key: cord-0874245-iz1bo218
authors: Shaw, Dominick; Portelli, Michael; Sayers, Ian
title: Asthma
date: 2014-05-02
journal: Handbook of Pharmacogenomics and Stratified Medicine
DOI: 10.1016/b978-0-12-386882-4.00028-1
sha: 844e5563f2255a5701d945729aa2d3962cf665cc
doc_id: 874245
cord_uid: iz1bo218

Asthma is a common respiratory disease with a complex etiology involving a combination of genetic and environmental components. Current asthma management involves a step-up and step-down approach based on asthma control with a large degree of heterogeneity in responses to the main drug classes currently in use: β(2)-adrenergic receptor agonists, corticosteroids, and leukotriene modifiers. Importantly, asthma is heterogeneous with respect to clinical presentation and the inflammatory mechanisms that underlie it. This heterogeneity likely contributes to variable results in clinical trials, particularly when targeting specific inflammatory mediators. These factors have motivated a drive toward stratified medicine in asthma based on clinical/cellular outcomes or genetics (i.e., pharmacogenetics). Significant progress has been made in identifying genetic polymorphisms that influence the efficacy and potential for adverse effects of all main classes of asthma drugs. Importantly an emerging role for genetics in phase II development of newer therapies has been demonstrated (e.g., anti-IL4). Similarly, the stratification of patients based on clinical characteristics (e.g., blood and sputum eosinophil levels) has been critical in evaluating newer therapies (e.g., anti-IL5). As a proof of concept, anti-IgE is the latest therapy to be introduced into clinical practice, although only for severe, allergic patients (i.e., in a stratified manner). As new asthma genes are identified using genome-wide association, among other technologies, new targets (e.g., IL33/IL33 receptor (IL1RL1)) will emerge and pharmacogenetics in these development programs will be essential. In this chapter we review the current understanding of asthma pathobiology and its clinical presentation, as well as the use of stratified medicine, which holds great promise for maximizing clinical outcomes and minimizing adverse effects in existing and new therapies.

umbrella definition will require integrating bench and bedside approaches using data from ongoing genomic and proteomic profiling studies of large, well-characterized asthma populations-such as the current EU-wide UBIOPRED study (formally, "Unbiased Biomarkers in Prediction of Respiratory Disease Outcomes")-with feedback to improved in vitro and in vivo animal models and human studies.

The International Consensus Report on the Diagnosis and Treatment of Asthma defines asthma as "a chronic inflammatory disorder of the airways in which many cells play a role, including mast cells and eosinophils. In susceptible individuals this inflammation causes symptoms which are usually associated with widespread, but variable, airflow obstruction that is often reversible either spontaneously or with treatment, and causes an associated increase in airway responsiveness to a variety of stimuli." The interaction of these features determines the clinical manifestations and severity of the disease and the response to treatment [3] . How these features relate to each other, how they are best measured, and how they contribute to clinical manifestations of the disease remains unclear. Age of disease onset also complicates understanding, although there are many shared features in the diagnosis of asthma in children and in adults, there are also important differences. The differential diagnosis, the natural history of wheezing illnesses, and the ability to perform certain investigations are all influenced by age [4] .

According to the World Health Organization (WHO), between 100 and 150 million people around the globe (roughly equivalent to the population of the Russian Federation) suffer from asthma. Worldwide, deaths from asthma have reached more than 180,000 annually. Overall, asthma affects 5-10% of the population in many developed countries. One study found that the annual estimated incidence of physician-diagnosed asthma ranged from 0.6-29.5 per 1000 persons. Risk factors for incident asthma among children include male gender, atopic sensitization, parental history of asthma, early-life stressors and infections, obesity, and exposure to indoor allergens, tobacco smoke, and outdoor pollutants. Risk factors for adult-onset asthma include female sex, airway hyper-responsiveness, lifestyle factors, and work-related exposures.

While the exact cost of asthma care is difficult to determine, a 2009 systematic review [5] found eight national studies reporting a national total cost. The total costs in 2008 in U.S. dollars were high and varied widely: Singapore, $49.36M; Canada, $654M; Switzerland, $1413M; Germany, $2740M; United States, $8256M. Although the costs of asthma care vary by country, it is estimated that worldwide the number of disability-adjusted life years (DALYs) lost from asthma is about 15 million per year. Worldwide asthma accounts for around 1% of all DALYs lost, which reflects the high prevalence and severity of asthma. The number of asthmacaused lost DALYs is similar to that for diabetes, cirrhosis of the liver, and schizophrenia. Costs are related to other factors, including asthma control, demographics, medical history, and doses of inhaled corticosteroid (ICS) prescribed [6] . Improving asthma control is associated with decreased cost.

Asthma classically causes wheeze, cough, chest tightness, and breathlessness (dyspnoea). The pattern of symptoms can be a clue to the diagnosis. Episodic symptoms worse at night and in the early morning, or in response to exercise, allergen and cold air exposure, or after taking aspirin or beta blockers are very suggestive of asthma. Moreover, symptoms that are relieved by bronchodilators point toward asthma as the underlying cause.

Symptoms that lower the probability of asthma include those of prominent dizziness, light-headedness, and peripheral tingling (all point toward dysfunctional breathing); a chronic productive cough in the absence of wheeze or breathlessness (more likely bronchiectasis); a repeatedly normal physical examination of chest when symptomatic; and voice disturbance, symptoms with colds only, or a significant smoking history. Taking a detailed patient history and recording spirometry when the patient is symptomatic is crucial to the accurate classification of the disease. There are several differential diagnoses that must be considered when making a diagnosis of asthma. These are outside the scope of this chapter. A good guide to the clinical diagnosis and management of asthma are the British Thoracic Society Asthma guidelines, which are updated regularly and are available at www.brit-thoracic.org.uk/clinical-information/ asthma/.

Triggers for asthma symptoms include allergens (proteins) derived from, among others, house dust mite (HDM), cat and dog dander, cockroach, and fungi, especially aspergillus fumigatus. There are numerous other nonallergic symptom triggers, including weather and atmospheric change, air pollution, exercise, menstruation, emotion and laughter, viral infection, gastroesophageal reflux, and rhinitis. Various occupations can both cause and worsen asthma (e.g., welding, baking, paint spraying).

Asthma symptoms often overlap with other allergies, including oral allergy syndrome and food allergies. Delineating between causes of symptoms can be difficult. If a specific aeroallergen is identified that triggers asthma (classically tree or grass pollen), a course of either sublingual or subcutaneous desensitization can be undertaken to reduce symptoms on exposure to that specific allergen.

There are multiple other symptom triggers that can include both allergic and non-allergic mechanisms; examples include exposure to chemicals, such as perfumes, paint and bleach, plants (e.g., Ligustrum vulgare-privet hedges), salicylates, and sulphites.

Multiple factors have been shown to either directly cause or worsen symptoms of asthma [4] . These are listed in Box 28.1. Other factors not directly related to the disease can worsen symptoms and are associated with more frequent exacerbations. These include dysfunctional breathing, vocal cord dysfunction, psychosocial problems, and poor adherence to treatment regimes [4] .

Other forms of asthma are recognized, among them occupational, exercise-induced, and pregnancy-related asthma. These subdivisions are somewhat arbitrary, but may have different causes, prognoses, and complexities. For example, occupational asthma includes that triggered by IgE-mediated mechanisms, that due to specific agents with unclear pathophysiology, and that secondary to irritants (also known as RADS-reactive airways dysfunction syndrome). Underlying asthma can also be made worse by occupational exposure (work-aggravated).

Most experts regard exercise-induced asthma as a different disease, albeit one resulting in similar symptoms. Exercise-induced asthma is often diagnosed with different investigations, including eucapnic voluntary hyperventilation or exercise testing. It is associated with certain sports (swimming and cross country skiing in particular) and has a unique pathophysiology directly related to damage of the airway epithelium.

Asthma control can deteriorate in pregnancy, due to both hormonal effects and the physical effects of a gravid uterus on diaphragm and respiratory muscle function. Pregnancy can also impact treatment decisions. Leukotriene modifiers are generally contraindicated, and the health of the fetus as well as that of the mother needs to be considered.

Poor adherence is probably the biggest single issue affecting asthma control. Studies have shown that people with asthma over-report use of ICS; one study found that although 95% of responders said they used their inhalers, only 58% actually did. Adherence to ICS therapy is associated with a lower rate of death, whereas increasing use of reliever (salbutamol/bricanyl) medication, which improves symptoms but does not treat underlying airway inflammation, is associated with increased mortality [7, 8] .

A 2010 report entitled Evaluation of the Scale, Causes and Costs of Waste Medicines: Report of DH funded national project (Trueman et al. York Health Economics Consortium and University of London School of Pharmacy, 2010) put figures to the possible cost benefits from improving asthma treatment adherence in the United Kingdom. It estimated the resulting net benefit associated with compliance to be £2250 per patient, with a reduction in expected annual treatment costs of approximately £75 per patient per year. Based on an asthma point prevalence of 5.8%, the findings suggest that there are almost 1.8 million asthmatics in the United Kingdom who are noncompliant. If interventions were available to change the behavior of all partially compliant medicine users to raise the percentage to 80% or more, the report estimated that over £130 million in treatment cost savings could be realized in the United Kingdom alone. A more modest target of doubling current compliance rates would result in savings of approximately £90 million per year.

The aim of asthma treatment is to improve symptoms, maintain lung function, and prevent exacerbations. From a patient's perspective, these are simple aims, but for the purposes of study design and treatment trials they do not address the complexity of the disease process. Accordingly, guidelines now specifically address asthma control (symptoms and lung function) and severity (need for treatment) separately. The current Global Initiative for Asthma (GINA) guidelines assess levels of control based on daytime symptoms, limitation of activities, nocturnal symptoms/awakening, need for reliever/rescue treatment, lung function (PEF or FEV 1 ), and exacerbations [9] ( Treatment steps are defined by the treatment level required to maintain asthma control (Table 28 .1), with patients at steps 1-2 having well-controlled asthma requiring little treatment, and patients at step 4 having poorly controlled asthma despite four or five different drugs. The step approach is similar to that used by the British Thoracic Society (BTS) Guidelines [4] .

Asthma is a variable disease both temporally and clinically. There is also overlap between symptoms and exacerbations. This overlap was the subject of a recent consensus document that defined asthma exacerbations as "events characterized by a change from the patient's previous status." Severe exacerbations are "events that require urgent action on the part of the patient and physician to prevent a serious outcome, such as hospitalization or death from asthma," and moderate asthma exacerbations are "events that are troublesome to the patient, and that prompt a need for a change in treatment, but that are not severe…clinically identified by being outside the patient's usual range of day-to-day asthma variation" [10] . The most practical definition of an asthma exacerbation is probably an episode of worsening symptoms not responding to increasing bronchodilator therapy. This definition was employed in an elegant study on the time course of peak flow changes [11] . For the purposes of clinical studies, most authors define exacerbation as the need for rescue courses of systemic corticosteroids (prednisolone/prednisone).

Although the majority of patients with asthma are treated and investigated in primary care settings, the main burden of asthma is due to its severe form (i.e., step 4) and exacerbations. Using the United Kingdom as an example (population 60 million), asthma is responsible for more than 1200 deaths per year and for over 50,000 hospital admissions, with an annual expenditure of £800 million on pharmaceutical costs alone.

It was recently estimated that a patient controlled at the mildest end of the spectrum (British Thoracic Society (BTS) guidelines: step 1 with no exacerbations) would cost 50 times less to provide his or her package of care than would a patient at the worst end of the spectrum (BTS guidelines: step 5 with exacerbations) [12] . It is also calculated that 5% of asthma patients are responsible for at least 50% of the total healthcare burden [4] ; the most expensive individual cost is an ICU admission (BTS level 3) due to a life-threatening exacerbation. In the U.S. more than half (53%) of asthma sufferers in 2008 had an asthma attack, and of these half of the children and one-third of the adults missed school or work because of it; on average, children missed four days of school and adults missed five days of work.

Exacerbation pathogenesis is not fully understood. Although research has focused on infective agents, especially viral, there may be other explanations.

Asthma exacerbations are associated with both inflammatory and immunological cell infiltration. The inflammatory cell infiltrate is composed of varying numbers of eosinophils, neutrophils, and lymphocytes. This airway inflammation, combined with smooth-muscle hypertrophy Step 1 (intermittent) Fewer than once/week; asymptomatic, normal PEF between attacks Two or fewer/month 80% or more Less than 20%

Step 2 (mild persistent) More than once/week, fewer than once/day; attacks may affect activity >2 times a month >/=80% 20-30%

Step 3 (moderate persistent) Daily; attacks affect activity More than once/week 60%-80% More than 30%

Step 4 (severe) Continuous; limited physical activity Frequent 60% or less More than 30%

and thickening of the lamina reticularis, is accentuated by mucus plugs, serum protein deposition, inflammatory cell and cellular debris, leading to blockage of the airway (airflow obstruction) and wheeze.

Approximately 80% of exacerbations are associated with respiratory tract viral infections, with rhinovirus responsible for about two-thirds of cases [13] . Asthmatic subjects also have much more severe lower respiratory tract illness with rhinovirus infection than do healthy control subjects. The mechanism for this is not fully understood. Infection induces inflammation, increasing levels of neutrophils, eosinophils, CD41+ cells, CD81+ cells, and mast cells through increased mRNA expression and translation of IL-6, IL-8, IL-16, eotaxin, IFN-γ-induced protein 10 (IP-10), Chemokine (C-C motif) ligand 5 (CCL5/RANTES), and other proinflammatory cytokines. Other viruses implicated in exacerbations include enterovirus, coronavirus, influenza, parainfluenza, respiratory syncytial virus, metapneumovirus, adenovirus and bocavirus [14] . Although influenza vaccination is recommended for all individuals with asthma, there is currently no hard evidence that this improves outcomes.

Bacterial infection has also been implicated in asthma exacerbation. Individuals with asthma have an increased risk of invasive pneumococcal disease [15] , and an increased frequency of detection of Chlamydophila pneumoniae [16] . One study found Mycoplasma pneumoniae infection in 20% of patients hospitalized for severe asthma [17] . Several trials have evaluated the role of macrolide antibiotics in treating and preventing asthma exacerbations. One study randomized adults with exacerbations to the ketolide antibiotic telithromycin or placebo; the telithromycin group had a small but significant improvement in symptoms and lung function from exacerbation to the end of treatment [18] . Further studies of macrolide antibiotics (which have immunomodulatory as well as antibiotic properties) are under way, but as yet macrolide antibiotics are not recommended by international asthma guidelines. New techniques for assessing the airway bacterial microbiota have established that the airways are not normally sterile, and may permit the role of bacterial infection in airways disease to be further delineated. A recent study found that pathogenic Proteobacteria, particularly Haemophilus species, were much more frequent in the bronchi of patients with asthma or COPD than in controls [19] .

Prolonged exposure to aeroallergens can result in chronic airway inflammation via Th2-driven IgE mechanisms. This immunological reaction may intensify airway inflammation, increase inflammatory cell activation, and stimulate mucus glands to hypersecrete, leading to airway obstruction. Studies of bronchoalveolar fluid before and after allergen challenge show eosinophilic inflammation as the major airway response, associated with late-phase responses of airflow obstruction. In addition, IL5 and IL13 are significantly raised. Exposure to seasonal allergens has been implicated in sudden asthma-related deaths; HDM, cat, and cockroach allergen sensitization are risk factors for emergency treatment [20] . Grass pollen sensitization, or ''thunderstorm asthma,'' has also been associated with epidemics of exacerbations [21] .

Fungal allergens are found both outdoors and indoors, and sensitivity to them is a risk factor for the development, persistence, severity, and mortality associated with asthma. Individuals with asthma are often sensitized to fungi such as Cladosporium species, Alternaria species, Penicillium species, and Candida species [22] .

Alternaria species sensitization and exposure are associated with symptoms and a 200-fold increased risk of respiratory arrest in subjects with asthma [23] . Fungal exposures may worsen disease by different mechanisms; fungalassociated proteases may lead to the development of allergic airway inflammation along with IgE-mediated responses; fungal allergen-induced asthmatic reactions evoke an IL5 response with increased eosinophil recruitment and degranulation [24] .

The relationship between fungal exposure and asthma is complicated by the degree of sensitivity to fungal allergens and the resultant allergy. A specific disease entity called allergic bronchopulmonary aspergillosis (ABPA) exists in which colonization of the respiratory tract with the ubiquitous Aspergillus fumigatus is associated with an increased allergic response (both type 1 and type 3) and severe asthma exacerbations. Trials in both ABPA and a related disease, severe asthma with fungal sensitization (SAFS), have shown some benefit with the oral antifungal itraconazole [25] , but good-quality trial data are lacking.

Unlike other chronic conditions, there is no gold standard for the diagnosis of asthma. Although tests of airway function can be used to diagnose and study different aspects of airway disease (airway inflammation/airway hyper-responsiveness), diagnosis still depends on the presence of specific symptoms, which include wheeze, cough, chest tightness, and difficulty breathing. These symptoms are often variable, worse at night, or worse in response to triggers. They often respond to asthma therapy and may be associated with atopy/allergy and a family history of similar problems.

The lack of a gold standard affects clinical provision and studies alike; most estimates suggest that asthma is wrongly diagnosed in approximately 30% of patients [26] . The development of a cheap and reliable noninvasive test that can diagnose asthma is seen as the holy grail for diagnosisbased research.

Measurements of airflow limitation provide an assessment of the severity, reversibility, and variability of airflow obstruction, and can help confirm the diagnosis of asthma. Spirometry (Forced Expiratory Volume in 1 second (FEV 1 ), Forced Vital Capacity (FVC), and FEV 1 /FVC) is the preferred method of measuring airflow limitation, as it is more repeatable and less effort dependent than peak flow. A FEV 1 /FVC ratio of less than 0.7 is consistent with a diagnosis of airflow obstruction, and an increase in FEV 1 of >12% (or >200 ml) after administration of a short-acting bronchodilator indicates reversible airflow limitation consistent with asthma [27] . Importantly, most asthma patients will not exhibit reversibility at each assessment, and repeated testing is advised to confirm a diagnosis. Consequently, an absent response to bronchodilators does not exclude asthma.

Peak expiratory flow (PEF) measurements can be used to monitor asthma. They are ideally compared to the patient's own previous best measurements using his or her own peak flow meter. An improvement of 60 L/min (or >20% of the prebronchodilator PEF) after inhalation of a bronchodilator, or diurnal variation in PEF of more than 20% (with twice-daily readings, more than 10%), suggests poorly controlled asthma.

Other approaches have been utilized to diagnose asthma, both to confirm a formal diagnosis and to identify corticosteroid response. These approaches include noninvasive measurements of airway inflammation or assessment of airway hyper-responsiveness. Measuring airway inflammation is relatively easy, either using techniques to induce sputum and then counting the differential eosinophil and neutrophil cell numbers present, or by measuring the fraction of nitric oxide (NO) present in the exhaled breath (F E NO) using portable chemical analyzers.

Airway hyper-responsiveness can be assessed by several different methods. All are designed to provoke bronchoconstriction, which is measured using spirometry. The concentration (or dose) at which bronchoconstriction occurs is then used to categorize the degree of airway responsiveness and the presence or absence of asthma. Although these tests are widely employed in research settings, their use in a clinical setting is hampered by their varying sensitivity and specificities for asthma diagnosis [4] , as both airway inflammation and airway hyper-responsiveness can be present in healthy individuals without symptoms.

The symptoms of asthma in children are recurrent wheezing, cough, difficulty breathing, and chest tightness. Evaluation of these symptoms is critical to the diagnosis, treatment, and outcome measures used in clinical studies; however, asthma triggers and inflammatory cell type are increasingly used to define childhood asthma phenotypes.

Pediatric asthma is complicated by wide variations in symptom prevalence. The majority of children with asthma have mild or moderate disease; 5% of all asthmatic children have chronic symptoms and/or recurrent exacerbations, despite maximum treatment with conventional medications [28] . Asthma in childhood is heterogeneous in several ways, including etiology, clinical presentation, and response to treatment. Several attempts have been made to stratify treatment on the basis of different asthma phenotypes in children, with varying degrees of success. Stratifying treatments in pediatric asthma is more complicated because of patient age and the inability (dependent on age) to perform more complicated or invasive tests.

In children, the onset and progression of asthma result from a complex interplay between genetic background and environmental exposures. The complexity is confounded by wheezing illnesses consisting of several distinct disease entities, with no general agreement on their number or underlying mechanisms [29] .

Among the environmental factors that influence the risk of asthma are viral and bacterial respiratory infections. Exposure to environmental tobacco smoke is also associated with increased rates of early viral illnesses. Other factors associated with the risk of wheeze include physical factors associated with increased breathing (exercise, laughing, crying, and excitement) or allergens (aeroallergens and food allergens). There are considerable age-related changes in the relative importance of trigger factors for wheeze in children.

Human rhinovirus (HRV) has been implicated in both the etiology of asthma and exacerbations; infants who have HRV infections with wheezing are at a significantly increased risk for subsequent asthma, and over 50% of childhood asthma exacerbations are triggered by HRV. It is important to note that exposure to HRV does not lead to wheezing illness in all children, nor does wheezing illness result in asthma in all cases, suggesting that the host genotype also plays a role. A recent study of two cohorts of children with asthma found that variants at the 17q21 locus were associated with asthma in children who had HRV wheezing illnesses, implicating an interaction between HRV wheezing illness in early childhood and the 17q21 genotype (see Section 28.3.2).

The presentation of asthma symptoms and exacerbations in childhood also varies. Some children have frequent exacerbations with few daily symptoms, while others have recurrent symptoms without airway inflammation and exacerbations. Other diseases coexist with childhood asthma, including gastroesophageal reflux disease, severe asthma with fungal sensitization, obesity, and vocal cord dysfunction. Targeting these coexisting disorders has met with varying clinical success [30] .

An important component of the development of asthma is the inflammatory cascade. This involves the infiltration of a number of inflammatory cells, such as eosinophils, neutrophils, B-and T-lymphocytes, macrophages, mast cells, dendritic cells, and basophils, into the airway, and the release of inflammatory mediators (e.g., leukotrienes, histamine, cytokines, and chemokines) by airway, structural cells (including epithelial, smooth-muscle, endothelial, and fibroblast cells). However, since the degree of inflammation in asthma is not directly related to asthma severity, this suggests that other factors, such as structural changes in the airways, also play a role in disease modulation and progression. These structural changes have been termed airway remodeling and may exist in the presence or absence of inflammatory mechanisms in the airway [31] .

The inflammatory response in asthma is a result of excessive activation of mast cells in the airway, where in the early response to allergen challenge, degranulation of mast cells results in the release of a number of proinflammatory factors such as IL4, leukotrienes, and histamine. This causes an immediate hypersensitivity response that in turn leads to airway narrowing. The concomitant release of other inflammatory factors such as cytokines and chemokines from the same mast cells provides an optimal environment for recruitment to the airways of other inflammatory cells such as eosinophils, basophils, neutrophils, and T-lymphocytes.

Once in the airways, these cells act in tandem with allergen-activated macrophages, resulting in what is termed the late asthmatic response (occurring at 4-8 hours postchallenge). Here, through the action of a number of released potent immunomodulators (e.g., TNFα, IL3, IL4, IL5, and IL13), airway narrowing occurs. This may occur in periods that last more than 24 hours. Of the inflammatory cells involved in the asthmatic response, it is the infiltration of eosinophils that is considered characteristic of the asthmatic phenotype, particularly that of mild/moderate allergic asthma ( Figure 28.1) . For a more extensive description of inflammatory mechanisms underlying allergic asthma see Hodge & Sayers [32] .

Although inflammation in asthma has been extensively studied, the relationship between inflammation and clinical symptoms of asthma is still unclear. This may be partially explained by a possible link between airway inflammation and airway hyper-responsiveness [33] . Here, airways become more susceptible to sensitizing agents, which would otherwise be unable to trigger an airway response, because of (a) increased release of mediators such as histamine and leukotrienes, (b) abnormal smooth-muscle behavior, and (c) airway thickening.

Airway remodeling plays an important role in asthma pathogenesis. Airway structural changes characteristic of airway remodeling occur because of prolonged airway inflammation. Here, prolonged release of inflammatory factors results in: l Thickening of the smooth-muscle bronchial walls, leading to airway narrowing l Denudation of the bronchial epithelium l Hyperplasia and hypertrophy of the airway smoothmuscle l Hyperplasia and hypertrophy of the epithelial goblet cells, resulting in the formation of large mucus plugs liable to occlude the airway and increase airway hyper-responsiveness Airway remodeling also involves changes occurring in the extracellular matrix and its constituent proteins (collagens I/III/IV, fibronectin, and laminin) and structural changes such as angiogenesis, vasodilation, increased airway blood flow, and changes in autonomic neurological function [34] .

The development of asthma results in a degree of airway remodeling. The structural changes that define airway remodeling are an important feature in asthmatic airways, where even in newly diagnosed asthma patients, a degree of structural change can be identified in the bronchial wall [35] . Airway remodeling can be defined as a process of sustained disruption and modification of structural cells and tissues leading to the development of a new airway wall morphology [36] . Airway remodeling ( Figure 28 .2) is initiated either through various inflammatory pathways, highlighting the importance of the inflammatory pathway [37] , or through bronchial hyper-responsiveness, where airway remodeling occurs independently of inflammation [38] .

Our understanding of the extent and nature of genetic variation in the human genome has dramatically improved since the publication of the first draft genome sequence in 2001, FIGURE 28.1 Asthma pathophysiology. Exposure to allergens-for example, Dermatophagoides pteronyssinus 1 (Der p1) and Dactylis glomerata (Dacg1)-originating from various sources, including pollen, house dust mite, and mold, are taken up by dendritic cells, B-lymphocytes, epithelial cells, and macrophages, which present the antigens to T-lympocytes. This activates the T-lymphocytes to produce cytokines that regulate immunoglobulin E (IgE) production by the B-lymphocytes. Bound IgE activates mast cells in the airways. Activated mast cells initially release a number of factors, including IL4, leukotrienes, and histamine, that cause airway hyper-responsiveness and result in an early response to allergen stimulus via bronchoconstriction. The concurrent release of cytokines and chemokines from the mast cell recruits eosinophils, basophils, and T-lymphocytes to the airway. These cells, in association with T cell-activated neutrophils and with antigen stimulated macrophages, release chemokines and leukotrienes over an extended period of time, resulting in a late response to allergen stimulus via bronchoconstriction. Prolonged stimulation of the late-response factors ultimately leads to the airway structural changes such as mucus hypersecretion, myofibroblast, and airway smooth-muscle proliferation, which are common to the asthmatic lung. and it continues to improve exponentially with initiatives such as the 100,000 Genomes project, which aims to genome-sequence this number of UK subjects, and the Encyclopedia of DNA Elements Consortium (ENCODE), an initiative to identify the genome's regulatory regions. Recent figures suggest that more than 6.9 million singlenucleotide polymorphisms (SNPs) or single-base-pair changes exist in a genome consisting of 3 billion base pairs. Similarly there is a growing realization that deletions, insertions, and expansions of tandem repeats also represent significant variation. These genetic polymorphisms are found throughout the genome and have the potential to influence gene function in several ways leading to human disease: (1) a coding-region nonsynonymous polymorphism can alter the amino acid sequence and structure/function of the protein; (2) a polymorphism can introduce a stop codon leading to the production of a nonfunctional protein; and (3) a polymorphism in a regulatory region can regulate the expression levels of a gene.

It has long been known that atopic diseases such as asthma run in families. In 1916, using 621 atopic probands and 76 nonatopic controls and their respective families, it was shown that 48.4% of atopic probands had a family history of atopy, compared with 14.5% in the control population [39] . Similarly, a study of 176 families showed a very high concordance of asthma, hayfever, and eczema in parents and children [40] . Twin studies have been useful in identifying a significant concordance of asthma that is higher in monozygotic twins (identical genotype) versus dizygotic twins (on average sharing half of their genes). In the largest study to date, using 11,688 Danish twin pairs, it was suggested that 73% of susceptibility to asthma was genetic, with a substantial environmental component [41] . More recent studies suggest hereditability estimates of 35-95% for asthma [42] . Therefore, asthma is considered a complex genetic disorder and, in contrast to single-gene disorders (e.g., cystic fibrosis), involves multiple genes, with expression influenced by both genetic and environmental factors. Multiple environmental factors are important in asthma development, including tobacco smoke exposure, respiratory viral infections, antibiotic use, diet, and allergen exposure. Gender and ethnic background also make a significant contribution. This complex mode of inheritance combined with the heterogeneity in the presentation of the disease has made gene discovery a challenge. Early studies investigated inheritance through families containing multiple affected children, using linkage analyses and candidate gene approaches based on biology or location in the genome. However, the FIGURE 28.2 Asthmatic airway pathology. This schematic comparison of a normal airway with that observed in severe chronic asthma indicates histological changes that accompany recurring inflammation seen over time. Unlike that from the unaffected individual, the bronchial mucosa from the severe asthmatic displays thickening of the basement membrane, airway smooth-muscle hypertrophy, leukocyte infiltration, epithelial cell desquamation, goblet cell hyperplasia in the epithelial lining accompanied by mucus hypersecretion, and plugging of the bronchial lumen, as well as edema and collagen deposition in the submucosal area. (Not drawn to scale.) Source: Reproduced with permission from Hodge and Sayers [32] .

reproducibility of these findings was disappointing primarily because of inadequate power, subject heterogeneity (different phenotype definition), population stratification, and multiple testing without correction.

More recently, hypothesis-free genome-wide association studies (GWAS), which examine association with typically 500,000+ common (>5% frequency) polymorphisms spanning the genome in cases and controls, using very stringent statistical thresholds ( Figure 28 .3), have been very successful. Via these multiple approaches, over 190 asthma genes have been described in more than 1000 publications. The major reproducible findings for asthma follow below (for more comprehensive reviews see [42, 43] ). Although this chapter focuses on the genetics of asthma diagnosis, subphenotypes in asthma (e.g., atopy, serum total IgE levels, and lung function (e.g., FEV 1 )) have also been investigated for genetic influences.

Positional cloning involves linkage analyses (region or whole genome) followed by fine mapping using association. Linkage analysis uses family data to follow the transmission of genetic information spanning the entire genome (commonly short tandem repeat markers) across generations. These data are used to determine if a genetic marker is close to, or linked with, a gene involved in a particular disease using families with multiple affected children.

Once a specific chromosomal region has been identified in this hypothesis-free approach, SNPs spanning the region are tested to see if they are more common in people with asthma compared to people without. The first gene to be identified with confidence using positional cloning was disintegrin and metalloprotease 33 (ADAM33) [44] .

Using a cohort of 460 families, linkage was demonstrated to a locus on chromosome 20p13 for asthma, bronchial hyper-responsiveness (BHR), and total IgE levels. Subsequently, polymorphisms spanning ADAM33 were associated with asthma in the second stage. ADAM33 is thought to contribute to airway remodeling via its enzymatic functions. Using positional cloning, multiple genes have been identified with functions varying from transcription factors to epithelial differentiation and tissue remodeling. These genes have provided a novel insight into potential mechanisms that are altered in asthma (Table 28. 2).

Polymorphisms spanning the urokinase plasminogen activator receptor (PLAUR or UPAR) were also associated with asthma susceptibility and were associated with rate of decline of lung function in asthma through linkage/association analyses on chromosome 19q13 [45] . uPAR is a serine protease receptor involved in multiple mechanisms, including cell proliferation, migration, and extracellular matrix degradations via plasmin generation. These data suggest that genetic factors may be important in determining airway structural changes or "remodeling" in asthma [46] . 

Genotype for typically 500,000+ SNPs in large numbers of cases and controls using arrays

Compare differences to identify disease-specific SNPs using stringent correction for multiple testing 

Candidate gene studies are hypothesis-driven and based on the suggested biology of a gene product or the location of the gene in a chromosomal region previously linked to asthma. Such studies commonly employ asthma cases versus controls, although family-based association has also been used. For excellent reviews of candidate gene studies in asthma, see Ober and Hoffjan [47] and Undarmaa et al. [48] . The primary highly reproducible genes identified using candidate gene approaches are multiple components of the interleukin 4/13 axis (see Table 28 .2), including the cytokine genes themselves (IL4/IL13); related receptor (IL4RA) and downstream signaling effector signal transducers and activators of transcription (STAT ) 6. The role of the IL4/IL13 axis in the pathogenesis of asthma has been extensively documented [49] . Human studies have identified elevated numbers of cells expressing IL13 mRNA in the bronchial tissue of atopic and nonatopic asthmatic subjects [50] ; administration of recombinant IL13 in mouse lungs resulted in an increase in airway mucus secretion, development of subepithelial fibrosis, airway hyper-responsiveness (AHR), and eosinophilic airway inflammation-that is, several key features of the human disease [51] . IL13 is produced by a variety of cells, including Th2 CD4+, Th1 CD4+, CD8+ T cells, mast cells, basophils, and eosinophils. IL13 mediates its effects by interacting with a complex receptor system comprising an IL-4Rα/IL13Rα1 heterodimer and the IL13Rα2 receptor [49] . The finding that genetic polymorphisms within IL13 that determine levels and/or structure and are associated with the development of asthma make biological sense.

The use of GWAS approaches to identify asthma susceptibility genes has revolutionized our understanding of asthma genetics. While positional cloning showed great success with simple Mendelian disorders, it did not perform well for complex diseases. Similarly, candidate gene studies are useful but very inefficient. The capacity to interrogate the entire genome for 500,000+ common SNPs in cases and controls has provided an excellent platform for gene discovery.

The first GWAS for asthma used a discovery cohort of 994 patients with childhood onset asthma and 1243 nonasthma controls, and identified significant association to a locus on chromosome 17q21 [52] . This locus includes genes for zona pellucida binding protein 2 (ZPBP2), gasdermin B (GSDMB) and orm1-like protein 3 (ORMDL3). The 17q21 locus has been reproduced as a locus associated with childhood onset asthma in many studies; however, the identification of the specific gene(s) underlying these effects remains to be resolved. ZPBP2, GSDMB, and ORMDL3 have been implicated in gene transcription, cell apoptosis, and sphingolipid synthesis, respectively.

This initial GWAS has now been superseded by several larger-scale studies. In particular, the GABRIEL consortium study, involving 10,365 asthma cases and 16,110 controls, identified association between polymorphisms spanning IL33, IL1RL1/IL18R1, HLA-DQ, SMAD3, and IL2RB [53] . Note: For candidate gene approaches, genes focused to replication in more than 10 studies [47, 48] . For GWAS, genes focused to those meeting conventional genome-wide significance (p < 5 × 10−8) and/or independent replication in the Caucasian population. To date, two of the most reproducible association signals are to the IL33 gene on chromosome 9p24 and the IL33 receptor (IL1RL1 or ST2) gene on chromosome 2, both of which highlight the significance of this ligand/receptor system in asthma. Interestingly, these loci are associated with severe asthma as well [54] . There is good evidence from biology that the IL33/ST2 axis may be of relevance in asthma; mice induced to develop allergic airway disease were treated with an antibody that blocks IL33 binding to its receptor. This treatment was shown to block many of the features of the disease, including reduced serum IgE and airway eosinophil and lymphocyte counts. Significantly, both induction and resolution of the disease were attenuated [55] .

In human studies, IL33 has been shown to be elevated in the airways of asthma patients, particularly in the airway structural cells, including the bronchial epithelium, and is induced in these cells by relevant stimulations (e.g., HDM [56] ). In addition, a soluble form of the ST2 receptor has been shown to be elevated during asthma exacerbation [57] . Overall, it is thought that the IL33/ST2 axis may be particularly important in attracting and activating inflammatory cells in the airways; however, the precise role(s) of this pathway remains to be resolved.

As outlined in Table 28 .2, other genes have been identified using GWAS, including those involved in diverse roles such as inflammatory cell function and airway smoothmuscle contraction, again providing a unique insight into potentially altered mechanisms in asthma. However, most GWAS determined single variants only confer a very modest odds ratio of 1.1-1.5 of developing asthma. For a more comprehensive review of GWAS findings in asthma, see Akhabir and Sandford [43] .

Over the last 40 years, there has been great progress in identifying asthma susceptibility genes. Current and future approaches include meta-analyses using cohorts of tens of thousands of asthma patients, the incorporation of gene expression/SNP analyses (the expression quantitative trait loci (eQTL)), the investigation of asthma subphenotypes using GWAS, and whole-genome sequencing.

The real challenge is the translation of these genetic findings into a deeper understanding of the biology of asthma and the potential identification of therapeutic targets. Already several of these newly identified genes (e.g., IL33 and IL33 receptor (IL1RL1/ST2)) represent excellent pharmacological targets, and it is anticipated that genetics in these loci will be essential to identifying patients most likely to gain clinical benefit from targeting this pathway.

The lack of concordance between approaches (e.g., linkage versus GWAS) can be explained by the fact that the methodologies are designed to detect different types of variants, e.g., linkage analysis has good power to detect high-risk disease-causing alleles but is not effective at identifying common alleles of modest effect, as GWAS does. It is reassuring that many of the genes identified in candidate gene approaches have been reproduced in GWAS (e.g., the IL13/IL4 locus on chromosome 5q31, a region that has also been linked to asthma). While there has been much success in recent years in identifying genes for asthma using GWAS, the overall genetic variation accounted for by these genetic polymorphisms is very small, leading to the concept of "missing hereditability." Possible explanations for missing hereditability include: (1) rare variants (<5% frequency) with larger effect size not measured on existing platforms;

(2) structural variation (e.g., copy number variation); (3) gene-environment contributions; (4) gene-gene interactions; (5) epigenetic mechanisms; and (6) overestimation of initial hereditability.

The aim of treatment is to achieve control of asthma and prevent exacerbations. Most guidelines adopt a stepwise approach to the management of asthma and advise stepping up treatment as necessary and stepping down when control is achieved for at least three months. Although there are several pharmacological options for asthma treatment (see Figure 28 .4), the two main classes remain bronchodilators (short-or long-acting) and corticosteroids.

Short-acting bronchodilators (e.g., salbutamol/terbutaline) are sympathomimetics (β 2 -adrenergic receptor agonists) that relax airway smooth-muscle, enhance mucociliary clearance, decrease vascular permeability, and possibly modulate mediator release from mast cells. They are used for symptom control, acute exacerbations, and exerciseinduced asthma. Frequent or regularly scheduled use of rapid-acting inhaled β 2 agonists for long-term management of asthma does not adequately control symptoms, PEF variability, or airway inflammation.

Long-acting bronchodilators (e.g., salmeterol/formoterol) activity is identical to that of short-acting β 2 -adrenergic receptor agonists, but should be used in combination with inhaled corticosteroids because of concerns regarding increased mortality when used as a monotherapy. Long-acting bronchodilators are normally introduced at treatment step 3.

Inhaled corticosteroids (ICS) form the backbone of treatment for all but the very mildest asthma. ICS improve lung function, symptoms, and quality of life, reduce exacerbations, and improve mortality. They have broad anti-inflammatory and immunosuppressive effects. Corticosteroids enter the cell cytoplasm and bind with the inactive glucocorticoid receptor complex. The activated glucocorticoid receptor binds to DNA at the glucocorticoid response element sequence and promotes synthesis of anti-inflammatory proteins (transactivation), and inhibits transcription and synthesis of many proinflammatory cytokines (transrepression). They also reduce the number of T-lymphocytes, dendritic cells, eosinophils, and mast cells in the airway, and reduce inducible nitric oxide production. ICS (beclomethasone dipropionate/fluticasone propionate/budesonide/ flunisolide/ciclesonide/mometasone) all have side effects related to their transactivation properties, including local effects on the oropharynx (hoarseness, candidiasis, cough, and dysphonia) and systemic effects (Cushing syndrome, osteoporosis, cataracts, dermal thinning and bruising, adrenal insufficiency, and growth suppression) in children [58] . Consequently, the lowest dose required to control symptoms and prevent exacerbations should be prescribed.

Other medications are normally reserved for second-or third-line treatment. These include methylxanthines (theophylline), leukotriene modifiers (montelukast), inhaled anticholinergics (ipratropium bromide/tiotropium), and sodium chromoglycate.

A wide range of interventions have been assessed for both the primary prevention of asthma and the secondary prophylaxis of the disease once present. These treatments have varying effects, but may be recommended in individual cases. Primary prevention measures include aeroallergen and food allergen avoidance, fish oil supplementation, avoidance of tobacco smoke and air pollutants, immunotherapy, and immunization.

Measures that have been assessed for secondary prophylaxis include HDM and allergen avoidance, smoking reduction, subcutaneous and sublingual immunotherapy, dietary manipulation, weight loss, breathing techniques, and exercise training. 

Some investigators argue that severe asthma is an inflammatory condition separate from mild/moderate asthma, rather than simply the severe end of the disease spectrum. This is difficult to prove, as underlying asthma may be complicated by other aspects (obesity/reflux/smoking/poor adherence), making it more difficult to treat and control. There are different definitions of severe asthma, but all require high-dose inhaled corticosteroids or oral corticosteroids and include recurrent exacerbations and poor control.

As high-dose ICS and oral corticosteroids are associated with side effects attempts have been made to design treatments to either reduce the amount of corticosteroid prescribed (steroid-sparing agents) or treat the underlying asthma in a different manner. Steroid-sparing agents trialled include methotrexate, gold, cyclosporin, and azathioprine. Currently, none have been recommended in any asthma guidelines. Box 28.2 lists other treatments used on an individual basis.

The need for a stratified-medicine approach to asthma has become apparent over the last decade. This has followed the realization that the term asthma covers a whole range of disease phenotypes. As techniques for investigating and researching asthma have evolved, it has become clear that different disease processes are present that require different treatments and approaches. Whether this represents disease phenotyping or just a better understanding of disease pathogenesis is unclear. Over the years, research focus has moved from the airway smooth-muscle via airway inflammation to airway immunology, and on to the epithelial-mesenchymal trophic unit. Future insight into disease pathogenesis and new treatment developments will follow with the use of the latest technologies, which include "omics" platforms and assessment of airway microbiota. Improving our understanding of stratified medicine for asthma-targeting the right therapy to the right patientaims to reduce hospital admissions due to poorly controlled asthma and fatalities. Although most asthma can be controlled by ICS and bronchodilators, there remains a cohort of patients with persistent symptoms, recurrent exacerbations, and worse quality of life. The current stepwise management approach (see Figure 28 .4) is inefficient, particularly in light of heterogeneity in clinical presentation and response to existing therapies, making a stratified approach essential.

Early work conceptualized asthma as extrinsic (allergic) and intrinsic (nonallergic) [59] . However, the recent use of unbiased approaches to classify disease using three large datasets and cluster analysis [60] [61] [62] has highlighted the different types of disease under the umbrella term asthma ( Figure 28 .5). Although these early studies were performed in different parts of the world and had statistical variations, their results were similar. Age at disease onset was found to be a key differentiating factor; early-onset disease was associated with more atopic and allergic conditions, whereas later-onset disease was associated with eosinophilic inflammation and obesity. These studies were all limited by a cross-sectional design and by the fact that treatment may have determined clusters. Longitudinal studies initiated in childhood are ongoing. For example, PACMAN (formally, "Pharmacogenetics of Asthma Medication in Children: Medication and Anti-Inflammatory Effects") aims to differentiate children with uncontrolled asthma, despite ICS treatment, using a range of clinical and cellular markers [63] .

The role of allergy (an inappropriate and harmful immune response to a normally nonharmful substance which requires sensitization) and that of atopy (the tendency to develop IgE antibodies to commonly encountered environmental allergens by natural exposure in which the route of entry is across intact mucosal surfaces) have been well studied in asthma. People with extrinsic asthma were thought to develop the disease earlier in life, be atopic, and have identifiable allergic triggers as well as other allergic diseases such as rhinitis or eczema. Intrinsic asthma was thought to develop later in life (after 40 years of age), and be associated with aspirin-exacerbated respiratory disease but not with allergic sensitization. When small studies in humans suggested that levels of Th2 cytokines were similar in extrinsic Anti-IgE Bronchial thermoplasty Macrolide antibiotics Antifungals (Itraconazole/Voriconazole) Anti-IL5 Anti-IL13 BOX 28.2 New and Emerging Therapies Reserved for Severe Asthma and intrinsic asthma, and that treatment with ICS was effective in the majority of mild to moderate asthma cases, the distinctions between extrinsic and intrinsic asthma fell out of favor [64] .

Heterogeneity in response to existing medication is exemplified by Malmstrom's study, which investigated individual response to leukotriene receptor antagonist (LTRA) (montelukast 10 mg once daily), and the inhaled steroid beclomethasone (200 μg twice daily) over a 12-week period in 895 chronic asthma patients [65] . Overall, there was a clear improvement in lung function (primary outcome FEV 1 ) for both treatments; however, when stratified based on individual patient data, there was a very large degree of heterogeneity for both treatment groups, with some patients showing up to 50% improvement in FEV 1 and others showing a decline in lung function with a 30% decrease in FEV 1 (Figure 28.6 ). This interindividual variability, a component of clinical trials that is seldom appreciated, is thought to have a strong genetic component.

The therapeutic regimen for the treatment of asthma has been well defined and was described earlier in this chapter. Despite such well-defined treatment regimens, which are constantly updated and streamlined, as well as the availability of high-quality medications, in many cases asthma is still not sufficiently controlled. While current asthma therapeutics are able to successfully treat 90-95% of the asthmatic population, around 50% still have daily symptoms and almost all patients report limitations of daily activities [66] . This also leaves around 5-10% of patients who do not respond to conventional treatments. This demonstrates that the current one-size-fits-all approach is not perfect, and emphasizes the need for a stratified medicine approach to asthma.

Variability in current therapy involves variation in drug efficacy and presentation of side effects, each of which presents challenges to successful maintenance of asthma. Variability in drug efficacy, especially when occurring for β 2 -adrenergic receptor agonists and glucocorticosteroids, and in the presentation of side effects, can result in patient therapy having to deviate from the standard therapeutic regimen for the treatment of asthma.

This high degree of interpatient variability in treatment response and in side effects can be attributed to a number of factors: (1) the interaction of genetic factors; (2) individual patient characteristics, such as weight, gender, and pregnancy; and (3) exposure to environmental insults, such as air pollution, allergens, and cigarette smoke. These factors interact, making the control of asthmatic episodes and the reduction of exacerbations more difficult. For example, asthma attacks can affect different age groups differently according to the season-children are more affected in summer while the older population suffers more in winter. When considering variation in treatment efficacy, one must also be mindful of individual patient factors such as inhaler technique and noncompliance (discussed earlier).

Currently asthma therapy is planned without consideration of genetic variance. However, genetic variance is arguably the most important of the factors listed previously-a claim supported by evidence gained from investigating the repeatability of response to therapy. This approach demonstrated that repeatability was between 60-80%, with a substantial proportion of the variance due to genetic factors [67] .

One important form of variance in therapeutic efficacy is non-response. Non-response is an important limitation in the maintenance and control of asthma, in that therapeutics expected to manage the disease at a certain level of severity fail outright, leading to the use of a trial approach in which different therapies are utilized with variable results until a suitable regimen is determined. A good example is severe asthma, where the patient is on permanent oral glucocorticosteroid therapy because of nonresponse to other therapies such as ICS, leukotriene receptor antagonists, and β 2 -adrenergic receptor agonists. Non-responders present with longer duration of asthma exacerbations and worse morning lung function, as well as a more frequent family history of asthma, when compared to those individuals responsive to glucocorticosteroid therapy [67] .

The recognition that allergen specific IgE activation of mast cells is central in driving allergic asthma lead to the discovery that the primary mast cell-signaling cascade could be inhibited by a monoclonal antibody toward the IgE binding site to the high affinity receptor (FcER1). Anti-IgE (omalizumab) became the first specific biologic to be used in the treatment of severe allergic asthma. Clinical trials revealed efficacy as well as almost total inhibition of early and late asthmatic responses to inhaled allergen [68] .

Following a rigorous examination of trial data by the National Institute for Clinical Excellence (NICE), omalizumab is now recommended as an option for treating severe

Individual Data persistent, confirmed allergic IgE-mediated asthma as an add-on to optimized standard therapy in people aged six and older who need continuous or frequent treatment with oral corticosteroids (defined as four or more courses in the previous year) in the United Kingdom. Although the drug is prescribed for cases of therapy-resistant asthma associated with allergy, it is also licensed for patients with evidence of allergy but a normal IgE. Omalizumab represents the first stratified treatment for asthma, but there is evidence that its use can be further stratified by using biomarkers of Th2 inflammation to predict response. In a recent paper studying the effect of omalizumab on uncontrolled severe persistent allergic asthma, treatment effect was analyzed in relation to F E NO, blood eosinophils, and serum periostin (an epithelial protein that is induced by IL13). After 48 weeks of omalizumab, reductions in protocol-defined exacerbations were greater in high versus low subgroups for all three biomarkers, suggesting a potential prognostic ability [69] .

This biomarker approach has helped divide asthma into and Th2 high and Th2 low disease. Asthma has traditionally been considered a Th2 process linked to atopy and allergy, type I hyper-sensitivity reactions, eosinophilic inflammation, and response to corticosteroids. "Th2 high disease," as evidenced by high F E NO and a sputum eosinophilia of >2%, has been associated with a better response to corticosteroids when compared to "Th2 low disease" [70] [71] [72] . Studies on lebrikizumab, a novel monoclonal antibody to IL13 [73] , have suggested that serum periostin may be a biomarker for a more general Th2 asthmatic phenotype. In one study a subgroup of individuals who had asthma and persistent elevation in serum periostin showed greater improvements in airway function and fewer exacerbations after lebrikizumab treatment than those with lower serum periostin. Interestingly, levels of F E NO, produced by inducible nitric oxide synthase, an enzyme induced in human airway epithelial cells by IL13, were as informative as periostin in identifying Th2-high individuals responsive to lebrikizumab.

Both the role of stratification based on phenotyping and the need for correct study end points in treatment trials have been informed by the salutary lesson of the anti-IL5 antibody mepolizumab. Although mepoluzimab effectively blocks eosinophilic inflammation, initial studies were disappointing in that no effect was demonstrated on spirometry or peak flow, despite a significant reduction in blood eosinophil numbers [74] . When the effect of the drug was studied on an outcome measure related to eosinophilia, namely severe exacerbations, there were significantly fewer exacerbations with a concomitant improvement in asthma-related quality of life [75] . Tellingly, there were no significant differences with respect to symptoms, postbronchodilator FEV 1 , or airway hyper-responsiveness, suggesting that different physiological processes determine these asthma outcomes. The study also helps researchers reflect upon the need to measure the correct end point in a disease that has different pathological mechanisms that all respond differently to treatment.

Phenotyping asthma has led to the identification of new disease targets, but aside from the use of anti-IgE, it has not yet led to a step change in management. Other approaches assessing the separate aspects of asthma pathophysiologynamely, airway inflammation, airway hyper-responsiveness, and airflow obstruction have been trialled with differing results. The key determinant in the design of these stratified approaches was the recognition that although asthma is a disease defined by symptoms, with treatment response also assessed by symptom improvement, there is no clear correlation between pathophysiology and symptomatology/ exacerbation risk.

In the FACET study (designed to evaluate the benefits of adding a long-acting β 2 -adrenergic receptor agonist to different doses of ICS), higher-dose ICS had a marked beneficial effect on exacerbation frequency, but relatively less effect on symptoms and peak expiratory flow. The opposite was true with the addition of long-acting β 2 -agonists [76] . This indicates that exacerbation frequency does not closely relate to symptoms and measures of disordered airway function, suggesting that the mechanisms responsible for these features are different [77] .

Numerous studies have demonstrated that in asthma the features of airway inflammation, airway hyper-responsiveness, variable airflow obstruction and associated symptoms can overlap, occur independently or change over time, in response to treatment or other external factors such as allergen exposure, or viral infection. One obvious example of this is eosinophilic bronchitis, a condition characterized by corticosteroid-responsive cough and the presence of a sputum eosinophilia occurring in the absence of variable airflow obstruction or airway hyper-responsiveness [78] .

Although asthma guidelines recommend the assessment of airflow obstruction (FEV 1 /PEF) and related symptoms for their primary treatment response outcomes, the realization is that this approach looks only at one aspect of asthma pathophysiology which is not closely related to exacerbation risk. This has led to new approaches aimed at reducing asthma exacerbations and symptoms while maintaining corticosteroid burden at the lowest possible level.

Airway hyper-responsiveness is defined as increased sensitivity to an inhaled constrictor agonist and a steeper slope of the dose-response curve. Two main forms of bronchoconstrictor stimuli exist: direct and indirect. Direct bronchoconstrictors, such as histamine or methacholine, stimulate receptors on the airway smooth-muscle, while indirect ones cause bronchoconstriction by secondary release of bronchoconstrictor mediators from mast cells or activation of neural pathways. Airway responsiveness is usually measured as the provocative dose of methacholine causing a 20% fall in FEV 1 by linear interpolation of the log dose-response curve (PC 20 ). In the general population, the distribution of airway hyper-responsiveness follows a continuous unimodal log-normal distribution, with asthma sufferers representing the hyper-responsive part of the distribution curve. A PC 20 is not usually measurable in non-diseased individuals, which suggests a large difference in airway responsiveness between non-diseased individuals and asthma patients.

The cut-off used to identify asthma is normally a methacholine concentration of <8 mg/ml. This value had a sensitivity of 100%, a specificity of 93%, and a negative predictive value of 100% in a study on a population of 500 college students with a diagnosis of current symptomatic asthma. The use of methacholine PC 20 to diagnose asthma has been evaluated; one study demonstrated that when asthma is defined as consistent symptoms with objective evidence of abnormal variable airflow obstruction, a positive methacholine challenge is more sensitive than PEF amplitude % mean and the acute bronchodilator response in diagnosis [79] .

The use of methacholine PC 20 to guide treatment has also been assessed. Although it was found that PC 20 -guided treatment resulted in a reduction in asthma exacerbations, this was at the expense of increased inhaled corticosteroid use [80] ; consequently, although PC 20 is often used for asthma diagnosis, or as a study end point, it is not routinely used to guide treatment decisions.

Asthma has been traditionally viewed as a condition where airway inflammation causes airway hyper-responsiveness, which in turn leads to variable airflow obstruction and symptoms; however, cross-sectional and longitudinal studies of airway inflammation using sputum induction in large populations with a diverse range of presentations suggest that this hypothesis requires modification. A recent bronchoscopy study demonstrated that bronchoconstriction is independent of airway inflammation and can lead to airway remodeling [38] .

The development of noninvasive techniques to assess airway inflammation, including induced sputum and F E NO, has made it possible to relate airway inflammation to objective measures of disordered airway function in larger and more heterogeneous populations than was possible with bronchoscopy studies. In general, these newer studies contradict findings in earlier studies and do not find a correlation between sputum eosinophil count and various markers of airway dysfunction.

It has been observed that a subset of patients with symptomatic asthma do not have sputum evidence of eosinophilic airway inflammation [81] . Many have sputum neutrophilia. This sputum profile is evident in corticosteroid-naïve as well as corticosteroid-treated subjects, suggesting that it is not always an artifact related to treatment. Importantly, patients with noneosinophilic asthma respond less well to inhaled budesonide than do a group with more typical sputum features [82] .

Similar sputum findings have been reported in more severe asthmatics; a subgroup of patients with refractory asthma have been identified who have bronchoscopic evidence of neutrophilic airway inflammation, normal eosinophil counts, and a normal basement membrane thickness. These findings suggest the presence of a distinct asthma phenotype characterized by a predominantly neutrophilic airway inflammatory response and relative corticosteroid resistance. Furthermore, there is evidence that neutrophilic asthma may result from activation of the innate immune system with the production of proinflammatory cytokines.

The use of noninvasive measures to guide treatment decisions by stratification based on the presence or absence of airway inflammation has been assessed. The strongest evidence for effect relates to the use of induced sputum to guide corticosteroid dose in moderate to severe asthma. Two studies have shown that using induced sputum differential counts to guide treatment results in fewer exacerbations for the same overall corticosteroid burden [81, 83] .

Although the evidence is clear that this approach works, induced sputum is not widely employed, possibly because it is seen as time consuming, requiring expertise both to induce the sputum and to perform the differential counts. The discovery that levels of F E NO, measured via a device similar to a breathalyzer, correlate well with sputum eosinophilia and relate to corticosteroid responsiveness, has driven the application of inflammometry (noninvasive measurement of inflammation in the airways). Inflammometry using F E NO measurements has also been used to guide and stratify treatment decisions. Though one study was positive in patients with asthma in pregnancy, with a reduction in exacerbations, most have failed to show an improvement when compared to guideline-driven (FEV 1 /PEF and symptoms) management. The hunt is still on for a reliable, inexpensive, and valid biomarker of airway inflammation.

FEV 1 and PEF measurements reflect changes in the caliber of the large airways. Our knowledge of anatomical and physiological changes in the small airways of patients with asthma is based on small case series of resected lung tissue from patients with asthma undergoing surgery for cancer, or on cases of fatal asthma.

These case series have demonstrated that there is significant inflammation present in the small airways (<2 mm diameter) in asthma. Fatal asthma is associated with peripheral airway inflammation and differences in the number of activated eosinophils in the distal lung. Other studies have revealed alterations in the epithelium and smooth-muscle, as well as mucous hypersecretion and distal airway plugging of the small airways. The presence of inflammation in the small airways in asthma may explain why small airways account for up to 50-90% of total airflow resistance in asthma, but only 10% of airflow resistance in normal airways.

Recently, the development of "small-particle" ICS, designed to target the peripheral lung, and the advent of new technologies-nitrogen washout, impulse oscillometry, and hyperpolarized noble gas magnetic resonance imaging, which allows assessment of peripheral lung function-have led to a resurgence of interest in the distal lung. Studies of small-particle ICS have been inconsistent; those comparing small-particle and standard-particle ICSs have failed to demonstrate improved asthma outcomes when administered in clinically comparable doses. Future asthma treatment may yet be stratified by the presence or absence of small airway inflammation.

There have been several attempts to base treatment decisions on measures of airway inflammation, including induced sputum and F E NO in children. Generally, they have been unsuccessful, although some experts do stratify treatment decisions on the response to oral or intramuscular steroids.

While methods of stratifying asthma patients to specific treatments based on nongenetic factors such as clinical outcomes, cellular measures, or protein biomarkers have shown some success, a large body of work has investigated the potential of genetic markers as predictors of patient responses to existing therapies, i.e., pharmacogenetics. Pharmacogenetics, the investigation of the effect of genetic polymorphisms on response to treatment or risk of adverse side effects, is one of the first steps in developing personalized prescribing. The use of genetic information to stratify patient prescribing is potentially more desirable compared to nongenetic stratification, as technology now allows a small amount of blood or saliva sample to be taken and a DNA test can be completed within hours.

To date, asthma pharmacogenetic studies have suffered from relatively small retrospective designs and a focus on only a few candidate genes; however, more recent, larger prospective studies have been completed that provide greater confidence in original findings and hypothesis-free approaches such as GWAS. These studies are primarily focused on pharmacodynamic aspects (e.g., improvement in lung function post-treatment), and only limited information is available about the impact of pharmacogenetics on adverse effects.

Pharmacogenetics in asthma is relatively advanced compared to that for other diseases. Genetic factors influencing the main treatment classes (i.e., β 2 -adrenergic receptor agonists, corticosteroids, and leukotriene modifiers) have been identified with some confidence (Table 28. 3). An overview of these findings follows; however, for more in-depth analyses, see Portelli and Sayers [84] .

β 2 -adrenergic receptor agonists carry out their function through the β 2 -adrenergic receptor-a 413-amino-acid G-protein-coupled receptor encoded by an intronless gene (ADRB2) located on chromosome 5q31.32. Binding of agonists to the receptor activates adenyl cyclase through stimulatory Gs proteins, which in turn activate protein kinase A. The latter phosphorylates several target proteins, resulting in a decrease in intracellular calcium that, importantly, causes smooth-muscle relaxation in the airways. It is not surprising that the majority of evidence of genetically driven effects on β 2 -adrenergic receptor-agonist therapy stem from the ADRB2 gene, which has therefore been extensively studied for pharmacogenetic effects on β 2 -adrenergic receptoragonist responses.

ADRB2 is a highly polymorphic gene containing 51 known and validated polymorphisms, of which 49 are SNPs and 2 are insertion/deletion variants. Most studies have focused on the role of four nonsynonymous coding-region polymorphisms: Arg16Gly (Arginine-to-Glycine substitution at position 16 in the protein), Gln27Glu, Val34Met, and Thr164Ile [85] . In a Caucasian population, the frequency of the polymorphisms at positions 16 and 27 were identified as 59% (Arg16Glu) and 29%, (Gln27Glu) [85] . The remaining Val34Met and Thr164Ile polymorphisms are rare, having approximate frequencies of <0.001% and 0.05%, respectively. The Arg16 variant has been shown to have pharmacogenetic potential through association with; An enhanced acute response to β 2 -adrenergic receptor agonists l A decline of asthma control following prolonged use of β 2 -adrenergic receptor agonists l A subsensitivity of response for bronchoprotection by β 2 -adrenergic receptor agonists However, several studies have failed to reproduce these effects, meaning that a common consensus on the contribution of Arg16 has yet to be reached. This lack of consensus can be explained when we consider that there are at least 12 different haplotypes (a specific combination of SNPs across the gene) in ADRB2. Investigations into the effect of a genotype in isolation, without consideration of its haplotype, is likely to introduce confounding into the association.

Functional effects of coding-region polymorphisms in ADRB2 have been identified through in vitro work carried out in cell lines. These have included the following: [86] . In one of the larger studies, Basu et al. identified an Arg16 copy numberdependent increase in disease exacerbations in 1182 patients with asthma aged 3 to 22 years on daily exposure to β 2 -adrenergic receptor agonist (regularly inhaled corticosteroid plus salbutamol on demand group, and regularly inhaled corticosteroid plus salmeterol and salbutamol on demand group) [87] .

It is important to note that this effect was driven by asthma patients who used salbutamol and/or salmeterol daily, supporting the suggestion that the Arg16 polymorphism has an integral role in the effectiveness of β 2 -adrenergic receptor-agonist therapy [87] . However, as with other studies, study limitations did not allow a clear association to be made between the ADRB2 gene (via its polymorphisms) and β 2 -adrenergic receptor-agonist efficacy. Namely, all participants were using β 2 -adrenergic receptor-agonists as a reliever; a study arm of non-β 2 -adrenergic receptoragonist reliever use would have provided a clearer interpretation of the detrimental effects of salbutamol/salmeterol in the Arg16 subjects.

Large studies have failed to observe a clinically relevant effect of these polymorphisms. These include longitudinal studies of β 2 -adrenergic receptor-agonist efficacy and the study of concomitant administration of corticosteroids.

A good example is a study of 2250 asthma patients randomly assigned to: (1) budesonide plus formoterol maintenance and reliever therapy; (2) fixed-dose budesonide plus formoterol; or (3) fixed dose fluticasone plus salmeterol for six months. No overall effect of the Gly16Arg genotype on clinical outcomes was found [88] . Another contributing factor to the pronounced differences in the conclusions from different studies investigating these ADRB2 polymorphisms is trial design variation.

The majority of studies have focused on the ADRB2 Arg16Gly polymorphism; however, other potential pharmacogenetically relevant polymorphisms also affect β 2 -adrenergic receptor-agonist efficacy. Multiple polymorphisms in the gene's regulatory regions that have potential clinical relevance are present in ADRB2. By studying eight common haplotypes based on 26 SNPs, a recent in vitro approach using "whole gene" transfection identified differential effects on receptor expression and downregulation that are haplotype-driven [89] . This study identified four common haplotypes with elevated receptor expression and two haplotypes with enhanced receptor downregulation. Another area recently investigated is polymorphisms occurring in the gene's untranslated regions, which may have an effect on gene expression and resultant drug efficacy.

Multiple genes are likely to be involved in the regulation of β 2 -adrenergic receptor-agonist response and expression of side effects, due to the agonists' known complex and multifactorial mechanism of action. One gene that has been associated with patient response to these agonists is Arginase 1 (ARG1), which was identified using a novel algorithm implemented in a family-based association test (FBAT) [90] . In this study of 209 children and their parents, the ARG1 SNP, rs2781659, was associated with bronchodilator response (BDR) when SNPs from 111 candidate genes (42 involved in β 2 -adrenergic receptor-signaling/regulation, 28 involved in glucocorticoid regulation, and 41 from prior asthma association studies) were investigated for their association with acute response to inhaled β 2 -adrenergic receptor-agonist in 209 children and their parents. In agreement with this study polymorphisms spanning ARG1 and influencing patient response to salbutamol have also been identified in a candidate gene study involving 221 asthma subjects [91] .

The ARG1 polymorphisms identified in both studies were in linkage disequilibrium (LD) (inherited together), suggesting a common causative mechanism involving potential transcriptional regulation due to the polymorphisms' location (predominantly 5′ to the gene). This alteration in transcription has now been confirmed in promoter-reporter studies, which found that the key ARG1 haplotype associated with improved BDR drives the highest level of ARG1 promoter activity [92] . In the study by Vonk et al., ARG2 SNPs were also associated with patient responses to salbutamol [91] , suggesting an integral role for the arginase family in β 2 -adrenergic receptor-agonist therapy.

Recently a S-nitrosoglutathione reductase (GSNOR) SNP (rs1154400, promoter region) was associated with a decreased response to salbutamol in 107 African-American children [93] . In the same study, a post hoc multilocus analysis discovered that a combination of rs1154400 with ADRB2 Arg16Gly, Gly27Glu, and the carbamoyl phosphate synthetase-1 (CPS1) SNP rs2230739 gave a 70% predictive value for lack of response to therapy [93] , implying that pharmacogenetic regulation of β 2 -adrenergic receptoragonist therapy may depend on several loci acting together via gene-gene interactions. In confirmation, 4/5 SNPs tested in GSNOR were associated with asthma patient responses to salbutamol in 168 Puerto Rican asthma patients [94] . These SNPs were also associated with asthma susceptibility, and the key risk haplotype was associated with increased transcriptional activity based on promoter-reporter studies [94] . GSNOR is an alcohol dehydrogenase that breaks down GSNO, an endogenous bronchodilator [95] . In addition, GSNO regulates nitrosylation of proteins, leading to alterations in function, including G-protein coupled receptor kinase 2 (GRK2), which phosphorylates and desensitizes the β 2 -adrenergic receptor [96] .

Other novel genes have recently been associated with β 2 -adrenergic receptor-agonist therapy including the spermatogenesis-associated, serine-rich 2-Like (SPATS2L) and collagen (COL22A1) genes [97] . The COL22A1 gene, through association with the intronic SNP rs6988229, was associated with acute bronchodilator response to inhaled salbutamol in a genome-wide association study in which ∼500,000 SNPs were tested in 403 Caucasian trios. This association was replicated in a pooled population of three asthma trial populations, as well as in three additional asthma populations [97] . SPATS2L was also identified as a modulator of β 2 -adrenergic receptor-agonist function through a GWAS. Here, the SNP rs295137, located near the SPATS2L gene, was significantly associated with percentage change in baseline FEV 1 in 1644 Caucasian asthma patients; this was replicated in two alternate Caucasian populations (n ∼ 500 each). Molecular biology techniques confirmed these results and identified that SPATS2L may be an important regulator of β 2 -adrenergic receptor downregulation.

The identification of patients at risk from potential adverse effects of β 2 -adrenergic receptor-agonists remains a critical clinical question, as does the targeting of this class of drug to patients most likely to benefit from it. While there has been clear progress with respect to study design (e.g., examining haplotypes instead of genotypes in isolation) and adequately powered studies using thousands of individuals, there is still a need for large prospective studies of asthma patients with matched phenotypes and carefully controlled covariates that include environmental influences. Similarly, these studies would benefit from GWA approaches and the identification of gene-gene interactions.

While investigations of the effect of genetic polymorphisms on β 2 -adrenergic receptor-agonist responses have predominantly focused on clinical end points (e.g., lung function parameters) in the different genotype groups, there has been recent interest in the real-life application of genetic knowledge. As outlined, the first-line treatment for asthma is a shortacting β 2 -adrenergic receptor-agonist (e.g., salbutamol) as needed (step 1); if symptoms persist, the addition of inhaled corticosteroid (e.g., beclomethasone) is considered (step 2); for further control, a long-acting β 2 -adrenergic receptor agonist (e.g., salmeterol) or a leukotriene receptor antagonist (LTRA) (e.g., montelukast) is added (step 3) (Figure 28.4) .

In a recent study, Lipworth et al. set out to use genetic information on the ADRB2 Arg16 polymorphism to inform the choice of prescribing salmeterol or montelukast as addon therapy [98] . Children with persistent asthma and homozygous for the Arg16 genotype (n = 62) were randomized to receive salmeterol (50 μg, bd) or montelukast (5 or 10 mg, once daily) as an add-on to inhaled fluticasone propionate for one year. The study tested whether carriers of the Arg16 genotype were more prone to adverse effects (e.g., prolonged β 2 -adrenergic receptor-agonist use associated exacerbations), and hence whether montelukast provided superior control for this preselected population.

Outcomes were school absences (primary outcome), exacerbation score, reliever use (salbutamol), morning dyspnoea, and Asthma Control Questionnaire (ACQ) qualityof-life scores. Montelukast provided superior benefit for all measures, with clinically relevant differences within three months. No significant difference in FEV 1 (% Pred) was observed, providing further evidence of limited correlation between lung function and symptom-based scores [98] . These results suggest that larger and longer prospective studies are warranted to provide more definitive data on the clinical utility of Arg16 stratification including a Gly16 study arm and the use of additional markers to define the population would provide clearer interpretation.

As outlined, leukotriene synthesis inhibitors (LTSIs) and leukotriene receptor antagonists (LTRAs) are commonly used as an add-on therapy in asthma to provide greater control or steroid-sparing effects. The cysteinyl leukotrienes (LTC 4 , LTD 4 , LTE 4 ) and dihydroxy leukotriene (LTB 4 ) contribute to the inflammatory process in asthma and are synthesized from arachidonic acid via the 5-lipoxygenase pathway (Figure 28.7) . Cysteinyl leukotrienes have been implicated in bronchoconstriction, mucus secretion, vascular permeability, inflammatory cell infiltration and cytokine production. Arachidonic acid is converted to 5-hydroperoxyeicosatetraenoic acid and leukotriene A 4 (LTA 4 ) by 5-lipoxygenase (5-LO/ALOX5) and 5-lipoxygenaseactivating protein (FLAP/ALOX5AP), which acts as an adaptor protein for this reaction. LTA 4 is then converted to leukotriene B 4 (LTB 4 ) by LTA 4 hydrolase (LTA4H) or, alternatively, is conjugated with reduced glutathione to form LTC 4 via the actions of leukotriene C4 (LTC 4 ) synthase (LTC4S). LTC 4 is next transported to the extracellular space via the multidrug resistance protein 1 (MRP1) ( Figure  28 .7). Leukotriene modifiers include LTSIs, which act by targeting 5-LO (e.g., zileuton), resulting in a decrease in all leukotriene biosynthesis (LTC 4 , LTD 4 , LTE 4 , and LTB 4 ) or LTRAs, which act by specifically blocking cysteinyl leukotrienes from binding to their primary receptor, CYSLTR1, which is found on many cell types including inflammatory cells and airway smooth-muscle cells, (e.g., montelukast and zafirlukast).

Data demonstrating the large degree of heterogeneity of asthma patient responses to this class of drug are summarized in Figure 28 .6. These data have led to extensive genetic studies investigating the roles of SNPs in a large number of candidate genes associated with leukotriene production and/or activity and acute responses to LTRAs and LTSIs. Associations have been described for polymorphisms in ALOX5, ALOX5AP, LTC4S, CYSLTR1, CYSLTR2, and MRP1; however, many of these studies have been small and so have led to a lack of reproducibility in findings [99] . More recent studies have looked at different aspects of leukotriene modifier functions, including the association of montelukast absorption (i.e., pharmacokinetics) with polymorphisms in OATP2PB1 [100] . However, others have failed to observe these associations [101] .

5-LO (along with 5-LO-activating protein) is the major regulatory switch for leukotriene production, and extensive studies have demonstrated that the level of 5-LO can have dramatic effects on both cysteinyl and dihydroxy leukotriene production. 5-LO catalyzes the conversion of arachidonic acid to LTA 4 , one of the early stages in leukotriene production ( Figure 28.7) . ALOX5 is found on chromosome 10q11.2 and is a large gene composed of 14 exons and 13 introns; it spans approximately 82 kb. Multiple studies suggest that ALOX5 polymorphisms can influence clinical responses to LTRAs and LTSIs-in particular, a functional repeat polymorphism in the promoter region (resulting in alterations in SP1 transcription factor binding) have been associated with altered gene transcription and with response to the LTSI, ABT-761 [99, 102] .

These early studies have now been extended to investigate multiple polymorphisms, spanning the entire gene, and association with LTRA and/or LTSI responses in asthma patients. For example, Tantisira et al. observed an association between ALOX5 intronic SNPs-rs892690, rs2029253, and rs2115819-and change in FEV 1 post-LTSI(zileuton)-in a cohort of 577 asthma patients [103] . The rs2115819 SNP was also a predictor of response to the LTRA, montelukast, in a previous study of 252 asthma subjects using percentage change in FEV 1 as the primary outcome [104] . In both cases, the GG versus GA or AA genotype had the greatest improvement in lung function (FEV 1 ) post-drug [104] . The lack of association for the rs892690, and rs2029253 SNPs in the montelukast study may be due to the reduced power of this study compared to the zileuton study. The functional significance of these SNPs remains to be resolved, and it is important to note that they may not be the causative genetic change, and that polymorphisms inherited at the same time (i.e., in LD) may be of relevance. Two ALOX5 SNPs (rs4987105 (synonymous Thr120Thr) and rs4986832 (5′ region)) were also associated with improvement in PEF on treatment with montelukast in a 12-week study in 174 asthma patients. This further confirmed the relevance of ALOX5 SNPs and this key enzyme in leukotriene production [105] . Additional support for the relevance of ALOX5 in asthma control is provided by a recent study of 270 children which found that the ALOX5 Sp1 promoter polymorphism determined urinary LTE 4 levels and was associated with reduced lung function and, potentially, an ACQ score in non-5 copy carriers [106] . In this study, there was an inverse relationship between urinary LTE 4 levels and FEV 1 .

Another important enzyme in the leukotriene pathway is leukotriene C 4 synthase (LTC4S), which specifically results in the generation of the first cysteinyl leukotriene, LTC 4 , by conjugating glutathione to LTA 4 . There has been intense research into the role of LTC4S in asthma susceptibility because this enzyme, like 5-LO, is thought to be a key regulatory switch for cysteinyl leukotrienes, which are thought to play a more prominent role in asthma (e.g., via bronchoconstriction) than the dihydroxy LTB 4 . LTC4S is relatively small, spanning 2.51 kb on chromosome 5q35. In particular, a promoter polymorphism (A-444C, rs730012) has been intensely investigated for pharmacogenetic effects as it is thought to alter LTC4S levels via transcription (higher levels in C allele carriers) and therefore alter LTC 4 production [107] .

Of the ten published studies investigating the role of LTC4S-444 A>C, five showed an improvement in outcome measures, including FEV 1 post-LTRA therapy for carriers of the C allele, as hypothesized, although results were not always statistically significant. In a more recent study of zileuton responses (FEV 1 change over time) in 577 asthma subjects, the LTC4S-444 polymorphism showed no effect; however, an alternative LTC4S polymorphism (rs272431, intron 1) was associated with improved response (mean FEV 1 ) [103] . As these results were obtained in different populations and using different end points, the relative contribution of the LTC4S polymorphism to LTRA or LTSI therapeutic responses in asthma are still unclear. Larger prospective studies using multiple ethnic groups are required.

Montelukast and other LTRAs target the cysteinyl leukotriene receptor 1, which is expressed in a variety of cells, including airway smooth-muscle and various inflammatory cells (e.g., eosinophils), therefore inhibiting the activation of this receptor by cysteinyl leukotrienes (Figure 28.7) . Cysteinyl leukotriene receptor 1 is thought to be the main receptor mediating cysteinyl leukotriene receptor smooth-muscle contraction and inflammatory cell cytokine production in asthma. Again, genetic variation in the target receptor for these compounds may influence how effective they are in carriers of these alleles by regulating receptor function/expression. The cysteinyl leukotriene receptor 1 gene, CYSLTR1, is intronless and found on chromosome Xq13-21; it generates a 337 amino acid G-protein-coupled receptor (GPCR).

A second receptor for cysteinyl leukotrienes has also been described and, while not the target of LTRAs, may modulate the effects of cysteinyl leukotrienes in vivo. The gene for cysteinyl leukotriene receptor 2 is found on chromosome 13q14 and encodes for a 346 amino acid GPCR. Interestingly, in combination with several ALOX5 SNPs, polymorphisms spanning CYSLTR1 have been identified as determinants of responses to montelukast, suggesting SNP-SNP interactions may be important [108] . However, direct evaluation of CYSLTR1 polymorphisms and responses to montelukast [104] or zileuton [102] have not identified a significant effect for them to date. Studies of CYSLTR2 are limited; however, SNPs rs91227 and rs912278 (3′UTR) have been associated with an improvement in morning PEF following administration of montelukast for 12 weeks [105] .

While the majority of studies have focused on candidate polymorphism analyses in genes directly involved in leukotriene production or activity, several studies have identified additional genes of relevance to LTSI and LTRA response. In addition to its effects on vasodilation and bronchoconstriction, the prostaglandin D2 receptor (gene: PTGDR) is thought to regulate, at least in part, levels of leukotriene C 4 . In a study of 100 asthmatic children prescribed montelukast (5 mg/day), a modest effect of the PTGDR-4441T/C was observed.

Similarly, in a more extensive study of 169 SNPs in 26 candidate genes in asthma patients prescribed fluticasone, fluticasone propionate plus salmeterol or montelukast, a significant association between four SNPs in the cholinergic muscarinic receptor 2 gene (CHRM2) and montelukast response (change in FEV 1 ) over 16 weeks was observed [106] . CHRM2 and CHRM3 are thought to mediate airway tone via regulation of contractile/relaxation responses and targeting of CHRM3 using antagonists; for example, tiotropium has shown clinical efficacy in multiple respiratory diseases. These data potentially suggest that the nature of the airway obstruction/tone determined by altered CHRM2 expression and/or activity may influence responses to montelukast.

While the majority of studies have investigated the association between gene polymorphisms and acute responses to asthma therapy, a recent study by Mougey et al. identified that solute carrier organic anion transporter family, member 2B1 (SLCO2B1, alternative name: organic anion transported SB1; OATP2B1), was able to mediate montelukast permeability using a model cell system engineered to express the OATP2B1 protein [100] . Subsequently, the same group also showed that an alteration in the protein structure of OATP2B1 that occurs naturallythat is, Arg312Gln (rs12422149)-was associated with reduced morning plasma concentrations of montelukast following an evening dose during one month or six months of treatment in 80 asthma patients [100] . More specifically, GA (Arg/Gln) genotype carriers had ∼20% lower montelukast concentration than the GG (Arg) group at one month, and ∼30% lower concentration at six months, and of clinical relevance A allele carriers did not demonstrate benefit from montelukast using a symptom-based score [100] .

The same researchers have now replicated these findings on montelukast absorption [109] ; however, another study did not replicate this finding and failed to identify any effect of the Arg312Gln polymorphism on montelukast plasma concentrations, albeit in a different study design involving fewer subjects [101] . These studies highlight the potential importance of genetic factors influencing drug transporters that may be anticipated to affect pharmacokinetics and the pharmacodynamics. Similarly, several SNPs in another transporter protein, multidrug resistance protein 1 (MRP1) (alternative name: ATP binding cassette, subfamily C, member 1 (ABCC1))-for example, rs119774have been associated with montelukast response (change in FEV 1 % predicted) [110] and zileuton response [103] . This association potentially confirms the pharmacogenetic significance of this MRP1 SNP marker. This SNP is intronic, and the underlying functional mechanism, including which genetic variant explains these effects, remains to be resolved.

Overall, there has been good progress in the pharmacogenetics of leukotriene modifier therapy in asthma, with SNPs in multiple leukotriene-synthesising enzymes (e.g., ALOX5) showing robust association with LTA and LTSI responses. Multiple genes in the leukotriene synthesis pathway and/or receptors have polymorphisms that have been associated with asthma susceptibility, which implies that there may be a more leukotriene-driven asthma that is therefore more amenable to treatment with LTSIs and LTRAs. Of interest in this drug class, preliminary data suggest a significant contribution of SNPs in drug transporter genes, ultimately determining both the pharmacokinetics and the pharmacodynamics. These data set the scene for larger prospective studies to provide accurate effect sizes and determine clinical implications with greater confidence.

Corticosteroids are an important pharmacogenetic target in asthma. Initial investigations into a pharmacogenetic approach for the corticosteroid element of asthma therapy focused on the glucocorticoid receptor gene (GR, alternative name: nuclear receptor subfamily 3, group C, member 1, NR3C1), which maps to chromosomal region 5q31-a region associated with multiple asthma phenotypes. As with other genes discussed in this chapter, several polymorphisms have been described in the GR gene that have functional consequences, such as a Val641Asp polymorphism that has been shown to influence the binding affinity for dexamethasone. However, similar to most polymorphisms shown to have potential pharmacogenetic effects, several of these polymorphisms are rare and their functional significance is questionable [84] . Despite this, the GR remains a tantalizing target for asthma pharmacogenetics, and it is surprising that there have not been further recent investigations into the role of GR polymorphisms in corticosteroid responses.

GR is not the only gene to be associated with geneticbased variance in corticosteroid efficacy in asthma treatment. A number of other genes have been implicated, including CRHR1, TBX1, NK2R, STIP1, DUSP1, and FCER2, and these are discussed below.

SNPs in the corticotrophin-releasing hormone receptor 1 (CRHR1) gene are associated with response to inhaled corticosteroid treatment, based on end point change in FEV 1 following 8 weeks of treatment in three asthmatic cohorts using a candidate gene approach (131 SNPs in 14 genes, n = 1117 asthma subjects). For example, the intronic CRHR1 SNP rs242941 exhibited genotype-specific changes in the percentage predicted change in FEV 1 in response to corticosteroid therapy in 470 adult asthma subjects [111] . This study was the first to show a pharmacogenetic effect for steroid efficacy in an asthmatic cohort, and it highlights the CRHR1 gene to be at least one of a number of factors determining corticosteroid efficacy. CRHR1 is thought to influence responses to exogenously administered corticosteroid through the regulation of endogenous levels of corticosteroid.

The T box 21 (TBX21) gene has been shown to be a predictor of improvement in bronchial hyper-responsiveness (BHR) (4 year change) post-corticosteroid treatment in children, based on a genotype-dependant variation of the nonsynonymous SNP rs2240017 (His33Gln), where the presence of the G allele gave greatest improvements [112] . However, the association of this gene remains under question because of its low allele frequency in Caucasians (MAF ∼0.04), which resulted in only a few subjects (n = 5) being available to contribute to this observation [112] . Additional data have given more confidence to the involvement of TBX21 because the His33Gln polymorphism was significantly associated with improved asthma control in the presence of corticosteroids, although the alternative allele to that described was associated with greater control in 53 Korean asthma patients during 5-12 weeks treatment [112] . Interestingly in the same study, Ye et al. identified a novel gene, the Neurokinin 2 receptor (NK2R), which was associated with improved asthma control in the presence of corticosteroids; the G allele (Gly) of the NK2R SNP rs77038916G/A, (Gly231Glu) was associated with the greatest improvement. Neurokinin A induces bronchoconstriction and inflammation, therefore modulation of the neurokinin receptor may influence the magnitude of these responses.

Multiple SNPs in the stress-induced phosphoprotein 1 gene (STIP1) were associated with variable FEV 1 responses to treatment with the inhaled corticosteroid flunisolide in 382 asthma subjects [103] . This study investigated SNPs spanning eight candidate genes and identified STIP1 SNPs rs4980524, rs6591838, and rs2236647 as SNPs affecting percent change in FEV 1 in response to flunisolide at 4 weeks and 8 weeks [103] . STIP1 codes for an adaptor protein that coordinates functions with HSP70 and may be involved in formation of the glucocorticosteroid receptor heterocomplex.

The dual-specificity phosphatase 1 (DUSP1) gene encodes a protein that has dual specificity for tyrosine and threonine and inactivates p38 mitogen-activated protein kinase (MAPK). Recently, several SNPs including a 5′ region SNP, rs881152, were associated with (1) bronchodilator responses and (2) asthma control in the presence of corticosteroid in a cohort of asthma patients (n = 430) [113] . The mechanism underlying these effects is unclear; however, it was suggested that corticosteroid induces DUSP1 expression, which may be altered in carriers of the rs881152 5′-region SNP. This influences the ability of DUSP1 to target the p38 MAPK signaling pathway.

In a cohort of 311 asthmatic children, the low-affinity IgE receptor gene (FCER2) has been highlighted as putatively involved in corticosteroid (budesonide) regulation of exacerbation by virtue of an association between the SNP T2206C (rs28364072, intronic) and the relative risk of severe exacerbations while taking budesonide [114] . Interestingly, the C allele was also associated with elevated IgE levels in the 311 asthma patients [114] . Importantly, this pharmacogenetic association has now been replicated in two other asthma cohorts (n = 386 and 939, respectively), with the 2206C allele being associated with increased hospital visits and uncontrolled asthma in patients receiving ICS [115] . It is not surprising that this polymorphic variation in an IgE receptor influences receptor expression, which alters IgE levels and is associated with more severe forms of asthma. In turn, these subjects cannot be adequately controlled by corticosteroid treatment.

The first GWAS in asthma, published in 2012, identified SNPs in the glucocorticoid-induced transcript 1 gene (GLCC1) as determinants of glucocorticoid response in 935 asthma subjects [116] . This study examined 534,290 SNPs using the Human Hap 550v3 Beadchip to identify 100 SNPs of interest in an initial cohort of 403 parent-child trios. The primary outcome was change in FEV 1 . Of interest was a SNP in the GLCC1 gene promoter region (rs37972C/T) that was associated with attenuation in FEV 1 improvement post-corticosteroid in three of four cohorts tested [116] .

A subsequent GWAS identified the T-gene as a novel player in lung function response (FEV 1 ) to inhaled corticosteroids. Analyses were carried out in 418 Caucasian asthmatics and replicated in a secondary asthmatic population (n = 407), where 3 out of 47 successfully replicated SNPs were associated under the same genetic model in the same direction, including 2 of the top 4 SNPs ranked by p value. These SNPs (rs3127412C/ T and rs6456042A/C) were in strong LD with a variant located in the T-gene, suggesting that this gene is a novel player in the pharmacogenetic regulation of corticosteroid efficacy. This was confirmed by a follow-up association, where the T-gene variant was associated with lung function response to inhaled corticosteroids in the initial GWAS analysis (an average 2-3-fold increase in FEV 1 response for homozygous wild-type subjects; rs3127412 TT; and rs6456042 CC) [117] .

Not all recent developments have been due to GWAS. For example, a recent hypothesis-driven study implicated the cytochrome P450 3A enzymes in glucocorticosteroid efficacy through modulations in corticosteroid metabolism. In this study, Roberts et al. showed, through mRNAdriven studies confirmed via microsomes and recombinant enzymes, that CYP3A4 and CYP3A5 metabolize corticosteroids (budesonide) into inactive metabolites [118] . They suggested that differences in the expression or function of these enzymes in the lung and/or liver could influence corticosteroid efficacy in the treatment of asthma.

To date, the majority of pharmacogenetic studies have focused on one class of asthma therapy. However, several recent studies have investigated combination therapy, which more accurately reflects the clinical situation. We have summarized the main findings regarding the influence of ADRB2 polymorphisms, including Arg16Gly, on combination therapies, particularly fluticasone/salmeterol and budesonide/formoterol in Section 28.5.5.1. However, it is clear that reports on associations between ADRB2 Arg16 and increased exacerbation in asthma patients using regular salbutamol in addition to their maintenance combination are mixed, with some studies failing to report the association, potentially because of differences in study design.

More recently, additional genes have been investigated for SNP associations with patient responses to combination therapy. In a study of 81 asthma patients regularly using an ICS/LABA combination, it was shown that polymorphisms in the nitrous oxide synthase 3 (NOS3) gene involving an amino acid change (G894T, Glu298Asp), were associated with post-drug improvement in lung function (change in FEV 1 (% Pred)) [119] . NOS3 codes for a key enzyme in the production of NO in the airways, and F E NO has been shown to be elevated in asthma and is considered a marker of ongoing inflammation.

Importantly, earlier data identified that the Glu298Asp polymorphism is functional in NOS3 and influences levels of F E NO with the TT genotype associated with lower F E NO levels. These data suggest that lower NOS3 activity and potentially lower F E NO levels identify patients more likely to respond to combination therapy; however, larger prospective trials are required to validate these findings because they are potentially counterintuitive if NO is a marker of inflammation and ICS target inflammation.

A recent 16-week study on the effect of genetic determinants (169 SNPs in 26 candidate genes) on asthma patient response to combined fluticasone propionate/salmeterol therapy identified three SNPs in the cholinergic muscarinic receptor 2 gene (CHRM2) that are associated with asthma ACQ scores, and found a single SNP, rs1461496 in heat shock 70kD protein 8 (HSPA8), to be associated with change in FEV 1 [106] . Of interest, CHRM2 SNPs were also associated with responses (lung function) to montelukast (see earlier), suggesting that the "tone" of the airways as determined by muscarinic receptors is important for therapies that target airway contraction. HSPA8 is involved in protein folding in the cell, but it is important to note that HSPA8 SNPs previously were not associated with response to ICS [120] .

Drug development in asthma has been slow to generate new classes, and the major advances in patient care have come from new compounds and/or indications for existing drug classes (e.g., improved duration of action for LABAs). One potential explanation for this limited progress is the design of and recruitment to Phase I and Phase II trials for the evaluation of new compounds. We have already outlined how initial trials of anti-IL5 demonstrated disappointing effects on lung function and asthma symptoms. Only after careful selection of patients, based on clinical and cellular parameters, were clinically relevant improvements observed. Researchers now understand that asthma is heterogeneous and, particularly for strategies targeting specific mediators, careful selection of patients is necessary to fully understand the therapeutic potential of new drugs.

Recent approaches to the treatment of asthma [121] include the following: The specific targeting of single mediators is unlikely to provide a therapeutic option for asthma in its broadest definition. However, targeting specific subpopulations has shown clinical efficacy. Very recently, a study investigating a human IL13-neutralizing monoclonal antibody (tralokinumab) used a carefully balanced population of atopic and nonatopic asthmatics, with exclusions for additional respiratory pathology, cigarette smoking ≥10 pack-years, recent infection, or treatment with immunosuppressive medication, in a randomized double-blind study [122] . Here, tralokinumab was associated with improved lung function based on an increase in FEV 1 and a decrease in daily β 2 -adrenergic receptor-agonist use [122] . These studies overall demonstrate that careful selection of asthma subjects, in this case based on phenotype, is critical in asthma drug development.

As novel asthma genes are identified, it is paramount that their evaluation as potential drug targets, particularly in the context of Phase II trials, takes into account pharmacogenetic factors. Preliminary data support genetic testing in Phase II trials of newer compounds, as shown in a recent report on a IL4/IL13 dual antagonist (pitrikinra). Polymorphic variation in the target receptor for this antagonist (i.e., IL4Rα) significantly influenced outcomes in allergic asthma subjects [123] . Pitrikinra is a recombinant form of IL4 differing at two amino acid residues (i.e., a mutein, R121D/Y124D). This has been shown to reduce late-phase antigen responses (LAR) to inhaled antigen as defined by changes in lung function (FEV 1 ) over 4-10 hours postantigen exposure.

In one study, subjects had been following a four-week period of twice daily active treatment of nebulized pitrikinra or placebo, with an increased LAR ratio correlating with a reduced FEV 1 [124] . Stratification of subjects (pitrikinra n = 15, placebo n = 14) into genotype groups for the nonsynonomous IL4RA SNP rs1801275 (Gln576Arg) identified Arg/Arg carriers as having an attenuated LAR (P < 0.0001) following pitrikinra treatment compared to Gln/Gln or Gln/Arg genotypes. Similarly, stratification based on rs1805011 (Glu400Ala) showed an attenuated LAR in the Glu/Ala group but not the Glu/Glu group [123] . Interestingly, the Arg576 variant (when in combination with Val75) was shown to be a risk factor for allergic asthma and to lead to enhanced IL4Rα signaling post IL4 stimulation [125] .

These preliminary data suggest that selecting a subgroup of patients with a particular genotype when it is anticipated that the receptor/pathway may have a more dominant role in that individual's asthma is critical to interpreting Phase II clinical trials.

Approaches to further define asthma subphenotypes (e.g., using cluster analyses) have been useful and further confirm this heterogeneity in clinical presentation. While current medications have been extremely successful in asthma management (e.g., ICS), it is clear that individual responses to these medications and to therapies in development are heterogeneous, with large variation in clinical benefits and/ or detrimental effects. The stratification of asthma treatment using clinical and/or genetic approaches therefore shows great potential for maximizing clinical outcomes and minimizing adverse effects, leading to improvement in the management of asthma.

There has been excellent progress in this area of research over the last five to ten years with the introduction of anti-IgE (omalizumab) therapy into mainstay asthma treatment, but only for those patients with clinical indications (i.e., severe allergic asthma) with an elevated serum IgE level. This stratification in prescribing for anti-IgE shows proofof-concept by targeting this expensive biologic to those patients most likely to show clinical benefit. Similarly, there is accumulating evidence that stratification of patients in Phase II trials of newer therapies-for example, anti-IL5 (mepoluzimab) and anti-IL13 (lebrikizumab), again based on clinical and cellular patient profiles, is essential to adequately evaluate these therapies and will be important if these therapies are introduced into clinical practice.

Clear progress in asthma pharmacogenetics has come with completion of newer studies that interrogate multiple SNPs in large prospective cohorts. To date, predominantly candidate gene/pathway approaches have been used, but several genetic variants have now been identified with confidence (e.g., ALOX5 and leukotriene modifier response; FCER2 and glucocorticoid response). Importantly, these studies show independent replication (the gold standard in genetic studies, including specific SNPs and direction of effect).

Also importantly, several GWASs have now been completed that provide integration of common variation spanning the entire genome (typically testing 500,000 SNPs) identifying novel gene variants in a hypothesis-free approach (e.g., the collagen (COL22A1) locus and SPATS2L and response to albuterol/salbutamol). These more recent findings still need further replication and validation; however, they potentially pave the way for a "responder" and "nonresponder" profile based on multiple SNPs in multiple genes, all measured on the same platform. A key question is the relative contribution of common variation (measured on these platforms) and rare variation (not currently measured to these outcomes).

Another remaining question relates to the magnitude of changes in clinical measures between genotype groups in asthma patients (e.g., FEV 1 ). Data so far suggests that these differences can be considered clinically relevant; however, further work with larger populations is required to accurately estimate effect sizes. Similarly, the genetic variation identified to date accounts only for a modest proportion of the overall variation of the trait (e.g., a recent estimate suggests the hereditability of bronchodilator response at ∼28.5%). It is likely that studies so far have been confounded by factors including SNP/haplotype analyses, gene-gene interactions, and the relative contribution of SNPs in different ethnic backgrounds. Therefore, while our knowledge has dramatically increased, there is a need for prospective trials of these current compounds that involve large numbers of subjects and hypothesis-free approaches including GWAS.

As with GWAS approaches to identifying disease susceptibility genes, "missing hereditability" will be an important issue in pharmacogenetics studies. This small proportion of hereditability currently assigned to identified loci (e.g., CRHR1 < 3%) of corticosteroid response needs to be addressed because asthma medications cannot be successfully personalized unless we resolve a larger degree of genetic variability and/or environmental factors.

As with clinical/cellular approaches to stratifying Phase II trials of newer asthma therapies, it is clear that genetic stratification has a role to play. For example, the recent Phase II trials of the IL4/IL13 dual antagonist (pitrikinra) suggests patient selection is critical to evaluating the clinical efficacy of this compound, with specific IL4Rα genotypes identifying responder groups. More pharmacogenetic integration is required for Phase II trials, which represents a clear challenge for drug development pipelines.

As our understanding of the genetic basis of asthma increases other therapeutic targets for asthma will become apparent (e.g., IL33 receptor antagonists). It is clear that the genetic variation associated with asthma susceptibility in these genes will influence pathways targeting methods, as they essentially identify individuals in which the pathway may be particularly important. It is also likely that existing therapies may be influenced by the genetic factors that underlie asthma susceptibility, particularly therapies that have multiple targets (e.g., ICS).

The impact of our genetic knowledge on asthma management and prescribing practice is limited at this time, as there is a need to further define the complex genetic basis of responders and adverse effects in large prospective studies. However, preliminary "real-life" studies with related outcome measures (e.g., days from school) have provided proof of concept that genetic information can have important implications for disease management (e.g., informing second-line therapy in asthma using ADRB2 Arg16) [98] .

There remain many unanswered questions regarding the potential of stratified approaches in the management of asthma. While progress has been made in identifying patient subgroups based on clinical measures, cell counts, or marker expression, more work is required using very large studies to extensively characterize asthma patients not only cross-sectionally, but also longitudinally over many years to identify the natural progression of the disease.

The field of genetics is rapidly progressing, with great technological developments moving at a dramatic pace, e.g., GWAS comprising a million common SNPs or several million rare SNPs, gene expression studies, and wholeexome or targeted resequencing. These approaches will allow the interrogation of genetic variation effect in patient subgroups. It is likely that a drug-specific genetic profile involving several genes (e.g., receptors, signaling mediators, transcription factors) will be a step toward personalized medicine in asthma, with associated benefits including avoidance of adverse side effects and adequate control of the disease. We anticipate that through these approaches many novel common variants/genes will be identified as underlying responses to current asthma medication. As the cost of genetic analyses continues to fall, the potential of a relatively simple test that will have an impact on asthma management costs becomes a real possibility.

If genetic or clinical information is going to influence clinical practice, there must be a simple, reproducible diagnostic test based on SNP combinations or simple clinical measures, with adequately high specificity and sensitivity that distinguishes responders and nonresponders. However, the introduction of routine genetic and/or clinical testing into practice will require a clear demonstration of health benefits and cost-effectiveness that outperforms the current stepwise approach to asthma management.

A plea to abandon asthma as a disease concept

A brief history of asthma and its mechanisms to modern concepts of disease pathogenesis

Guidelines for the diagnosis and management of asthma. National asthma education and prevention program: expert panel report III. NIH Publication No. 08-4051

British Thoracic Society and Scottish Intercollegiate Guidelines Network: British Guideline on the Management of Asthma. A national clinical guideline

Economic burden of asthma: a systematic review

Asthma-related resource use and cost by GINA classification of severity in three European countries

Inhaler competence in asthma: common errors, barriers to use and recommended solutions

Low-dose inhaled corticosteroids and the prevention of death from asthma

Global initiative for asthma (GINA) and its objectives

An official American thoracic society/European respiratory society statement: asthma control and exacerbations: standardizing endpoints for clinical asthma trials and clinical practice

Differences between asthma exacerbations and poor asthma control

Risk factors and costs associated with an asthma attack

Neutrophil degranulation and cell lysis is associated with clinical severity in virus-induced asthma

Viruses and bacteria in acute asthma exacerbations-a GA2LEN-DARE* systematic review

Asthma as a risk factor for invasive pneumococcal disease

Increased frequency of detection of Chlamydophila pneumoniae in asthma

Mycoplasma pneumoniae and asthma in children

The effect of telithromycin in acute exacerbations of asthma

Disordered microbial communities in asthmatic airways

Etiology of asthma exacerbations

Thunderstorm outflows preceding epidemics of asthma during spring and summer

The link between fungi and severe asthma: a summary of the evidence

Exposure to an aeroallergen as a possible precipitating factor in respiratory arrest in young patients with asthma

Asthma exacerbations 2: aetiology

Randomized controlled trial of oral antifungal treatment for severe asthma with fungal sensitization: the fungal asthma sensitization trial (FAST) study

A cross-sectional study of patterns of airway dysfunction, symptoms and morbidity in primary care asthma

Interpretative strategies for lung function tests

Severe asthma in childhood: assessed in 10 year olds in a birth cohort study

Exclusive viral wheeze and allergic wheeze: evidence for discrete phenotypes

Phenotypes of refractory/severe asthma. Paediatr Respir Rev

The effect of airway remodelling on airway hyperresponsiveness in asthma

Airway hyperresponsiveness in asthma: not just a matter of airway inflammation

Is asthma a nervous disease? The Parker B. Francis lectureship

Airway mucosal inflammation even in patients with newly diagnosed asthma

Airway remodelling

Structural and inflammatory changes in COPD: a comparison with asthma

Effect of bronchoconstriction on airway remodeling in asthma

Human sensitisation

The familial incidence of allergic disease

Genetic and environmental influence on asthma: a population-based study of 11,688 Danish twin pairs

The genetics of asthma and allergic disease: a 21st century perspective

Genome-wide association studies for discovery of genes involved in asthma

Association of the ADAM33 gene with asthma and bronchial hyperresponsiveness

PLAUR polymorphisms are associated with asthma, PLAUR levels, and lung function decline

Evidence of a genetic contribution to lung function decline in asthma

Asthma genetics 2006: the long and winding road to gene discovery

Replication of genetic association studies in asthma and related phenotypes

Interleukin-13 in asthma pathogenesis

Elevated expression of messenger ribonucleic acid encoding IL-13 in the bronchial mucosa of atopic and nonatopic subjects with asthma

Interleukin-13: central mediator of allergic asthma

Genetic variants regulating ORMDL3 expression contribute to the risk of childhood asthma

A large-scale, consortium-based genomewide association study of asthma

Genome-wide association study to identify genetic determinants of severe asthma

Resolution of allergic inflammation and airway hyperreactivity is dependent upon disruption of the T1/ST2-IL-33 pathway

House dust mite allergen induces asthma via Toll-like receptor 4 triggering of airway structural cells

Elevated soluble ST2 protein levels in sera of patients with asthma with an acute exacerbation

Inhaled corticosteroids in lung diseases

Intrinsic asthma

Cluster analysis and clinical asthma phenotypes

Identifying adult asthma phenotypes using a clustering approach

Identification of asthma phenotypes using cluster analysis in the severe asthma research program

Inflammatory phenotypes underlying uncontrolled childhood asthma despite inhaled corticosteroid treatment: rationale and design of the PACMAN2 study

The immunopathology of extrinsic (atopic) and intrinsic (non-atopic) asthma: more similarities than differences

Oral montelukast, inhaled beclomethasone, and placebo for chronic asthma. A randomized, controlled trial. Montelukast/Beclomethasone Study Group

Measuring pulmonary function in practice

Heterogeneity of therapeutic responses in asthma

The effect of an anti-IgE monoclonal antibody on the early-and late-phase responses to allergen inhalation in asthmatic subjects

Exploring the effects of omalizumab in allergic asthma

Pathological features and inhaled corticosteroid response of eosinophilic and non-eosinophilic asthma

T-helper Type 2-driven inflammation defines major subphenotypes of asthma

Characterization of within-subject responses to fluticasone and montelukast in childhood asthma

Lebrikizumab treatment in adults with asthma

Effects of an interleukin-5 blocking monoclonal antibody on eosinophils, airway hyper-responsiveness, and the late asthmatic response

Mepolizumab and exacerbations of refractory eosinophilic asthma

Exacerbations of asthma: a descriptive study of 425 severe exacerbations. The FACET international study group

Sputum analysis, bronchial hyperresponsiveness, and airway function in asthma: results of a factor analysis

Eosinophilic bronchitis: clinical features, management and pathogenesis

A comparison of the validity of different diagnostic tests in adults with asthma

Clinical control and histopathologic outcome of asthma when using airway hyperresponsiveness as an additional guide to long-term treatment. The AMPUL study group

Asthma exacerbations and sputum eosinophil counts: a randomised controlled trial

Analysis of induced sputum in adults with asthma: identification of subgroup with isolated sputum neutrophilia and poor response to inhaled corticosteroids

Determining asthma treatment by monitoring sputum cell counts: effect on exacerbations

Genetic basis for personalized medicine in asthma

Mutations in the gene encoding for the beta 2-adrenergic receptor in normal and asthmatic subjects

Polymorphisms of the beta2-adrenergic receptor correlated to nocturnal asthma and the response of terbutaline nebulizer

Adrenergic beta(2)-receptor genotype predisposes to exacerbations in steroid-treated asthmatic patients taking frequent albuterol or salmeterol

Effect of ADRB2 polymorphisms on response to longacting beta2-agonist therapy: a pharmacogenetic analysis of two randomised studies

Common ADRB2 haplotypes derived from 26 polymorphic sites direct beta2-adrenergic receptor expression and regulation phenotypes

ARG1 is a novel bronchodilator response gene: screening and replication in four asthma cohorts

Arginase 1 and arginase 2 variations associate with asthma, asthma severity and beta2 agonist and steroid response

Regulatory haplotypes in ARG1 are associated with altered bronchodilator response

Genetic variants of GSNOR and ADRB2 influence response to albuterol in African-American children with severe asthma

GSNO reductase and beta2-adrenergic receptor gene-gene interaction: bronchodilator responsiveness to albuterol

Endogenous nitrogen oxides and bronchodilator S-nitrosothiols in human airways

Regulation of beta-adrenergic receptor signaling by S-nitrosylation of G-protein-coupled receptor kinase 2

A genome-wide association study of bronchodilator response in asthmatics

Tailored second-line therapy in asthmatic children with the Arg(16) genotype

Leukotriene pathway genetics and pharmacogenetics in allergy

Absorption of montelukast is transporter mediated: a common variant of OATP2B1 is associated with reduced plasma concentrations and poor response

SLCO2B1 c.935G>A single nucleotide polymorphism has no effect on the pharmacokinetics of montelukast and aliskiren. Pharmacogenet

Pharmacogenetic association between ALOX5 promoter genotype and the response to anti-asthma treatment

The glucocorticoid receptor heterocomplex gene STIP1 is associated with improved lung function in asthmatic subjects treated with inhaled corticosteroids

Influence of leukotriene pathway polymorphisms on response to montelukast in asthma

Pharmacogenetics of the 5-lipoxygenase biosynthetic pathway and variable clinical response to montelukast

Pharmacogenetics of asthma controller treatment

Variant LTC(4) synthase allele modifies cysteinyl leukotriene synthesis in eosinophils and predicts clinical response to zafirlukast

Asthma pharmacogenetic study using finite mixture models to handle drugresponse heterogeneity

Effect of citrus juice and SLCO2B1 genotype on the pharmacokinetics of montelukast

Treatment heterogeneity in asthma: genetics of response to leukotriene modifiers

Corticosteroid pharmacogenetics: association of sequence variants in CRHR1 with improved lung function in asthmatics treated with inhaled corticosteroids

Pharmacogenetic study of the effects of NK2R G231E G>A and TBX21 H33Q C>G polymorphisms on asthma control with inhaled corticosteroid treatment

Dualspecificity phosphatase 1 as a pharmacogenetic modifier of inhaled steroid response among asthmatic patients

FCER2: a pharmacogenetic basis for severe exacerbations in children with asthma

FCER2 T2206C variant associated with chronic symptoms and exacerbations in steroidtreated asthmatic children

Genomewide association between GLCCI1 and response to glucocorticoid therapy in asthma

Genome-wide association identifies the T gene as a novel asthma pharmacogenetic locus

Metabolism of beclomethasone dipropionate by cytochrome P450 3A enzymes

G894T polymorphism of eNOS gene is a predictor of response to combination of inhaled corticosteroids with long-lasting beta2-agonists in asthmatic children

Pharmacogenetics of asthma

Ancestry, ancestry-informative markers, asthma, and the quest for personalized medicine

A phase II placebo-controlled study of tralokinumab in moderate-to-severe asthma

IL-4 receptor alpha polymorphisms are predictors of a pharmacogenetic response to a novel IL-4/IL-13 antagonist

Effect of an interleukin-4 variant on late phase asthmatic response to allergen challenge in asthmatic patients: results of two phase 2a studies

V75R576 IL-4 receptor alpha is associated with allergic asthma and enhanced IL-4 receptor function

Research in the authors' laboratory is funded by Asthma UK, the Medical Research Council, and the British Medical Association. We thank Dr Emily Hodge for originally generating Figure 28 .2.