progression at 1 yr). AUC = area under the curve.

Figure 6.

Twelve-month change in home- but not hospital-based spirometry is strongly predictive of subsequent outcome. Subjects were dichotomized into those with progressive (>10% change in hospital-based FVC between baseline and 12 months or >10% annual rate of change in home-based FVC) and relatively stable (<10% change in hospital-based FVC between baseline and 12 months or <10% annual rate of change in home-based spirometry) disease. (A) Kaplan–Meier plot of stable and progressive patients based on hospital-measured FVC (n = 37; 13 patients died before 12 months and were therefore unevaluable). (B) Kaplan–Meier landmark analysis of 12-month rate of change in home FVC (n = 37 as the 13 deaths before 12 months are excluded from the landmark analysis).