Corticosteroid insensitive alveolar macrophages from asthma patients; synergistic interaction with a p38 mitogen-activated protein kinase (MAPK) inhibitor

Abstract

Aims

Some asthma patients remain symptomatic despite using high doses of inhaled corticosteroids (ICS). We used alveolar macrophages to identify individual patients with insensitivity to corticosteroids and to evaluate the anti-inflammatory effects of a p38 mitogen-activated protein kinase (MAPK) inhibitor combined with a corticosteroid on these cells.

Methods

Alveolar macrophages from 27 asthma patients (classified according to the Global Initiative for Asthma (GINA) treatment stage. Six GINA1, 10 GINA2 and 11 GINA3/4) were stimulated with lipoploysaccharide (LPS) (1 μg ml⁻¹). The effects of dexamethasone (dex 1–1000 nm), the p38 MAPK inhibitor 1-(5-tert-butyl-2-p-tolyl-2Hpyrazol-3-yl)-3(4-(2-morpholin-4-yl-ethoxy)naphthalen-1-yl)urea (BIRB-796 1–1000 nm) and both drugs combined at all concentrations on supernatant TNFα, IL-6 and CXCL-8 concentrations were analyzed by ELISA. Dose-sparing and efficacy enhancing effects of combination treatment were determined.

Results

Dexamethasone reduced LPS-induced TNFα, IL-6 and CXCL-8 in all groups, but maximum inhibition was significantly reduced for GINA3/4 compared with GINA2 and GINA1 (P < 0.01). A subgroup of corticosteroid insensitive patients with a reduced effect of dexamethasone on cytokine secretion were identified. BIRB-796 in combination with dexamethasone significantly increased cytokine inhibition compared with either drug alone (P < 0.001) in all groups. This effect was greater in corticosteroid insensitive compared with sensitive patients. There were significant synergistic dose-sparing effects (P < 0.05) for the combination treatment on inhibition of TNFα, IL-6 and CXCL-8 in all groups. There was also significant efficacy enhancing benefits (P < 0.05) on TNFα and IL-6.

Conclusions

p38 MAPK inhibitors synergistically enhance efficacy of corticosteroids in macrophages from asthma patients. This effect is greater in corticosteroid insensitive asthma patients, suggesting that this class of drug should be targeted to this patient phenotype.

What is Already Known about this Subject

• A proportion of asthma patients remain symptomatic despite inhaled corticosteroid (ICS) treatment.

• It is possible to measure corticosteroid insensitivity using macrophages as a biomarker.

• Combining corticosteroids and p38 mitogen-activated protein kinase (MAPK) inhibitors causes greater anti-inflammatory effects on macrophages compared with either drug alone.

What This Study Adds

• Using alveolar macrophages as a biomarker, a subset of asthma patients display corticosteroid insensitivity.

• A p38 MAPK inhibitor combined with a corticosteroid gives synergistic inhibition in alveolar macrophages from asthma patients.

• The synergistic combination effect is greatest in corticosteroid insensitive asthma patients.

Introduction

Inhaled corticosteroids (ICS) are used to treat persistent airway inflammation in asthma. These drugs are effective therapies in many asthma patients, but a proportion of patients remain symptomatic despite high dose ICS treatment 1.

The p38 mitogen-activated protein kinase (MAPK) signalling pathway controls inflammatory mediator production, through the activation of transcription factors such as NF-κB 2 and post-transcriptional mechanisms such as mRNA stabilization and enhanced protein translation. A range of stimuli evoke p38 MAPK signalling, including Toll-like receptor agonists, cytokines and osmotic stress 3. Drugs designed to inhibit p38 MAPK have been tested in inflammatory diseases such as rheumatoid arthritis and chronic obstructive pulmonary disease (COPD) 4–6. There is increased p38 MAPK activation in the airways of severe asthma patients 7,8, suggesting that p38 MAPK inhibitors may also be effective in asthma.

Alveolar macrophages secrete a range of pro-inflammatory cytokines and chemokines 9. The in vitro effects of corticosteroids on cytokine production from alveolar macrophages are reduced in patients with severe asthma 7,10. This in vitro insensitivity to the effects of corticosteroids mirrors the clinical situation where many patients with severe asthma respond poorly to ICS. The in vitro assessment of alveolar macrophages may be a surrogate for the clinical response to corticosteroids. Such biomarkers of drug response may be useful as part of a personalized medicine approach, where treatment is tailored according to specific individual characteristics 11.

p38 MAPK inhibitors reduce cytokine production from human alveolar macrophages 12–14. p38 MAPK activation in alveolar macrophages is corticosteroid insensitive. Corticosteroids have no effect on the phosphorylation of p38 MAPK or its downstream target, heat shock protein 27, in lipopolysaccharide (LPS) stimulated alveolar macrophages 14.

It is known that combining corticosteroids and p38 MAPK inhibitors causes greater anti-inflammatory effects on alveolar macrophages and peripheral blood mononuclear cells from asthmatics compared with either drug alone 15,16. However, to determine properly if this combination effect is additive or synergistic, it is necessary to perform full dose−response curves with both drugs alone and with the drugs combined 14,17. We have used this methodology to demonstrate that corticosteroids and p38 MAPK inhibitors have additive and synergistic effects on cytokine production from COPD alveolar macrophages 14.

The aim of this paper was to identify corticosteroid insensitive patients with asthma and to study the potential anti-inflammatory benefits of p38 MAPK inhibition in these patients. We used alveolar macrophages as a biomarker of corticosteroid sensitivity and studied the effects of combining a p38 MAPK inhibitor with corticosteroids. We have investigated whether an additive or synergistic interaction occurs between these drugs in corticosteroid insensitive and sensitive macrophages by using full dose−response curves with both drugs alone and with the drugs combined 14,17.

Methods

Study subjects

Patients with a previous physician diagnosis of asthma were recruited. All subjects were required to be lifelong non-smokers. Patients were categorized into GINA groups based on treatment; short acting β-adrenoceptor agonist use only (GINA stage 1; n = 8), ICS use (GINA stage 2; n = 10), and ICS and long acting β₂-adrenoceptor agonist (LABA) use (GINA stage 3 or 4; n = 12). Patients performed spirometry for measurement of FEV₁ and reversibility to inhaled salbutamol (200 μg), the asthma control questionnaire (ACQ), skin prick testing using house dust mite, cat and grass allergens and exhaled nitric oxide (eNO) at 50 ml s⁻¹ (Niox, Aerocrine, Sweden). All subjects gave written informed consent. The study was approved by the local research ethics committee NRES Committee North West – Greater Manchester South (MAIN REC REF: 06/Q1403/156).

Bronchoscopy

Bronchoscopies were performed as previously described 18 with a total instilled volume of 480 ml. Broncho-alveolar lavage (BAL) fluid was placed on ice. Cytospins were prepared by cytocentrifugation at 7000 g, air-drying for 30 min, followed by methanol fixation and staining with RAPI-DIFF II (Triangle Biomedical Science, Durham, NC, USA). Four hundred non-squamous cells were counted and differential cell counts obtained as a percentage of total non-squamous cells.

Cell culture

Alveolar macrophages were isolated from BAL as previously described 18 and cultured with or without dexamethasone alone (at 1, 10, 100, 300 and 1000 nm), the p38 MAPK inhibitor BIRB-796 alone (1–1000 nm) and both drugs combined at all of these concentrations for 2 h followed by LPS (1 μg ml⁻¹) stimulation (serotype O26:B6) for 24 h. Both drugs were reconstituted with dimethylsulfoxide (DMSO) and diluted in supplemented RPMI 1640 media to a final concentration of 0.005% DMSO. Supernatants were stored at −80°C.

Cytokine and chemokine assays

Supernatants were analyzed by ELISA according to the manufacturer's instructions (R & D systems, Abingdon, UK) to quantitate TNFα, IL-6 and CXCL-8 concentrations. The lower limits of quantification for the TNFα, IL-6 and CXCL-8 ELISAs were 15.6 pg ml⁻¹, 31.3 pg ml⁻¹ and 9.4 pg ml⁻¹, respectively.

Data analysis

Normality was assessed using the Kolmogorov−Smirnov test. BAL cell counts were compared between all subject groups using analysis of variance (anova). All ELISA data were normally distributed. Data were compared between all subject groups using anova followed by unpaired t-tests. Within group comparisons were performed using paired t-tests. P < 0.05 was considered significant. IC₅₀ values were determined in GraphPad Prism (GraphPad Software http://www.graphpad.com) using a four parameter non-linear iterative curve fitting analysis.

The equivalence of response between subject groups was assessed by fitting dose−response curves with separate parameters for each subject group, fitting dose−response curves restricting the parameter of interest, either IC₅₀ or maximal inhibition, to be common across the subject groups and comparing the goodness of fit using an F-statistic. Two analyses were performed to assess whether a combination of dexamethasone and BIRB-796 exhibited synergy, a dose-sparing analysis to assess whether equivalent responses can be achieved at lower doses of compound than expected given the monotherapy response of the two compounds and an efficacy enhancing analysis to assess whether the combination results in a significantly greater maximal effect than either compound alone as monotherapies. The dose-sparing analysis calculates a combination index with confidence intervals using the method described 17. The efficacy enhancing analysis fits Hill dose−response curves to the monotherapy and combination results using both common and separate parameters for maximal response, and tests for the improvement in fit from allowing the parameter to vary by using an F-test. Both analyses were performed assuming a slope parameter in the Hill dose−response equation equal to 1. Robustness analyses were also performed estimating the slope parameters and found to give the same conclusions.

Results

The clinical characteristics of the patients are shown in Table 1. GINA 3/4 patients had the lowest FEV₁ % predicted and highest ACQ score, whilst taking more ICS than patients in the GINA 1 or 2 groups. BAL cytospin cell differential counts did not differ between groups.

Table 1.

Subject demography

Clinical characteristics	GINA 1 (n = 6)	GINA 2 (n = 10)	GINA 3/4 (n = 11)	P value
Gender Male/Female	5/1	5/5	4/7	–
Age (years)	45 (18–63)	43 (19–68)	36 (19–60)	0.5
Daily ICS dose* (μg)	0 ± 0	330 ± 600	868 ± 605	0.0001
Atopy	5/6	8/10	7/11	–
FEV1 % predicted (%)	88.5 ± 10.4	79.1 ± 18.7	78.0 ± 17.6	0.5
Reversibility (ml)	223 ± 198	363 ± 213	256 ± 250	0.4
Reversibility (%)	7.74 ± 7.27	16.1 ± 13.0	12.0 ± 11.4	0.3
ACQ-7 score	0.80 ± 0.21	1.34 ± 0.69	1.73 ± 0.78	0.04
NO ppb	59.9 ± 40.8	43.4 ± 20.8	27.0 ± 31.1	0.05
BAL cell count
Neutrophil %	4.4 ± 7.2	5.8 ± 6.0	7.0 ± 7.4	0.8
Macrophage %	82.6 ± 12.8	83.5 ± 6.3	78.3 ± 15.7	0.7
Eosinophil %	4.1 ± 7.9	3.2 ± 3.8	7.7 ± 11.5	0.9
Lymphocytes %	7.8 ± 5.3	6.7 ± 7.8	5.8 ± 6.7	0.7
Epithelial %	0.6 ± 1.3	0.3 ± 0.7	0.3 ± 0.4	0.9

^*Beclometasone equivalent dose. Data are presented as means± SD, median (range) for age. ICS, inhaled corticosteroids; NO, nitric oxide; ppb, parts per billion.

LPS-stimulated protein production

The secretion of TNFα, IL-6 and CXCL-8 from unstimulated alveolar macrophages was similar between the GINA groups (anova P > 0.05 for all comparisons; see Figure S1). LPS increased the secretion of these proteins, with no difference between groups observed (anova P > 0.05 for all comparisons; Figure S1).

Effects of dexamethasone

Dexamethasone significantly reduced LPS stimulated secretion of TNFα, IL-6 and CXCL-8 from alveolar macrophages in a concentration-dependent manner in all three GINA groups (Figure 1). The top of the dose−response curve was observed by 300 nm. The magnitude of cytokine inhibition at this concentration (maximal inhibition) was reduced in GINA 3/4 patients compared with GINA 1 and GINA 2 patients as shown in Table 2, e.g. IL-6 maximal inhibition was 87%, 71% and 47% in GINA 1, 2 and 3/4, respectively. The drug effect was significantly lower in GINA 3/4 patients compared with GINA 1 and GINA 2 for TNFα (P = 0.02 and P = 0.03, respectively), compared with GINA 1 for CXCL-8 (P = 0.005) and approaching significance compared with GINA 1 for IL-6 (P = 0.05). A separate analysis of maximal inhibition using the fitted dose−response curves also showed a significant difference between groups for TNFα (P = 1.14 × 10⁻⁸), IL-6 (P = 3.03 × 10⁻⁸) and CXCL-8 (P = 3.31 × 10⁻⁸). The IC₅₀ values were also greater in GINA 3/4 patients compared with GINA 1 and GINA 2 patients (see Table 2).

Figure 1.

Effect of dexamethasone on inhibition of LPS induced TNFα (A), IL-6 (B) and CXCL-8 (C) in alveolar macrophages from asthma patients with GINA 1 (n = 6), GINA 2 (n = 10) or GINA 3/4 (n = 11) classification. Data shown are mean ± SEM percentage inhibition of LPS induced TNFα, IL-6 and CXCL-8 measured by ELISA. , GINA 1 ICS naive; , GINA 2 ICS only; , GINA 3/4 (ICS + LABA)

Table 2.

Maximal inhibitions and IC₅₀ of dexamethasone alone on pro-inflammatory mediators

Subject group	TNFα		IL-6		CXCL-8		Difference between % inhibition (300 nm) between cytokines
	% inhibition (300 nm)	[IC50] (nm)	% inhibition (300 nm)	[IC50] (nm)	% inhibition (300 nm)	[IC50] (nm)	ANOVA P value
GINA 1 (n = 6)	87	3	87	3	73	7	0.11
GINA 2 (n = 10)	83	6	71	4	47	NA	0.002
GINA 3/4 (n = 11)	63	18	48	NA	33	NA	0.059
Difference between % inhibition (300 nm) between groups	P = 0.0076		P = 0.04		P = 0.0089

Data are presented as means. NA, not able to calculate IC₅₀ from the curve as 50% inhibition not reached.

The effect of dexamethasone varied between cytokines; The maximal inhibition of CXCL-8 was numerically lower compared to IL-6 and TNFα in all GINA groups, with this difference reaching significance in GINA 2 patients (P < 0.01 and P < 0.001 respectively; Table 2), and approaching statistical significance in GINA 1 patients (anova P = 0.11) and GINA 3/4 patients (anova P = 0.059). Furthermore, the dexamethasone IC₅₀ for CXCL-8 was greater than TNFα and IL-6 in all GINA groups.

The dexamethasone dose−response curves for individual patients showed a range of different responses, with a subset of patients displaying a reduced maximal inhibition (Figure 2). Cut-off levels of maximal inhibition of 70%, 70% and 60% for TNFα, IL-6 and CXCL-8, respectively, appeared to identify a subgroup of patients with reduced sensitivity to corticosteroids. Five subjects were below these cut-off levels for all three proteins, with seven, eight and 13 subjects below the cut-off levels for TNFα, IL-6 and CXCL-8, respectively. Table S1 shows that the number of corticosteroid insensitive patients was greatest in the GINA 3/4 group.

Figure 2.

Effect of dexamethasone on inhibition of LPS induced TNFα (A and B), IL-6 (C and D) and CXCL-8 (E and F) in alveolar macrophages from 27 individual asthma patients. A, C and E patients are highlighted as GINA 1 (red), GINA 2 (green) or GINA 3/4 (blue) classification. B, D and F patients are highlighted as corticosteroid sensitive (grey) or corticosteroid insensitive (purple) with relative cut-off for TNFα (70%), IL-6 (70%) and CXCL-8 (60%) shown. Data shown are percentage inhibition for LPS induced TNFα, IL-6 and CXCL-8 measured by ELISA. , GINA 1; , GINA 2; , GINA 3/4; , steroid sensitive; , steroid insensitive

There was no difference in ICS usage between corticosteroid sensitive and insensitive groups for patients taking ICS (GINA 2/3/4; see Table S2). Table S3 shows the clinical characteristics of all the patients categorized according to whether they were corticosteroid sensitive or insensitive. The ACQ score was significantly increased in the insensitive patients (P < 0.05). There were no differences in any other clinical characteristics between groups.

Basal and LPS stimulated levels of inflammatory mediators were similar between corticosteroid sensitive and insensitive patients (anova P > 0.05 for analyses; Figure S2).

Effects of p38 inhibition

BIRB-796 significantly reduced LPS stimulated TNFα, IL-6 and CXCL-8 production in a concentration dependent manner in all three GINA groups (Figure 3). There were no significant differences in maximal inhibition caused by BIRB-796 between GINA groups apart from between GINA 1 and GINA 2 patients for IL-6 (P < 0.05). There were no significant differences in maximal inhibition with BIRB-796 between corticosteroid sensitive and insensitive patients (anova P > 0.05), and the dose−response effects of BIRB-796 were similar between these groups (Figure 5).

Figure 3.

Effect of BIRB-796 on inhibition of LPS induced TNFα (A), IL-6 (B) and CXCL-8 (C) in alveolar macrophages from asthma patients with GINA 1 (n = 6), GINA 2 (n = 10) or GINA 3/4 (n = 11) classification. Data shown are mean ± SEM percentage inhibition of LPS induced TNFα, IL-6 and CXCL-8 measured by ELISA. , GINA 1 ICS Naive; , GINA 2 ICS only; , GINA 3/4 (ICS + LABA)

Figure 5.

Combination effect of dexamethasone and BIRB-796 on inhibition of LPS induced TNFα (A–C), IL-6 (D–F) and CXCL-8 (G–I) in alveolar macrophages from corticosteroid sensitive (grey) and insensitive (purple) asthma patients. A, D and G show dexamethasone alone (1–1000 nm) response. B, E and H show BIRB-796 alone (1–1000 nm) response. C, F and I show dexamethasone (1–1000 nm) in combination with BIRB-796 (1000 nm) response. Corticosteroid sensitivity groups are specific to each individual cytokine. Corticosteroid sensitive patients n = 20, 17 and 12, insensitive patients n = 7, 8 and 13 for TNFα, IL-6 and CXCL-8 respectively. Data shown are mean ± SEM percentage inhibition of LPS induced TNFα, IL-6 and CXCL-8 measured by ELISA. , steroid sensitive; , steroid insensitive

Combination effects of dexamethasone and BIRB-796

BIRB-796 in combination with dexamethasone caused greater maximal inhibition of cytokine production compared with dexamethasone or BIRB-796 alone (Figure 4A–C). Using the concentration of dexamethasone that caused maximal inhibition (300 nm), the addition of BIRB-796 1 nm caused a further significant reduction in cytokine production, with progressive inhibition of cytokine production with increasing concentrations of BIRB-796 (see Table 3). There were differences in the size of the effect according to the cytokine measured and GINA group. Significant dose-sparing synergy was observed for the combination treatment on TNFα, IL-6 and CXCL-8 inhibition in all three subject groups (P < 0.05 for all analyses; Table S4).

Figure 4.

Combination effect of dexamethasone and BIRB-796 on inhibition of LPS induced TNFα (A), IL-6 (B) and CXCL-8 (C) in alveolar macrophages from asthma patients with GINA 1 (i), GINA 2 (ii) or GINA 3/4 (iii) classification. Data shown are mean ± SEM percentage inhibition of LPS induced TNFα, IL-6 and CXCL-8 measured by ELISA. GINA 1 (n = 6), GINA 2 (n = 10) and GINA 3/4 (n = 11) shown. , BIRB-796; , Dex; , Dex + BIRB-796 1 nm; , Dex + BIRB-796 10 nm; , Dex + BIRB-796 100 nm; , Dex + BIRB-796 300 nm; , Dex + BIRB-796 1000 nm

Table 3.

Maximal inhibitions of dexamethasone and BIRB-796 in combination on pro-inflammatory mediators

Subject group	Maximum % inhibition (300 nm dexamethasone)
	Dex	BIRB-796	Dex + BIRB-796 1 nm	Dex + BIRB-796 10 nm	Dex + BIRB-796 100 nm	Dex + BIRB-796 300 nm	Dex + BIRB-796 1000 nm
	TNFα
GINA 1 (n = 6)	87	75	90	91	95** ##	95 ** ##	95 ** ##
GINA 2 (n = 10)	83	62	89** ##	91** ##	95** ##	95** ##	95** ##
GINA 3/4 (n = 11)	63	71	67	79 **	81 ** #	87** ##	89** ##
Corticosteroid sensitive (n = 20)	86	69	89** ##	92** ##	94** ##	95** ##	96** ##
Corticosteroid insensitive (n = 7)	47	67	55	70**	73**	83** #	84** #
	IL-6
GINA 1 (n = 4)	87	68	89	91#	93* #	94** #	94** #
GINA 2 (n = 10)	71	36	80** ##	87** ##	87** ##	87** ##	87** ##
GINA 3/4 (n = 11)	48	53	52	67**	71** #	77** ##	79** ##
Corticosteroid sensitive (n = 17)	78	44	82*##	89** ##	90** ##	91** ##	91** ##
Corticosteroid insensitive (n = 8)	29	45	38**	56**	60**	68** #	70** #
	CXCL-8
GINA 1 (n = 6)	73	48	73#	81##	80##	86##	84##
GINA 2 (n = 8)	46	23	62** ##	63** ##	73** ##	73** ##	71** ##
GINA 3/4 (n = 11)	30	33	38	44*	48**	50** #	54** ##
Corticosteroid sensitive (n = 12)	69	39	76** ##	77** ##	81** ##	82** ##	82** ##
Corticosteroid insensitive (n = 13)	28	28	39	46** #	53** ##	55** ##	56** ##

^*, ^**, ^***, Significantly above dexamethasone alone (P < 0.05, 0.01, 0.001, respectively).

^{#, ##, ###} Significantly above BIRB-796 alone (P < 0.05, 0.01, 0.001, respectively). Data are presented as means. Corticosteroid sensitivity groups are specific to each individual cytokine.

Combination effects in corticosteroid sensitive and insensitive individuals

Table 3 shows that BIRB-796 added to dexamethasone 300 nm caused greater reductions in cytokine production compared with either drug used alone in both corticosteroid sensitive and insensitive patients.

The combination treatment of dexamethasone and BIRB-796 provided significant efficacy-enhancing effects for TNFα (P < 0.001) and IL-6 (P < 0.05) inhibition in both groups of patients (Table 4). This analysis could not be performed for CXCL-8 data due to high variance in the data. This efficacy-enhancing effect was greater in corticosteroid insensitive patients compared with sensitive patients. Figure 5 shows the difference in the dexamethasone response curves between insensitive and sensitive patients, and the increased cytokine suppression in both groups using BIRB-796 1000 nm.

Table 4.

Efficacy-enhancing combination index of dexamethasone and BIRB-796 treatment on pro-inflammatory mediators

Subject group	Efficacy-enhancing combinations
	Efficacy-enhancing benefit	Confidence interval		P value
	TNFα
Corticosteroid sensitive	10.8	5.6	16.0	6.78 × 10−5
Corticosteroid insensitive	36.4	17.7	55.1	2.30 × 10−4
	IL-6
Corticosteroid sensitive	9.6	2.3	16.8	0.010
Corticosteroid insensitive	45.1	7.0	83.1	0.021

Corticosteroid sensitivity groups are specific to each individual cytokine.

Significant dose-sparing synergy was also observed for the combination treatment upon TNFα (P < 0.001) and IL-6 (P < 0.001) inhibition in both groups (Table S4). For CXCL-8, this analysis could not be performed. There was no difference between corticosteroid sensitive and insensitive patients (P > 0.05).

Discussion

In clinical practice, some asthma patients respond poorly to ICS therapy 1. We used alveolar macrophages to identify asthma patients with in vitro insensitivity to corticosteroids. Corticosteroid insensitive patients were most prevalent within the GINA 3/4 group. The combination of a p38 MAPK inhibitor and corticosteroid caused greater anti-inflammatory effects compared with either drug alone. The interaction between these drugs was synergistic, with a greater synergistic effect in the corticosteroid insensitive population.

p38 MAPK inhibitors have been in clinical development for inflammatory diseases for a number of years. One of the problems with this drug class is that the therapeutic index has been limited by side effects. There are few anti-inflammatory treatment options for asthma patients who are symptomatic despite ICS therapy. Our results show that the synergistic interactions between p38 MAPK inhibitors and corticosteroids were greatest in individuals who displayed cellular insensitivity to corticosteroids. This suggests that the optimum use of p38 MAPK inhibitors in clinical practice is to enhance corticosteroid effects in patients on corticosteroids. This approach might optimize the therapeutic index.

Bhavsar et al. showed a reduced effectiveness of corticosteroids on LPS stimulated IL-1β, IL-6 and MIP-1α secretion in alveolar macrophages and peripheral blood mononuclear cells (PBMCs) from severe compared with non-severe asthma patients 15. Here we observed similar findings, as alveolar macrophages from GINA 3/4 asthma patients were less sensitive to corticosteroid inhibition of LPS induced TNFα, IL-6 and CXCL-8 production. The relative corticosteroid insensitivity of the GINA 3/4 patients was driven by a subgroup of individuals with a reduced response to this drug. A personalized medicine approach has been advocated in asthma, so that specific therapies are tailored towards individual patients with discrete characteristics 11. Our alveolar macrophage results highlight the potential to identify corticosteroid insensitive asthma patients using an in vitro assay. This may be used as a biomarker to direct additional anti-inflammatory therapy, such as p38 MAPK inhibitors.

Mercado et al. also showed reduced inhibition of CXCL-8 by dexamethasone in PBMCs from severe asthma patients compared with controls 16,19. Furthermore, it was shown that a greater proportion of patients with severe asthma were corticosteroid insensitive (33%) compared with controls (10%) 16. This is consistent with our data showing an increased proportion of macrophages to be corticosteroid insensitive in the GINA 3/4 group compared with GINA 1 and 2.

We used cut-off levels for corticosteroid insensitivity based on visualization of dose−response curves. This approach has been previously used 16,20, as there is no accepted value that should be used. Corticosteroid sensitivity is a continuous variable, rather than simply categorical based on fully sensitive (100% inhibition) and fully resistant (0% inhibition). The cut-off levels used here may not accurately represent the spread seen in the wider population, but served to demonstrate the concept that some patients have lower corticosteroid sensitivity than others.

The anti-inflammatory effects of BIRB-796 alone were lower than dexamethasone in corticosteroid sensitive patients. This agrees with previous publications showing that corticosteroids have a greater effect than p38 MAPK inhibitors on macrophage cytokine production 13–15. However, in corticosteroid insensitive asthma patients, the relative effects of these drugs are altered due to the reduced corticosteroid effect. Similarly, Ratcliffe & Dougall reported that corticosteroids have a poor effect in a subset of COPD patients, and that BIRB-796 was as effective as corticosteroids in this subset 20.

The effects of a p38 MAPK inhibitor combined with a corticosteroid in vitro have previously been studied in macrophages and PBMCs from asthma patients 15,19 using single concentrations, not full dose−response curves. These studies reported inhibitory effects that were possibly more than additive 15. The analysis of full dose−response curves reported here demonstrates a synergistic interaction between these drugs. As with our previous data in COPD macrophages 14, we showed that this synergistic effect enhanced the efficacy and had a dose-sparing effect on corticosteroid treatment. While this effect was observed in all patient groups, importantly the combination effect was greatest in the corticosteroid insensitive patients.

The potential mechanisms of corticosteroid insensitivity include up-regulation of NF-κB signalling, increased oxidative stress 1 and reduction in glucocorticoid receptor (GR) function due to phosphorylation by p38 MAPK or c-Jun N-terminal kinases (JNK) 16,19,21. Phosphorylation of the GR at serine 226 causes nuclear export of the receptor 22. PBMCs from patients with severe asthma showed a reduction of GR nuclear translocation associated with increased phosphorylation of the GR at serine 226 compared towith healthy volunteers 16. Furthermore, IL-2 and IL-4 caused p38 MAPK dependent phosphorylation of the GR at serine 226, indicating that inflammatory cytokine signalling can decrease corticosteroid effectiveness through p38 MAPK activation. p38 MAPK activation is increased in alveolar macrophages, airway smooth muscle and airway epithelial cells from severe asthma patients compared with controls 7,23. Increased p38 MAPK activity in severe asthma may therefore be responsible for reduced corticosteroid effects in asthma due to GR phosphorylation at serine 226.

The effects of corticosteroids on alveolar macrophages varies between cytokines 13,14,18,24–26. This is due to differences in the signalling pathways involved in the production of different cytokines. These signalling pathways may show varying degrees of sensitivity to corticosteroids. We observed a similar pattern here, with variation in the corticosteroid effects between cytokines.

There are some limitations to this study. There were differences between groups in some clinical parameters, such as ACQ score which showed that GINA 3/4 patients had the lowest degree of asthma control. However, the limited samples sizes of the groups made it difficult to show statistical significance for differences between groups for other clinical parameters such as lung function. There is a possibility that GINA 3/4 patients had relatively less corticosteroid sensitive alveolar macrophages due to previous ICS treatment, but this is unlikely as there was no significant difference in the average dose of ICS between the corticosteroid sensitive and insensitive patients. Another limitation was that we were unable to analyze differences between corticosteroid insensitive and sensitive patients for dose sparing and synergistic effects for CXCL8 production due to the high variance in the data. However, TNFα and IL-6 had less variance and, therefore, were better bioassays for evaluating the effects of combination treatment.

An important question is whether the methods used here can be used in clinical practice to guide pharmacotherapy. Bronchoscopy is a relatively invasive procedure and it would be preferable if the assays reported here could be performed using less invasive sampling methods, such as induced sputum to obtain macrophages. Some methodology work would be required to determine whether sputum macrophages are able to demonstrate corticosteroid insensitivity as well as BAL macrophages. Similarly, the use of blood derived monocytes for this purpose should be explored. Nevertheless, bronchoscopy is routinely used in clinical practice and the methods reported here could be used to evaluate corticosteroid sensitivity on an individual basis.

In summary, we have shown that a subset of asthma patients have corticosteroid insensitive alveolar macrophages. p38 MAPK inhibitors synergistically enhanced the efficacy of corticosteroids. This effect was greater in patients with corticosteroid insensitive cells. Identifying subsets of asthma patients with corticosteroid insensitive characteristics may be the optimum way to develop p38 MAPK inhibitors for the treatment of asthma.

Conflicts of Interests

This work was funded by AstraZeneca.

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and SL, NK, GB and JA declare no support from any organization for the submitted work. CH and DS had support from AstraZeneca for the submitted work. SL, NK, GB and JA declare no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years. CH is currently an employee of AstraZeneca.

DS has received sponsorship to attend international meetings, honoraria for lecturing or attending advisory boards and research grants from various pharmaceutical companies including Almirall, AstraZeneca, Boehringer Ingelheim, Chiesi, CIPLA, Forest, Genetech, GlaxoSmithKline, Merck, Novartis, Pfizer and Takeda in the previous 3 years. SL, CH, NK, GB and JA declare no other relationships or activities that could appear to have influenced the submitted work

We thank Keith Wregget for his intellectual support.

Contributors

Participated in research design: Lea, Harbron, Armstrong and Singh

Conducted experiments: Lea, Khan, Booth and Armstrong

Contributed new reagents or analytic tools: Harbron

Performed data analysis: Lea and Harbron

Wrote or contributed to the writing of the manuscript: Lea, Harbron and Singh

Supporting Information

Additional Supporting Information may be found in the online version of this article at the publisher's web-site: