INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cyclosporin A (CsA) has been previously shown to improve lung function, decrease exacerbations, and decrease oral corticosteroid requirements in chronic severe asthmatics (1, 2). However, the mechanism of these effects in asthma is unclear. In a recent double-blind placebo-controlled study, Sihra and coworkers demonstrated attenuation of the allergen-induced late (LAR), but not the early asthmatic reaction (EAR), by CsA. In addition CsA inhibited the associated increase in peripheral blood eosinophils after allergen challenge (3).

Inhaled allergen challenge in atopic asthmatic subjects provokes an early asthmatic reaction, which in some subjects is followed by a sustained, delayed-in-time LAR (4). The LAR has provided a useful model for studying the inflammatory changes in chronic asthma whereas the EAR is believed to reflect immediate IgE-dependent mast cell events. As with chronic asthma, the LAR is associated with increases in peripheral blood (5) and bronchial (6, 7) eosinophils as well as the numbers of interleukin-5 (IL-5) and granulocyte macrophage colony-stimulating factor (GM-CSF) messenger RNA- positive (mRNA⁺) cells in bronchoalveolar lavage fluid (BALF) (8). Eotaxin protein⁺ cells were also elevated in induced sputum after allergen challenge (9). Many of these variables correlated with the magnitude of the LAR.

In view of our previously reported study on the effect of CsA on the LAR and peripheral blood eosinophil count after allergen challenge (3), we hypothesized that attenuation of the LAR by CsA would be associated with inhibition of eosinophils, and eosinophil-associated cytokines or chemokines, in bronchial biopsies and/or bronchoalveolar lavage (BAL). We also postulated that basophil infiltration of atopic asthmatic airways would be inhibited by CsA, as basophils and eosinophils originate from a common myeloid lineage (10), and share expression of CCR3 (11), IL-5R (12), and ₄₁ (very late appearing antigen-4 [VLA-4]) (13).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-four atopic individuals with a history of wheezy breathlessness and rhinoconjunctivitis on exposure to timothy grass pollen, house dust mite, or cat dander were recruited from the London Chest Hospital or the Royal Brompton Hospital, London, UK. All subjects had an FEV₁ > 70% predicted and required intermittent use of inhaled ₂-agonists (for control of symptoms). All patients had documented reversible airways obstruction either spontaneously or after inhalation of -agonists. None had received orally administered corticosteroids in the 6 mo, or inhaled steroids in the 2 wk, preceding the study. Patients with seasonal symptoms were studied outside the pollen season of the United Kingdom. None were smokers or had a history compatible with respiratory infection in the 3 wk before or during the study.

Written informed consent was obtained from all subjects and the study was approved by both of the two Hospitals' Ethics Committees.

Study Design

The study was conducted over 4 wk. At an initial assessment a full clinical history was taken and examination performed. Atopy was defined by positive skin-prick test (wheal diameter of over 3 mm after subtraction of the negative control to one or more aeroallergen extracts [house dust mite, cat dander, and timothy grass pollen, Soluprick; ALK, Hørsholm, Denmark]). A full blood count, urea and electrolytes, clotting profile, and hepatitis B serology were analyzed. Following this, subjects underwent allergen challenge. If both an EAR and LAR were demonstrated, the subject was selected for randomization. After 3 wk patients had a baseline bronchoscopy (Monday 9:00 A.M.), allergen challenge (Wednesday 9:00 A.M.), and then repeat bronchoscopy (Thursday 9:00 A.M.). In a double-blind manner each subject received either CsA (500 mg, n = 13) or placebo (n = 11), 12 h and just before allergen challenge. Trough whole blood CsA concentrations (12 h postdose) were determined with an ethylenediamine tetraacetic acid (EDTA)-treated blood sample. Urine dipstick and blood pressure measurements were performed at all visits. Female participants were advised not to become pregnant and to use effective contraception during the study. Pregnancy tests were performed at the beginning and end of the trial period.

Allergen Challenge

Sixty-seven subjects were screened by allergen challenge to yield the 24 subjects who were randomized. The allergen used for challenge was selected on the basis of the clinical history and a positive skin-prick test. On the screening day an allergen challenge protocol was standardized using a modification described by Chai and coworkers (14). Increasing strength serial doubling dilutions of allergen extract (Aquagen, ALK) commencing at 200 squillion unit (SQ-U) (manufacturer's [ALK] measurement) were inhaled at 30-min intervals under identical conditions using a breath-activated dosimeter (Mefar; S. P. A., Borezzo, Milan, Italy). After the FEV₁ had decreased by 20% from the baseline (saline challenge) value, further FEV₁ readings were performed at 10-min intervals for an hour, followed by 30-min intervals for a further 7 h. At the end of the observation period patients were allowed home if the FEV₁ had returned to within 20% of baseline values. If the FEV₁ was less than 20% of baseline, nebulized salbutamol was administered as required until discharge. Patients were selected to continue the study if they had demonstrated an EAR of at least 20% (0-2 h postchallenge) and a LAR of at least 15% (2 to 8 h postchallenge). The dose of allergen required to produce a LAR of 15% was predicted from the dose-response curve and magnitude of the LAR on the screening day. On the second challenge day, the cumulative dose was administered as a bolus using the same nebulizer, dosimeter, and dosimeter settings. The FEV₁ was monitored at the same time points as the baseline challenge day.

Fibreoptic Bronchoscopy (FOB)

Each subject received nebulized salbutamol (2.5 mg) and atropine (600 µg) intravenously 15 min before FOB. Sedation was achieved with intravenously administered diazepam and supplemental oxygen was given throughout the procedure. Patient oxygenation was monitored by pulse oximetry. FOB was performed by the same operator for all subjects. After 1% lignocaine spray was applied to the nose and throat of the subject, the bronchoscope (Olympus mode OSE with a 2.2-mm-width biopsy channel) (Olympus Corp., Lake Success, NY) was introduced. Local anesthesia of the larynx was produced with topical 4% lignocaine, and 2% lignocaine was used below the vocal cords. After inspection of the bronchial tree, the bronchoscope tip was wedged in a segmental or subsegmental division of either the middle or lingula lobes (sides chosen for lavage or biopsies were randomized and the contralateral segments used for the repeat bronchoscopic study). Two 60-ml aliquots of warmed, sterile saline solution were introduced, and returned fluid was collected by gentle machine suction, followed by a further 60-ml aliquot. Aspirated fluid was collected into a sterile siliconized glass bottle and cooled gradually to 4° C. Endobronchial biopsies were then taken from either the right middle or lower lobe on segmental carinae. After the procedure the patient was given an additional 2.5 mg of salbutamol and observed on the ward for at least 4 h.

Biopsy Samples

One biopsy was taken into phosphate-buffered saline (PBS) and then frozen within 30 min using OCT as before. Sections were cut as previously described. They were air-dried for 30 min, fixed in acetone/ methanol (60/40, 10 min) and further air-dried for 1 h. They were then wrapped in pairs in foil and stored at 80° C. Three further biopsies were taken for archive use.

BAL

The volume of returned fluid from the BAL was measured. The BAL was passed through sterile gauze to remove any mucus. Cells were pelleted by centrifugation at 300 g for 10 min at 4° C. The supernatant fluid was carefully decanted and stored at 80° C in aliquots. The cell pellet was washed in RPMI 1640 (Sigma-Aldrich Co. Ltd.) medium, then resuspended in 10 ml of RPMI. A total cell count was performed with a Neubauer hemocytometer (Merck, Lutterworth, UK) and Kimura stain. The cells, at a concentration of 1 × 10⁶ (90 µl) were then used to make cytocentrifuge preparations with a Shandon 2 cytospin (Shandon Southern Products Ltd., Runcorn, Cheshire, UK) at 800 rpm for 5 min. Cytospins were air-dried and either frozen wrapped in foil for eosinophil enumeration by chromotrope or fixed in paraformaldehyde (as before) for 30 min followed by two washes in sucrose (as before) for 30 min and a quick rinse in PBS and then incubated at 37° C for 12 h before being frozen in the presence of silica gel.

Immunohistochemistry

Immunohistochemistry was performed in tissue sections. Monoclonal antibody staining was detected by modifications of the alkaline phosphatase-antialkaline phosphatase (APAAP) method. Rabbit anti-mouse IgG₁ and APAAP were purchased from Dako (High Wycombe, UK). Normal human serum was used to prevent nonspecific binding of the second and third layer antibodies where and as necessary. The monoclonal antibodies used were anti-CD4, anti-CD8, anti-CD25, anti-CD68 (Dako, High Wycombe, UK) and EG2 (Pharmacia Upjohn, Milton Keynes, Bucks., UK) recognizing the cleaved form of eosinophil cationic protein (ECP), representing activated eosinophils. Monoclonal anti-human antibody 2G6 to eotaxin (Leukosite, Cambridge, MA) was used as stated (15) with minor modifications. BB1, a basophil-specific monoclonal antibody was a kind gift from Dr. Andrew Walls (University of Southampton, Southampton, UK) and was used as previously described (16). Sections were counted blind by a single observer (per antibody). Positive cells stained red after developing with the alkaline phosphatase substrate Fast Red. The numbers of positively stained cells were counted in a zone 220 µm deep as defined by a squared eyepiece graticule (Olympus Corp.) along the entire length of the epithelial basement membrane of each section. The cell counts were expressed as the number per unit length of basement membrane (positive cells per millimeter of basement membrane).

In Situ Hybridization (ISH)

ISH was performed as previously described. [³⁵S]-uridine triphosphate (UTP)-labeled RNA probes were prepared from complementary DNA (cDNA) for IL-5 and granulocyte-macrophage colony-stimulating factor (GM-CSF). To avoid nonspecific binding of ³⁵S-labeled RNA probes, incubation with N-ethyl maleimide, iodoacetamide, and triethanolamine was included in prehybridization steps, and dithiothreitol (100 nM) was included in the hybridization mixture (Sigma, Poole, UK). Counting was performed without knowledge of treatment groups, and hybridization was assessed by counting positive cells per 1,000 total BAL cells for both cytokines.

IL-5 ELISA

Concentrations of IL-5 in unconcentrated BALF were measured in the laboratory of Dr. D. Huston using murine monoclonal antibody as previously described in detail (17).

Statistical Analysis

The magnitude of the EAR and LAR were defined as the area under the curve (AUC) of the percentage change from baseline (or saline) for two periods: zero to 1 h and 2 to 8 h after challenge, respectively (18). The AUC values were calculated by the trapezoid method.

Nonparametric statistics were used to compare the effect of CsA with placebo on the parameters we had chosen to evaluate. Wilcoxon matched pairs test was used to examine changes within groups and the Mann-Whitney U test for between-group variation. Correlation coefficients were calculated by Spearman's method with correction for tied values. Statistics were performed using Minitab (Minitab for Windows, Minitab Release 9.2; Minitab Inc., State College, PA) software. Values of p < 0.05 were considered significant. Some of the baseline biopsies (preplacebo or pre-CsA) were also used in a separate study (16).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient characteristics are described in Table 1. Both groups were well matched for age and baseline FEV₁. Thirteen patients received CsA and 11 received placebo. In the CsA group the mean trough whole blood CsA concentration was 296.5 µg/L (SEM, 73.8). Paired data for analysis was obtained from at least 23 patients for immunohistochemistry and all 24 subjects for ISH. The performance of allergen inhalation challenges and fibreoptic bronchoscopic procedures was uneventful. In the CsA group, three patients felt unwell after the first dose of treatment. This took the form of generalized aches and pains associated with a sensation of nausea and in one case "burning." No patient required additional medication. There were no reported side effects in the placebo group. No subject had a decrease in FEV₁ in response to saline challenge.

The LAR was significantly attenuated in the treatment group when compared with the placebo group (p = 0.048) but there were no significant differences in the magnitude of the EAR (Figure 1).

View larger version (18K): [in this window] [in a new window]   Figure 1.   Effects of CsA and placebo on mean (SE) percentage changes in FEV1 from baseline (postsaline) values, after allergen challenge (Time 0). The statistical analysis is described in METHODS.

Sufficient amounts of bronchial biopsy material, BAL cells, and BALF were obtained to perform all cellular and protein markers in all subjects. CsA significantly inhibited the allergen-induced increases in EG2⁺ eosinophils (Figure 2) and eotaxin⁺ cells (Figure 3) in bronchial biopsies and IL-5 mRNA (Figure 4) and GM-CSF mRNA (Figure 5) in BAL (between-group p values of p = 0.227, p = 0.038, p = 0.02, and p = 0.0028, respectively). For some measurements (i.e., IL-5 [Figure 6] and eosinophils in BAL, and CD4⁺ cells in bronchial biopsies) the within-group values showed significant increases on the placebo, but not the CsA day. However, the between-group values were nonsignificant (Table 2). CsA had no effect on the allergen-induced changes in basophils (Figure 7). There were no significant changes in the numbers of CD8 or CD68+ cells with either placebo or CsA (Table 1).

View larger version (19K): [in this window] [in a new window]   Figure 2.   The effect of CsA on the numbers of EG2+ eosinophils in bronchial biopsies pre- and postallergen challenge. = 0.0227.

View larger version (18K): [in this window] [in a new window]   Figure 3.   The effect of CsA on the numbers of eotaxin+ cells in bronchial biopsies pre- and postallergen challenge. = 0.038.

View larger version (17K): [in this window] [in a new window]   Figure 4.   The effect of CsA on IL-5 mRNA+ cells in BAL pre- and postallergen challenge. = 0.02.

View larger version (16K): [in this window] [in a new window]   Figure 5.   The effect of CsA on GM-CSF mRNA+ cells in BAL. = 0.0028.

View larger version (13K): [in this window] [in a new window]   Figure 6.   The effect of CsA on IL-5 protein in BAL (pg/ml) pre- and postallergen challenge. = ns.

View larger version (17K): [in this window] [in a new window]   Figure 7.   The effect of CsA on the numbers of BB1+ basophils in bronchial biopsies pre- and postallergen. = ns.

No significant correlations were found between either the magnitude of the LAR and eosinophil numbers, or between eosinophils and eosinophil active cytokines or eotaxin.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study we have found that the attenuation of the LAR by CsA was associated with the inhibition of bronchial eosinophilia and accumulation of IL-5 mRNA and protein, GM-CSF mRNA, and eotaxin protein. There was no effect on the allergen-induced influx of basophils into bronchial mucosa.

Eosinophils are believed to be major proinflammatory cells in asthma through the elaboration of inflammatory mediators such as cysteinyl leukotrienes and platelet-activating factor (PAF) as well as release of basic proteins from the crystalloid granules. Eosinophil maturation, activation, and survival are under the control of IL-3, IL-5, and GM-CSF. IL-5 is essential for terminal differentiation of committed precursors (19). The mechanism of eosinophil recruitment appears to involve the expression of C-C chemokines, in particular eotaxin (20).

Although the clinical efficacy of CsA has been demonstrated in severe corticosteroid-dependent asthmatics (1, 2) and in an allergen challenge model (3), there have been no systematic human studies evaluating its in vivo effect on airway inflammation.

Several animal studies have demonstrated that CsA attenuates the LAR and EAR (23, 24) as well as inhibiting the influx of eosinophils in BAL (25, 26) and bronchial mucosa (27) after allergen challenge.

A case report of a corticosteroid-dependent asthmatic showing a reduction in T lymphocytes and activation markers in lavage and bronchial mucosa after 12 mo of CsA has recently been described (28). Furthermore, Alexander and coworkers demonstrated a decrease in serum IL-2R in a group of corticosteroid-dependent asthmatics taking CsA as compared with placebo (29). Our present double-blind, placebo-controlled study has allowed us to directly demonstrate an anti-inflammatory effect of CsA in terms of both eosinophil influx as well as the local expression of the eosinophil active cytokines and eotaxin. We have also demonstrated an inhibition of the increase in CD4 counts in the CsA group, but not the placebo group.

Although the effect of CsA is predominantly on T cells (30, 31) and T cells are a major source of eosinophil active cytokines (32), we cannot exclude an effect on other proinflammatory cells such as mast cells and basophils which also have the ability to elaborate T helper cell, type 2 (Th2) cytokines and have also been shown in vivo to be inhibited by CsA (33, 34).

It is notable in our study that the influx of basophils was not affected by CsA despite the attenuation of the LAR. As in the earlier study by Sihra and coworkers (3), CsA had no effect on the EAR, again indicating that CsA is likely to be having a clinical effect by its action on cytokine synthesis rather than on the immediate release of preformed mediators.

The lack of effect of CsA on the basophil influx suggests that the control mechanisms for this cell type may also be distinct from those of the eosinophil. This is in agreement with a previous study showing that C-C chemokine expression in cutaneous late-phase reactions correlated with the eosinophil, but not basophil numbers (22).

Given the effect of CsA on other cell types in vitro it is possible that a more protracted course of CsA may have resulted in a more profound effect on the late-phase reactions. To improve the safety profile of CsA it would also be interesting to evaluate topical therapy because inhaled CsA was shown to be well tolerated in normal human volunteers (35). In addition, other T-cell-specific immunomodulators may be equally effective but with less adverse effects. Drugs already being evaluated in the context of transplantation and autoimmune disease such as FK506, lefunomide, and mycophenolic acid may be potentially useful as corticosteroid-sparing agents for severe asthmatics.

In conclusion, this study has demonstrated that attenuation of the allergen-induced LAR by CsA was associated with a decrease in bronchial eosinophils and an inhibition of the expression of eosinophil active cytokines and chemokines. The use of inhibitors of T-cell function and T-cell-derived eosinophilic cytokines may therefore be of clinical benefit in asthma.