Severe asthma phenotypes

INTRODUCTION — Asthma is defined and diagnosed through a combination of clinical symptoms and physiologic abnormalities, generally without reliance upon pathologic or biologic markers. However, the physiologic definition of asthma is relatively nonspecific, consisting of airway hyperreactivity and airflow limitation during expiration, which is variable and/or reversible with bronchodilators. In most asthma patients, the presence of bronchial hyperreactivity is never objectively confirmed.

The data suggesting that multiple phenotypes exist within the classification of "severe asthma" are reviewed here. Details regarding the classification, evaluation, diagnosis, and treatment of severe asthma are provided separately. (See "An overview of asthma management" and "Evaluation of severe asthma in adolescents and adults" and "Asthma in adolescents and adults: Evaluation and diagnosis" and "Treatment of severe asthma in adolescents and adults" and "Mechanisms and clinical implications of glucocorticoid resistance in asthma".)

DEFINITION OF SEVERE ASTHMA — The diagnosis of asthma is based upon the presence or history of symptoms consistent with asthma (most commonly episodic cough, wheezing, or dyspnea) provoked by typical triggers, combined with the demonstration of variable expiratory airflow obstruction. After confirming a diagnosis of asthma and addressing comorbidities, severe asthma is that which requires treatment with high dose inhaled glucocorticoids (table 1) plus a second controller (eg, long-acting beta-agonist, leukotriene modifier, theophylline) and/or systemic glucocorticoids for 50 percent or more of the year to prevent asthma from becoming “uncontrolled” or that which remains “uncontrolled” despite this therapy (table 2) [1,2]. Other conditions must have been evaluated and excluded, and potential exacerbating factors remediated.

RATIONALE FOR CHARACTERIZING ASTHMA PHENOTYPES — Patients with severe asthma present with a variety of clinical histories, physiologic changes (beyond changes in forced expiratory volume in one second [FEV₁]), and airway inflammation, suggesting that severe asthma is not a single disease or is a single process that produces widely varying host responses. Patients with severe asthma may have manifested severe disease all their life, developed progressively more severe disease over time, or may never have had asthma (to the best of their recollection) until some point in their adult life, after which the disease progressed at a rapid pace [3]. Some patients with severe asthma have a clearly defined atopic history, while others give little indication of an allergic component.

In addition to the variability in clinical features, patients with asthma do not respond in a uniform fashion to asthma medications, particularly glucocorticoids and other nonspecific anti-inflammatory/cytotoxic medications. Furthermore, it is increasingly clear that molecular pathways can have different, clinically recognizable phenotypes [4-6]. It is hoped that exploration of asthma phenotypes will translate into improved understanding of asthma pathophysiology and optimized medication selection [1,7].

SEVERE ASTHMA PHENOTYPES — Among patients who meet the definition of severe asthma (table 2), a variety of clinical presentations, physiological characteristics, and responses to therapy are evident [1,3,8,9]. Disease characteristics (phenotypes) that are most commonly used to differentiate patients with severe asthma are described in this section. However, substantial overlap exists; as an example, adult onset asthma does not exclude atopy, although it makes it less likely.

In order to characterize the various presentations of severe asthma, two statistical cluster analyses were performed on patients participating in the Severe Asthma Research Program (SARP), including those with mild, moderate, and severe asthma [3]. The first cluster analysis only utilized clinical and physiologic data [3], while the second analysis also included bronchoscopic and inflammatory variables [10]. While similar, some differences were noted. Both cluster analyses identified childhood onset, generally allergic asthma from mild to severe, and very severe asthma with a mixed inflammatory process. The inflammatory clusters also identified a very distinct group with adult onset disease and nasal polyposis. (See 'Hypereosinophilic adult onset asthma and aspirin exacerbated respiratory disease' below.)

The following sections describe characteristics that have been identified (with greater or lesser frequency) among patients with severe asthma and potential ramifications in terms of treatment choices.

Childhood onset allergic asthma — Data from case series of patients with severe asthma suggest that childhood-onset (before the age of 12) versus later onset (after age 12) distinguishes two distinct subgroups [3,11-13]. Childhood-onset asthma represents a relatively homogeneous group of patients, often with a strong allergic history and family history of asthma. In contrast, adult-onset asthmatics are a very mixed group of patients.

Adult onset atopic asthma — Among patients with adult onset asthma, those with severe disease are less likely to be atopic (34 percent) than those with mild-to-moderate persistent asthma (52 percent) [14]. While only applicable to a minority of severe asthma patients, atopic/allergic asthma can arise in adulthood, often in patients with a history of allergic rhinitis in childhood. Identifying the presence of atopy has implications for selection of therapy and potentially for environmental modifications (eg, dust mite mitigation). (See "Trigger control to enhance asthma management".)

As an example, treatment with the monoclonal antibody to IgE (omalizumab) is indicated for patients with severe asthma, elevated IgE levels (eg, 35 to 700 international units/mL), and positive allergy skin tests to perennial allergens. Treatment with omalizumab can reduce the frequency of exacerbations and reduces the inhaled glucocorticoid requirement [15,16]. As improvements are limited primarily to a reduction in exacerbations of approximately 30 percent, even in those with allergic asthma, a retrospective analysis of responders to omalizumab using Type-2 biomarkers (eg, fraction of exhaled nitric oxide [FeNO], blood eosinophils) was undertaken. Elevations over the median in any of these Type-2 biomarkers was a better predictor of response than serum IgE or specific IgE testing [17,18]. (See "Anti-IgE therapy".)

Adult onset nonatopic asthma — Many patients with adult onset asthma describe the onset of asthma symptoms following a viral illness, occupational exposure, or the ingestion of aspirin, while such histories are less frequent among those with childhood-onset asthma. In adult-onset asthma, an allergic component may be more difficult to confirm, as significantly fewer asthmatics with onset of disease after the age of 12 demonstrate positive allergy skin tests or specific allergic symptoms compared to early onset [19,20].

Hypereosinophilic adult onset asthma and aspirin exacerbated respiratory disease — A subset of patients with adult onset eosinophilic asthma require systemic glucocorticoids to maintain control of their asthma and often early in the course of disease [10-12,14]. They are less likely to be atopic/allergic but have high levels of blood and tissue eosinophils, as well as high levels of exhaled nitric oxide. These findings are often seen in association with severe sinus disease, nasal polyposis, and in a minority, aspirin exacerbated respiratory disease (AERD).

AERD refers to the combination of asthma, chronic rhinosinusitis (CRS) with nasal polyposis, and acute upper and lower respiratory tract reactions to ingestion of aspirin (acetylsalicylic acid, ASA) and other cyclooxygenase-1 (COX-1) inhibiting nonsteroidal antiinflammatory drugs (NSAIDs). (See "Aspirin-exacerbated respiratory disease".)

AERD is good example of the complex overlap between the various phenotypes of severe asthma, as illustrated by the following observations. Peripheral blood eosinophilia occurs in approximately 50 percent of patients with AERD, although not all patients with eosinophilic asthma have AERD. Depending on the case series, between 30 and 70 percent of patients with AERD are atopic.

Patients with AERD have a baseline dysregulation of arachidonic acid metabolism with over-production of leukotrienes C4, D4, and E4, which are potent bronchoconstrictors and promoters of airway inflammation. In the presence of COX-1 inhibitors such as aspirin (ASA) and NSAIDs, COX-1 inhibition leads to a further increase in production of leukotrienes C4, D4, and E4. Leukotriene modifying agents (eg, montelukast, zafirlukast, zileuton) are used to address the underlying dysregulation of leukotriene production and also protect patients from severe exacerbations due to accidental NSAID exposure. (See "Aspirin-exacerbated respiratory disease", section on 'Abnormalities in AERD'.)

Aspirin desensitization and daily aspirin therapy can be helpful in patients with AERD (with primary benefits in the upper airway), although it is not used in all patients, often for safety concerns. (See "Aspirin-exacerbated respiratory disease: NSAID challenge and desensitization".)

Agents targeting Type 2 inflammation may also have a role in treating nasal polyposis and AERD. Dupilumab, mepolizumab, and omalizumab are all approved by the US Food and Drug Administration (FDA) for the treatment of nasal polyps [21-24]. Similar findings have been observed with both anti-IL-5 and anti-IgE agents in this population [22,23], although some data suggest that anti-IL5R approaches have the most marked benefit on asthma outcomes in patients with concurrent asthma and nasal polyps [24,25]. (See "Aspirin-exacerbated respiratory disease", section on 'Biologic agents' and "Chronic rhinosinusitis with nasal polyposis: Management and prognosis", section on 'Biologic therapies'.)

Not all patients with adult onset hypereosinophilic asthma have nasal polyps or AERD. However, they are often refractory to inhaled therapy alone. Studies with anti-IL-5 monoclonal antibodies suggest that patients with hypereosinophilic adult onset asthma may respond favorably to anti-IL-5 directed therapy, including an improved ability to taper oral glucocorticoids [26-28]. (See "Treatment of severe asthma in adolescents and adults", section on 'Anti-IL-5 therapy'.)

Asthmatic granulomatosis — Some patients with severe asthma have granulomatous inflammation on lung biopsy. It is not known whether this inflammation is an intercurrent disease process, a consequence of therapy, or different asthma phenotype. In a case series that included 19 patients with severe asthma who underwent video-assisted lung biopsy, 10 were found to have interstitial non-necrotizing granulomas with asthma-like submucosal inflammation and mucus plugging in the small airways, and without evidence of hypersensitivity pneumonitis [29]. Peripheral blood eosinophilia was present in most, and the fraction of exhaled nitric oxide (FE_NO) was elevated. Atopy was present in 6 of 10. Computed tomography was abnormal in some patients, and a large majority had a family history of autoimmune disease. (See "Evaluation of severe asthma in adolescents and adults", section on 'Assessing conditions that mimic asthma'.)

Rheumatoid arthritis — Emerging epidemiologic data suggest that the autoimmune disease rheumatoid arthritis may be associated with asthma, but the association with severity of disease is as yet unknown [30,31].

CLINICAL CHARACTERISTICS ASSOCIATED WITH SEVERE ASTHMA — Certain clinical characteristics have been identified that are associated with severe asthma but do not define a particular phenotype.

Glucocorticoid resistance — By definition, most severe asthmatics receive oral or high doses of inhaled glucocorticoids (GC). Thus, most severe asthma has an element of GC resistance. However, even with these very potent anti-inflammatory drugs, the response to therapy is often variable [32]. In some patients with severe asthma, the response appears to be shifted, so that although a response occurs, it requires extremely high doses. These patients would appear to be "glucocorticoid-dependent." In others, even high doses of glucocorticoids may lead to little or no response, or even worsening. These patients may be truly glucocorticoid-resistant, but are likely in a minority. (See "Mechanisms and clinical implications of glucocorticoid resistance in asthma".)

Many, if not most studies of glucocorticoid resistance have been done in moderate rather than severe asthmatics, and extrapolation from those results is often difficult [33,34]. Some severe asthmatics may not demonstrate an improvement with glucocorticoids on a chronic basis, but may demonstrate improvement with glucocorticoids during a period of exacerbation. This difference in the acute versus chronic response has not been studied to date. In addition, patients with persistent sputum eosinophilia appear to respond better to high dose glucocorticoid therapy than those with no evidence for eosinophilia [35]. Thus, GC resistance is a rather vague term. Generally, this term has been replaced with more specific phenotypes below.

Exacerbation frequency — Some patients with severe asthma have frequent episodes of extreme airflow limitation (brittle asthma), while others do not [36]. No consistent definition of "frequent exacerbation" has been agreed upon, but more than two or three exacerbations during a year is commonly used [1]. Approximately 30 percent of patients with severe asthma meet this definition of frequent exacerbations [37]. Patients with frequent exacerbations tend to have greater peripheral airway obstruction on pulmonary function tests, persistent eosinophilia in blood and bronchoalveolar lavage, and more comorbidities (eg, chronic rhinosinusitis, recurrent respiratory infections, obesity, obstructive sleep apnea, and psychosocial issues) [10,38,39].

Among patients enrolled in the Severe Asthma Research Program (SARP), frequent exacerbations were associated with persistent eosinophilic inflammation in peripheral blood despite high doses of systemic glucocorticoids as well as high plasma interleukin (IL)-6 levels [10,40]. However, similar to GC resistant asthma, the clinical characteristics of frequent exacerbations may be due to multiple different contributors, like smoking, obesity, and psychologic factors. Therefore, this clinical phenotype has generally been replaced with the phenotypes below.

Asthma associated with obesity — Several studies have identified an increased prevalence of asthma among obese individuals compared with those of normal weight, although the exact reason for the association is not known. In a Dutch cohort, the prevalence of obesity (body mass index ≥30 kg/m²) among a cohort with severe asthma was 21 percent [41]; in a similar cohort in the United Kingdom, the prevalence of obesity was 48 percent [42]. The risk of obesity-associated asthma is higher for women than men and among nonatopic individuals than atopic. Asthma associated with obesity is often more difficult to control and less likely to respond to traditional asthma therapy. (See "Obesity and asthma".)

While obesity can be present in association with any asthma phenotype, there appears to be a particular phenotype of obese asthma associated with metabolic syndrome and elevated blood IL-6 levels [43]. In one study of patients with severe asthma, those with insulin resistance demonstrated more rapid decline in lung function and increased resistance to beta-agonist and oral glucocorticoid therapies compared with patients having normal insulin sensitivity [44]. These metabolic effects were more strongly associated with asthma outcomes than obesity itself. Clinical trials evaluating asthmatic participants with insulin resistance are needed to determine if improving insulin resistance improves these asthma outcomes.

In general, studies of weight loss interventions show improvements in asthma control, asthma-related quality of life, and lung function if a sufficient amount of weight loss (at least 5 percent) is attained [45]. (See "Obesity and asthma", section on 'Weight loss interventions'.)

Perimenopausal onset asthma — Onset of asthma within a year of a woman's last menstrual period may define a separate asthma phenotype [19,20]. In a case series, induced sputum from 40 women with menopause onset asthma showed a higher percentage of neutrophils and slightly lower percentage of eosinophils compared with sputum from 35 women with premenopausal onset asthma [20]. None of the women with menopausal onset asthma were atopic, while 76 percent of the premenopausal asthma onset group were atopic. Whether menopausal onset asthma is different from adult onset asthma in men requires additional study. How much overlap there is with the adult onset less atopic, generally obese women from the cluster analysis of the Severe Asthma Research Program population remains to be determined, but a large percentage of that population was also post-menopausal [3].

PHENOTYPING BASED ON BIOMARKERS OF INFLAMMATION — It is hoped that identifying biomarkers for the various asthma phenotypes will aid the selection of therapies that are most likely to work for individual patients. Considerable progress has been made in this area, with most asthma experts now classifying asthma patients based on the amount of Type 2 (T2) inflammation.

T2 inflammation is mediated by eosinophils, mast cells, basophils, T helper-2 lymphocytes (secrete interleukin [IL]-4, IL-5 and IL-13), group 2 innate lymphoid cells (ILC2s), and IgE-producing B cells [46]. Some patients have evidence of high levels of T2 inflammation (T2-high) and others have little or no evidence for T2 inflammation. This division of asthma phenotypes based on biomarkers is increasingly important, at least in the case of T2 biomarkers, as specific T2-targeted therapies with considerable efficacy have been approved for use in T2-high patients (eg, dupilumab, mepolizumab, reslizumab, benralizumab).

The exact promoters of severe asthma in patients with T2-low inflammation are less well-characterized than the T2-high phenotype, but likely involve neutrophilic, smooth muscle, or metabolic-related processes [43]. Mediators shown to be present in non-type 2 asthma (but not exclusive to non-type 2 asthma), include IL-1beta, IL-6, IL-17, IL-18, interferon gamma, and tumor necrosis factor alpha [43,47-51]. Despite this, the utility of these as reliable and consistent biomarkers remains unclear. The mediator thymic stromal lymphopoietin (TSLP) that is associated with both non-type 2 and type 2 asthma holds promise but is not reported in peripheral blood samples. Inhibition of this airway epithelial cell product, also called an "alarmin," improves type 2 and non-type 2 asthma, although greater benefits are seen in patients with type 2 markers [52-54]. (See "Treatment of severe asthma in adolescents and adults", section on 'Anti-thymic stromal lymphopoietin (tezepelumab)' and "Treatment of severe asthma in adolescents and adults", section on 'Persistently uncontrolled asthma'.)

Type 2/eosinophilic asthma — Approximately 70 percent of severe asthma is associated with persistent elevation in markers of Type 2 inflammation (blood eosinophils and fraction of exhaled nitric oxide [FeNO]) and response to targeted Type 2 biologics, likely arising from residual elevations in the Type 2 cytokines IL-4, 5, and 13. Data in support for this concept arise from the Severe Asthma Research Program (SARP) and other clinical networks, as well as clinical trials of antibodies directed towards these cytokines [26,55-58].

As noted above, in severe asthma, the eosinophilic/Type-2 Hi phenotype appears more common among late onset asthma than childhood onset, which is more typically associated with atopy (see 'Severe asthma phenotypes' above). However, data from clinical trials suggest that blood levels of eosinophils between 200 and 300/microL or FeNO levels above 24 ppb support an underlying active Type 2 immune process, which will respond to Type 2 specific therapy [27,55,59]. The ideal method for assessing Type 2 inflammation is not known [60]. Sputum eosinophilia and measurement of T2 cytokine mRNA in sputum predict airway eosinophilia, but are time consuming tests and observer dependent [61]. Statistical associations have been noted between sputum eosinophilia and peripheral blood eosinophil counts and FeNO, but the accuracy of these biomarkers for predicting sputum eosinophilia is imperfect [62,63].

The European Respiratory Society/American Thoracic Society and the Global Initiative for Asthma (GINA) guidelines suggest that treatment of severe asthma be guided by clinical criteria and biomarkers such as blood eosinophil levels or FeNO, rather than by clinical criteria alone [2,64].

Peripheral blood eosinophils appear to be good predictors of response to Type-2 targeted therapy as shown in studies of the monoclonal antibodies to IL-5 (ie, mepolizumab and reslizumab) and also anti-IL-4RA [27,28,65,66]. However, it is likely that additional Type-2 biomarkers will be identified, which subdefine Type-2 asthma even further, and which will eventually better guide therapy than the current blood eosinophils and FeNO. (See "Treatment of severe asthma in adolescents and adults", section on 'Anti-IL-5 therapy' and "Treatment of severe asthma in adolescents and adults", section on 'Anti-lL-4 receptor alpha subunit antibody (dupilumab)'.)

Fraction of exhaled nitric oxide — Measurement of FeNO has been used as a biomarker for Type-2 inflammation. Increased FeNO levels (eg, greater than 50 ppb in adults or greater than 35 ppb in children) correlate with eosinophilic airway inflammation and responsiveness to inhaled glucocorticoids. (See "Exhaled nitric oxide analysis and applications", section on 'Clinical use of FENO in asthma'.)

In a study of bronchial brushings and sputum obtained from subjects with severe asthma, mild/moderate asthma (on or off inhaled glucocorticoids), and healthy controls, FeNO correlated with sputum eosinophils, while expression of inducible nitric oxide synthetase (iNOS), the enzyme that generates exhaled NO, differentiated severe asthma from the other groups better than FeNO, arginase2 mRNA/protein, or nitrotyrosine (NT) protein [67]. In a separate study from SARP, FeNO was a better predictor of systemic glucocorticoid use among patients with severe asthma than forced expiratory volume in one second (FEV₁) [68]. FeNO is also consistently decreased in response to IL-4/IL-13 targeted therapies suggesting that at least a portion of the FeNO is driven by these cytokines. In patients with moderate-to-severe asthma, the addition of FeNO levels to blood eosinophil counts improved prediction of future exacerbations as well as response to dupilumab therapy [59,69].

However, a number of other factors have been identified that affect FeNO values, such as age, sex, atopy, and cigarette smoking, making it an imperfect biomarker for guiding therapeutic decisions. Thus, the European Respiratory Society/American Thoracic Society guidelines [2], the National Asthma Education and Prevention Program (NAEPP) guidelines [70], and the Global Initiative for Asthma (GINA) guidelines [64] all suggest that additional research is needed to clarify when and how FeNO should be used (eg, in addition to blood eosinophils) to guide diagnosis and treatment decisions in severe asthma.

Neutrophilic asthma — The existence of a neutrophilic asthma phenotype (eg, 40 to 60 percent neutrophils in induced sputum) is controversial [71-73]. The specificity of neutrophilic inflammation for a particular subtype of asthma is complicated by the many confounding factors that can contribute to neutrophilia in sputum, including the use of inhaled glucocorticoids, air pollution, respiratory infection, sensitization to aspergillus, and gastroesophageal disease. In one study, sputum neutrophilia was associated with pre and post-bronchodilator FEV₁, suggesting that neutrophilic airway inflammation may have a role in persistent airflow limitation in asthma [74].

Analysis of patients in SARP identified two subgroups with moderate-to-severe asthma and frequent health care use despite treatment with high doses of inhaled or oral glucocorticoids; one of these subgroups also had reduced lung function [3,10,75]. The majority (>83 percent) of those with reduced lung function had sputum neutrophilia alone or in combination with sputum eosinophilia. The combination of neutrophilia and eosinophilia in sputum appears to identify a more severe phenotype [75]. (See 'Rationale for characterizing asthma phenotypes' above.)

Several long-term cohort studies have demonstrated that elevated blood neutrophilia is also associated with an increased risk of asthma exacerbations [76-78]. In one cohort, asthma patients with blood neutrophil counts >4400 cells/microL demonstrated higher inhaled corticosteroid and antibiotic use, increased BMI (body mass index), elevated C-reactive protein, and lack of lung function decline compared with patients without neutrophilia [78]. As with sputum neutrophilia, it is unclear whether these characteristics are a cause or consequence of heavy corticosteroid use.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Severe asthma in adolescents and adults".)

SUMMARY

●A consensus definition of severe asthma requires that patients have needed therapy with high dose inhaled glucocorticoids (table 1) and a long-acting beta-agonist (LABA) or leukotriene modifier/theophylline for the previous year and/or have needed systemic glucocorticoids for 50 percent or more of year to prevent asthma from becoming uncontrolled; other conditions must have been excluded, and exacerbating factors treated (table 2). (See 'Definition of severe asthma' above.)

●Severe asthma represents a heterogeneous population. This heterogeneity is obvious to clinicians who treat asthma routinely and notice that responses to therapy vary among patients. (See 'Severe asthma phenotypes' above.)

●Characteristics that appear to differentiate subgroups or "phenotypes" of severe asthma include age, sex, age of asthma onset, atopic status, obesity, exacerbation frequency, eosinophilia, aspirin exacerbated respiratory disease, and glucocorticoid resistance, although substantial overlap exists among these. (See 'Severe asthma phenotypes' above.)

●A better understanding of asthma phenotypes has the potential to lead to improved treatment of this group of patients. As examples, patients with atopic severe asthma may appear more likely to benefit from anti-IgE therapy (omalizumab) and those with adult onset, eosinophilic disease with or without aspirin exacerbated respiratory disease may benefit from therapy with anti-interleukin (IL)-5/IL-5R antibodies or IL-4RA antibodies as well as leukotriene modifiers and aspirin desensitization. (See 'Severe asthma phenotypes' above.)

●It is hoped that identifying biomarkers for asthma phenotypes will improve the selection of therapies that are most likely to work for individual patients. Ideally, biomarkers would be reproducible, accurate, and accessible noninvasively. International guidelines suggest that treatment of severe asthma be guided by clinical criteria and peripheral blood or sputum eosinophil counts (performed in centers experienced in this technique) rather than by clinical criteria alone. (See 'Phenotyping based on biomarkers of inflammation' above.)

●Fraction of exhaled nitric oxide (FeNO) levels, in combination with blood eosinophil counts, may help guide treatment decisions, but additional research is needed to clarify when and how FeNO should be used. (See 'Fraction of exhaled nitric oxide' above.)