α₁-Antitrypsin inhalation reduces airway inflammation in cystic fibrosis patients

Study population

Inclusion criteria were: the diagnosis of CF by clinical symptoms and positive sweat tests or disease inducing mutations; age ≥8 yrs; forced expiratory volume in one second (FEV₁) >25% of predicted value; free elastase activity levels detected in the sputum sample at visit 1 (free elastase activity levels two sds above the negative blank sample); having tested positive at least three times for P. aeruginosa in the previous 2 yrs and positive testing in induced sputum at visit 1; stable concomitant therapy ≥2 weeks prior to visit 1 and during the study; and written informed consent.

Exclusion criteria were: a history of lung transplant; lung surgery within the previous 2 yrs; being on any thoracic surgery waiting list; severe concomitant disease (serious malignant disease, congestive heart failure New York Heart Association III/IV, cor pulmonale with the need of oxygen therapy); severe liver cirrhosis with ascites, hypersplenism or grade III/IV oesophageal varices; selective IgA deficiency with anti-IgA antibodies; active pulmonary exacerbation within the 4 weeks prior to screening; being a current smoker; being pregnant or breastfeeding; and being a female of child-bearing age without adequate contraception.

Study outline

At the first visit (visit 1), eligible subjects received a random number. Accordingly, each subject received an individually programmed SMART CARD (Inamed Gmbh, Gmünden, Germany) to be used in conjunction with the AKITA® device (Inamed GmbH), which determined the respective inhalation pattern with bronchial or peripheral deposition 19. During the following 2-week run-in period, the subjects inhaled an isotonic saline solution once daily in order to get used to the inhalation pattern. At visit 2, baseline measurements were taken and a 4-week treatment period began where AAT was inhaled. After 2 weeks (visit 3) and 4 weeks (visit 4) all measurements were repeated. The study treatment was added to the regular therapy of the patients. Sputum induction was performed only at the study visits in the hospital, as described in detail hereafter.

Treatment administered

In order to determine the optimal region for AAT deposition, two inhalation modes were programmed in a double-blind fashion on a SMART CARD connected to the inhalation device (AKITA® inhalation device with a Pari LC Plus (Pari, Starnberg, Germany) or an Pari LC Star nebuliser (Pari)). Patients were randomly assigned to either bronchial or peripheral deposition, according to the method described by Brand et al. 19.

Peripheral deposition was achieved by a slow inhalation flow (200 mL·s^-1) and an individualised inhalation volume calculated from the individual inspiratory capacity determined by spirometry. Therefore, the following formula inhalation volume was used 19:

2×e^{(-1.5/inspiratory capacity)} + 0.25 (1)

The nebuliser of choice for peripheral deposition was a Pari LC Star, delivering an aerosol with a mass median diameter (MMD) of 3.5 µm.

Bronchial deposition was achieved by a very slow inhalation flow of 100 mL·s^-1 and a low inhalation volume of 60% of the individual inspiratory capacity, with a maximum value of 0.5 L. The nebuliser used was a Pari LC Plus, generating an aerosol with an MMD of 5.0 µm. The two nebulisers had an identical outer appearance.

The SMART CARDs were programmed to deposit 25 mg AAT (Prolastin®; Bayer Corporation, Clayton, NA, USA) during the treatment phase. Prolastin® is made from human blood plasma and is primarily monomeric in solution (>99.5%). It consists of purified AAT in a buffer containing 100–210 mM sodium, 60–180 mM chloride and 15–28 mM sodium phosphate. During the run-in phase, an equal amount of the isotonic saline solution instead of AAT was inhaled according to the respective inhalation pattern. The daily inhalations took place in the evening between 18:00 and 23:00 h. The time required for the inhalations was 5−15 min, depending on the lung function of the patient. Treatment compliance (date and time of inhalation, number of breaths and completeness of each breath) was monitored on the SMART CARD.

Induced sputum

Induced sputum was obtained at the study visits in the hospital between 9:00 and 13:00 h, following the AAT inhalation the previous evening. Before sputum was induced, subjects underwent physiotherapy and spirometry, then two puffs of salbutamol were inhaled and 5 mL of a 5.85% sodium chloride solution were nebulised for 15 min with the Pari LC Plus connected to a Pari Master compressor (Pari). A physician was present during sputum induction.

The pooled sputum was divided into three different samples: one for bacteriology (0.5–1 mL), one for AAT, elastase, cytokine and neutrophil assessments (3 mL) and one for IgG determination (3 mL). The samples for bacteriological analysis were shipped overnight at 4°C to the Max von Pettenkofer Institute, a south German reference laboratory for P. aeruginosa at the University of Munich, Munich. The samples for AAT, elastase, cytokines, neutrophil and IgG analysis were incubated 1:1 with dithiothreitol (DTT; Sputolysin; Calbiochem Biosciences, Beeston, UK), filtered through Nitex gaze-filters (BD Biosciences, San Diego, CA, USA) and washed with Hanks solution. The sputum suspension was centrifuged at 500×g for 10 min at 4°C and the supernatant was centrifuged at 4,000×g for 20 min at 4°C. The supernatant was then mixed with a protease inhibitor (0.5 mM EDTA, 500 µM Pefabloc (Merck, Darmstadt, Germany), 5 µM E-64, 50 µM Bestatin (both Roche, Basel, Switzerland)) in order to block free elastase activity and to avoid in vitro proteolysis. Aliquots of the supernatant were frozen at -70°C until final analysis. Cytospin slides were prepared with the cell pellet using 200,000 cells per slide. In total, ≥400 cells were differentiated by May-Grünwald-Giemsa staining. The viability was assessed by the trypan blue dye exclusion test. The remaining cells were suspended in Hanks buffer at 4°C and were immediately processed for flow cytometry.

The levels of AAT were measured by ELISA. The levels of free elastase activity were analysed by a chromogenic assay according to standard protocols as described by Hilliard et al. 20. Numbers of P. aeruginosa were quantified according to the method described by Hogardt et al. 21.

Flow cytometry

CD45-allophycocyanin (APC) mouse IgG1 (Pharmingen, Heidelberg, Germany) and mouse IgG1-APC (Immunotech, Marseille, France) as isotype control were used. Calculations were performed with Cell Quest analysis software (Cell Quest Pro; Becton-Dickinson, Heidelberg, Germany). Re-suspended sputum cells were blocked with human IgG for 20 min to avoid nonspecific binding. They were then incubated with monoclonal antibodies for 40 min, washed twice and finally analysed by flow cytometry (FACS Calibur; Becton-Dickinson) as previously described elsewhere 22. Prior to the study, the discrimination between neutrophils and alveolar macrophages was optimised. Gating of neutrophils was based on light scatter properties and positive expression for CD45. In the present experimental setting, macrophages were less of a disturbance than the considerable amount of apoptotic or dead cells present in the sputum of CF patients. Therefore, propidium iodide (5 μg·mL^-1; Sigma, St Louis, MO, USA) and Annexin V-fluorescein isothiocyanate (5 μg·mL^-1, dilution 1/100; Boehringer Mannheim GmbH, Mannheim, Germany) were used to discriminate intact viable leukocytes (Annexin V^-, PI^-) from apoptotic (Annexin V⁺, PI^-) and necrotic (Annexin V⁺, PI⁺) cells. Only viable neutrophils were included in the analysis. These gates were used to analyse 10,000 neutrophils per sample.

Immunoglobulin G

For the assessment of IgG fragments, a modified method according to Fick et al. 23 was used. A total of 10 µL of the sputum supernatant pre-treated with a proteinase inhibitor were lyophilised, dissolved in 20 µL of sample buffer and reduced for 10 min at 70°C (NuPAGE Reducing Agent; Invitrogen, Madison, WI, USA). The sample, together with 500 µL of antioxidant (NuPAGE antioxidant; Invitrogen), was subjected to electrophoresis (10% Bis-Tris Gel) and 500 ng human IgG was used as standard. The effect of human leukocyte elastase (HLE) on IgG was assessed in vitro by incubation of 500 ng IgG isolated from human serum (Sigma) with HLE (40 U·mL^-1; EPC, St. Louis, MO, USA) for 2 h at 37°C. The proteins were transferred on nitrocellulose membranes (Amersham Pharmacia Biotech, Uppsala, Sweden) for semi-dry Western blotting (Xcell II blot). IgG fragments were identified by incubation for 12 h with 2.5 µL of a specific goat anti-human IgG antibody. After intensification of the signal by chemiluminescence, the blots were scanned with a computing densitometer (Fluor-S MultiImager; BioRad, Richmond, CA, USA) and were semi-quantitatively analysed using Quantity One (BioRad). In order to detect corresponding protein spots between gels, all blots were stacked and spots were matched according to their molecular weight.

Cytokine protein levels

Levels of IL-8, tumour necrosis factor (TNF)-α and IL-1β were analysed in sputum supernatant in duplicate by a sandwich ELISA according to the manufacturer's instructions (R&D Systems, Minneapolis, MN, USA). The levels of LTB₄ were measured in sputum supernatant by a sandwich ELISA according to the manufacturer's instructions (Amersham Pharmacia Biotech). The detection limits were 3.5 pg·mL^-1 for IL-8, 0.12 pg·mL^-1 for TNF-α, 1 pg·mL^-1 for IL-1β and 1.25 pg·mL^-1 for LTB₄. Furthermore, cytokine levels in sputum were analysed with and without DTT pre-treatment at the same dilution used for sputum samples.

Cytokine mRNA levels

mRNA levels of IL-8, TNF-α and IL-1β were analysed in induced sputum cell pellets in duplicate by a quantitative real-time RT-PCR system (Icycler; Biorad, Hercules, CA, USA) according to the manufacturer's instructions. The following forward (F) and reverse (R) primer pairs were used. IL-8F: TCTTGGCAGCCTTCCTGATT; IL-8R: TCCAGACAGAGCTCTCTTCCATC; TNF-αF: AGGAACAGCACAGGCCTTAGTG; TNF-αR: AAGACCCCTCCCAGATAGATGG; IL-1βF: CAGGGACAGGATATGGAGCAA; and IL-1βR: ATGTACCAGTTGGGGAACTG. Induced sputum cells were lysed in Trizol LS Reagent (Invitrogen, Karlsruhe, Germany). Total RNA (200–500 ng) was isolated according to the manufacturer's instructions and reverse transcribed into cDNA. Expression levels of cytokines were determined in duplicate by real-time RT-PCR using SYBR green and the iCycler iQ detection system (Biorad). Threshold cycle values for genes of interest were normalised to glyceraldehyde-3-phosphate dehydrogenase and used to calculate the relative quantity of mRNA expression.

Statistics

The primary analysis population was a modified intention-to-treat (mITT) population, including all randomised subjects who received any amount of study medication (AAT) and had at least one evaluation of the primary efficacy variable, i.e. free elastase activity in induced sputum at baseline (visit 2). The primary efficacy comparison for the change in free elastase activity in induced sputum from baseline to end-point was a two-way ANCOVA with treatment group and centre as fixed factors (main effect model) and baseline measurement of free elastase activity in induced sputum as covariate in the mITT population. Based on the sd of elastase activity levels in sputum as previously described elsewhere 3, 24, a sample size of 24 subjects per treatment group was calculated to detect a change in free elastase activity in induced sputum in the magnitude of the sd at a 5% two-sided alpha and a power of 80%. All further calculations were made in an exploratory way, including the evaluation of the overall treatment effect as pre- and post-comparisons for the combined deposition groups (Wilcoxon and sign test with Bonferroni correction for multiple comparisons).

Study population, deposition modes and safety

In total, 72 CF subjects were enrolled in the present study. A total of 37 subjects were randomised to the group for peripheral deposition and 35 to the group for bronchial deposition (fig. 1⇓). Eight subjects dropped out during the run-in phase prior to AAT treatment due to the following protocol violations: inability to produce sputum (two patients), diagnosis of IgA deficiency after inclusion into the study (one patient), inability to perform daily inhalations with the AKITA® device due to personal breathing pattern (one patient) and withdrawal of consent or cessation of the study for personal reasons (four subjects). Another five patients incurred other protocol violations that excluded them from the study. Therefore, 59 subjects received at least one dose of study drug and were valid for safety and intention-to-treat (ITT) analysis. Seven subjects were excluded due to missing or invalid data for the primary efficacy variable (free elastase activity in induced sputum), because no sputum was produced or the sputum was unavailable for analysis. The group of subjects remaining (mITT group) was defined before the study as the primary analysis population. Thus, the analysis was based on 28 subjects using inhalation for peripheral deposition and 24 subjects using inhalation for bronchial deposition. The baseline characteristics were not statistically different between the two deposition groups (table 1⇓). Electronically recorded compliance with the study drugs was good.

The number of subjects with adverse events was 10 out of 30 (seven out of 28 in the treatment phase) in the peripheral deposition group and 13 out of 29 (eight out of 24 in the treatment phase) in the bronchial deposition group. With the exception of one severe adverse event (sinusitis), all other events were of mild or moderate intensity. No serious adverse events occurred. All adverse events had resolved or improved by the end of the study. Three subjects in each group experienced drug-related adverse events. One of these (fatigue) was attributable to the inhalation manoeuvre or the saline, as it had already occurred during the run-in phase. Two other drug-related adverse events (fatigue and haemoptysis) arose in the bronchial deposition group and three in the peripheral deposition group (pruritus, influenza-like illness, gastro-intestinal pain).

No significant differences between the peripheral and bronchial deposition group were noted at baseline or after 2 and 4 weeks of AAT inhalation with respect to free elastase activity, AAT level, the percentage of neutrophils, P. aeruginosa counts, cytokine levels, IgG fragments in sputum (table 2⇓), lung function (FEV₁, functional vital capacity, maximum expiratory flow at 25% of vital capacity), number of exacerbations and overall rate of adverse effects (data not shown). As no difference between peripheral or bronchial AAT deposition site was found for any of the parameters analysed, the overall effect of inhaled AAT was assessed in an explorative analysis by combining the two deposition groups and making before and after comparisons.

α₁-Antitrypsin and free elastase activity

Treatment with 25 mg of AAT increased the concentration of AAT after 2 (p<0.001) and 4 weeks of treatment (p<0.001; fig. 2a⇓) and slightly reduced the levels of free elastase activity in induced sputum after 4 weeks of treatment (p<0.05; fig. 2b⇓). Furthermore, the AAT inhalation decreased percentages of sputum neutrophils after 4 weeks of treatment (p<0.05; fig. 3a⇓) and reduced P. aeruginosa counts after 4 weeks of treatment (p<0.05; fig. 3b⇓), while FEV₁ did not change (fig. 3c⇓).

In conclusion, these data suggest that 4 weeks of α₁-antitrypsin inhalation reduce airway inflammation in cystic fibrosis patients. Although no effect on lung function was observed, reduced airway inflammation may precede pulmonary structural changes. Longer, placebo-controlled studies aiming to deposit optimal amounts of α₁-antitrypsin into the lungs of cystic fibrosis patients are worthwhile undertaking.

Pro-inflammatory cytokines

DTT treatment had no significant effect on IL-8, TNF-α, IL-1β or LTB₄ levels (data not shown). Sputum protein levels of IL-8 (fig. 4a⇓), IL-1β (fig. 4b⇓), TNF-α (fig. 4c⇓) and LTB₄ (fig. 4d⇓) and sputum mRNA levels of IL-8 (fig. 5a⇓), IL-1β (fig. 5b⇓) and TNF-α (fig. 5c⇓) decreased significantly after 2 and 4 weeks of AAT inhalation when compared with the baseline. The changes of IL-8 (r = 0.48, p<0.01), IL-1β (r = 0.39, p<0.05), TNF-α (r = 0.4, p<0.05) and LTB4 (r = 0.45, p<0.05) protein levels and the changes of neutrophils (r = 0.52, p<0.01) correlated positively with the changes of free elastase levels.

Immunoglobulin G

CF-related proteases were found to cleave IgG into two antigen-binding fragments (F(ab)₂) and crystallisable fragments (Fc) 23, 25, 26. In sputum samples of the CF patients in the present study, IgG bands of 26 kDa, 32 kDa, 36 kDa, 43 kDa and 54 kDa were detected under reducing conditions (fig. 6⇓). In accordance with previous studies analysing proteolytically cleaved IgG fragments 3, 23, 27–29, the 26-kDa band corresponded to the reduced form of the IgG light chain, the 32-kDa band to the cleaved IgG Fc fragment 27, 28 and the 43-kDa band to the cleaved F(ab)₂ fragment 28. The 36-kDa band could not be identified. The 54-kDa band represented the reduced form of the regular IgG heavy chain of 168 kDa 27–29. The treatment of IgG with HLE cleaved IgG into F(ab)₂ and Fc fragments. The percentages of 54-kDa IgG increased significantly after 4 weeks of AAT inhalation (fig. 7a⇓), whereas no significant change was found for the other IgG bands. The percentages of 54-kDa IgG correlated inversely with levels of free elastase activity in sputum before and after AAT treatment (fig. 7b⇓).