Evidence That Asthma Is Heritable

A genetic basis for asthma was recognized as early as 1860 when Henry Salter reported that asthma was a heritable disease with “distinct traces of inheritance” among related individuals. Salter’s observation was confirmed in studies demonstrating that asthma prevalence was between 20% and 25% among first-degree relatives of asthma subjects, significantly higher than the approximately 5% prevalence in the general population. The clustering of asthma in families is estimated by comparing the risk for developing asthma in those related to an asthmatic proband with the risk for asthma in the general population as a lambda ratio (λ _R ). A higher λ _R indicates that the disease has greater heritability and, therefore, a stronger genetic component. For instance, asthma has a λ _R of 2 to 4 while an autosomal recessive disease such as cystic fibrosis has a much higher λ _R . The lower λ _R of a common, complex disease such as asthma is likely caused by multiple gene variants each with small effects interacting with factors not detected in heritability studies, including a diverse group of environmental exposures and disease triggers. The observed heritability for asthma demonstrated in these early studies provided the rationale for the comprehensive genetic studies discussed in this chapter.

The First Genome-Wide Screens for Susceptibility Loci: Family-Based Studies

The first genetic studies for asthma susceptibility, which were performed before the first genomes were sequenced, used family-based approaches. They used genetic variants that are equally spaced throughout the genome to identify chromosomal susceptibility regions for asthma or associated phenotypes (e.g., bronchial hyperresponsiveness, serum IgE levels, skin test atopy). In family-based genetic studies, linkage analysis is used to identify genetic loci that cosegregate or are coinherited with a trait such as asthma in families. The degree of linkage is measured as an LOD score, the “likelihood of the odds” that a genomic region and trait cosegregate ( Table 45-1 ). An LOD score of 3 or more is considered significant and is equal to a 1000 : 1 odds that a genomic region is in linkage or coinherited with a specific trait or phenotype. Family-based studies have been successful in identifying genetic diseases with an autosomal dominant or recessive mode of inheritance, but are less powerful for complex or multifactorial diseases wherein multiple genes interact to determine disease susceptibility or severity. Transmission disequilibrium testing is another family-based association test that compares the transmission of a genetic marker from both parents to their affected offspring with a disease or trait of interest (see Table 45-1 and Fig. 45-2 ).

Two commonly used association testing methods: family-based and case-control.

A, Family-based association testing. The basic fundamental unit of family-based association testing are the trios of affected (cases in red) probands. Trios consist of cases or affected probands and their parents, whose affected status is not necessarily known (in gray). In family-based studies, the over transmission of an allele of a polymorphism to affected probands is measured ( red arrows ). This demonstrates how a single nucleotide polymorphism (SNP) with alleles “A” or “G” shows overtransmission of the “A” allele to affected probands. The strength of the overtransmission or association at this locus is estimated using the P- value. B, Case-control association testing. Case-control association testing compares the frequency of the different alleles for this SNP in unrelated cases (in red) and controls (in blue) to determine an odds ratio (OR, red double-arrow) or genetic effect. The basic illustration demonstrates that the “A” allele is more frequent among cases compared to controls. The strength of the association at this locus is estimated using the P- value.

Early family-based genome-wide studies demonstrated that multiple genes are associated with development of asthma and of related phenotypes such as atopy and elevated serum immunoglobulin E (IgE). These studies, including those performed by the National Heart, Lung, and Blood Institute (NHLBI) Collaborative Study on the Genetics of Asthma (CSGA), provided evidence for linkage with asthma and related phenotypes in multiple different chromosomal regions: chromosome 5q, 6p, 7q, 11q, 12q, 14q, and 16q. The genetic markers identified by these family-based studies represent chromosomal regions that have subsequently been studied to identify specific risk genes responsible for these linkages ( eTable 45-1 ).

These early linkage studies demonstrated gene-gene interactions in asthma susceptibility between chromosomes 5q31, 8p23, and 12q22, and 15q13. In addition, these studies provided evidence for gene-by-environment interactions between passive cigarette smoking and asthma susceptibility at chromosomes 1p, 5q, and 9q, and 17q21. These gene-by-environment interactions provided the earliest examples of important epigenetic effects in asthma in which environmentally induced DNA modifications alter transcription without a change in gene sequence.

Early family-based linkage studies also provided evidence for genetic determinants of altered pulmonary function in asthma. In these studies, a quantitative trait locus on chromosome 5q33 and 2q32 was linked with important lung function measures. The quantitative trait locus in chromosome 2q32 was adjacent to the region linked to forced expiratory volume in 1 second/forced vital capacity (FEV ₁ /FVC) in families with early-onset COPD suggesting that there may be similar genes that affect the development of both asthma and COPD.

Linkage Studies Provide Important Insight into the Complexity of Asthma Pathogenesis

The limitations of family-based linkage studies include the challenge of recruiting and characterizing related subjects and the use of genetic markers that cover genomic regions that contain hundreds of genes, requiring more detailed analysis with fine mapping or positional cloning to identify individual susceptibility genes. Another important limitation of family-based linkage studies is that they were designed to identify loci containing gene variants with a strong effect on disease susceptibility and are underpowered to detect common gene variation with a weak or modest effect. Despite these limitations, novel asthma genes have been identified (see eTable 45-1 ) which provide insight into the polygenic pathogenesis and the importance of gene-environment interactions in asthma.

Linkage Analysis and Positional Cloning Reveal Novel Asthma Susceptibility Loci

A disintegrin and metalloprotease-33 gene (ADAM33) was the first gene identified for asthma susceptibility using a family-based linkage study followed by fine mapping and association studies with cases and controls (i.e., positional cloning). It was identified by positional cloning on chromosome 20p13. ADAM33 encodes for a membrane-anchored glycoprotein that controls cell-matrix interactions and is involved in regulation of lung growth and morphogenesis. The ADAM33 protein is expressed in airway smooth muscle cells and lung fibroblasts, contains a catalytic domain with proteolytic activity and affects airway remodeling. ADAM33 is among the most replicated genes for asthma susceptibility in different racial and ethnic groups. ADAM33 gene variants have been associated with accelerated decline in lung function in asthma and in COPD, suggesting a role of this gene in the pathogenesis of progressive airflow obstruction. This is further supported by the observation that bronchoalveolar lavage concentrations of soluble ADAM33 correlate with asthma severity and inversely correlate with lung function.

Other family-based whole-genome studies used positional cloning to identify the dipeptidyl peptidase-10 ( DPP10 ) gene on chromosome 2q14 as a locus for asthma susceptibility. A s ingle nucleotide polymorphism (SNP) in DPP10 was recently identified in a GWAS of African Americans and African Caribbeans, providing additional evidence that DPP10 is an asthma susceptibility locus. Dipeptidyl peptidases cleave the terminal peptides of cytokines and chemokines and enhance airway inflammation. Several other asthma susceptibility genes have been identified using fine mapping or positional cloning techniques and include the protocadherin-1 gene (PCDH1) on chromosome 5q31, human leukocyte antigen complex G (HLA-G) on chromosome 6p22, the n europeptide S receptor-1 gene (GPRA) on chromosome 7p14, and PHD finger protein-11 (PHF11) on chromosome 13q14 (see eTable 45-1 ). The identification of asthma susceptibility loci is an example of the use of unbiased genetic techniques, that is, without preconceived notions of mechanism, to delineate novel biologic mechanisms that are important in the pathogenesis of asthma.

Highly Replicated Candidate Genes on Chromosome 5q31 and Evidence for Gene-Gene Interactions

Linkage analyses of chromosome 5q31-35 have consistently provided evidence that this region is linked to asthma or closely related phenotypes. This inflammatory gene–rich region contains multiple candidate genes related to allergic inflammation, including IL3, IL4, IL5, IL13, CD14, serine peptidase inhibitor Kazal type 5 (SPINK5), the β ₂ -adrenergic receptor (ADRB2), and leukotriene C4 synthase (LTC4S). The gene encoding for IL-13 and IL-4 contains promoter SNPs and variants consistently associated with asthma and related phenotypes in multiple populations. IL-13 and IL-4 bind to the α chain of the IL-4 receptor encoded by IL4R on chromosome 16p12, which also contains variants associated with asthma and related phenotypes. The interaction of two variants in IL4R has been associated with risk for life-threatening asthma exacerbations in persistent asthmatics suggesting a role of this pathway in determining disease severity. In addition, interactions between genetic variants in IL4, IL13, and IL4R and other pathway-related genes such as STAT6 on chromosome 12q13 cumulatively determine asthma risk and serum IgE at a greater level than does each individual locus, supporting the role of multiple genes interacting in asthma susceptibility and pathogenesis. Finally, one of the first GWAS in a population of severe or difficult-to-treat asthmatics demonstrated SNPs in IL13 and a neighboring gene encoding for a DNA repair protein (RAD50) as a locus for asthma susceptibility. The interaction of variants in different Th2 inflammatory pathway signaling genes has been critical to understanding the nature of gene-gene interactions in determining asthma susceptibility and severity. Furthermore, these positive associations for the IL-4–IL-13 pathways have supported the current development of biologic therapies targeting these cytokines.

Limitations of Candidate Gene Association Studies: Lessons Learned and the Road to GWAS

All association studies, including candidate gene studies, are limited by the need to ascertain well-characterized case-control populations. An issue with early association studies was the small sample sizes (<200 cases) resulting in studies underpowered to identify gene variants with small effects and prone to false-positive results. Unrelated cases and controls have variable ancestral backgrounds, which may lead to false-positive results. SNP allele frequencies vary in different populations based on ancestry (population stratification), which may cause spurious differences in allele frequencies between cases and controls if different ethnic groups are analyzed (i.e., false positives).

The need to account for potential population stratification in genetic association studies is true for all types of association studies but was well recognized and implemented in candidate gene studies. Another limitation of all association studies is the underlying correlation of alleles in genetic loci located in a chromosomal region (i.e., linkage disequilibrium or LD). For example, if two genetic loci are physically close together, they will be inherited as one unit through multiple generations. In this case, an association study would result in positive results for both genes, making it difficult to isolate the gene or variant of interest. This is especially true for SNPs within a gene. However, in older populations of more generations or admixed population, there have been more opportunities to break the correlation, thus allowing the true risk allele or loci to be identified.

GWAS Identifies ORMDL3 as a Susceptibility Locus

In 2007, a GWAS for asthma susceptibility was performed in a European cohort (GABRIEL). SNPs in the ORM1-like 3 gene (ORMDL3) region on chromosome 17q12 were significantly associated with the diagnosis of childhood asthma and this finding was replicated in two independent cohorts. A second larger GWAS by the European GABRIEL consortium validated ORMDL3 as a locus for asthma risk while identifying additional susceptibility loci in the neighboring gasdermin-like genes ( GSDMB and GSDMA ) as well as other loci in the genome, including IL1RL1 / IL18RL1 (between the IL1 and IL18 receptor genes, rs3771166) on chromosome 2q12, HLA-DQB1 (rs9273349), IL33 on chromosome 9p24 (rs1342326), SMAD3 on chromosome 15q22 (rs744910), and IL2RB on chromosome 22q12 (rs2284033).

Gene variants in the ORMDL3 locus have been associated with asthma in multiple subsequent GWAS, making ORMDL3 the most replicated locus using genome-wide screening methods in case-control populations. The challenge of interpreting genetic associations in chromosome 17q12 is the strong linkage disequilibrium (LD) that exists between the variants in multiple genes that span this region. It remains unclear whether the causative gene is ORMDL3 or an adjacent gene, such as GSDMB or GSDMA .

This novel asthma susceptibility gene (ORMDL3) is a member of a class of genes that encodes transmembrane proteins anchored in the endoplasmic reticulum; however, its specific biologic role in asthma pathogenesis remains unknown. GWAS in asthma have demonstrated a predilection for genes that regulate relevant molecular pathways and are related to epithelial damage and adaptive immune response. It is likely that altered production of cytokines such as thymic stromal lymphopoietin (TSLP) and IL-33 from damaged or disrupted epithelial cells promotes Th2-mediated inflammatory responses by activating IL1RL1 receptors on mast cells, Th2 lymphocytes, and regulatory T cells.

GWAS in African Americans and Other Asthma Populations

Subsequent GWAS were performed in multiethnic populations including non-Hispanic white and African American subjects to understand the genetic diversity of populations from different ancestral backgrounds (i.e., ethnic groups) as well as to address the genetic causes of differences in the frequency and severity of asthma in different ethnicities. This is particularly important for ethnic groups such as African Americans and Puerto Ricans who have a greater asthma incidence associated with higher asthma-related morbidity and mortality. These groups may have unique genetic loci due to a higher frequency of risk alleles that determine disease susceptibility or severity. For example, African Americans are at low risk for cystic fibrosis because of the low frequency of CFTR mutations in this ancestral population while at higher risk for sickle cell disease due to increased frequency of mutations in that gene.

The EVE consortium combined 5416 asthma cases in a large, multiethnic population representing Americans of European descent, African Americans, African Caribbeans, and Hispanics (Mexican Americans and Puerto Ricans) with replication in 12,649 subjects. This large-scale multiethnic GWAS for asthma susceptibility identified genome-wide significant associations with SNPs on chromosome 17q12 (spanning IKZF3 / ZPBP2 / GSDMB / ORMDL3 ), IL1RL1, and TSLP on chromosome 5q22. Associations at these loci and a SNP in IL33 were replicated in independent cohorts. This GWAS validated the susceptibility loci identified on the chromosome 17q12 locus and IL1RL1 confirming their importance in asthma risk (see eTable 45-2 ).

The EVE consortium also identified SNPs in the pyrin and HIN domain family member 1 ( IFIX, interferon inducible nuclear protein X gene, PYHIN1 ) associated with asthma in African American and African Caribbean cohorts. The first GWAS of a population from a primarily African ancestry also identified novel susceptibility loci in the α _1B -adrenergic receptor on chromosome 5q33 (ADRA1B) and prion-related protein on chromosome 20p12 (PRNP). In addition, the DPP10 locus was associated with asthma, confirming earlier positional cloning studies.

GWAS in Asian populations confirm the importance of previously described asthma susceptibility loci in HLA-DQ, HLA-DPA1, HLA-DPB1, and TSLP in Japanese and Korean populations. In addition, five polymorphisms in chromosome 17q12 encompassing ORMDL3, GSDMB, ZPBP2, and IKZF3 were associated with asthma in a Han Chinese population (see eTable 45-2 ).

Admixture Mapping as an Alternative Approach in Admixed Populations

The colonization of the Americas by the Europeans and the subsequent African slave trade has resulted in the mixing of genetic ancestries and the complex population structures encountered in the genomes of African Americans, African Caribbeans, Mexican Americans, and Puerto Ricans. There is marked variability in the demographic histories of these admixed populations of African origin, resulting in a complex population structure and reduced LD that is not fully captured with GWAS that use genotyping platforms designed for non-Hispanic white populations. The recent mixing of these ancestries provides the rationale for genome-wide ancestry-based approaches. African Americans, on average, have an estimated 20% European ancestry (80% African) while Hispanic ethnic groups such as Puerto Ricans and Mexican Americans have a combination of three different ancestries.

Mapping by admixture linkage (admixture mapping, Fig. 45-3 ) is a genome-wide approach that is based on the principle that many SNPs show marked variable allele frequencies between different ancestral populations. In admixture mapping, estimates of ancestry at each SNP are tested for association with a phenotype, in contrast to GWAS which compares allele frequencies. Admixture mapping requires a substantially smaller number of genetic markers compared to GWAS and can evaluate regions with rare variants ( Fig 45-4 ) and other variants such as polynucleotide insertion-deletions. The optimal setting for the use of admixture mapping is in admixed populations where there are marked racial disparities in disease phenotype not attributed completely to environmental factors.

Illustration of admixture mapping.

A, The hypothesis behind mapping by admixture linkage disequilibrium or admixture mapping is that chromosomes from an admixed population (shown with red and blue genetic regions from different ancestries) contain a susceptibility allele that is more frequent in the red ancestral region versus the blue. Admixture mapping identifies increased ancestry at a susceptibility locus in cases compared to controls (region intersected by thick black line). B, Loci with significant associations between ancestry and disease risk are represented by admixture mapping peaks or chromosomal regions with an overrepresentation of ancestry from the ancestral population with the highest proportion of risk alleles at the locus containing the risk-invoking variant.

( A, Adapted from Montana G, Pritchard JK: Statistical tests for admixture mapping with case-control and cases-only data. Am J Hum Genet 75:771–789, 2004. B, From Patterson N, Hattangadi N, Lane B, et al: Methods for high-density admixture mapping of disease genes. Am J Hum Genet 74:979–1000, 2004, Fig. 1.)

Impact of genetic variants in human disease.

The “common disease–common allele” hypothesis states that multiple common genetic variants with small to modest effect sizes contribute to common disease susceptibility in an additive fashion. Genome-wide association studies (GWAS) have identified multiple common variants associated with risk for different common diseases. In contrast, the “common disease–rare allele” hypothesis states that rare genetic variants with a large effect size contribute to susceptibility for common diseases. Rare variants can only be evaluated with family-based genetic studies, DNA sequencing, and admixture mapping, and are not easily evaluated with GWAS.

(Adapted from Tsuji S: Genetics of neurodegenerative diseases: insights from high-throughput resequencing. Hum Mol Genet 19:R65–70, 2010.)

Studies of African ancestry in African Americans from the general population have suggested a role for gene variation related to African ancestry on lung function and asthma severity. Estimates of African ancestry are inversely associated with baseline lung function measures in three independent African American populations and Puerto Ricans. In African Americans, each percentage of African ancestry was associated with a 3 to 5 mL lower FEV ₁ . This finding has implications for calculating predicted normal lung function in admixed ethnic groups. For example, an African American with 50% African Ancestry is likely to have an FEV ₁ up to 200 mL higher compared to an African American of the same age, sex, and height with 90% African ancestry. Estimates of African ancestry have also been associated with risk for self-reported asthma and asthma exacerbations in African Americans from the United States.

Genome-wide admixture mapping studies have the potential to identify the genetic loci that account for the increased asthma incidence in those with significant African ancestry. Admixture mapping for susceptibility loci has been performed in Puerto Ricans and Mexican Americans from the Genetics of Asthma in Latino Americans (GALA) cohort, and SNPs in nearly 62 loci associated with asthma (i.e., admixture mapping peaks, see Fig. 45-3 ) have been identified. Admixture mapping in African American and Puerto Rican asthma subjects from the NHLBI-sponsored Severe Asthma Research Program (SARP), CSGA, and GALA cohorts has also identified a novel asthma susceptibility locus on chromosome 6q14.1.

Rationale for Pharmacogenetic Research in Asthma

Pharmacogenetics examines the role of gene variation in therapeutic responses to pharmacologic therapies. Pharmacogenetics is a gene-environment interaction wherein the therapeutic agent is the environmental exposure. There is evidence that genetic factors play an important role in the observed variability in therapeutic responses to different drugs. Approximately, 70% to 80% of asthmatics show varying responses to common antiasthma therapies, resulting in a marked variance in therapeutic drug responses. Genetic variation might contribute to a large percentage of the variability in drug response, whether beneficial or adverse. The rationale for pharmacogenetic studies in asthma is that genetic markers could be used to identify subgroups of nonresponders, responders, or a subgroup at risk for adverse responses.

Pharmacogenetic research in asthma is driven by two major challenges related to asthma management. First, a small subgroup of approximately 5% to 10% of asthmatics experiences uncontrolled symptoms and recurrent exacerbations despite treatment with multiple antiasthma therapies, including high doses of inhaled or oral corticosteroids. This population with refractory asthma experiences substantial morbidity and generates a major financial burden compared to those with controlled mild or inter­mittent asthma. Second, adverse effects have been associated with some asthma therapies, particularly rare, life-threatening events associated with the use β ₂ -adrenergic receptor agonists (i.e., β-agonists). Pharmacogenetic studies have the potential to identify genetic markers that would personalize therapeutic approaches in individual asthmatics to optimize therapeutic response and prevent adverse side effects.

Pharmacogenetic studies of asthma therapies have been based on the primary clinical end points used in asthma clinical trials. In most pharmacogenetic studies, these predetermined clinical end points are analyzed for genetic associations with candidate gene studies or GWAS after the clinical trial is completed. These pharmacogenetic studies are important for the identification of genes or gene variants that influence drug responses. A minority of pharmacogenetic studies employs a prospective, genotype-stratified study design in which DNA is collected and genotyped for a risk variant of interest and subjects are randomized based on their genotype to treatment or placebo groups. The advantage of a prospective, genotype-stratified approach is that it can ensure sufficient statistical power to analyze less common genetic variants because the population is recruited based on a predefined risk gene variant. However, a genotype-stratified pharmacogenetic trial is feasible if only a limited number (one or two) genotypes are being considered.

Glucocorticoid Pharmacogenetics

Inhaled corticosteroids (ICS) have consistently been shown to be the most effective therapy to control persistent asthma. However, some asthmatics respond poorly to ICS. The pharmacogenetics of the glucocorticoid pathway is based on genes that code for a complex pathway consisting of the glucocorticoid biosynthetic pathway, chaperone proteins, and the cytosolic receptor heterocomplex ( eTable 45-3 ).

A candidate gene analysis of the corticotropin-releasing hormone ( CRHR1 ) gene in more than 1000 asthma subjects from the National Institutes of Health and industry clinical trial cohorts identified two SNPs (rs242941 and rs1876828) that were associated with a change in lung function during ICS treatment. In another study, the gene coding for a glucocorticoid receptor complex chaperone protein, the heat shock organizing protein (STIP1), identified three SNPs (rs2236647, rs6591838, and rs1011219) that were also associated with a significant change in lung function during ICS treatment. The corticosteroid pathway interacts with other gene pathways that have been evaluated using a candidate gene approach in different clinical trial cohorts. The “T-box expressed in T-cells transcription factor” is encoded by TBX21 and is an important regulator of the naive T-lymphocyte development pathway. TBX21 contains a coding SNP (rs2240017, His ³³ Gln) that was associated with “bronchoprotection,” that is, a reduction in bronchial hyperresponsiveness, during ICS treatment in the NHLBI Childhood Asthma Management Program (CAMP) and a Korean cohort. Adenylyl cyclase type 9 is a key enzyme in the β ₂ -adrenergic receptor pathway and is encoded by ADCY9, which contains a coding SNP (rs2230739, Met ⁷⁷² Ile) that has been associated with an albuterol bronchodilator response in the CAMP cohort during ICS treatment.

Use of GWAS approaches has the potential to identify novel pharmacogenetic loci for ICS response. The first GWAS evaluating ICS treatment response was performed using family-based testing in CAMP followed by association studies for replication in more than 900 asthma subjects from four independent clinical trial cohorts. A promoter SNP in the glucocorticoid-induced transcript-1 gene (rs37972 in GLCCI1 ) was associated with changes in lung function in response to ICS treatment. An in vitro study was also performed in which cells were transfected with another promoter variant in LD with rs37972 (rs37973) resulting in decreased gene expression. These studies were primarily performed in children with fewer years of exposure to corticosteroids compared to adult asthmatics. Interestingly, more recent studies using large clinical trial cohorts of adult asthmatics were not able to replicate this association. Functionally, GLCCI1 is an important regulator of apoptosis in response to glucocorticoids; therefore, this promoter variant may delay apoptosis of eosinophils during ICS therapy, thereby modulating therapeutic responses in asthma.

Another GWAS for ICS response was performed in asthma subjects from the NHLBI CAMP, Asthma Clinical Research Network (ACRN), and Childhood Asthma Research and Education Network. This GWAS identified SNPs in the T gene (rs3127412 and rs6456042) that were associated with changes in FEV ₁ during ICS treatment. Subsequent detailed genotyping identified SNPs (rs3099266, rs2305089, rs1134481) within functional regions of the T gene. Pharmacogenetic loci in the glucocorticoid pathway, GLCCI1, and the T gene account for a small proportion of the interindividual variability for ICS therapeutic responses in asthma. It has yet to be determined whether these loci have additive effects with other pathway-related gene variants or are independent determinants of ICS response. Future pharmacogenetic studies in independent, large clinical trial cohorts or genotype-stratified trials are necessary to confirm and characterize these genes associated with ICS therapeutic responses and develop a genetic profile that may be used to predict therapeutic responsiveness to corticosteroids.

Cysteinyl Leukotriene Pharmacogenetics

The cysteinyl leukotriene pathway is an inflammatory pathway containing potential pharmacogenetic loci for two classes of antiasthma therapies: 5-lipoxygenase (5-LO) inhibitors (e.g., zileuton) and cysteinyl leukotriene antagonists (e.g., montelukast and zafirlukast). The cysteinyl leukotriene biosynthetic pathway is initiated by 5-LO (encoded by ALOX5 ) followed by leukotriene A ₄ hydrolase (LTA4H), and leukotriene C4 synthase (LTC4S). Synthesized leukotrienes are transported to the extracellular space by multidrug resistance protein 1 ( MRP1 ) and activate cysteinyl leukotriene receptors ( CYSLTR1 and CYSLTR2 ).

ALOX5 has a tandem repeat polymorphism in its promoter that has been associated with changes in lung function during treatment with leukotriene antagonists. An analysis of the cysteinyl leukotriene pathway in asthmatics treated with zileuton also identified SNPs in ALOX5 (rs892690, rs2029253, and rs2115819), LTC4S (rs272431), and MRP1 (rs215066 and rs119774) that were associated with changes in lung function during 5-LO inhibition. Variation in leukotriene pathway genes has been reported in additional, smaller studies. These results support the finding that some of the variation in therapeutic responses to leukotriene modifiers is regulated by pharmacogenetic mechanisms (see eTable 45-3 ).

β ₂ -Adrenergic Receptor Pharmacogenetics

Inhaled β-agonist treatment for asthma includes short-acting β-agonists (SABA) and long-acting β-agonists (LABA). β-Agonists activate a G protein–coupled receptor pathway via adenylyl cyclase type 9 that regulates airway smooth muscle relaxation. Inhaled β-agonists are the most commonly used treatment for asthma despite having been associated with very rare life-threatening adverse responses since the 1960s.

While various analyses have attempted to explain these “mini-epidemics” of asthma mortality, the U.S. Food and Drug Administration issued a boxed warning for all LABA-containing inhalers regarding the risk for life-threatening exacerbations. This LABA safety controversy is being evaluated in 46,800 asthma subjects who are currently being recruited for the U.S. Food and Drug Administration–mandated international LABA safety study. Pharmacogenetic studies have attempted to identify genetic markers for β-agonist response to identify a subgroup that is susceptible to these serious adverse effects. These studies have primarily focused on genes related to the β ₂ -adrenergic receptor and nitric oxide synthetic pathways (see eTable 45-3 ).

The gene encoding the β ₂ -adrenergic receptor (ADRB2) has a single exon and is located on chromosome 5q31 with more than 49 polymorphisms identified. The most studied of these ADRB2 polymorphisms is the common coding variant, Gly ¹⁶ Arg. In vitro, receptors expressing the Gly ¹⁶ variant show increased receptor down-regulation in response to β-agonist stimulation compared to Arg ¹⁶ .

Two early pharmacogenetic studies of ADRB2 in asthmatic children demonstrated that Arg ¹⁶ homozygotes experienced a greater bronchodilator response to a single dose of albuterol compared to Gly ¹⁶ homozygotes. This effect of Gly ¹⁶ Arg genotypes on acute SABA bronchodilator response was confirmed in small asthma populations. In contrast, pharmacogenetic studies of long-term, regular SABA treatment in two independent clinical trial cohorts demonstrated effects on therapeutic responses: Arg ¹⁶ homozygotes experienced a decline in peak flow rate (PEFR) during chronic SABA treatment while Gly ¹⁶ homozygotes did not experience changes in PEFR.

In a genotype-stratified, crossover trial, the ACRN, Gly ¹⁶ and Arg ¹⁶ homozygotes were randomized to treatment with regular albuterol or placebo. Gly ¹⁶ homozygotes showed an increase in PEFR during regular albuterol treatment while Arg ¹⁶ homozygotes showed no changes in PEFR with regular albuterol but had an increase in PEFR during as-needed albuterol treatment ( Fig. 45-6 ). In addition, regular albuterol treatment resulted in decreased rescue inhaler use and asthma symptom scores in Gly ¹⁶ homozygotes, while Arg ¹⁶ homozygotes experienced a deterioration of these secondary outcomes. Although the adverse effects of regular SABA treatment in Arg ¹⁶ homozygotes are important to understand, asthma guidelines do not recommend regular scheduled SABA in asthma. However, regular scheduled LABA is used to control asthmatics who are symptomatic on ICS alone. Thus, pharmacogenetic effects related to LABA exposure would have major therapeutic implications

The Asthma Clinical Research Network Beta Agonist Response by Genotype (BARGE) Trial.

The BARGE trial evaluated the effect of genotypes in response to regular short-acting β-agonist (albuterol) therapy or placebo given four times daily for 16-week treatment periods. For patients with the Gly/Gly genotype, the PEFR improved significantly more when they were treated with albuterol (red line) compared with placebo. In contrast, patients with the Arg/Arg genotype improved more when treated with placebo (blue line) compared with albuterol. PEFR, peak flow rate.

(Adapted from Israel E, Chinchilli VM, Ford JG, et al: Use of regularly scheduled albuterol treatment in asthma: genotype-stratified, randomised, placebo-controlled cross-over trial. Lancet 364:1505–1512, 2004. Figure 3.)

The genotypic effects of Gly ¹⁶ Arg during long-term LABA therapy were evaluated in two small trial arms from larger NHLBI ACRN clinical trials. Results from these trials showed a potential for reduced effectiveness of LABA in Arg ¹⁶ homozygotes. Subsequent prospective clinical trials and pharmacogenetic analyses with larger cohorts including two genotype-stratified trials did not show evidence for reduced therapeutic effectiveness of LABA therapy in Arg ¹⁶ homozygotes compared to Gly ¹⁶ homozygotes. Interestingly, in one of these prospective genotype-stratified trials, Gly ¹⁶ homozygotes experienced a more prolonged protection from methacholine-induced bronchoconstriction than was observed in Arg ¹⁶ homozygotes. Another genotype-stratified study randomized 62 Arg ¹⁶ homozygotes to either montelukast or LABA in addition to ICS therapy and demonstrated greater symptom control with the addition of montelukast. This study suggested that the Gly ¹⁶ Arg locus could interact with other therapies in asthma.

In contrast to a common variant such as Gly ¹⁶ Arg that likely exhibits smaller effects, it is possible that rare genetic variants with strong effects might be responsible for some of the uncommon and severe adverse responses associated with LABA treatment in asthma (see Fig. 45-4 ). A rare variant within the fourth transmembrane domain of ADRB2 , Thr ¹⁶⁴ Ile, results in a protein with decreased β ₂ -adrenergic receptor ligand binding, coupling to G _s protein, and sequestration in response to SABAs and LABAs in vitro. This rare variant also significantly impairs the binding of salmeterol to its “exosite,” or secondary binding site, on the β ₂ -adrenergic receptor. The Ile ¹⁶⁴ allele was associated with an increased risk of airflow obstruction and reduced lung function in a cross-sectional population-based study of nearly 60,000 subjects. Thr ¹⁶⁴ Ile and a 25 base pair insertion variant found only in African Americans, have also been associated with severe exacerbations requiring hospitalization in asthma patients treated with a LABA. Thr ¹⁶⁴ Ile has also been associated with poor symptom control during LABA treatment in two independent cohorts. This finding of a rare genetic variant and adverse LABA effects has the potential to identify a very important at-risk asthma subpopulation. Thus, rare variants such as those in ADRB2 are potential biomarkers for more personalized, precise guideline-based treatment strategies in the small subset of asthmatics that have altered responsiveness to the combination therapy of a LABA with ICS. In addition, a recent study combining admixture mapping and GWAS in Puerto Ricans and Mexican asthmatic children from GALA has identified rare variants in solute carrier genes associated with bronchodilator response to SABAs, providing further evidence for the role of rare genetic variation as pharmacogenetic loci.

Association studies of candidate genes related to the β ₂ -adrenergic receptor pathway have identified additional common polymorphisms that have been associated with the acute bronchodilator response to SABA. ADCY9 is a canonical β ₂ -adrenergic receptor pathway gene with a common coding variant, Ile ⁷⁷² Met, associated with increased acute FEV ₁ bronchodilator response to albuterol during ICS therapy in children from the CAMP cohort. The effect of Met ⁷⁷² Ile genotypes on the bronchodilator response to β-agonists during ICS therapy was also reported in a Korean population treated with a LABA (formoterol). CRHR2 is another pathway-related gene with SNPs that have been associated with the SABA response. CRHR2 encodes the corticotropin-releasing hormone receptor-2, a G-coupled protein receptor that regulates relaxation of airway smooth muscle through a signaling pathway similar to that of the β ₂ -adrenergic receptor (via activation of adenylyl cyclase and protein kinase A). Five SNPs in CRHR2 were associated with acute bronchodilator responses to SABA in asthma subjects from three independent clinical trial cohorts.

Association studies of genes indirectly related to the β ₂ -adrenergic receptor pathway have identified variants within the nitric oxide biosynthetic pathway associated with acute SABA bronchodilator response. An association study of 111 genes from the β ₂ -adrenergic receptor and glucocorticoid pathways in subjects from the CAMP cohort and three independent trial cohorts identified a SNP in arginase 1 (ARG1), rs2781659, which was associated with an acute bronchodilator response. SNPs in ARG2 have also been associated with a SABA bronchodilator response. Arginase 1 and 2 metabolize l -arginine, which is a natural substrate for nitric oxide synthase, resulting in the production of nitric oxide. Nitric oxide is an endogenous bronchodilator; therefore, ARG1 and ARG2 polymorphisms may alter airway smooth muscle relaxation during β-agonist treatment.

Finally, a GWAS reported by Himes and colleagues identified a novel genetic locus for acute SABA bronchodilator responses. The primary GWAS analyzed 1644 asthmatics from six clinical trial cohorts with replication assessed in 1051 subjects from SARP and a Dutch asthma cohort. A SNP within the promoter region of the spermatogenesis associated, serine-rich 2–like gene (rs295137 in SPATS2L ) was associated with the acute bronchodilator response to albuterol. In human airway smooth muscle cells, the functional importance of SPATS2L in the β ₂ -adrenergic receptor pathway was confirmed after a knockdown of this gene resulted in increased β ₂ -adrenergic receptor expression. The investigators also confirmed previous associations for three candidate genes for albuterol bronchodilator response: ADCY9, CRHR2, and ARG1 .

Limitation of Current Pharmacogenetic Associations and Future Directions

Most of the pharmacogenetic associations for therapeutic response do not account for a significant amount of the variability in drug responses. In part, this is because multiple genes influence these responses. In addition, pharmacogenetic analysis was not the primary outcome of most of these studies and sample sizes were too small to assess single or multiple gene interactions. Furthermore, when the combination of a LABA and ICS are used, polymorphisms in the β ₂ -adrenergic receptor pathway might result in a negative treatment response that interacts with polymorphisms in the same pathway or an alternative (i.e., glucocorticoid) pathway to obscure associations.

Pharmacogenetic studies have the potential to identify the subgroup of asthmatics who are more or less responsive to biologic therapies currently under development. Biologic therapies are more expensive and have the potential to cause unwanted adverse responses. Thus, pharmacogenetic therapeutic “biomarkers” that identify susceptible asthmatics should facilitate development and registration of biologic therapies. Pitrakinra is a molecular inhibitor of the IL4α receptor subunit that inhibits both IL-4 and IL-13, which are important in the regulation of Th2 allergic inflammation. In a phase 2b clinical trial, specific variants in the gene coding for the IL4α receptor subunit (IL4RA) have been associated with changes in antigen-induced bronchial hyperresponsiveness in atopic asthmatics treated with pitrakinra, and improvements in asthma exacerbation frequency. These results are an excellent example of the use of a pharmacogenetic “biomarker” to identify the subset (approximately 35 percent) of responders to this IL4α receptor antagonist.

Pharmacogenetic studies in asthma had initially been limited to candidate gene studies in smaller clinical trial cohorts; however, investigators have more recently performed GWAS which have identified novel pharmacogenetic loci (see eTable 45-3 ). In vitro studies have been used to validate the function of these SNPs. Future pharmacogenetic studies in independent, large clinical trial cohorts or genotype-stratified trials will be necessary to evaluate the importance of these gene variants, to confirm previous associations, and to determine whether these polymorphisms act independently or have additive effects on therapeutic responses to ICS, inhaled β-agonists, leukotriene modifiers, and other, newer asthma therapies. Understanding the pharmacologic and genetic basis of therapeutic responsiveness in asthma may lead to the development of more effective therapeutic agents. These approaches, when performed in adequately powered pharmacogenetic trials using unbiased methodology (GWAS), have delineated genetic determinants of pharmacologic treatment, with both beneficial and adverse responses, in other complex diseases such as those currently under investigation in the National Institutes of Health Pharmacogenomics Research Network. Thus, future precise therapeutic approaches in asthma will include the use of genetic pharmacogenetic profiles.