Mechanisms and clinical implications of glucocorticoid resistance in asthma

INTRODUCTION — Glucocorticoids (GCs) have potent antiinflammatory actions and are the most effective anti-inflammatory agents in the treatment of asthma. It is evident that asthma is a syndrome with many distinct and overlapping phenotypes with severe GC-refractory asthma at one end of a spectrum of GC responsiveness [1]. Patients with GC-refractory asthma account for a large percentage of the overall costs for asthma worldwide, but also provide unique insights into the mechanisms of GC action. These patients should not be confused with those who either do not take their anti-inflammatory medication or do not have access to the correct treatments [2,3].

The European Respiratory Society/American Thoracic Society (ERS/ATS) and the Global Initiative for Asthma (GINA) guidelines provide pragmatic approaches to treatment strategy for severe asthma based on patient stratification and expert opinion [3,4].

The basic mechanisms of glucocorticoid resistance in asthma and clinical implications for diagnosis and management of severe asthma will be reviewed here. GC use in GC-sensitive asthma and GC resistance in other disease states are reviewed separately. (See "An overview of asthma management".)

DEFINITION — Glucocorticoid (GC) responsiveness represents a continuous spectrum, with GC-resistant individuals falling at one end of a unimodal distribution. Patients with severe asthma who are poorly responsive to high doses of GCs and are without confounding factors (table 1) have been termed GC-resistant [4]. A larger subset of patients with asthma that is poorly-controlled despite optimal treatment or who experience worsening of asthma control during GC withdrawal have severe asthma and are considered relatively GC-insensitive [4]. High doses of GCs usually indicate a daily dose of 1000 mcg or more of inhaled fluticasone propionate or 2000 mcg or more of triamcinolone, the equivalent (table 2).

MECHANISMS OF GLUCOCORTICOID ACTION — GCs exert their major effects by influencing target cell transcriptional regulation and protein synthesis (figure 1).

●Upon entering the cell, GCs bind to a cytoplasmic glucocorticoid receptor (GR or NR3C1), which is a member of the nuclear receptor superfamily, to form a GC-GR complex.

●GR is expressed in all cells in the airway to a varying extent [5].

●GR can exist in several isoforms including GRalpha, the major form, as well as GRbeta which acts as a dominant negative isoform of GRalpha [6].

●The GC-GR complex then undergoes a conformational change revealing a nuclear localization signal and translocates into the nucleus. Within the nucleus, the activated GR binds to cis-acting DNA sequences (glucocorticoid-response elements [GREs]) and generally upregulates the transcription of specific target genes (figure 1) [6].

●Binding of the activated GR-ligand complex to DNA can also occur at a negative glucocorticoid-response element (nGRE). These sequences are implicated in the control of over 1000 genes in man and can result in repression of transcription [7]. The induction or repression of target genes ultimately results in the altered expression of GC-regulated proteins, such as increased transcription of antiinflammatory genes such as the cytokine interleukin (IL)-10 [6].

●GC-induced gene transcription is associated with changes in chromatin structure as evidenced by enhanced acetylation of core histones around which DNA is wound.

●Administration of glucocorticoids, systemically or through inhalation, results in a significant increase in systemic or local IL-10 synthesis, respectively. In vitro studies have shown a concentration-dependent induction of T lymphocyte IL-10 expression by glucocorticoids [8]. IL-10 inhibits proinflammatory cytokine production, antigen presentation, T cell activation, and also mast cell and eosinophil function [8].

●The glucocorticoid receptor-ligand complex can also modulate transcription of genes in a hormone-dependent manner through tethering to other DNA-bound transcription factors eg, cyclic adenosine monophosphate (AMP) response element binding protein (CREB), activation protein (AP)-1, nuclear factor kappa B (NF-kappaB), and signal transducers and activators of transcription (STATs) [6,9], rather than by direct binding to GREs. Activity of these other transcription factors is largely regulated by extracellular signals received by cell surface receptors.

●Activated GRs (ie, the GC-GR complex) can recruit histone deacetylase (HDAC) complexes to sites of NF-kappaB-driven proinflammatory gene transcription and thereby decrease inflammatory gene transcription [10].

●The ability of activated GRs to associate with DNA and co-repressor/activator proteins is regulated by its phosphorylation status [11].

●These molecular actions of GCs are sufficient in most asthmatics to repress airway inflammation, reduce airway hyperresponsiveness and improve lung function. However, they are not a cure for asthma as symptoms return following withdrawal.

MECHANISMS OF GLUCOCORTICOID RESISTANCE — Glucocorticoid (GC) resistance is probably produced by the summative effect of a number of heterogeneous mechanisms (figure 2). As the preceding section implies, modulation of GC activity theoretically could occur through changes in binding kinetics of the cytosolic or nuclear receptors, or through alterations in the interactions with other transcription factors. Defects in most aspects of intracellular glucocorticoid receptor (GR) activity, such as GR expression levels, binding of GCs to GR and nuclear translocation of GC-GR complex, have been implicated as being defective in patients with severe asthma. However, is it likely that the major mechanisms for GC resistance occur distal to the nuclear translocation step, such as in splicing of RNA, cytokine expression, binding to GC response elements on DNA, and interactions of the GC-GR complex with other nucleoproteins. Thus far, GC resistance in asthma cannot be explained by malabsorption or pharmacokinetic mechanisms.

Genetic determinants — There are few data addressing whether GC insensitivity is genetically determined although genome-wide association studies indicate other genetic links to severe asthma (eg, ORMDL3, GSDMB, interleukin [IL]-1RL1) [12]. Given the marked variability in the dose of glucocorticoids needed to inhibit T-cell proliferative responses in normal individuals, it is possible that there are genetic determinants of glucocorticoid responsiveness that only become manifest clinically with the development of a disease process requiring GC therapy. A functional polymorphism in glucocorticoid-induced transcript 1 (GLCCI1) has been linked to extremes of GC responsiveness in a genome wide study [13]. Additionally, there may be differences between racial groups as a GR D641V variant is associated with GC-resistant asthma in the Chinese Han population [14]. Finally, the 3-beta-hydroxysteroid dehydronease-1 (HSD3B1; 1245) allele regulates the adrenal production of potent androgens and is associated with GC resistance in the Severe Asthma Research Program (SARP) cohort [15].

Altered splicing of GC receptor premessenger RNA — Alternative splicing of GR RNA can produce C-terminal isoforms termed GC receptor-alpha (GRalpha) and GC receptor-beta (GRbeta) along with several N-terminal truncated isoforms of both GRalpha and GRbeta [16]. These isoforms may contribute to GC-resistance in severe asthma by attenuating the ability of GRalpha to translocate to the nucleus and modulate gene expression. However, conflicting data exist. The expression of GRbeta is significantly increased in some patients with GC-resistant asthma and may contribute to GC resistance [17,18]. Autopsies of patients who suffered fatal asthma have revealed increased expression of GRbeta, although it is uncertain if this is due to GC treatment or if GRbeta causes severe asthma [19]. (See "Severe asthma phenotypes".)

GRbeta expression is enhanced by bacterial superantigens [20] which may account for the relative GC refractory nature of asthma exacerbations. This refractoriness is also seen in viral-induced exacerbations [21,22] where the mechanism may involve inhibition of GR DNA binding [23].

Altered cytokine expression — Peripheral blood mononuclear cells (PBMCs), particularly the CD4+ T lymphocytes and monocytes, and airway structural cells exhibit a defective GC response in GC-resistant asthma. Airway cells from patients with GC-sensitive compared with GC-resistant asthma have different patterns of cytokine gene expression and distinct responses to GC therapy.

●IL-2 and IL-4 – Alterations in the cytokine milieu may strongly influence responsiveness to GCs. For example, the number of bronchoalveolar lavage (BAL) cells expressing IL-2 and IL-4 messenger RNA was greater in GC-resistant asthmatics compared to GC-sensitive patients [24]. PBMCs collected from normal subjects exposed to a combination of recombinant IL-2 and IL-4 demonstrate dose-dependent GC resistance [25].

IL-2 and IL-4 activate a specific kinase (p38 mitogen activated protein kinase [MAPK]) whose expression is enhanced in GC-resistant asthma. Inhibitors of p38 MAPK are able to restore GC sensitivity in peripheral blood cells, BAL macrophages, and airway smooth muscle cells from patients with severe asthma [26,27].

●IL-10 – Synthesis of IL-10 is deficient in the airways of asthmatics as compared with controls, and polymorphisms of the IL-10 gene, leading to impaired expression of IL-10, are associated with a more severe disease phenotype [28]. CD4+ T lymphocytes from GC-resistant asthmatics have a marked deficiency in the capacity to synthesize IL-10 following in vitro stimulation in the presence of dexamethasone with IL-10 levels being restored by vitamin D3 treatment of cells from both adults and children [29,30].

●Additional T-cell and cytokine profiles in glucocorticoid-resistant asthma – Several additional subsets of patients with GC-refractory asthma have been characterized based on cytokine or T-cell responses. One of these subsets is notable for elevated levels of Th17 cytokines, such as IL-17 [31,32], which are not reduced by glucocorticoids and can drive glucocorticoid insensitivity in animals and human airway epithelial cells [33]. Vitamin D3, however, is able to attenuate IL-17 production in cell culture models, reducing inflammation and enhancing steroid sensitivity [34]. The clinical implications of these findings have not been determined.

Another novel glucocorticoid-insensitive cytokine, IL-33, has also been implicated in airway remodeling in children with severe treatment refractory asthma [35]. One clinical trial of an IL-33 inhibitor and basic research into IL-33 activation pathways suggest particular utility of targeting this pathway in patients with severe GC-resistant non-T2 asthma [36-39].

Another commonly seen cytokine pattern is heightened interferon gamma activity. Reduced Th2 and IL-17 responses and enhanced interferon (IFN)-gamma expression were reported in the BAL of individuals with GC-resistant asthma [40] as well as increased expression of IFN-stimulated genes (ISG) in epithelial brushings [41]. The effects of IFN-gamma-related activation on airway hyperresponsiveness are modulated by CCR5 [42].

In one cohort, elevated plasma IL-6 levels were associated with worse lung function and more frequent exacerbations in both obese and nonobese subjects [43]. Activation of IL-6 on epithelial cells via trans-signaling leads to a transcriptional program associated with frequent exacerbations, blood eosinophilia, and submucosal infiltration of T cells and macrophages in the absence of type 2 inflammation [44].

Finally, reduced numbers of BAL Treg and mucosal associated invariant T (MAIT) cells are found in severe asthma [45]. (See "The adaptive cellular immune response: T cells and cytokines", section on 'Cytokine profiles and functions of CD4+ T helper cell subsets'.)

These and other studies highlight the importance of defining the airway compartment being studied in severe asthma. Single cell RNA-sequencing analysis of immune cells in BAL and/or biopsies from these patients will provide better insight into the contribution of these pathways and T-cell subtypes to GC-insensitive asthma.

GR binding to GC response elements — Early studies indicated that the activated GR in peripheral blood cells of GC-insensitive asthmatics were less able to bind to DNA than cells from patients with GC-sensitive asthma and non-atopic control subjects (figure 3) [46]. It is unlikely that this defect lies at the level of GR polymorphisms, and probably reflects differences in GR post-translational modifications and association with co-factors and chromatin

[47]

Interactions with other nucleoproteins and epigenetics — Interaction of the nuclear GR-ligand complex with other nucleoproteins modulates GC activity and may be responsible for the DNA binding abnormalities noted above. In one study, the interaction between the GC receptor and the pro-inflammatory transcription factor activator protein (AP)-1 was significantly reduced in GC-resistant patients, although interaction with NF-kappaB was unaffected [48]. An increase in the basal levels of AP-1 DNA binding was also detected in the nuclei from patients with GC-resistant asthma [9]. In addition, a bioinformatic analysis of regulatory networks from cells from asthmatics who were either good or poor responders to ICS, identified AP-1 as a key hub gene for poor GC response [49].

Epigenetic changes also appear to contribute to GC sensitivity. Reduced HDAC activity and increased histone acetylation have been reported in bronchial biopsies of children with severe GC-resistant asthma [50] and in peripheral blood cells of affected adults [51]. This decrease in HDAC activity may be a result of increased IL-17 since IL-17 reduces GC function in primary human epithelial cells by reducing HDAC [33]. Differences in DNA methylation profiles in both airway smooth muscle cells and airway epithelial cells have also been reported to be liked to asthma severity using weighted gene coexpression network analysis [52,53].

Expression quantitative trait loci (eQTL) analysis has been performed in epithelial and BAL cells of severe asthmatics to explore the genetic influence on gene expression [54,55]. Many of these eQTLs reside in regulatory elements of genes enriched in the Th2 lymphocyte and IL-33 immune responses and of MUC5AC, particularly in airway epithelial cells [56,57]. Interestingly, IL-13 resets the DNA methylome in human airway epithelial cells to modulate asthma severity–associated pathways [52].

A small study looking at peripheral blood transcriptomic profiling of 13 children with GC-resistant asthma identified genes such as RAR-Related Orphan Receptor alpha (RORA) which were known asthma genes by genome wide association studies [58]. This study also highlighted the importance of decreased GR signaling and increased MAPK and Jun N-terminal kinase (JNK) activity in these children with GC-resistant asthma. Decreased GR expression along with impaired nuclear translocation is associated with attenuated GC-responsiveness in airway smooth muscle [59] and peripheral blood monocytes from patients with GC-resistant asthma [60]. In blood cells this has been linked to reduced histone acetylation at the MAPK phosphatase-1 (MKP-1) promoter and MKP-1 induction by dexamethasone [60]. This in turn may result in the reduced ability to attenuate mitogen-induced phospho-p38 MAPK activity which has been reported in peripheral blood CD14+ monocytes, but not CD4+ or CD8+ T lymphocytes, from GC-resistant asthmatics [61]. (See "Genetics of asthma".)

Analysis of transcriptomic profiles in bronchial epithelial cells and brushings indicated a group of patients with high levels of tissue eosinophilia and Th2 cytokines who also expressed high levels of steroid-sensitive genes despite being treated with high doses of inhaled and oral GCs [62]. Specific pathways associated with airway remodelling such as matrix metalloprotease 10 (MMP10) and hepatocyte growth factor (MET) provide potential targets for intervention and the control of airway hyperresponsiveness which is important clinically [63]. In SARP, asthmatics with persistently high Th2 inflammation were older and had more severe disease. Body weight had a major impact on blood biomarker expression in these subjects [64].

Oxidative and nitrosative stress levels are high in patients with GC-insensitive asthma and oxidative stress has been shown to reduce GR translocation in primary human airway fibroblasts [65]. This effect of oxidative stress may be important in asthmatic subjects who smoke as these patients do not respond well to either high dose inhaled or oral GCs [66].

Viral and bacterial infections — The role of viral and bacterial infection in modulating GC responsiveness is becoming increasingly evident. Both rhinovirus and respiratory syncytial virus can induce a degree of GC insensitivity in primary bronchial epithelial cells following activation of pro-inflammatory signalling pathways [67].

The airway microbiome in patients with severe asthma differs from that seen in patients with milder disease and may be an additional contributor to abnormalities in GC sensitivity [68]. In a study of bacterial RNA in bronchial brushings, samples from patients with GC-resistant asthma were significantly enriched in Actinobacteria and a Klebsiella genus member although dysbiosis was also associated with different sub-phenotypes of GC-resistant asthma. Thus, patients with worse Asthma Control Questionnaire (ACQ) scores and higher sputum total leukocyte values had a greater predominance of Proteobacteria, whereas those with a high body mass index (BMI) were associated with greater levels of Bacteroidetes/Firmicutes.

In addition, bacterial clearance from the airways of severe GC-resistant asthmatics may be defective. Monocyte-derived macrophages from patients with GC-resistant asthma are less able to phagocytose Haemophilus influenzae and Staphylococcus aureus than those from non-severe asthmatics [69].

Importantly, the influence of the airway microbiome in glucocorticoid responsiveness in asthma has also been reported [70]. In a study that compared bacteria in the bronchoalveolar lavage samples from subjects with GC-resistant and GC-sensitive asthma, expansion of specific gram-negative bacteria including Haemophilus parainfluenzae, which induce GC resistance through p38 MAPK activation was noted in a subset of subjects with GC-resistant asthma, but not in subjects with GC-sensitive asthma. However, inhibition of transforming growth factor-beta-associated kinase-1 (TAK1), but not p38 MAPK, restored cellular sensitivity to GCs.

Unbiased hierarchical clustering of differentially expressed genes in sputum of severe and moderate asthmatics identified three transcriptome-associated clusters (TACs). TAC1 was characterized by IL33R, CCR3, and TSLPR receptors and highest enrichment of the Th2high signature; TAC2 was characterized by IFN and inflammasome-associated genes; and TAC3 was characterized by metabolic genes and normal mitochondrial function [71]. Inflammasome activation in TAC2 was linked with sputum neutrophilia, asthma severity and with high sputum IL-1beta protein levels [72]. The presence of neutrophils and inflammasome activation suggests the potential presence of sub-clinical infection or microbial dysbiosis in these patients [73]. Interestingly, the macrolide antibiotic azithromycin given for 48 weeks reduced exacerbations and improved asthma quality of life and may be a useful therapy for these patients [74].

Topological data analysis of asthma sputum proteomics and transcriptomics revealed 10 subdivisions of asthma phenotypes and highlighted three distinct groups of eosinophilic and neutrophilic asthmatics with separate protein, gene, and clinical features.

CLINICAL FEATURES — Patients with glucocorticoid (GC)-resistant asthma were first reported in 1967 [75]. These patients typically have severe asthma, but extrapulmonary glucocorticoid function is largely normal.

Asthma — Patients with GC-resistant asthma have a phenotype of severe asthma (table 2) [4] and require high-dose inhaled glucocorticoids and sometimes chronic use of oral glucocorticoids. (See 'Definition' above and "Evaluation of severe asthma in adolescents and adults".)

In both the Severe Asthma Research Program (SARP) and the Unbiased BIOmarkers in PREDiction of respiratory disease outcomes (U-BIOPRED) cohorts, at least 30 percent of patients with severe asthma required regular use of oral GCs to maintain asthma control [76]. These patients appeared to be relatively, rather than absolutely, GC insensitive based on the presence of a response to higher doses of systemic glucocorticoids (ie, triamcinolone 120 mg, intramuscularly) [77]. The features of severe asthma in general are described separately. (See "Evaluation of severe asthma in adolescents and adults", section on 'History and physical'.)

Among patients who require high-dose inhaled glucocorticoids or chronic oral glucocorticoids to maintain asthma control, approximately 50 percent have persistent airflow limitation and 69 percent have sputum eosinophilia (defined as ≥2 percent eosinophils) [76,78]. A subset of patients with frequent oral GC use has late onset asthma, obesity, and moderate reductions in the forced expiratory volume in one second (FEV₁). A case series of 58 asthmatic subjects who were clinically resistant to prednisolone therapy described a longer duration of asthma, poorer morning lung function, and a greater degree of bronchial reactivity than GC-sensitive asthmatics [79].

Clustering on clinical features of severe asthmatics has revealed subsets of severe asthmatics but no specific group was associated with GC-insensitive asthma or identified key driver mechanisms [80,81]. GC responsiveness may vary over time in a given individual depending on factors such as cigarette smoking, viral infections, and allergen exposure. A portion of patients who appear GC-insensitive have been found to be nonadherent with oral GC therapy [82].

Extrapulmonary glucocorticoid responses — Abnormalities associated with GC resistance may be elicited in the skin but tend not to be significant in other extrapulmonary sites (eg, adrenal, bone). As an example, the cutaneous vasoconstrictor response to beclomethasone dipropionate is significantly reduced in patients with GC-resistant asthma [83]. This pattern of reduced sensitivity in the lung and in skin blanching, but not in other extrapulmonary sites suggests that this is a defect in the anti-inflammatory effects of GCs, but not in metabolic or endocrine effects. In addition, these patients still develop Cushingoid symptoms.

The hypothalamic-pituitary-adrenal axis appears normal among patients with GC-resistant asthma, as reflected by the levels of urinary free cortisol and plasma adrenocorticotropic hormone (ACTH). Cortisol measurements and dexamethasone suppression responses are normal. (See "Dexamethasone suppression tests".)

GCs cause similar changes in bone metabolism in GC-sensitive and GC-resistant patients. One study compared the effects of five days of treatment with daily prednisolone upon markers of bone turnover in six persons with GC-sensitive and six patients with GC-resistant asthma [84]. No significant differences before or following treatment were observed between the two groups in serum osteocalcin, tartrate resistant acid phosphatase, and alkaline phosphatase, or urinary free deoxypyridinoline.

Before a diagnosis of GC-resistant asthma is made, a number of other entities that are in the differential diagnosis of severe asthma must be excluded and patients must be assessed for compliance with therapy [82,85]. The evaluation of these disorders is discussed separately. (See "Evaluation of wheezing illnesses other than asthma in adults" and "Evaluation of severe asthma in adolescents and adults".):

EVALUATION — Patients with severe asthma who require oral glucocorticoids (GCs) to achieve asthma control or whose asthma remains poorly controlled despite oral GCs should be evaluated following guidelines for severe asthma [4,86] to exclude alternative diagnoses, comorbid diseases, compliance and confounding issues that may exacerbate asthma. The term GC-resistant asthma refers to patients at the less responsive end of a spectrum of GC sensitivity, but is not an actual diagnosis. (See "Evaluation of severe asthma in adolescents and adults".)

The majority of patients who meet criteria for severe asthma are relatively insensitive to GCs, rather than completely GC-resistant. As GC-resistance is almost always relative, and because such a designation does not clearly affect the treatment approach, specific testing to establish GC-insensitivity is not needed. (See 'Definition' above.)

For the rare patients with suspected true GC-resistance on the basis of an absence of Cushingoid features despite ongoing oral GCs, laboratory testing may be helpful in determining whether the patient is nonadherent or GC-resistant. Alternatively, assessing response to intramuscular triamcinolone is sometimes used to determine whether nonadherence is contributing to a lack of response.

Laboratory testing — Specific laboratory testing for glucocorticoid resistance is largely a research tool, and complete GC resistance is rare.

●For patients who have no response to two weeks of oral GCs (eg, prednisone 40 mg/day), one way to assess nonadherence is by checking a serum level of prednisolone or morning serum cortisol, but this is rarely necessary.

●Research suggests that GC resistance in asthma cannot be explained by malabsorption or pharmacokinetic mechanisms, so assessment of absorption and clearance kinetics is unlikely to be helpful. (See 'Mechanisms of glucocorticoid resistance' above and "Measurement of cortisol in serum and saliva", section on 'Measurement of serum cortisol'.)

●Familial GC resistance due to a missense mutation of the glucocorticoid receptor (GR) hormone binding domain causes global insensitivity to GC action and leads hyperandrogenism, but is not associated with GC-resistance in asthma [10,87]. Thus, we do not test for this abnormality. (See "Adrenal hyperandrogenism".)

●Assessment of sputum eosinophils or exhaled nitric oxide may help to predict which patients would be expected to respond to increased GC doses, although data are conflicting [85]. As an example, a fraction of exhaled nitric oxide (FE_NO) >50 ppb suggests GC responsive airway inflammation, while a FE_NO <25 ppb implies an absence of eosinophilic airway inflammation and a low likelihood of GC response. (See "Exhaled nitric oxide analysis and applications", section on 'Clinical use of FENO in asthma'.)

IMPLICATIONS FOR TREATMENT — Management of patients with glucocorticoid (GC)-resistant asthma previously presented unique challenges because of a paucity of effective and well-tolerated alternatives to GCs but the advent of new biological therapies targeting Th2high asthma has proven effective in many subjects [88]. Treatment strategies include avoidance of asthma triggers, use of higher doses and a longer duration of systemic GCs, and optimization of non-GC medications (eg, beta adrenergic agonists, antileukotriene agents, muscarinic antagonists, omalizumab, mepolizumab, benralizumab, reslizumab, dupilumab) and nonpharmacologic methods. (See "An overview of asthma management" and "Treatment of severe asthma in adolescents and adults" and "Allergen avoidance in the treatment of asthma and allergic rhinitis" and "Trigger control to enhance asthma management" and "Antileukotriene agents in the management of asthma" and "Anti-IgE therapy".)

Oral and parenteral glucocorticoids — Because resistance to inhaled and systemic GCs tends to be relative rather than absolute, patients may respond to systemic GCs when given at higher doses or for a longer duration than required by other patients with asthma [4]. While systemic GCs remain the treatment of choice for asthma exacerbations, every effort should be made to avoid long-term use of systemic GCs for asthma management [89]. (See "Acute exacerbations of asthma in adults: Emergency department and inpatient management", section on 'Systemic glucocorticoids'.)

For patients who do not appear to respond to high-dose inhaled GC or oral GC, a therapeutic trial of triamcinolone can be administered 120 mg, intramuscularly, as a single dose [77]. Studies in adults and children show a variable response with 20 percent showing a 10 percent improvement in lung function but indicating a non-responsive pathobiology [90]. Importantly, there appears to be an ethnic difference in the fraction of exhaled nitric oxide (FE_NO) and exacerbation responses to triamcinolone, which are less evident in Black children compared with White children [91]. The mechanism for this difference is unknown. (See "Treatment of severe asthma in adolescents and adults", section on 'Systemic glucocorticoids'.)  

While the mechanisms for GC insensitivity are heterogeneous, inflammation may contribute to a lack of response to GCs through increases in the proinflammatory transcription factor activation protein (AP)-1 and elevated levels of interleukin (IL)-2 and IL-4. If this hypothesis is correct, early use of GCs might suppress inflammation and preserve GC sensitivity [92]. In addition, a course of systemic GCs may reduce airways inflammation enough that the patient becomes responsive to inhaled GCs.

Inhaled glucocorticoids — Patients with severe asthma are relatively refractory to inhaled GCs and require high-dose inhaled GCs to maintain asthma control. When asthma is not well controlled on high dose inhaled GC (eg, about 1000 mcg per day), but appears to be GC responsive, we consider a further increase in dose or use of an inhaled GC with a smaller particle size. The evidence for increasing the dose of inhaled GC above 1000 mcg per day is limited and not all patients will improve with higher inhaled GC doses. (See "Treatment of severe asthma in adolescents and adults", section on 'Inhaled glucocorticoids'.)

Some work has shown that GC sensitivity may vary over time, so it is clinically prudent to institute a trial of a higher dose of inhaled GC at regular intervals even if they have not been effective when tried previously [93]. At some point, assessment of airway eosinophilia (eg, sputum eosinophils, increased FE_NO) may become a useful tool to assess whether a patient might respond to a higher dose of inhaled GCs, although further study is needed to determine the exact biomarkers and parameters [85,94,95]. (See "Severe asthma phenotypes", section on 'Phenotyping based on biomarkers of inflammation'.)

Non-glucocorticoid agents — Patients with relative GC-resistance may respond better to non-GC treatments for asthma than to GCs, depending on their asthma phenotype. Non-glucocorticoid agents include long-acting beta agonists, long-acting muscarinic antagonists, leukotriene-modifying agents, anti-immunoglobulin E therapy (omalizumab), anti-interleukin-5 (anti-IL-5) therapy (mepolizumab, reslizumab, benralizumab), and anti-IL-4/13 therapy (dupilumab). (See "Severe asthma phenotypes" and "Treatment of severe asthma in adolescents and adults", section on 'Adjusting controller therapy' and "Treatment of severe asthma in adolescents and adults", section on 'Persistently uncontrolled asthma' and "Anti-IgE therapy".)

Allergen immunotherapy has not proved effective in these patients and has considerable side effect issues [96].

Research to identify phenotypes among patients with severe asthma may help guide selection of appropriate agents. (See 'Future directions' below and "Severe asthma phenotypes".)

Nonpharmacologic management — Proper nonpharmacologic management is critical in GC-resistant individuals. Close follow-up and frequent communication between patient and clinician are advisable.

●Precipitants of bronchospasm should be identified, and specific recommendations regarding trigger avoidance should be given (table 3). For patients who continue to smoke cigarettes, every effort should be made to achieve smoking cessation. (See "Trigger control to enhance asthma management" and "Allergen avoidance in the treatment of asthma and allergic rhinitis" and "Overview of smoking cessation management in adults".)

●Patients should be educated regarding the nature of their disease, and correct inhaler technique should be periodically reinforced. (See "Asthma education and self-management" and "Delivery of inhaled medication in adults" and "The use of inhaler devices in adults".)

●Bronchial thermoplasty involves targeted application of heat (via radiofrequency waves) to the airways and may be of benefit in selected adults with severe asthma that is not well-controlled with inhaled GCs and long-acting beta agonists. (See "Treatment of severe asthma in adolescents and adults", section on 'Bronchial thermoplasty'.)

FUTURE DIRECTIONS — Current research is focused on identifying non-glucocorticoid (GC) immunomodulatory agents that target pathways of inflammation in asthma and determining which patients are most likely to respond to specific therapies.

●Investigational therapies – New drugs are being developed to target other pathways of asthma inflammation [97], such as interleukins and their receptors, chemokine receptor homologous molecule expressed on T helper type 2 cells (CRTh2), and also chemokine receptors, such as the C-C chemokine receptor, CCR3 [97]. Investigational agents for asthma are discussed in detail separately. (See 'Altered cytokine expression' above and "Investigational agents for asthma".)

Another line of investigation is to identify agents that work alone or in combination with glucocorticoids to overcome GC resistance, and thus re-establish essential anti-inflammatory mechanisms. Examples include agents that restore histone deacetylase activity (HDAC), or agents that inhibit phosphorylation of the glucocorticoid receptor (GR) via inhibiting p38 mitogen activated protein kinase (p38 MAPK) or other kinases [98]. (See 'Mechanisms of glucocorticoid resistance' above.)

Calcitriol, the active form of vitamin D (1alpha,25-dihydroxyvitamin D3), influences IL-10 production and may play a role in GC sensitivity in asthma [29,99]. Calcitriol restores the impaired IL-10 synthesis observed in GC refractory asthma patients, both in vitro and after oral administration for one week [29]. Children with GC-resistant asthma have reduced serum levels of vitamin D, and this correlates with the forced expiratory volume in one second (FEV₁) percent predicted, Asthma Control Test (ACT test) scores, and bronchodilator reversibility, as well as inversely correlating with exacerbations and inhaled GC usage [100]. (See 'Altered cytokine expression' above.)

Specific agents that improve asthma by targeting mechanisms of neutrophilic asthma have not been identified [101]. (See "Severe asthma phenotypes", section on 'Neutrophilic asthma'.)

●Use of biomarkers to guide treatment – Asthma is a syndrome with many potential phenotypes or endotypes that may help to determine the most effective therapeutic regimen suitable for each patient; it is therefore necessary to stratify patients accordingly [102]. Several large groups such as the Severe Asthma Research Program (SARP) and Unbiased BIOmarkers in PREDiction of respiratory disease outcomes (U-BIOPRED), as well as individual centers, have been instrumental in defining these subgroups of asthma and severe asthma [103]. These subgroups include for example the T helper lymphocyte (Th2) high and Th2 low groups [104]. Th2 high asthma is inversely correlated with the presence of IL-17/Th17 high [31,32] and IL-6 trans-signaling (IL6TS) high groups [44] and a group with high expression of IFN-stimulated genes (ISGs) [41] in the absence of viral infection. There is also a group of obese asthmatics with severe disease [105]. The American Thoracic Society/European Respiratory Society (ATS/ERS) guidelines highlight the need for greater understanding of severe asthma, disease stratification, and determination of treatment efficacy [4]. (See "Severe asthma phenotypes".)

Some studies have assessed the efficacy of certain biomarkers to guide selection of asthma therapy. Patients who are not well-controlled despite high dose inhaled or oral GCs and who exhibit high sputum and blood eosinophils and high levels of FE_NO may respond better to anti-IL-5-directed therapies or omalizumab with respect to exacerbation frequency and oral GC use [106-109]. The potential role of novel biologic agents in the treatment of asthma is discussed separately. (See "Severe asthma phenotypes" and "Investigational agents for asthma", section on 'Biologic agents'.)

Measurements of exhaled metabolites (electronic nose) and urinary eicosanoids may help characterize the inflammatory profile of patients with severe asthma and differentiate those that may respond to anti-Th2 or non-Th2 therapies [110,111].

SUMMARY AND RECOMMENDATIONS

●Definition of GC-resistance – Glucocorticoid (GC) responsiveness represents a continuous spectrum; patients with chronic asthma who are unresponsive to high doses of GCs and are without confounding factors have been termed GC-resistant. (See 'Definition' above.)

●Clinical features of GC-resistance

•Among patients who require high-dose inhaled glucocorticoids or chronic oral glucocorticoids to maintain asthma control, approximately 50 percent have persistent airflow limitation and 69 percent have sputum eosinophilia (defined as ≥2 percent eosinophils). While the cutaneous vasoconstrictor response to topical glucocorticoid is significantly reduced, other tissues (eg, adrenal, bone) do not appear to have an abnormal GC response. (See 'Clinical features' above.)

•In patients with GC-resistant asthma, the function of the hypothalamic-pituitary-adrenal axis appears normal. Serum cortisol measurements and dexamethasone suppression responses are normal, and GC-induced side-effects are a major issue. (See 'Clinical features' above.)

●Evaluation – Patients with severe asthma who require oral GCs to achieve asthma control or whose asthma remains poorly controlled despite oral GCs should be evaluated following guidelines for severe asthma to exclude alternative diagnoses, comorbid diseases, and confounding issues that may exacerbate asthma. As GC-resistance is almost always relative and does not clearly affect the treatment approach, specific testing to establish GC-insensitivity is not needed. (See 'Evaluation' above.)

●Mechanism

•GC resistance in asthma cannot be explained by malabsorption or pharmacokinetic mechanisms. Defects in GC-glucocorticoid receptor (GR) binding, GR nuclear translocation, and intranuclear actions of GR have all been proposed as major mechanisms for GC resistance (figure 1). (See 'Mechanisms of glucocorticoid action' above.)

•The exact mechanism of GC resistance in asthma is not known. Possible mechanisms include altered splicing of the GR premessenger RNA, altered proinflammatory cytokine expression, a reduced number of GRs, and epigenetic changes (figure 2). (See 'Mechanisms of glucocorticoid resistance' above.)

●Implications for treatment – The optimal treatment for GC-resistant asthma is not known. Treatment strategies should aim to reduce the total exposure to systemic GCs using a pulsed approach and optimization of non-GC agents (eg, long-acting beta-adrenergic agonists, long-acting muscarinic antagonists, antileukotriene agents), nonpharmacologic therapies (eg, trigger avoidance and possibly bronchial thermoplasty), and targeted biological therapies (omalizumab, mepolizumab, benralizumab, reslizumab, dupilumab, tezepelumab). (See 'Implications for treatment' above and "Treatment of severe asthma in adolescents and adults" and "Investigational agents for asthma".)

●Future directions

•New drugs are being developed that target other pathways of asthma inflammation, such as interleukins and their receptors and chemokine receptors (eg, C-C chemokine receptor, CCR3), or that prevent oxidative stress. Biomarkers of adherence to therapy, eosinophilic and noneosinophilic phenotypes of severe asthma, and predictors of response to novel therapeutic agents are also in development. (See 'Future directions' above.)

•Biomarkers of compliance and predictors of response to therapeutic agents are needed. Examples of biomarkers being developed are dipeptidyl-peptidase (DPP)4 and FK506 Binding Protein (FKBP)5. (See 'Future directions' above.)