Two studies report genome-wide association studies for lung function, using cross-sectional spirometric measurements in healthy individuals. They identify six genetic loci newly associated to natural variation in lung function, which may have implications for the related airway diseases of asthma and chronic obstructive pulmonary disease.
Asthma is a clinical syndrome defined by spontaneous or chemically induced increased airway responsiveness (bronchoconstriction) and reversible airflow obstruction (bronchodilation) with airway inflammation, often allergic in nature. In contrast, chronic obstructive pulmonary disease (COPD) is an airway disease with inflammation, but here the source of the inflammation is usually cigarette smoking, and it is defined by fixed, rather than reversible, airflow obstruction1. On pages 36 and 45 of this issue, two studies report genome-wide association studies (GWAS) identifying loci associated with lung function in predominantly healthy individuals as measured by spirometry2,3. Together, these studies report six loci newly associated with natural variation in lung function, bringing new insight into our understanding of the genetic basis of lung development and the related airway disorders asthma and COPD.
Asthma is primarily a disease of early childhood, with 80% of all cases diagnosed by 6 years of age. In contrast, COPD is a disease of later life, with most cases being diagnosed after age 60 years. Lung function abnormalities are present in both disorders. The primary environmental causes of both asthma and COPD are respiratory infections (viral and, in the case of COPD, bacterial) and cigarette smoking (passive and active). Although only 15% of active cigarette smokers develop COPD, 80% of individuals with COPD have increased airway responsiveness, one of the phenotypes defining asthma3. Thus, many investigators believe that smoking’s effect on COPD risk is largely defined by the concomitant presence of asthma or increased airway responsiveness and by the ability of these intermediate phenotypes to reduce growth and accelerate decline in lung function.
Lung function measurements
Spirometry is defined by the dynamic maneuver of taking in a maximal deep breath and exhaling it as hard and as fast as one can. The forced expiratory volume in 1 second (FEV1) is the amount of air that is blown out in 1 second after this maximal deep inhalation. The forced vital capacity (FVC) is defined as the volume of air expired after a maximal inhalation and then a maximal forced exhalation of 6 seconds duration or greater. The ratio of these two measures (FEV1/FVC) is a measure of airflow obstruction. An FEV1/FVC ratio less than 70% of that predicted after administration of a short-acting bronchodilator (beta-2 agonist) defines COPD. Repapi et al.2 and Hancock et al.3 include measurements of both FEV1 and FEV1/FVC, but do not include measurements on the change in FEV1 after the individual has taken either a bronchodilator (a measure of reversibility) or a bronchoconstrictor (a measure of airway responsiveness). These pharmacologically enhanced spirometric tests are measured as the change in FEV1, at 15 minutes after two puffs of a short-acting beta agonist (bronchodilator) and the provocative dose of histamine or methacholine inducing a 20% decrease in FEV1 from baseline (PD20, increased airway responsiveness). Specific values of these tests (bronchodilator response >10% and PD20 <12.5 mg methacholine) define asthma and are also present in most individuals with COPD and some healthy individuals. These measures of lung function are correlated with each other and have been found to be heritable, with values of h2 (the narrow-sense heritability due to genetic factors), as estimated from twin and family studies, ranging from 0.7 for FEV1 and FVC to 0.35 for increased airway responsiveness—indicating that there is a substantial genetic component for all lung function measures4,5.
Repapi et al.2 performed a meta-analysis of 14 studies totaling 20,288 individuals as part of the SpiroMeta consortium, with replication in the Health 2000 study and the CHARGE Consortium study (21,209 individuals). Hancock et al.3 in the CHARGE Consortium performed a similar meta-analysis of 20,890 individuals with replication in the SpiroMeta consortium (20,228 individuals). These two studies each involved cross-sectional lung function data (FEV1 and FEV1/FVC measured at a single point in time), without use of a bronchodilator or a bronchoconstrictor. The study cohorts included predominantly individuals with normal lung function, although there were a small number of subjects with self-described asthma or COPD (the associations remained significant even after excluding individuals with these diseases)2,3. They tested replication between these traits, performed an analysis stratified by ever- or never-smoking status, and also repeated analyses in a subset of the cohorts after excluding individuals with self-reported asthma and COPD. Finally, Repapi et al.2 also performed gene expression studies on several of their candidate genes.
Genetics of lung function
The first GWAS report on lung function was reported by the Framingham Heart Study based on Affymetrix 100K data, suggesting an association to GSTO2 on chromosome 10 to FEV1, although this has not been replicated6. Subsequently, a study reported a GWAS to the FEV1/FVC ratio in 7,691 participants in the Framingham Heart Study, with replication in an additional 835 participants of the Family Heart Study7. They reported association of a locus on chromosome 4q31 with the percent of expected FEV1/FVC ratio defined in a fashion identical to the way it was defined in the two studies reported in this issue7. The associated variants were in an intergenic region upstream of HHIP, a hedgehog pathway gene with a known role in development. This locus was also reported to be associated with COPD, as defined by reduced FEV1/FVC ratio, in an independent GWAS8. SERPINE2 was selected from mouse lung–development microarray studies and then tested in a candidate gene association study with 127 probands from a family-based study of severe, early-onset COPD, with replication in 304 cases from the National Emphysema Treatment Trial and 441 controls from the Normative Aging Study9. Subsequent association studies confirmed this association for FEV1/FVC in individuals with COPD10. Thus, two of the genes previously identified as associated to lung development have also been shown to be associated to COPD.
The first finding from the current studies was replication of the HHIP association to the FEV1/FVC ratio2,3. The two studies also report six new genetic loci associated with lung function and implicate a number of biologically plausible candidate genes. This includes associations at 6p21 near two genes with immune functions, AGER and PPT2; at 2q35 near TSN1, encoding an actin-binding protein; at 6q24.1 within GPR126, which encodes a G protein–coupled receptor; and at 15q33 within a thrombospondin family gene, THSD4. The associated locus at 5q32–33 includes HTR4, a serotonin pathway gene, and the chromosome 4q24 locus includes four genes including GSTCD. Five of the six candidate genes (AGER, TNS1, HTR4, THSD4 and GSTCD) were shown to be expressed in lung tissue2. These loci are all biologically plausible candidates that are involved in pathways known to be important in lung biology, including immune function, muscle function or inflammation. These studies did not replicate the SERPINE2 association and also did not find evidence for association to the 15q CHRNA5-CHRNA3-CHRNB4 locus previously associated with nicotine dependence, lung cancer and COPD11,12,13.
Lung function levels depend strongly on age, height and gender, with distinct changes from childhood to adulthood (Fig. 1). Thus, because of study-design considerations, some gene variants identified in the two studies reported here may simply influence lung function in nondiseased subjects and will have no independent effect on patients with airways disease. However, the high within-individual consistency of spirometric lung function measures over time, and the links of the identified candidate genes to lung development, suggest that genetic variations associated with lung function development might be important genetic determinants of lung function in both healthy individuals and those with airways disease (asthma and COPD). To test this hypothesis, the newly associated variants should be tested in both individuals with asthma and those with COPD. These studies should also be extended to include bronchodilator and bronchoconstrictor responses.
These loci account for a small proportion of the variation in FEV1 and the FEV1/FVC ratio. Repapi et al.2 estimates that the five new loci identified in their study account for ∼0.14% of the variation in FEV1/FVC ratio. Translating this to clinical prediction remains further away, and the first steps toward this goal include additional validation, refinement of the genetic loci and determination of the specific genes and their functional variants, as well as more accurate estimation of effect sizes from additional studies of unaffected and diseased subjects.
About this article
Is Overall Mortality the Right Composite Endpoint in Clinical Trials of Acute Respiratory Distress Syndrome?*
Critical Care Medicine (2018)