Epidemiology, genetics, and subtyping of preserved ratio impaired spirometry (PRISm) in COPDGene

Abstract

Background: Preserved Ratio Impaired Spirometry (PRISm), defined as a reduced FEV1 in the setting of a preserved FEV1/FVC ratio, is highly prevalent and is associated with increased respiratory symptoms, systemic inflammation, and mortality. Studies investigating quantitative chest tomographic features, genetic associations, and subtypes in PRISm subjects have not been reported.

Methods: Data from current and former smokers enrolled in COPDGene (n = 10,192), an observational, cross-sectional study which recruited subjects aged 45-80 with ≥10 pack years of smoking, were analyzed. To identify epidemiological and radiographic predictors of PRISm, we performed univariate and multivariate analyses comparing PRISm subjects both to control subjects with normal spirometry and to subjects with COPD. To investigate common genetic predictors of PRISm, we performed a genome-wide association study (GWAS). To explore potential subgroups within PRISm, we performed unsupervised k-means clustering.

Results: The prevalence of PRISm in COPDGene is 12.3%. Increased dyspnea, reduced 6-minute walk distance, increased percent emphysema and decreased total lung capacity, as well as increased segmental bronchial wall area percentage were significant predictors (p-value <0.05) of PRISm status when compared to control subjects in multivariate models. Although no common genetic variants were identified on GWAS testing, a significant association with Klinefelter's syndrome (47XXY) was observed (p-value < 0.001). Subgroups identified through k-means clustering include a putative "COPD-subtype", "Restrictive-subtype", and a highly symptomatic "Metabolic-subtype".

Conclusions: PRISm subjects are clinically and genetically heterogeneous. Future investigations into the pathophysiological mechanisms behind and potential treatment options for subgroups within PRISm are warranted.

Trial registration: Clinicaltrials.gov Identifier: NCT000608764.

Trial registration: ClinicalTrials.gov NCT00608764 NCT00608764.

Figure 1

Distribution of spirometry in COPDGene. Legend: Forced expiratory volume in the first second (FEV₁) is plotted on the x-axis while FEV₁/FVC ratio is plotted on the y-axis. Dashed lines represent fixed-threshold criteria used to delineate Preserved Ratio Impaired Spirometry (PRISm) subjects (highlighted in blue-upper left quadrant), control (upper right quadrant), and COPD (lower left quadrant) subjects.

Figure 2

Overview of cluster analysis in subjects with Preserved Ratio Impaired Spirometry (PRISm).

Figure 3

Results of k-means clustering in fixed-threshold defined P reserved R atio I mpaired S pirometry (PRISm) in non-Hispanic whites. Legend: Unsupervised k-means clustering was performed in non-Hispanic white subjects with PRISm. The first two principal components generated using the scaled 6 key input variables used for clustering (body mass index, FEV₁%, predicted, FEV₁/FVC ratio, percent emphysema, total lung capacity, and segmental wall area) are plotted on the x- and y-axes respectively.

Figure 4

Results of k-means clustering in fixed-threshold defined P reserved R atio I mpaired S pirometry (PRISm) in African Americans. Legend: Unsupervised k-means clustering was performed in African American subjects with PRISm. The first two principal components generated using the scaled 6 key input variables used for clustering (body mass index, FEV₁%, predicted, FEV₁/FVC ratio, percent emphysema, total lung capacity, and segmental wall area) are plotted on the x- and y-axes respectively.