INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Formoterol (FM) is a potent, fast-acting ₂-agonist with a long duration of action (1). There is, however, continuing concern about the regular use of long-acting ₂-agonists, such as salmeterol and FM, in the treatment of asthma (2). In particular, tolerance may develop during regular use of long-acting ₂-agonists.

There is evidence that tolerance to the bronchoprotective (3) and bronchodilator (3) properties of FM occur with regular use, and similar findings have been reported for salmeterol (6). In certain studies (5, 8, 10, 11), bronchodilator and bronchoprotective subsensitivity, as well as ₂-adrenoreceptor (₂-AR) downregulation, have been demonstrated despite the fact that patients were also receiving inhaled corticosteroid. It is not known whether systemic corticosteroid, as administered in an acute attack of asthma, might modulate such ₂-AR downregulation and associated subsensitivity.

There are good reasons to believe that systemic corticosteroids might reverse ₂-AR subsensitivity, in that they exhibit a facilitatory role in reversing lymphocyte ₂-AR downregulation in normal (12, 13) and asthmatic subjects (14) exposed to ₂-agonists. Earlier studies have suggested reversal of in vivo subsensitivity by systemic corticosteroid (15, 16), although the time-course of action of this effect and whether upregulation of ₂-AR in vitro correlates with reversal of in vivo responses are both unclear.

The aim of the present study was to investigate the acute modulating effects of systemic corticosteroid on in vitro and in vivo ₂-AR function in asthmatic patients treated with regular inhaled FM (i.e., in the presence of established subsensitivity).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Twelve asthmatic patients (four female and eight male), with a mean age of 44 yr (range: 20 to 63 yr) were recruited for and completed the study. All gave written informed consent before being randomized according to a double-blind, placebo-controlled, crossover protocol approved by the Tayside Medical Ethics Committee. A full physical examination, 12-lead electrocardiogram (ECG), and measures of biochemical and hematologic parameters were normal prior to inclusion of the patients. All had asthma according to the criteria of the American Thoracic Society (ATS) (17), and all were nonsmokers. At the initial screening visit, subjects were required to have an FEV₁ of 40 to 80% of the predicted normal value, with at least 15% reversibility of FEV₁ with inhaled FM at a dose of 24 µg. The mean (SEM) FEV₁ in liters and percent predicted were 2.14(0.15) and 66(3)% predicted (range: 1.34 to 2.971 L; % predicted range: 47 to 75%). All patients were receiving inhaled corticosteroid (either budesonide or beclomethasone dipropionate; one patient received fluticasone propionate) at a median dose of 1,000 µg/d (range: 400 to 1,600 µg/d). All had been inhaling short-acting ₂-agonists, as required prior to recruitment, in doses of < 800 µg/d. Five subjects were inhaling salmeterol 100 µg/d. In addition, four subjects were taking slow-release oral theophylline preparations. None of the subjects had received oral corticosteroid for at least 3 mo, and none had had a recent exacerbation of asthma in the month preceding the study. Before entry into the study, all subjects were supervised in the use of a metered-dose inhaler (MDI), using a Vitalograph aerosol inhalation monitor (Vitalograph Ltd, Buckingham, UK).

Protocol

After the initial screening visit, the subjects had a 2-wk washout period without any ₂-agonists, during which they used ipratropium bromide at 40 µg per actuation (Atrovent Forte; Boehringer Ingelheim, Bracknel, UK) as a substitute for rescue requirements. After 2 wk, subjects attended for a randomization visit, and FEV₁ was measured, yielding a mean ± SEM value of 2.20 ± 0.15 L, or 67 ± 3% predicted. Subjects were then randomized to receive concurrent administration in crossover fashion of either inhaled placebo or inhaled FM at 24 µg twice daily (12 µg per actuation, Foradil; Ciba-Geigy AG, Basel, Switzerland), with both administered via an MDI for 4 wk, while they maintained their inhaled steroid and other antiasthma therapy at a constant dose (Figure 1). Six subjects received FM as first treatment, and six received placebo first. During the treatment period, an ipratropium bromide inhaler was also available for rescue purposes in order to ensure that ₂-agonists were not used during the placebo period. The study treatment was taken twice daily, between 7:00 A.M. and 9:00 A.M. and again between 7:00 P.M. and 9:00 P.M. The subjects were also asked to keep morning and evening PEF readings, made with a Wright peak flow meter (Airmed, London, UK), in a diary.

View larger version (12K): [in this window] [in a new window]   Figure 1.   Overview of study design.

Subjects attended the laboratory after 3 wk and 4 wk of treatment at 8:00 A.M., having withheld their study medication for 24 h, ipratropium bromide for at least 8 h, and oral theophylline for 48 h. At each visit, an intravenous cannula was inserted and kept patent with bolus injections of heparinized saline. A cannula dead space of 2 ml was withdrawn prior to blood sampling. A dose-response curve (DRC) for inhaled FM was constructed 1 h after ingestion of placebo tablets and intravenous placebo injection at the 3-wk visit, and 1 h after oral prednisolone 50 mg and intravenous hydrocortisone 200 mg at the 4-wk visit. The rationale for the administration of corticosteroid was to mimic what would normally be given in an acute asthma attack (i.e., an initial bolus of hydrocortisone and a first dose of oral prednisolone to provide cover for the next 24 h). The administration of placebo at Week 3 and systemic corticosteroid at Week 4 was done so as to avoid any possibility of carryover from one DRC to another.

The DRC was constructed with inhaled FM given via an MDI, using doses of 12 µg, 24 µg, 24 µg and 48 µg (i.e., a cumulative dose of 108 µg after the last dose), with the doses separated by 40 min. Measurements of FEV₁, FEF_25-75, serum potassium (K), heart rate (HR), and postural finger tremor (Tr) were undertaken over a 10-min period at baseline (before placebo or steroid tablets and injection), 1 h after placebo or steroid tablets and injection (i.e., at the start of the DRC), and 30 min after each dose, and were repeated at 1, 2, 4, and 6 h after the last dose. Forty milliliters of blood for lymphocyte ₂-AR parameters (B_max, E_max, and K_d) was withdrawn at 1 and 3 h after administration of placebo or steroid. The next day, 24 h after administration of placebo or steroid, subjects attended the laboratory for measurement of lymphocyte ₂-AR parameters and of FEV₁ and FEF_25-75 before and 30 min after a dose of 12 µg FM. The purpose of the latter measurement was to observe whether the effects of the steroid were still present after 24 h in terms of modulating ₂-AR function in vitro and in vivo, although a full DRC was not generated.

Ten milliliters of whole blood was also taken, for the study ₂-AR polymorphism, and was stored in ethylenediamine tetraacetic (EDTA) acid at 20° C. Unfortunately, one sample was lost, and we analyzed ₂-AR polymorphism in 11 of 12 subjects.

Measurements

Airway responses. Measurements of FEV₁ and FEF_25-75 were made according to ATS criteria (18), with a Vitalograph Compact Spirometer (Vitalograph Ltd, Buckingham, UK). Forced expiratory maneuvers were performed from TLC to RV. The FEV₁ and FEF_25-75 values were taken from the best of three consistent measurements. A coefficient of variation (CV) of less than 3% was considered acceptable.

Systemic responses. Serum potassium was measured with flame photometry, HR from the R-R interval on the ECG, and finger tremor with an accelerometer transducer, all as previously described (5, 8). The within- and between-assay coefficients of variation (CVs) for analytical imprecision for serum potassium were 0.93% and 0.79%, respectively.

Lymphocyte ₂-AR parameters were measured as previously described (5, 8). In brief, B_max and K_d were determined with a radioligand binding method with (-)[¹²⁵I]iodocyanopindolol (ICYP; NEN-du-Pont[UK] Ltd, Stevenage, UK), and E_max was determined through the cyclic AMP (cAMP) response to isoprenaline 10⁴ M. cAMP was measured with a radioimmunoassay (Incstar Ltd, Wokingham, UK). The within-assay CVs for analytical imprecision were 10.29% and 5.85% for B_max and K_d, respectively; and the within- and between- assay CVs for E_max were 5.36% and 9.3%.

Identification of ₂-AR polymorphisms. ₂-AR polymorphisms were identified as previously described (19). In brief, genomic DNA was extracted from whole blood, and a 234-bp fragment that spanned the regions of interest was generated with the polymerase chain reaction (PCR). The primers used were 5'-CCCAGCCAGTGCGCTTACCT and 3'-CCGTCTGCAGACGCTCGAAC. Genotype was determined with allelle-specific oligonucleotide (ASO) hybridization, using probes homologous for the Arg-16, Gly-16, Gln-27 or Glu-27 forms of receptor. A random selection of PCR fragments was also directly sequenced to confirm the specificity of genotype determination by ASO.

Statistical Analysis

Data for finger tremor and B_max were transformed with logarithms to base 10, since neither variable was normally distributed. All variables were then analyzed as delta responses from the baseline at 1 h after injection of placebo or hydrocortisone. Comparisons for the DRC were made in terms of area under the curve (AUC), in order to obviate multiple comparisons at several time points. For all parameters, comparisons were made through multifactorial analysis of variance (MANOVA) using subjects, treatments, visits, and period as factors for analysis. When significant overall differences between treatments were found to occur with MANOVA, Duncan's multiple-range testing was applied to identify where these differences occurred. A value of p < 0.05 (two-tailed) was considered as being significant, and 95% confidence intervals (95% CIs) for mean treatment differences were calculated where significant. Data were analyzed with a Statgraphics Statistician Software Package (STSC Software Publishing Group, Rockville, MD).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Baseline values for airway and systemic parameters were not signficantly different after treatment with placebo or FM, either before or after administration of placebo or steroid tablets and injection (Table 1).

Bronchodilator and Systemic DRCs

The bronchodilator and systemic DRCs (as AUC) after regular administration of placebo and treatment with FM, with and without systemic steroid, are summarized in Table 2 and shown graphically in Figures 2, 3, and 4.

View larger version (21K): [in this window] [in a new window]   Figure 2.   Dose-response curves and time profile after last dose for FEV1 and FEF25-75 after regular treatment with placebo (PL) and formoterol (FM) with or without acute administration of systemic corticosteroid (S). Values are shown as mean (SEM) change from baseline. *p < 0.05 versus formoterol for AUC.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Absolute values for FEV1 and FEF25-75 at baseline and 30 min after a single 12 µg dose of formoterol, with or without acute systemic corticosteroid, at 24 h after administration, following regular treatment with formoterol (FM) and placebo (PL). *p < 0.05 versus formoterol at baseline. No signficant differences were seen after formoterol 12 µg.

View larger version (18K): [in this window] [in a new window]   Figure 4.   Dose-response curves and time profile for heart rate (HR), finger tremor (Tr) and serum potassium (K) after last dose of regular treatment with placebo (PL) and formoterol (FM) with or without systemic corticosteroid (S). Values are shown as mean (SEM) change from baseline. *p < 0.05 versus formoterol; +p < 0.05 versus placebo for AUC.

The DRCs (as change from baseline) after either administration of placebo or pretreatment with FM showed dose- dependent increases in FEV₁ and FEF_25-75, and a plateau in response was not attained within the dose range, 40 min after the last dose (i.e., at t = 2 h and 40 min) (Figure 2). There was a rightward shift in the DRC for both FEV₁ and FEF_25-75 after treatment with FM as compared with placebo: p = 0.01 for FEV₁ and FEF_25-75. Following administration of systemic corticosteroid, there was significant reversal of bronchodilator tachyphylaxis for FEV₁ (p = 0.03) and FEF_25-75 (p = 0.005). There was no significant difference in response to FM plus steroid (FM + S) and placebo plus steroid (PL + S).

On the following day (i.e., 24 h after injection of placebo or steroid) the baseline FEV₁ and FEF_25-75 were lower after pretreatment with FM than with placebo (p = 0.02 for FEV₁ and p = 0.004 for FEF_25-75). FM 12 µg was then administered, and the 30-min postbronchodilator values for FEV₁ and FEF_25-75 were not signficantly different.

There were dose-dependent increases in Tr and HR, and a decrease in serum K (Figure 4). There were significant rightward shifts in Tr and HR after treatment with FM as compared with placebo: p < 0.001 for HR and p = 0.04 for Tr. There was no significant rightward shift in K (p = 0.13). After the administration of systemic corticosteroid, tachyphylaxis in HR response was reversed (p < 0.001). There was no significant reversal in the DRC for Tr following FM plus steroid compared with FM alone (p = 0.20). There was augmentation of the HR response with placebo plus steroid compared with placebo alone (p < 0.001).

Lymphocyte ₂-AR Parameters

There were no significant differences between any of the treatments in log B_max (fmol/10⁶ cells) at 1 h after injection of placebo or steroid (Figure 5). However, at 3 h after injection, there was a significant increase in log B_max for FM plus steroid as compared with FM alone (p = 0.01). There was also a significant increase with systemic steroid after placebo pretreatment (p < 0.001). After 24 h, the effect or corticosteroid had diminished, and log B_max had fallen to levels below baseline. The trend for time-profile changes in E_max (pmol/10⁶ cells) mirrored that of log B_max in that there was a significant increase at 3 h after corticosteroid administration in the placebo pretreated group (E_max at 1 h versus E_max at 3 h: p = 0.003). Values at 3 h were (PL + S versus PL) 4.62 versus 3.74, but these were not significantly different. For binding affinity, K_d, the time-profile changes were different from those for B_max and E_max. At 1 and 3 h after administration of corticosteroid, there were no significant differences in K_d. However, by 24 h there was a significant (p < 0.05) decrease in K_d in subjects treated with FM or placebo, except in those subjects given corticosteroid.

View larger version (12K): [in this window] [in a new window]   Figure 5.   Lymphocyte 2-adrenoceptor parameters: density (Bmax), maximal cAMP response to isoprenaline 104 M (Emax), and binding affinity (Kd) with or without acute systemic corticosteroid at 1, 3, and 24 h after administration, following regular treatment with placebo (PL) and formoterol (FM). *p < 0.05 versus formoterol; +p < 0.05 versus placebo)

PEFR Diary Cards

Mean peak expiratory flow rates (L/min) in the morning were higher with FM than with placebo, and this finding was sustained for 4 wk of treatment: (PL versus FM) for Week 1: p = 0.05; for Week 2: p = 0.008; for Week 3: p = 0.008, and for Week 4: p = 0.005 (Figure 6). Likewise, evening PEFR was higher with FM than with placebo throughout the 4 wk of treatment.

View larger version (18K): [in this window] [in a new window]   Figure 6.   Mean (SEM) peak-expiratory-flow diary recordings (12 h after each dose) in the morning and evening during 4 wk of treatment with placebo (PL) and formoterol (FM). *p < 0.05 FM versus PL.

₂-AR Polymorphism

Eleven subjects were analyzed for ₂-AR polymorphism at loci 16 and 27. Seven subjects were homozygous for Gly-16, three were heterozygous for Arg-16/Gly-16, and one was homozygous for Arg-16; three were homozygous for Glu-27, six were heterozygous for Gln-27/Glu-27, and two were homozygous for Gln-27. The individual responses for AUC FEV₁ and AUC FEF_25-75 in terms of ₂-AR polymorphism at loci 16 and 27 are shown in Figures 7 and 8, respectively. The data were not analyzed in terms of differing subsensitivity because of insufficient numbers of subjects with different polymorphisms. Hence, only a qualitative assessment could be made of these results. After the development of bronchodilator subsensitivity following treatment with FM, the administration of systemic corticosteroid produced reversal in the majority of subjects irrespective of ₂-AR polymorphism (Figures 7 and 8).

View larger version (16K): [in this window] [in a new window]   Figure 7.   Individual responses for AUC FEV1 and AUC FEF25-75 after treatment with placebo (PL) and formoterol (FM), compared with reversibility following steroid (FM + S), according to 2-AR polymorphism at locus 16.

View larger version (15K): [in this window] [in a new window]   Figure 8.   Individual responses for AUC FEV1 and AUC FEF25-75 after treatment with placebo (PL) and formoterol (FM), compared with reversibility following steroid (FM + S), according to 2-AR polymorphism at locus 27.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results in this study demonstrate that subsensitivity of bronchodilator and systemic ₂-agonist responses occurs after treatment with regular inhaled FM as compared with placebo. These results are consistent with the findings in our previous studies with long-acting ₂-agonists with dry powder (4) and aerosol (5) preparations of FM and salmeterol (8). In contrast, bronchodilator subsensitivity does not appear to occur with short-acting ₂-agonists in placebo-controlled studies (20, 21), although Turki and colleagues have reported downregulation of lung-cell ₂-AR after 24 h with metaproterenol (22).

It is important to note that baseline values for FEV₁ and FEF_25-75 in the present study were virtually identical between treatments both before and after injection of placebo or steroid, making it unlikely for differences between treatments to be due to baseline effects. There was no period effect between visits, which excludes the possibility that the difference between FM alone and FM plus steroid was a sequence phenomenon.

The use of FM in constructing DRCs in the present study has clinical relevance in that FM is fast acting, and it is conceivable that patients who are on regular therapy with FM might also use it repeatedly for rescue purposes, as might occur during an acute attack of asthma.

As in previous studies (4, 5, 8), morning and evening expiratory flow rates from diary cards at 12 h after dosing were significantly greater with FM than with placebo, and this effect was sustained during the 4-wk treatment period. This clearly illustrates the importance of constructing proper DRCs and ensuring at least a 24 h washout prior to DRC in order to detect subsensitivity. It is, however, conceivable that improved peak flow rates may lull patients into a false sense of security and delay them in seeking medical attention. It is also possible that regular long-acting ₂-agonist therapy, although improving symptoms and PEFR, might mask worsening inflammation, as might occur before an acute asthmatic attack. It is also relevant to note in this study that bronchodilator subsensitivity was demonstrated despite the use of regular inhaled corticosteroid by all of the subjects.

Our results also show rapid reversal of bronchodilator subsensitivity within 1 h after injection of hydrocortisone. This finding is similar to that in a previous study by Ellul-Micallef and Fenech in 1975 (15). In their open, non-placebo-controlled study, intravenous prednisolone restored the bronchodilator response to a single 200-µg dose of isoprenaline given 60 min after injection to 10 patients who had previously been shown to be nonresponsive. Also, Holgate and colleagues, in a study with normal subjects, showed reversal of the salbutamol DRC (in terms of specific airway conductance, sGaw) at 16 h after hydrocortisone (16).

There is conflicting evidence as to whether lymphocyte ₂-AR can be used as a surrogate for following changes in airway smooth-muscle ₂-AR (23, 24). In our subjects, systemic corticosteroid upregulated lymphocyte ₂-AR, in keeping with a reversal of bronchodilator subsensitivity. This is also in keeping with the findings in previous studies with lymphocytes, in which upregulation was shown to occur in normal (12, 13) and asthmatic (14) subjects with systemically administered corticosteroid. The mechanism of this effect of corticosteroids may involve an increase in the rate of synthesis of receptors through a process of increased ₂-AR gene transcription (25), or a reversal or inhibition of internalization of receptors from the cell surface (26). Corticosteroids are also thought to promote the formation of the coupled, high-affinity state of the receptor, which in turn increases receptor function (26).

Systemic corticosteroid also upregulated lymphocyte ₂-AR that had not previously been downregulated (i.e., after placebo pretreatment). Mean values for log B_max after placebo were similar to previously reported control values for normal volunteers (0.29 versus 0.28 fmol/10⁶ cells) (27). This upregulation of lymphocyte ₂-AR by corticosteroid was mirrored in its effect on HR, for which steroid administration augmented the HR response after placebo pretreatment. Indeed, even after FM, the presence of steroid increased the HR response as compared with placebo alone. This therefore suggests that systemic corticosteroid is capable of supernormalizing ₂-AR density. This is a phenomenon that we have previously shown in patients with mild asthma, with essentially naive ₂-AR, in whom a single 50-mg dose of oral prednisolone but not inhaled fluticasone at 2-mg produced upregulation and increased the cAMP response to isoprenaline (27). This, along with the lack of protection against ₂-AR downregulation and subsensitivity with inhaled corticosteroid, would suggest that a facilitatory effect of corticosteroid is conferred only with the systemic route.

We studied a group of asthmatic subjects who were heterogenous in terms of ₂-AR polymorphism. In particular, it is known that homozygous Gly-16 and homozygous Glu-27 polymorphism respectively confers susceptibility to and protection against downregulation of the response to ₂-agonists (28). The results of the present study show that in general, systemic corticosteroid produced reversal of bronchodilator subsensitivity independent of ₂-AR polymorphism. This is clinically relevant because subjects homozygous for Gly-16 are most susceptible to desensitization, but appear to be responsive to the facilitatory effects of systemic corticosteroid. Our patients had stable, moderately severe asthma, and it is conceivable that even a small degree of subsensitivity would assume greater clinical importance in persons with more severe asthma in the setting of an acute attack.

There were also some interesting findings at 24 h after administration of corticosteroid. The baseline level of FEV₁ at 24 h was lower after pretreatment with FM than with placebo, irrespective of administration of steroid. However, this is unlikely to reflect persistent subsensitivity, since the 30-min response to FM 12 µg was preserved. This does not appear to relate to lymphocyte ₂-AR density or isoprenaline cAMP response, both of which were diminished at 24 h irrespective of prior steroid administration. This decrease in lymphocyte ₂-AR density and cAMP responsiveness between 3 h and 24 h may represent the legacy of the repeated FM doses used in generating the DRC. Lymphocyte receptor affinity as measured by the dissociation constant, K_d, showed a different pattern, in that corticosteroid prevented a decrease in K_d, which occurred irrespective of FM or placebo pretreatment, inferring an effect of the previous FM DRC.

What might be the clinical relevance of our study? In acute asthma, in which be desensitization may occur as a consequence of regular long-acting ₂-agonist therapy, the study demonstrates the importance of the early administration of systemic corticosteroid in order to restore normal airway ₂-AR responsiveness. This suggests that systemic corticosteroid has a dual action in acute asthma in terms of an early effect on ₂-AR response and a later effect on the inflammatory response.