Airway function , inflammation and regulatory T cell function in subjects in asthma remission

Centre de recherche, Institut universitaire de cardiologie et de pneumologie de Québec, Québec, Québec Correspondence: Dr Louis-Philippe Boulet, Institut universitaire de cardiologie et de pneumologie de Québec, 2725, Chemin Sainte-Foy, Québec, Québec G1V 4G5. Telephone 418-656-4747, fax 418-656-4762, e-mail lpboulet@med.ulaval.ca Asthma is characterized by variable airway obstruction and hyperresponsiveness (AHR), which is attributed to airway inflammation and structural changes (1,2). We and others have provided evidence that airway inflammation and remodelling begin many years before the onset of asthma symptoms; however, once asthma has been diagnosed, it is considered to persist for life in the majority of patients (3-5). Although relatively rare in adults, some patients may ‘grow out’ of asthma and, will therefore, be considered to be in asthma remission. There is evidence that even if symptoms have remitted in former mildto-moderate asthmatic patients, residual airway inflammation on bronchoalveolar lavage or bronchial biopsies (6-8) can be observed in many subjects. Van den Toorn LM et al (7,8) and Warke et al (9) demonstrated persistant subepithelial fibrosis in subjects with symptomatic remission of asthma compared with healthy controls. The mechanisms involved in symptomatic or complete remission of asthma, particularly in adulthood, are not well understood and need further investigation. The study of this phenomenon could help in determining how to possibly induce the remission of asthma. There is, therefore, a need to characterize these patients in so-called ‘asthma remission’ to determine the factors associated with remission and with recurrence of asthma in adulthood. In the present study, we compared pulmonary function, airway response to methacholine (MIT) and adenosine monophosphate (AMP), perception of respiratory symptoms, immunoglobulin (Ig) E levels, airway inflammation on induced sputum, and T cell function in patients in symptomatic or complete asthma remission and mild asthmatic patients.

confirmed by a physician diagnosis and follow-up, in addition to previous asthma symptoms and asthma medication use, were enrolled (Table 1).Six patients also had a positive methacholine test during their asthma period dating back to before the occurrence of remission.Furthermore, 16 asthmatic subjects in symptomatic remission for more than two years, but still presenting hyper-reactive airways (11,12), 29 subjects with mild asthma not using anti-inflammatory agents and 15 healthy controls with normal airway responsiveness (PC 20 >16 mg/mL) were studied.All subjects provided informed written consent; the protocol was reviewed by the institutional ethics committee (Clinical Trial registra- tion number: NCT 00526019).
The primary end point of the study was airway inflammation, assessed by airway responsiveness to AMP, and induced sputum eosinophil and neutrophil counts.Secondary end points were the following: airway response to methacholine (in symptomatic remission and current mild asthma compared with normal controls) and baseline FEV 1 , blood IgE levels and blood eosinophils, diurnal variation in peak expiratory flows (PEF) and the profile of peripheral blood regulatory T cells (Tregs).

Procedures
All patients visited the laboratory at study entry and one week later.They completed a respiratory questionnaire based on the European Community Respiratory Health Survey Questionnaire (13).Patients underwent physical examination, skin prick tests, baseline spirometry and a methacholine challenge according to American Thoracic Society criteria (14) using the method described by Juniper et al (15), with additional measures of modified Borg perception scores for breathlessness and chest tightness.Complete blood count and serum IgE levels were performed on blood samples.Sputum was induced using the method described by Pin et al (16) and modified by Pizzichini et al (17) by inhalation of hypertonic saline.
On the second visit, after baseline spirometry, an AMP inhalation test was performed using the 2 min tidal breathing method described by Cockcroft et al (18) for histamine and applied to AMP by Oosterhoff et al (19).

induced sputum processing
Sputum was processed within 2 h of induction.Briefly, mucus plugs were selected from saliva and treated with four times their volume of dithiothreitol.An equal volume of 1×Dulbecco's phosphate-buffered saline was then added and the suspension was filtered.Total cell count and viability were determined using the trypan blue exclusion method.Slides were prepared and stained with Diff-Quik (Dade Diagnostics Inc, USA) for differential cell count.

Regulatory T cell isolation
To study regulatory T cell function, eight healthy nonatopic controls, nine subjects with current mild asthma and nine asthmatic subjects in complete remission of asthma underwent peripheral blood sampling (150 mL) out of the allergy period.Peripheral blood mononuclear cells were isolated using lymphocyte separation media (Wisent, St-Bruno, Canada) density gradient centrifugation.Monocytes were separated from total lymphocytes by adhesion for 1 h at 37°C.CD4 + CD25 − and CD4 + CD25 + CD127 − cells were isolated using a commercially available CD4 + T cell isolation Kit II and CD4 + CD25 + CD127dim/ − Regulatory T cell Isolation Kit according to the manufacturer's instructions (Miltenyi Biotec, USA).The purity of the CD4 + CD25highFoxp3 + CD127 − cells was >90%.

Statistical analysis
Results were expressed as mean ± SD, geometric means for PC 20 methacholine and AMP, or percentages according to the distribution of data.PC 20 values for methacholine and AMP were log transformed for analysis.For statistical comparisons, a PC 20 >16 mg/mL was considered to be equal to 16 mg/mL.The subjects' demographic characteristics were analyzed using Fisher's exact test or one-way ANOVA for categorical and qualitative data, respectively.Posteriori comparisons were performed using Tukey's multiple comparison test.Relationships between pulmonary function parameters and airway inflammation (sputum eosinophils) were measured using Pearson's or Spearman's correlation coefficients.The univariate normality assumption was verified with the Shapiro-Wilk test.The Brown and Forsythe's variation of Levene's test statistic were used to verify the homogeneity of variances.P≤0.05 was considered to be statistically significant.All analyses were performed using SAS version 9.2 (SAS Institute Inc, USA).

Results
Of the 92 subjects assessed for eligibility, six did not meet the inclusion criteria.A study flow diagram is presented in Figure 1.Table 2 summarizes the demographic characteristics of the subjects at study enrollment.Atopy was more prevalent in patients with previous or current asthma diagnosis than in controls.Family history of asthma and/or allergy or their absence did not appear to be a predictor of asthma remission nor of symptomatic remission (P=0.31).
The age at time of asthma diagnosis was not significantly different among the groups (P=0.47 [Figure 2]).Patients in the complete remission group were generally younger than those of the symptomatic remission group at time of asthma remission (P=0.03).The duration of asthma was not significantly different between asthma remission and symptomatic remission groups.
When questionned about the circumstances surrounding asthma remission and the possible reasons why they became asymptomatic, patients (complete remission, n/symptomatic remission, n) mentioned the following:

MiT and airway response to AMP
The mean MIT was similar in control and asthma remission groups (both PC 20 MIT >16 mg/mL [inclusion criterion]).Patients in the asthma group demonstrated significant hyper-responsiveness (geometric mean 1.1mg/mL; P<0.0001) compared with control and asthma remission groups (Figure 3).The symptomatic remission of asthma group (3.5 mg/mL) was significantly different from the control (P=0.0006) and the complete remission groups (P<0.0001), and from the asthma group (P=0.0003).
Mean airway responsiveness to AMP, expressed as the PC 20 AMP (the provocative dose inducing a 20% fall in FEV 1 ), was similar in control (312 mg/mL) and asthma remission groups (287 mg/mL); patients in the asthma group demonstrated airway hyper-responsiveness, with a mean PC 20 AMP of 25 mg/mL (P<0.0001compared with remission and control groups).The symptomatic remission of asthma group had a mean PC 20 of 106 mg/mL, significantly different from the asthma (P=0.0005) and the complete remission groups (P=0.02),but not statistically different from the control group (P=0.07)(Figure 3).

Correlation of airway response to AMP and MiT
MIT was correlated to AMP response in patients with asthma (r=0.6;P=0.001) and symptomatic remission (r=0.67;P<0.0001).Correlations for patients in complete remission and for controls could not be calculated due to the inclusion criteria of a PC 20 methacholine >16 mg/mL.

FEv 1 and FvC
Baseline FEV 1 and FVC, and FEV 1 /FVC ratio before methacholine inhalation tests varied slightly but not significantly among the groups (Table 3).Borg scores for perception of respiratory symptoms at 20% fall of FEV 1 on methacholine inhalation test were not significantly different in the groups in symptomatic remission of asthma and asthma for breathlesness (median [95% CI]: symptomatic remission: 0 (0 to 1.6) and asthma: 0.5 (0.5 to 1.7; P=0.27) and chest tightness (0.5 [0 to 2.0] and 1.2 [0.8 to 2.1], respectively; P=0.65).

Diurnal variation in PEF
Daily measurements of morning (AM) and afternooon (PM) PEF over one week are summarized in Table 3.The magnitude of variations of AM and PM PEF was not significantly different between groups (P=0.54 and P=0.48, respectively).

inflammatory cells in induced sputum
Evaluation of inflammation was performed using induced sputum.Total cell number was not significantly different among the groups (ie, P>0.05, [Table 4]).There were no significant differences in neutrophil or eosinophil percentages among the groups.

Correlation with airway response to AMP
The patients in the symptomatic remission and current mild asthma groups who had negative AMP challenges had lower blood IgE levels and lower blood and sputum eosinophils; PC 20 AMP correlated with blood IgE levels (r=−0.347;P<0.0001), blood eosinophils (r=−0.27;P=0.002) and sputum eosinophils ( r=−0.58;P<0.0001).

Treg function
There were no significant differences in the percentage of CD4 + CD25highFoxp3 + CD127 − cells among controls, asthma and complete asthma remission groups (9.1±1.8%,7.8±0.6%and 5.2±0.6%,respectively).The suppressive function of Tregs was analysed by co-culture of these Tregs with CD4 + CD25 − effector cells.The capacity of Tregs to suppress proliferation of effector T cells was higher in healthy controls (83.1±3.0% of inhibition) than in the asthma group (51.3±4.9%;P=0.0006) or the complete remission of asthma group (34.0±10.9%;P=0.0002).There were no differences in the   suppressive function of Tregs between the complete asthma remission and asthma group (Figure 5).

DiSCuSSiOn
In the present study, patients in asthma remission presented features similar to controls with normal airway hyper-responsiveness. Remission was associated with shorter disease duration.Patients in symptomatic asthma remission appeared to be in an intermediate state between complete remission of asthma and symptomatic asthma.Asthma remission may be observed in adolescence, but occurs less frequently in adulthood.There appears to be a continuum of remission, with patients having no symptoms but still evidence of airway hyper-responsiveness or inflammation.In this regard, persistent airway obstruction, increased PEF variability and AHR to methacholine or cold air have been observed during symptomatic asthma remission (20)(21)(22)(23)(24).
Sears et al (25) investigated factors predicting persistence or relapse of wheezing in a New Zealand birth cohort, and showed that sensitization to house dust mites, airway hyper-responsiveness, female sex, smoking, and early age at onset were factors involved in the persistence of wheezing or relapse.Furthermore, Taylor et al (26) reported data from the latter cohort, showing that one of three asthmatic children was considered to be in remission at 18 years of age, although relapses were not easily predicted.Other studies showed that some children may outgrow their asthma at puberty (27,28), but 30% to 80% redevelop symptomatic asthma in adulthood (28)(29)(30)(31)(32). Regarding the factors associated with a recurrence of asthma symptoms, development of chronic airflow limitation in childhood and persistent airway inflammation have been proposed (6)(7)(8)33,34).In the present study, airway inflammation evaluated by induced sputum showed no significant difference between asthma and symptomatic or complete asthma remission.This is consistent with the study by Van den Toorn et al (35), who found that residual inflammation exists in asthma remission.There is evidence that even if symptoms have remitted in former mild-to-moderate asthmatic patients, residual airway inflammation often persists.Pumputiene et al (6) reported increased airway T cell and eosinophil activity (assessed by serum interleukin-2 and eosinophil cationic protein levels) in atopic and nonatopic children despite the absence of asthma symptoms.Warke et al (9) showed that eosinophil numbers were raised in the bronchoalveolar lavage fluid of atopic children who had apparently outgrown their asthma compared with healthy controls.However, there was no relationship between duration of remission and degree of airway eosinophilia, but the authors suggested that the latter could be a risk factor for future relapses (9).Although there were a slightly higher number of inflammatory cells in asthmatic subjects in our study, globally, the number of these cells was not different among the groups.This was also shown in a recent study by Broekema et al (36), in which asthma remission and asthma without an accelerated decline in lung function demonstrated a similar sputum inflammatory profile.
Tregs contribute to the maintenance of immunological tolerance in the airways, evidenced by impaired function of these cells in asthma and allergy (37).Thus, peripheral blood CD4 + CD25 + T cells derived from atopic donors were shown to be suppressive in allergen-stimulated cultures.This suppressive activity was similar to healthy controls.Ling et al (38) showed that FOXP3 Treg cells were defective in atopic individuals.This defect was particularly marked during seasonal pollen exposure.Verhagen et al (39) reported an absence of CD4 + CD25 + FOXP3 + cells in allergic dermatitis, suggesting a dysregulated control of inflammation.We evaluated the function of these cells in a subgroup of subjects in complete asthma remission and compared them with asthmatic patients and controls.Although the number of Tregs was similar in all groups, we found that subjects in complete remission still had decreased suppressive function of Tregs.In asthmatic subjects, this is a feature that may indicate an increased risk of future recurrence of the disease.
To characterize subjects in asthma remission, Vonk et al (40) reported outcomes in a cohort of 119 allergic asthmatic children initially evaluated at five to 14 years of age and followed for a mean of 30 years.In this group, complete remission of asthma was present only  in a small subset of asthmatic patients, while one-half of the subjects showed symptomatic remission.A higher FEV 1 in childhood and improvement in FEV 1 over the years were associated with both complete and symptomatic asthma remission when older (11,40).Panhuysen et al (41) also investigated the outcome of asthma in 181 patients 13 to 44 years of age.When retested 25 years later, 40% reported no respiratory symptoms, 25% showed an FEV 1 >90% predicted while 21% showed no AHR.Complete remission of asthma after 25 years was associated with younger age and less severe airway obstruction when compared with the initial evaluation.Absence of AHR was associated with younger age, a higher FEV 1 , a shorter untreated period between onset of asthma symptoms and treatment from the first visit, and a lower total serum IgE level.Neither sex nor atopy were significant determinants of the outcome for both asthma and AHR.Results of the present study support the hypothesis that milder disease and earlier intervention are associated with a beneficial outcome (41).Finally, in a study by De Marco et al (42) involving 18,156 subjects zero to 44 years of age who attended the clinical stage of the European Community Respiratory Health Survey, family history of asthma or allergy was associated with a reduced possibility of remission throughout life, while early contact with older children was associated with an increased possibility.
Indirect challenges may more closely reflect acute changes in airway inflammation (43)(44)(45)(46).Agents such as adenosine 5′-monophosphate (AMP) act through the activation of bronchial mast cells; thus, 5′AMP could better reflect airway inflammation than methacholine response (42).In the present study, airway responsiveness to AMP decreased progressively from asthma to asymptomatic AHR, and even further in the complete remission of asthma group, with the latter group exhibiting an AMP response similar to controls.
We do not know whether patients in symptomatic or complete asthma remission who participated in the present study will experience a recurrence of the disease.They had many features similar to controls: baseline FEV 1, and FVC and FEV 1 /FVC ratio, low PEF variations, and similar levels of blood IgE and blood eosinophils.Family history of asthma and/or allergy was similar in all groups.However, although there were few indications of airway inflammation, atopy, which is suspected to be a factor predisposing to asthma recurrence (47), was more prevalent in the remission groups than in the controls.Moreover, evaluation of Treg function in the complete asthma remission group showed that suppressive function of these cells was reduced and was comparable with the symptomatic asthma group, suggesting that the mechanism of immune regulation in these patients is still abnormal and may predispose to disease recurrence.
Poor perception of symptoms has been implicated in asymptomatic subjects with AHR, although this is an unlikely explanation (48).However, perception during symptomatic asthma remission appears to be different.Van den Toorn et al (8) investigated perception scores of dyspnea after a 20% fall in FEV 1 during methacholine and AMPinduced bronchoconstriction challenges, and concluded that they were not different between subjects in symptomatic remission and subjects with symptomatic asthma (as in our study), thereby making blunted dyspnea perception an unlikely explanation for the stated discrepancy.
We previously observed that subjects with asymptomatic AHR had a lesser degree of airway inflammation and remodelling on bronchial biopsies compared with symptomatic asthma subjects -features that became similar to patients with asymptomatic asthma when they subsequently developed symptomatic asthma (3).These data suggest that combined increases in airway inflammation and remodelling may reach a threshold supporting the development of respiratory symptoms.Consequently, the opposite may be true in asthma remission, with asthma symptoms decreasing with reduced airway inflammation and remodelling.

COnCLuSiOn
We showed that asthma remission presents as a continuum, some previously asthmatic subjects being in asymptomatic remission but showing a persistence of both AHR and airway inflammation while others (so-called 'complete asthma remission') have neither.However, even in complete remission of asthma, Treg function appears to remain abnormal compared with normal controls and may predispose to disease recurrence.

ACknOWLEDgEMEnTS: This study was supported by
GlaxoSmithKline Canada (CRIF-FRIC) and by a local funds program.

Figure 2 )
Figure 2) Characteristics of asthma and remission.Age at time of diagnosis, duration of asthma, age at time of asthma remission and duration of remission are compared among groups.Closed circles represent atopic patients, open circles represent nonatopic patients

Figure
Figure3) Airway responsiveness to methacholine (MIT) and adenosine monophosphate (AMP).Airway response to challenges was not significantly different within groups.Patients in symptomatic asthma remission were significantly less responsive to MIT and AMP than asthma patients, and significantly more responsive to both challenges than patients in complete asthma remission, whose responsiveness was similar to that of controls.Bar represents mean.PC 20 Provocative response to methacholine causing a 20% fall in forced expiratory volume in 1 s

TaBle 1 Characteristics of subjects in complete asthma remission during asthma period Subject asthma period Physician asthma symptoms asthma medication PC 20 methacholine † , mg/ml emergency room visit for asthma Diagnosis Follow-up*
† When available.B Breathlessness; Bd Inhaled bronchodilator; CT Chest tightness; C Cough; ICS Inhaled corticosteroids; PC 20 Provocative concentration of methacholine causing a 20% fall in forced expiratory volume in 1 s; Ph Phlegm production; W Wheezing

TaBle 2
Data presented as n or n/n, unless otherwise indicated.*χ 2 or Fisher's exact test categorical data, ANOVA for quantitative data.BD Bronchodilator; ex Exsmokers; NS Never smokers; prn As needed; S Smokers