American Journal of Respiratory and Critical Care Medicine

Rationale: Several studies suggest that patients with asthma who are homozygous for arginine at the 16th position of the β2-adrenergic receptor may not benefit from short-acting β-agonists.

Objectives: We investigated whether such genotype-specific effects occur when patients are treated with long-acting β-agonists and whether such effects are modified by concurrent inhaled corticosteroid (ICS) use.

Methods: We compared salmeterol response in patients with asthma homozygous for arginine at B16 (B16Arg/Arg) with those homozygous for glycine at B16 (B16Gly/Gly) in two separate cohorts. In the first, subjects were randomized to regular therapy with salmeterol while simultaneously discontinuing ICS therapy. In the second, subjects were randomized to regular therapy with salmeterol while continuing concomitant ICS.

Results: In both trials, B16Arg/Arg subjects did not benefit compared with B16Gly/Gly subjects after salmeterol was initiated. In the first cohort, compared with placebo, the addition of salmeterol was associated with a 51.4 L/min lower A.M. peak expiratory flow (PEF; p = 0.005) in B16Arg/Arg subjects(salmeterol, n = 12; placebo, n = 5) as compared with B16Gly/Gly subjects (salmeterol, n = 13; placebo, n = 13). In the second cohort, B16Arg/Arg subjects treated with salmeterol and ICS concurrently (n = 8) had a lower A.M. PEF (36.8 L/min difference, p = 0.048) than B16Gly/Gly subjects (n = 22) treated with the same regimen. In addition, B16 Arg/Arg subjects in the second cohort had lower FEV1 (0.42 L, p = 0.003), increased symptom scores (0.2 units, p = 0.034), and increased albuterol rescue use (0.95 puffs/d, p = 0.004) compared with B16Gly/Gly subjects.

Conclusions: Relative to B16Gly/Gly patients with asthma, B16Arg/Arg patients with asthma may have an impaired therapeutic response to salmeterol in either the absence or presence of concurrent ICS use. Investigation of alternate treatment strategies may benefit this group.

β2-Adrenergic agonists are the most commonly prescribed medications for asthma due to their quick relief of symptoms of airflow obstruction (1). Whether frequent or regularly scheduled use of these medications actually reduces asthma control is controversial (25). Several studies have suggested that short-acting β-agonists such as albuterol may produce adverse outcomes (e.g., decreased peak flow or increased risk of exacerbation) in patients homozygous for arginine (Arg/Arg) at the 16th amino acid position of the β-adrenergic receptor gene (herein referred to as B16) compared with patients homozygous for glycine (Gly/Gly) at B16 (see Figure E1 of the online supplement) (68).

Large, controlled, prospective studies show that long-acting β-agonists are more effective than albuterol for asthma treatment, resulting in their widespread use, especially with concurrent inhaled corticosteroid (ICS) therapy (9). However, a number of studies have also suggested an association between use of long-acting β-agonists with adverse outcomes such as asthma exacerbation and, possibly, increased risk of death (1013). Whether this phenomenon with long-acting β-agonists is related to a genotype-specific susceptibility to adverse effects, as observed with regular use of short-acting β-agonists, or whether the bronchodilating and symptom-relieving effects of salmeterol mask increasing inflammation and delay awareness of worsening asthma is unclear (14).

We sought evidence of a genotype-specific difference in the response to long-acting β-agonists by analyzing two studies in which patients were treated with salmeterol. In the first, subjects had received salmeterol or placebo while inhaled steroids were withdrawn as part of the National Heart Lung and Blood Institute's (NHLBI) Salmeterol or Corticosteroid (SOCS) trial (15). In the second, subjects received salmeterol and a constant dose of ICS as part of the Salmeterol ± Inhaled Corticosteroid (SLIC) trial (16). We hypothesized that if a genotype-specific difference between B16Arg/Arg and B16Gly/Gly subjects were to occur in response to regular salmeterol use, independent of concurrent ICS use, a large proportion of patients with asthma could benefit from modification of asthma treatment regimens.

We genotyped subjects who provided written, informed consent to participate in two published NHLBI Asthma Clinical Research Network trials: the SOCS (15) and SLIC trials (16). On the basis of the data from our previous studies showing a genotype-specific difference in a.m. PEF between B16Arg/Arg and B16Gly/Gly subjects in response to β-agonist therapy (6, 8), we focused our primary analysis on the effect of salmeterol in a.m. PEF in these two genotypes. We secondarily examined the effect of salmeterol on other outcomes in these studies, as well as the effect of other previously identified β-adrenergic receptor gene polymorphisms. The original studies are reviewed below.

Trial Overviews
Trial 1: SOCS trial.

The SOCS trial was a multicenter, 28-wk, randomized, double-blind, placebo-controlled, parallel-group trial comparing long-acting β-agonist monotherapy with continued therapy with ICS in subjects with persistent asthma (see Figure 1) (15). After 6 wk of treatment with open-label triamcinolone acetonide (400 μg twice daily), 164 subjects with well-controlled asthma (FEV1 > 80% predicted, and PEF variability < 20%) were randomized in a double-blind manner for 16 wk to the following groups: (1) to continue inhaled triamcinolone (400 μg twice daily), (2) to substitute ICS with salmeterol xinafoate (42 μg twice daily) via metered-dose inhaler as monotherapy, or (3) to substitute ICS with placebo. The trial concluded with a 6-wk, single-blind, placebo run-out. Albuterol sulfate was used as needed as rescue medication. Although no significant differences between the salmeterol and triamcinolone groups were observed with respect to a.m. PEF, p.m. PEF, asthma symptom scores, rescue albuterol use, or quality of life, the salmeterol group had more asthma exacerbations, as well as greater increases in sputum eosinophils, eosinophilic cationic protein, and tryptase. It was concluded that patients with persistent asthma well controlled by low doses of triamcinolone cannot be switched to salmeterol monotherapy without risk of clinically significant loss of asthma control.

To assess the effect of β2-adrenergic receptor (β2AR) genotype on response to salmeterol as monotherapy, we analyzed the B16 genotype–specific effect of treatment with salmeterol compared with placebo.

Trial 2: SLIC trial.

The SLIC trial was a multicenter 24-wk, randomized, controlled, blinded, double-dummy, parallel group trial examining the effects of adding salmeterol to triamcinolone (see Figure 2) (16). After a 6-wk run-in period with open-label triamcinolone acetonide (400 μg twice daily), 175 subjects whose asthma was not well controlled (FEV1 ⩽ 80% predicted, or average PEF variability ⩾ 20%) received add-on therapy with 42 μg of salmeterol xinafoate (two puffs) twice daily via metered-dose inhaler for 2 wk plus 400 μg of inhaled triamcinolone acetonide twice daily. Half the subjects were then randomly assigned to maintain triamcinolone dosage throughout the trial or undergo a blinded, one-step 50% reduction in triamcinolone for 8 wk followed by an 8-wk triamcinolone elimination phase of salmeterol monotherapy. Albuterol sulfate was used throughout as needed as rescue medication. This study demonstrated that, in patients with persistent asthma suboptimally controlled by ICS therapy alone but whose asthma symptoms improve after addition of salmeterol, a substantial reduction (50%) in ICS dose could occur without a significant loss of asthma control. However, total elimination of triamcinolone therapy results in a significant deterioration in asthma control and, therefore, could not be recommended.

To assess the effect of salmeterol in the setting of concomitant ICS use, we analyzed polymorphisms of the β2AR gene in individuals in the SLIC trial randomized to salmeterol and a steady dose of triamcinolone.

Sample Collection and Genotypic Analysis

Blood was drawn and genetic material collected for genotyping from 53 subjects receiving salmeterol and 43 placebo subjects in the SOCS trial, and from 74 individuals randomized to receive salmeterol and a steady dose of triamcinolone in the SLIC trial. Genomic DNA was prepared for genotypic analysis by standard methods (17). The primary polymorphism studied, A+46G, coding for the 16th amino acid position of β2AR, consisted of either glycine (Gly) or arginine (Arg). The three possible genotypes are termed B16Arg/Arg, B16Arg/Gly, and B16Gly/Gly. Genotypes at B16 were assessed by the amplification refractory mutation system (18, 19) and confirmed in all subjects by restriction fragment length polymorphism by individuals unaware of the clinical trial results (6, 20). We also genotyped seven other polymorphic loci on the β-adrenergic receptor gene (the previously reported polymorphisms C-709A, G-654A, C-47T, C+79G, G+252A, C+491T, and C+523A; see Table E1 for genotype success rates) (21, 22), and performed haplotype analysis as previously described (23). Additional β-adrenergic receptor gene loci genotypes were assessed by single-base extension followed by matrix-assisted laser desorption/ionization–time of flight mass spectrometry (Sequenom, San Diego, CA) (24). We performed additional secondary analyses for single single nucleotide polymorphism (SNP) associations between a.m. PEF and variants at SNP +79 and +523.

Statistical Analysis

In view of our prior results in B16Arg/Arg as compared with B16Gly/Gly patients (6, 8), we set out to compare a.m. PEF primarily, and the other outcomes secondarily, in these two genotype-defined groups. In both trials, patients who met preassigned criteria for treatment failure or asthma exacerbation received protocol-defined rescue treatment with prednisone, inhaled triamcinolone, or both. Because rescue therapy could affect outcomes, data were analyzed by the intent-to-treat method, and the last observation carried forward method, in which the last value observed before treatment failure, asthma exacerbation, or drop-out was carried forward to the end of the 16-wk randomized treatment period in SOCS (Week 22) and to the end of the 18-wk randomized treatment period in SLIC (Week 24). In the SOCS trial, patients were randomized to placebo or salmeterol; the effect was compared with placebo. Statistical analysis included comparison of each response variable to baseline*

* Baseline was defined as the average of Weeks 5 and 6, the 2 wk before randomization, for those response variables that were measured daily (a.m. and p.m. PEF, PEF variability, average symptom scores, and average puffs of β-agonist). Baseline is defined as the measurement at Week 6, the week of randomization, for the remaining response variables (FEV1, FEV1 %predicted, PC20, exhaled nitric oxide, and AQOL score). Change from baseline was calculated for all response variables each available week after baseline by taking the measurement that week and subtracting the baseline value as defined above.

stratified by genotype at the end of treatment (both SOCS and SLIC trials), and in the case of the SOCS trial, at the end of run-out, on the basis of our findings from our prior trial that suggested that a significant effect occurred after withdrawal of the β-agonists (6). In the SOCS trial, because we performed analysis of our primary outcome, change in a.m. PEF, at two endpoints (Weeks 22 and 28), we applied a Bonferroni correction with a significant p value ⩽ 0.025 for the primary endpoint. In the SLIC trial, the primary analysis outcome, change in a.m. PEF, was examined at only one point in time (Week 24). Other outcomes in both trials were examined secondarily in an exploratory manner, and no correction was applied to those outcomes.

In the SOCS trial, for all response variables, except asthma-related quality-of-life (AQOL) score (data available for only select weeks), a two-slope (first six treatment weeks, remaining 10 treatment weeks) segmented model was fitted to the data. A single-slope model was used to evaluate the run-out phase because active therapy was discontinued. In the SLIC trial, slopes were fitted for the first 2 wk of the salmeterol treatment phase, followed by two 8-wk periods of constant triamcinolone therapy. Model estimates of change from baseline for each response variable by genotype were calculated and compared at the end of treatment and run-out. Because AQOL score was not distributed normally, comparisons for change from baseline at the end of treatment and run-out used signed rank and Wilcoxon rank sum tests. All outcomes were adjusted for age and baseline lung function by including them as covariates in the model.

Genetic Analysis

Each of the polymorphic loci was tested for deviations from Hardy-Weinberg equilibrium using the exact test in the computer software Genetic Data Analysis (25). Haplotype blocks were identified using two linkage disequilibrium–based algorithms (26, 27) and one recombination-based algorithm (28) as implemented in the computer program Haploview (27). Association between haplotypes and a.m. PEF was tested using the statistical package haplo.stats (23), which implements a score test to test for statistical significant association between haplotypes and outcome measurements, allowing for adjustment for nongenetic covariates. Adjusting for treatment arms, association between haplotypes formed by all contiguous marker loci (e.g., all adjacent two-marker haplotypes, all adjacent three-marker haplotypes, etc.) and a.m. peak flow changes were tested in both trials. In the SOCS trial, the changes considered were from baseline (Week 6) to Week 22 and from baseline to Week 28. In the SLIC trial, the a.m. peak flow change was from baseline (Week 6) to Week 24. In addition, stratified analyses were also done for several treatment arms, including the SOCS salmeterol arm, the SOCS placebo arm, and the SLIC salmeterol + triamcinolone arm. In the stratified analyses, association between the haplotypes and a.m. peak flow at the end of treatment and run-out (Week 24 in the SLIC trial, Weeks 22 and 28 in the SOCS trial) was tested with adjustment for a.m. peak flow at baseline. Haplotypes formed by all possible marker combinations were considered (e.g., all possible two-marker haplotypes, all possible three-marker haplotypes, etc.).

Trial 1: SOCS Trial

Fifty-four subjects in the SOCS trial were randomized to receive salmeterol and simultaneously have their ICS withdrawn, whereas 56 were randomized to receive placebo. Of the subjects who gave blood samples for genetic analysis, 48 salmeterol subjects (Arg/Arg, n = 12; Gly/Gly, n = 13), and 43 placebo patients (Arg/Arg, n = 5; Gly/Gly, n = 13) were successfully genotyped at the B16 locus. The genotypic distribution at this locus was consistent with the Hardy-Weinberg equilibrium (p = 0.918). Table 1 summarizes the prerandomization baseline characteristics of the Arg/Arg and Gly/Gly subjects who received either salmeterol or placebo. Subjects of different genotypes were well matched except that B16Gly/Gly subjects receiving salmeterol were younger than subjects in the other groups.

TABLE 1. MEAN BASELINE CHARACTERISTICS OF B16Arg/Arg AND B16Gly/Gly SUBJECTS RANDOMIZED TO SALMETEROL AND PLACEBO ARMS OF THE SOCS TRIAL



Arg/Arg Placebo vs. Salmeterol

Gly/Gly Placebo vs. Salmeterol

Combined Arg/Arg vs. Gly/Gly
Characteristic
Placebo (n = 5)
Salmeterol (n = 12)
p Value
Placebo (n = 13)
Salmeterol (n = 13)
p Value
Arg/Arg (n = 17)
Gly/Gly (n = 26)
p Value
Male, n (%)1 (20)5 (42)0.603 (23)5 (38)0.676 (35)8 (31)0.76
Minority, n (%)2 (40)5 (42)1.004 (31)2 (15)0.647 (41)6 (23)0.21
African American, n (%)1 (20)1 (8)0.510 (0)1 (8)1.002 (12)1 (4)0.32
Age, mean (SD), yr33.44 (15.11)38.47 (11.88)0.4733.17 (10.48)25.10 (9.39)0.05**36.99 (12.64)29.14 (10.59)0.03**
a.m. PEF, mean (SD), L/min*446.2 (66)434.8 (125)0.85431.3 (82)480.2 (121)0.24438.2 (109)455.7 (104)0.60
PM PEF, mean (SD), L/min*456.2 (67)444.4 (119)0.84436.9 (82)496.4 (131)0.18447.9 (104)466.6 (112)0.58
PEF variability, mean (SD), %*14.9 (5.9)10.1 (4.5)0.0910.6 (3.7)9.7 (5.4)0.6311.5 (5.3)10.2 (4.6)0.38
FEV1, mean (SD), L2.88 (0.71)2.99 (0.75)0.792.95 (0.61)3.30 (0.96)0.282.96 (0.71)3.13 (0.81)0.49
FEV1% predicted, mean (SD)92.8 (9.4)92.7 (9.5)0.9893.5 (8.8)95.4 (7.9)0.5892.7 (9.2)94.5 (8.2)0.52
PC20, mean (CV), mg/ml1.01 (1.37)0.97 (1.97)0.960.42 (1.45)0.66 (1.61)0.330.98 (1.76)0.54 (1.54)0.11
a.m. symptom scores (SD)§0.5 (0.6)0.3 (0.5)0.570.3 (0.4)0.3 (0.3)0.980.36 (0.50)0.29 (0.33)0.56
p.m. symptom scores (SD)§0.5 (0.5)0.5 (0.5)0.890.4 (0.4)0.4 (0.4)0.920.46 (0.46)0.39 (0.40)0.63
Average rescue puffs β-agonist/d, mean (SD)1.7 (2.9)1.9 (2.2)0.842.6 (3.0)2.4 (2.5)0.851.8 (2.4)2.5 (2.7)0.40
Exhaled nitric oxide, mean (SD), ppb17.5 (7.1)20.5 (18.2)0.8018.4 (11.6)21.3 (12.3)0.6319.5 (14.9)20.0 (11.7)0.93
Asthma quality-of-life score (Q1, Q3)
2.9 (2.5, 3.0)
1.9 (1.4, 2.6)
0.13
1.9 (1.7, 2.84)
1.9 (1.6, 2.)
0.49
2.13 (1.56, 2.97)
1.90 (1.61, 2.53)
0.71

Definition of abbreviations: CV = coefficient of variation; PEF = peak expiratory flow; Q1, Q3 = interquartile range; SOCS = Salmeterol or Corticosteroid trial.

Sex, minority, and African-American status are summarized with frequencies and percentages and are compared via an exact χ2 test. a.m. and p.m. PEF, PEF variability, FEV1, FEV1% predicted, a.m. and p.m. symptom scores, average puffs of β-agonist, exhaled nitric oxide, and PC20 were compared via a two-sample t test.

* Values represent averages for the fifth and sixth weeks of the run-in period.

PEF variability was calculated as ([evening PEF − morning PEF]/evening PEF) × 100.

Geometric mean.

§ Asthma symptoms were graded by the subject each day, from 0 for no symptoms to 3 for incapacitating symptoms.

Median.

Asthma-specific quality-of-life questionnaires were completed by the subjects during clinical-center visits. A score of 1.0 indicates that asthma had no effect on the overall quality of life; a score of 2.0, “a little limited” by asthma; 3.0, “some limitation”; and 7.0, “total limitation.”

** p < 0.05.

During the treatment period, compared with those treated with placebo, a.m. PEF increased in salmeterol-treated Gly/Gly subjects (21.5 L/min, p = 0.050; Table 2; Figure 3). In contrast, it declined by 29.9 L/min (p = 0.040) in salmeterol-treated Arg/Arg subjects compared with placebo, producing a difference between genotypes of 51.4 L/min (p = 0.005; Figure 3; Table 2). At the end of the run-out, this genotype-specific difference had decreased to 32.5 L/min (p = 0.179).

TABLE 2. CHANGE FROM BASELINE TO THE END OF TREATMENT PERIOD OF THE SOCS TRIAL FOR EACH OUTCOME IN B16Arg/Arg AND B16Gly/Gly SUBJECTS RANDOMIZED TO SALMETEROL TREATMENT ARM, COMPARED WITH PLACEBO



A/A Sal-Plac (n = 12, salmeterol; n = 5, placebo)

G/G Sal-Plac (n = 13, salmeterol; n = 13, placebo)

A/A Sal-Plac–G/G Sal-Plac
Outcomes
Difference
p Value
Difference
p Value
Difference
p Value
a.m. PEF, L/min*−29.9 (−58.5, −1.3)0.0421.5 (−0.0, 43.0)0.05−51.4 (−87.2, −15.6)0.005
p.m. PEF, L/min*−11.7 (−40.2, 16.9)0.424.2 (−17.8, 26.1)0.71−15.8 (−51.8, 20.2)0.39
PEF variability, %*2.4 (−1.5, 6.3)0.22−3.0 (−6.0, −0.1)0.055.4 (0.6, 10.3)0.03
FEV1, L*0.06 (−0.23, 0.34)0.690.11 (−0.11, 0.34)0.32−0.05 (−0.42, 0.30)0.76
a.m. symptom scores*0.1 (−0.1, 0.3)0.480.0 (−0.1 0.2)0.830.1 (−0.2, 0.3)0.66
p.m. symptom scores*0.2 (−0.1, 0.4)0.150.0 (−0.2, 0.2)0.990.2 (−0.1, 0.5)0.25
Average rescue puffs β-agonist/d*0.3 (−0.7, 1.3)0.580.1 (−0.7, 0.9)0.840.2 (−1.1, 1.5)0.75
Exhaled nitric oxide, ppb*8.9 (−4.9, 22.6)0.20−4.1 (−14.6, 6.4)0.4413.0 (−4.3, 30.3)0.14
Asthma quality-of-life score§
0.2 (−0.8, 1.1)
0.75
−0.5 (−1.3, 0.4)
0.26
0.6 (−0.7, 1.9)
0.34

Definition of abbreviations: PEF = peak expiratory flow; SOCS = Salmeterol or Corticosteroid trial.

* Two-slope segmented model, mean (95% confidence interval).

PEF variability was calculated as ([evening PEF − morning PEF]/evening PEF) × 100.

Asthma symptoms were graded by the subject each day, from 0 for no symptoms to 3 for incapacitating symptoms.

§ Asthma-specific quality-of-life questionnaires were completed by the subjects during clinical-center visits. A score of 1.0 indicates that asthma had no effect on the overall quality of life; a score of 2.0, “a little limited” by asthma; 3.0, “some limitation”; and 7.0, “total limitation.”

Signed rank test, Wilcoxon rank sum test, median (Q1, Q3).

p < 0.05.

We also compared the change in physiologic outcomes, symptoms, exhaled nitric oxide, albuterol rescue use, and AQOL in the B16Arg/Arg and B16Gly/Gly subjects (Table 2). All of these secondary outcomes (Table 2) tended to deteriorate in B16Arg/Arg subjects compared with B16Gly/Gly subjects at the end of active treatment, but the only outcome that approached statistical significance was change in PEF variability (5.43%; p = 0.030).

There was no significant difference in asthma exacerbations or treatment failures when individuals of different genotypes were treated with salmeterol or placebo: 3 of 12 (25%) B16Arg/Arg subjects and 5 of 13 (38%) B16Gly/Gly subjects treated with salmeterol (p = 0.38) and 1 of 5 (20%) B16Arg/Arg subjects and 5 of 13 (38%) B16Gly/Gly subjects treated with placebo (p = 0.62) experienced an exacerbation or treatment failure by the end of the treatment phase. There was no difference between any B16Arg/Arg and all B16Gly/Gly subjects experiencing an exacerbation or treatment failure (p = 0.888).

In a secondary analysis, we examined the response of the heterozygotes in this study. During the treatment period, heterozygotes treated with salmeterol (n = 23) as compared with those treated with placebo (n = 25) experienced an improvement in a.m. PEF compared with baseline (+18.18 L/min increase over baseline, p = 0.025), which was significantly different from the decline observed with in B16Arg/Arg subjects (p = 0.004; see Table E2).

We also secondarily examined the association of a.m. PEF at the end of treatment and run-out periods with other polymorphic loci in the β2AR gene and haplotypic combinations of these loci (see Methods). Only one test generated a p value less than 0.05 (p = 0.045 for haplotype formed by c47t–c523a with a.m. peak flow at Week 28 in the SOCS trial).

Trial 2: SLIC Trial

Seventy-four subjects in the SLIC trial were randomized to have salmeterol added to ICS therapy (triamcinolone). A total of 58 of 70 subjects from whom blood was obtained were successfully genotyped at B16: 8 were Arg/Arg and 22 were Gly/Gly, consistent with Hardy-Weinberg equilibrium (p = 0.95). Table 3 summarizes the prerandomization characteristics of the B16Arg/Arg and B16Gly/Gly subjects.

TABLE 3. MEAN BASELINE CHARACTERISTICS OF B16Arg/Arg AND B16Gly/Gly SUBJECTS RANDOMIZED TO SALMETEROL AND CONSTANT DOSE OF TRIAMCINOLONE ARM OF SLIC TRIAL


Characteristic

Arg/Arg (n = 8)

Gly/Gly (n = 22)

p Value
Male, n (%)3 (38)12 (55)0.68
Minority, n (%)3 (38)5 (23)0.64
African American, n (%)2 (25)2 (9)0.55
Age, mean (SD), yr28.6 (9.1)36.1 (11.4)0.11
a.m. PEF, mean (SD), L/min*406.3 (87)404.5 (140)0.97
p.m. PEF, mean (SD), L/min*412.8 (83)423.1 (124)0.83
PEF variability, mean (SD), %*11.6 (4.2)13.6 (5.3)0.34
FEV1, mean (SD), L2.46 (0.37)2.53 (0.63)0.77
FEV1% predicted, mean (SD)71.1 (7.6)71.1 (8.4)0.10
PC20, mean (CV), mg/ml1.57 (2.01)2.10 (1.42)0.58
a.m. symptom scores, mean (SD)*§0.2 (0.2)0.3 (0.3)0.33
p.m. symptom scores, mean (SD)*§0.2 (0.2)0.4 (0.3)0.08
Average rescue puffs β-agonist/d, mean (SD)*0.9 (1.3)1.5 (1.1)0.25
Exhaled nitric oxide, mean (SD), ppb14.9 (6.7)17.1 (8.2)0.59
Asthma quality-of-life score, median (Q1, Q3)
1.8 (1.3, 3.3)
2.5 (1.9, 2.7)
0.25

Definition of abbreviations: CV = coefficient of variation; PEF = peak expiratory flow; Q1, Q3 = interquartile range; SLIC = Salmeterol ± Inhaled Corticosteroid trial.

Sex, minority, and African-American status are summarized with frequencies and percentages and are compared via an exact χ2 test. a.m. and p.m. PEF, PEF variability, FEV1, FEV1% predicted, a.m. and p.m. symptom scores, average puffs of β-agonist, exhaled nitric oxide, and PC20 were compared via a two-sample t test.

* Values represent averages for the fifth and sixth weeks of the run-in period.

PEF variability was calculated as ([evening PEF − morning PEF]/evening PEF) × 100.

Geometric mean.

§ Asthma symptoms were graded by the subject each day, from 0 for no symptoms to 3 for incapacitating symptoms.

Asthma-specific quality-of-life questionnaires were completed by the subjects during clinical-center visits. A score of 1.0 indicates that asthma had no effect on the overall quality of life; a score of 2.0, “a little limited” by asthma; 3.0, “some limitation”; and 7.0, “total limitation.”

Figure 4 shows the time course of the change in a.m. PEF during combination therapy. Although B16Arg/Arg subjects (n = 8) appeared to initially benefit from the addition of salmeterol, by the end of 18 wk of randomized treatment, a.m. PEF deteriorated. In contrast, B16Gly/Gly individuals (n = 22) had a sustained beneficial effect from salmeterol, resulting in a genotype-specific difference of 36.8 L/min (p = 0.048) between the two groups. There was no significant difference in asthma treatment failures between the genotypes (2 of 8 [25%] B16Arg/Arg subjects and 1 of 22 [5%] B16Gly/Gly subjects; p = 0.17).

We also examined the change in physiologic outcomes, symptoms, exhaled nitric oxide, albuterol rescue use, and AQOL in the B16Arg/Arg and B16Gly/Gly subjects (Table 4). The p.m. PEF rates, PEF variability, FEV1, rescue puffs of β-agonist, and AQOL all improved in B16Gly/Gly subjects when salmeterol was added to their ICS regimen, although nitric oxide, a marker of airway inflammation, increased. There was no improvement in Arg/Arg subjects in these parameters. The difference between genotypes in the change in these outcomes after salmeterol therapy was significant (p < 0.003 to p < 0.037) for FEV1, a.m. symptom scores, and average β-agonist rescue use (Table 4). There was also a trend (p < 0.1) toward greater PEF variability, and worse AQOL scores in B16Arg/Arg patients compared with B16Gly/Gly patients.

TABLE 4. CHANGE FROM BASELINE TO THE END OF TREATMENT PERIOD OF THE SLIC TRIAL FOR EACH OUTCOME IN SUBJECTS RANDOMIZED TO SALMETEROL AND CONSTANT DOSE OF TRIAMCINOLONE ARM IN B16Arg/Arg AND B16Gly/Gly SUBJECTS



Arg/Arg (n = 8)

Gly/Gly (n = 22)

Arg/Arg vs. Gly/Gly
Outcomes
Difference Compared with Baseline
p Value
Difference Compared with Baseline
p Value
Difference between Genotypes
p Value
a.m. PEF, L/min*−18.1 (−49.8, 13.6)0.2618.7 (0.6, 36.9)0.04**−36.8 (−73.4, −0.3)0.05**
p.m. PEF, L/min*13.8 (−13.7, 41.1)0.3215.3 (−0.2, 30.8)0.04**−1.5 (−33.0, 30.0)0.92
PEF variability, %*−0.1 (−3.9, 3.8)0.97−4.3 (−6.4, −2.2)0.0002**4.2 (−0.2, 8.6)0.06
FEV1, L*−0.14 (−0.37, 0.10)0.240.28 (0.14, 0.41)<0.0001**−0.42 (−0.68, −0.15)0.003**
a.m. symptom scores*0.1 (0.0, 0.3)0.15−0.1 (−0.2, 0.0)0.070.20 (0.0, 0.4)0.03**
p.m. symptom scores*0.1 (−0.1, 0.2)0.45−0.1 (−0.00, 0.01)0.050.2 (−0.0, 0.4)0.11
Average rescue puffs β-agonist/d*0.46 (−0.08, 1.0)0.09−0.49 (−0.80, −0.18)0.003**0.95 (0.33, 1.58)0.004**
Exhaled nitric oxide, ppb*−3.92 (−14.21, 6.36)0.455.58 (0.04, 11.12)0.05**−9.50 (−21.19, 2.18)0.11
Asthma quality-of-life score§
−0.04 (−0.47, 1.06)
0.78
−0.41 (−0.72, −0.13)
0.001**
0.37 (0.25, 1.19)
0.08

Definition of abbreviation: PEF = peak expiratory flow.

* Three-slope segmented model, mean (95% confidence interval).

PEF variability was calculated as ([evening PEF − morning PEF]/evening PEF) × 100.

Asthma symptoms were graded by the subject each day, from 0 for no symptoms to 3 for incapacitating symptoms.

Asthma-specific quality-of-life questionnaires were completed by the subjects during clinical-center visits. A score of 1.0 indicates that asthma had no effect on the overall quality of life; 2.0, “a little limited”; 3.0, “some limitation”; and 7.0, “total limitation.”

§ Signed rank test, Wilcoxon rank sum test, median (Q1, Q3).

p < 0.1.

** p < 0.05.

We performed a secondary analysis of the effect of salmeterol on a.m. PEF in heterozygotes in SLIC (see Table E2). Heterozygotes who received salmeterol (n = 27) had a 44.3 L/min increase in a.m. PEF (p < 0.0001 compared with baseline), which was significantly different from the B16Arg/Arg patient response (p = 0.004). We also found no association between other polymorphisms of the β2AR gene or their haplotypic combination and a.m. PEF.

Although the long-acting β-agonist salmeterol improves asthma control in non–genotype-stratified groups (9, 2931), our current analyses, from two separate trials, suggest that in a subset of patients with asthma (the one-sixth of patients who are B16Arg/Arg), airway function and indices of asthma control do not improve to the same degree with salmeterol treatment, either with or without concurrent ICS use. These results are similar to those of previous studies that suggested that regular use of the short-acting β-agonist albuterol may produce adverse effects in B16Arg/Arg subjects with asthma (6, 7) and that they may benefit by withdrawal from β-agonists (6, 8).

In the first trial (SOCS), in which subjects with milder asthma had their ICS treatment withdrawn and were treated instead with salmeterol or placebo, B16Arg/Arg subjects who received salmeterol alone, compared with placebo, had a significantly worse response with respect to a.m. PEF compared with B16Gly/Gly individuals. In the second trial (SLIC), in which subjects with more severe asthma received salmeterol while remaining on ICS, a.m. PEF in B16Gly/Gly subjects improved significantly compared with B16Arg/Arg subjects. In this cohort, there were also genotype-specific differences in FEV1, rescue albuterol use, and symptoms. In both trials, similar to a prior retrospective analysis (6), heterozygotes harboring the B16Arg/Gly genotype responded in a manner similar to that of B16Gly/Gly subjects.

It is clear from our data that B16Arg/Arg subjects do not do as well as B16Gly/Gly subjects when treated with salmeterol. Although we were only able to detect genotype-specific differences in a.m. PEF and a trend toward a difference in PEF variability in our first cohort, the fact that we observed differences in symptoms and medication use in our second cohort suggests that the effect is not restricted to PEF alone. Furthermore, the genotype-specific differences in a.m. PEF in these two trials (51.4 and 36.8 L/min differences in a.m. PEF in the SOCS and SLIC trials, respectively) were substantial and were similar in both magnitude and pattern to those reported with short-acting β-agonists in retrospective studies (6, 7) (see Figure E1), as well as in a randomized prospective trial (8). Smaller differences in a.m. PEF (e.g., < 20 L/min) have been associated with clinically important changes in asthma control (32, 33). Of interest, we did not observe a significant genotype-specific difference in p.m. PEF or in some of the other outcomes. This may relate to the sample size or to changes in duration of salmeterol effect with continued use (34).

The reason that B16Arg/Arg patients failed to achieve as salutary a response to salmeterol as B16Gly/Gly subjects is unclear. Some have speculated that it may be due to differences in receptor down-regulation between polymorphic variants of β2AR (35) or to genotype-specific difference in loss of bronchoprotection (36). Others have suggested that persistent β2AR-induced activation of extracellular signal-regulated kinases in airway smooth muscle cells contributes to mitogenesis and exaggerated inflammatory cytokine expression (37). These effects may be affected by polymorphisms of the receptor. A recent report that alterations in the β2AR gene can affect the signaling and function of other receptors that control airway contractility (38) suggests an additional mechanism. These mechanisms might explain the slow onset reported here and in our prior findings with short-acting β-agonists (6, 8) and are consistent with an apparent initial salutary response to salmeterol in B16Arg/Arg patients in the SLIC trial. Furthermore, it is also possible that the B16Arg/Arg polymorphism may be in linkage disequilibrium with the true etiologic polymorphism. In this regard, similar to recent findings that failed to demonstrate an association between acute bronchodilator response and β2-adrenoceptor haplotype in patients with asthma (39), we were unable to detect stronger associations using combinations of additional polymorphisms (haplotypes) in the gene. Last, we cannot rule out the possibility that the genotype-specific effects associated with salmeterol are related to a secondary increase in albuterol use in B16Arg/Arg patients. However, the fact that B16Arg/Arg subjects receiving salmeterol used more albuterol than B16Gly/Gly subjects suggests that they had sufficient symptoms to warrant the increased use of this rescue medication.

Prior studies with short-acting β-agonists have demonstrated an increased risk of asthma exacerbations in B16Arg/Arg subjects receiving albuterol on a regular basis (7). Although salmeterol appears to have a negative impact in B16Arg/Arg patients not receiving ICS, the question of whether salmeterol is indeed deleterious (with declining lung function or increasing exacerbations, as occurs with short-acting β-agonists) in these subjects when used with ICS is less clear from our data. In the SOCS trial, when ICS were discontinued, salmeterol in and of itself was associated with a decline in a.m. PEF compared with placebo in B16Arg/Arg subjects. In SLIC, although the outcomes trended in a negative direction, the decline did not reach statistical significance. Whether our failure to demonstrate a deleterious effect in SLIC may have been due to the small number of B16Arg/Arg patients in this arm or due to the concurrent use of ICS is unclear.

This documented decline in B16Arg/Arg subjects treated with salmeterol compared with placebo in SOCS raises the question of whether the reports of a possible small increased rate of death and/or severe deteriorations in asthma symptoms in subjects receiving salmeterol (12, 13) (or other long-acting β-agonists such as formoterol [40]) are related to the effect we observed by genotype. Of note, there were no such severe deteriorations or deaths observed in our studies. Although such potential β-agonist–related adverse effects were reported to occur more frequently in African Americans (13), who have an increased frequency of B16Arg/Arg (41), and although ethnic-specific pharmacogenetic differences have been demonstrated in individuals with different β-adrenergic genotypes (42), it is important to note that inclusion of African Americans did not bias our study finding. There were no more than two African-American subjects of each genotype in any arm of our studies (Tables 1 and 3), and their exclusion had little effect on our results. Furthermore, we did not observe differences between the genotypes in rates of asthma exacerbation as has been previously observed with short-acting β-agonists (7). This may be due to the fact that the rate of exacerbations was not high enough in our cohorts to detect such a difference.

In summary, in two separate trials, including one with concomitant ICS treatment, we demonstrated that B16Arg/Arg patients have a substantially diminished therapeutic response to the regular use of salmeterol, compared with those harboring the B16Gly/Gly genotype. In fact, in the setting of ICS withdrawal, B16Arg/Arg patients may experience adverse effects with salmeterol. These findings suggest that B16Arg/Arg patients (∼ one-sixth of whites and ∼ one-fifth of African Americans) might benefit from alternate asthma treatment strategies. Before implementation, our findings would benefit from prospective confirmation and further investigation of concurrent factors that might affect this altered response.

The authors thank the following individuals who were previous Asthma Clinical Research Network (ACRN) principal investigators and who participated in the SOCS and SLIC trials: James Fish, M.D., and Jean Ford, M.D. They also thank Reuben Cherniack, M.D., Chairman of the ACRN Steering Committee, and Hector Ortega, M.D., and Gene Pesola, M.D., Ph.D., who have served on the ACRN Steering Committee, for their assistance.

1. Nelson HS. Drug therapy: beta-adrenergic bronchodilators. N Engl J Med 1995;333:499–506.
2. Sears MR, Taylor DR, Print CG, Lake DC, Li QQ, Flannery EM, Yates DM, Lucas MK, Herbison GP. Regular inhaled beta-agonist treatment in bronchial asthma. Lancet 1990;336:1391–1396.
3. Drazen JM, Israel E, Boushey HA, Chinchilli VM, Fahy JV, Fish JE, Lazarus SC, Lemanske RF, Martin RJ, Peters SP, et al. Comparison of regularly scheduled with as-needed use of albuterol in mild asthma. N Engl J Med 1996;335:841–847.
4. Sears MR. Is the routine use of inhaled beta-adrenergic agonists appropriate in asthma treatment? No. Am J Respir Crit Care Med 1995;151:600–601.
5. Wanner A. Is the routine use of inhaled beta-adrenergic agonists appropriate in asthma treatment? Yes. Am J Respir Crit Care Med 1995;151:597–599.
6. Israel E, Drazen JM, Liggett SB, Boushey HA, Cherniack RM, Chinchilli VM, Cooper DM, Fahy JV, Fish JE, Ford JG, et al., for the National Heart Lung and Blood Institute's Asthma Clinical Research Network. The effect of polymorphisms of the β2-adrenergic receptor on the response to regular use of albuterol in asthma. Am J Respir Crit Care Med 2000;162:75–80.
7. Taylor DR, Drazen JM, Herbison GP, Yandava CN, Hancox RJ, Town GI. Asthma exacerbations during long term beta-agonist use: influence of beta2 adrenoceptor polymorphism. Thorax 2000;55:762–767.
8. Israel E, Chinchilli VM, Ford JG, Boushey HA, Cherniack RM, Craig TJ, Deykin A, Fagan JK, Fahy JV, Fish J, et al., for the National Heart Lung and Blood Institute's Asthma Clinical Research Network. Genotype stratified prospective trial of regularly scheduled albuterol treatment in asthma. Lancet 2004;364:1505–1512.
9. Pearlman DS, Chervinsky P, LaForce C, Seltzer JM, Southern DL, Kemp JP, Dockhorn RJ, Grossman J, Liddle RF, Yancey SW. A comparison of salmeterol with albuterol in the treatment of mild-to-moderate asthma. N Engl J Med 1992;327:1420–1425.
10. Bisgaard H. Effect of long-acting β2 agonists on exacerbation rates of asthma in children. Pediatr Pulmonol 2003;36:391–398.
11. D'Alonzo GE, Nathan RA, Henochowicz S, Morris RJ, Ratner P, Rennard SI. Salmeterol xinafoate as maintenance therapy compared with albuterol in patients with asthma. JAMA 1994;271:1412–1416.
12. Castle W, Fuller R, Hall J. The Serevent Nationwide Surveillance study. BMJ 1993;306:1034–1037.
13. U.S. Food and Drug Administration. Medwatch Safety Information and Adverse Event Reporting Program: 2003 safety alert—Serevent (salmeterol xinafoate). Cold Spring Harbor, NY: U.S. Food and Drug Administration; 2003.
14. McIvor RA, Pizzichini E, Turner MO, Hussack P, Hargreave FE, Sears MR. Potential masking effects of salmeterol on airway inflammation in asthma. Am J Respir Crit Care Med 1998;158:924–930.
15. Lazarus SC, Boushey HA, Fahy J, Chinchilli VM, Lemanske RF, Sorkness C, Kraft M, Fish J, Peters SP, Craig TJ, et al.; Asthma Clinical Research Network. A randomized study of long-acting beta-agonists in patients with persistent asthma: I. Monotherapy. JAMA 2001;285:2583–2593.
16. Lemanske RF, Sorkness CA, Mauger EA, Lazarus SC, Boushey HA, Fahy JV, Drazen JM, Chinchilli VM, Craig TJ, Fish JE, et al.; Asthma Clinical Research Network. Inhaled corticosteroid reduction and elimination in patients with persistent asthma receiving salmeterol: a randomized controlled trial. JAMA 2001;285:2594–2603.
17. Maniatis T, Fritsch EF, Sambrook J. Molecular cloning: a laboratory manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1982.
18. Newton CR, Graham A, Heptinstall LE, Powell SJ, Summers C, Kalsheker N, Smith JC, Markham AF. Analysis of any point mutation in DNA: the amplification refractory mutation system (ARMS). Nucleic Acids Res 1989;17:2503–2516.
19. Newton CR, Heptinstall LE, Summers C, Super M, Schwarz M, Anwar R, Graham A, Smith JC, Markham AF. Amplification refractory mutation system for prenatal diagnosis and carrier assessment in cystic fibrosis. Lancet 1989;2:1481–1483.
20. Martinez FD, Graves PE, Baldini M, Solomon S, Erickson R. Association between genetic polymorphisms of the beta 2-adrenoceptor and response to albuterol in children with and without a history of wheezing. J Clin Invest 1997;100:3184–3188.
21. Silverman EK, Kwiatkowski DJ, Sylvia JS, Lazarus R, Drazen JM, Lange C, Laird NM, Weiss ST. Family-based association analysis of β2-adrenergic receptor polymorphisms in the childhood asthma management program. J Allergy Clin Immunol 2003;112:870–876.
22. Drysdale CM, McGraw DW, Stack CB, Stephens JC, Judson RS, Nandanbalan K, Arnold K, Ruano G, Liggett SB. Complex promoter and coding region beta 2-adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness. Proc Natl Acad Sci USA 2000;97:10483–10488.
23. Schaid DJ, Rowland CM, Tines DE, Jacobson RM, Poland GA. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet 2002;70:425–434.
24. Ross P, Hall L, Smirnov I, Haff L. High level multiplex genotyping by MALDI-TOF mass spectrometry. Nat Biotechnol 1998;16:1347–1351.
25. Lewis PO, Zaykin D. Genetic Data Analysis: computer program for the analysis of allelic data. Version 1.0 (d16c). Available from: http://lewis.eeb.uconn.edu/lewishome/software.html (accessed July 2003).
26. Gabriel S, Schaffner S, Nguyen H, Moore J, Roy J, Blumenenstiel B, Higgins J, Defelice M, Lochner A, Faggart M, et al. The structure of haplotype blocks in the human genome. Science 2002;296:2225–2229.
27. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 2005;21:263–265.
28. Hudson R, Kaplan H. Statisticial properties of the number of recombination events in the history of a sample of DNA sequences. Genetics 1985;111:147–164.
29. Greening AP, Ind PW, Northfield M, Shaw G; Allen and Hanburys Limited U.K. Study Group. Added salmeterol versus higher-dose corticosteroid in asthma patients with symptoms on existing inhaled corticosteroid. Lancet 1994;344:219–224.
30. Woolcock A, Lundback B, Ringdal N, Jacques LA. Comparison of addition of salmeterol to inhaled steroids with doubling of the dose of inhaled steroids. Am J Respir Crit Care Med 1996;153:1481–1488.
31. Pauwels RA, Lofdahl C-G, Postma DS, Tattersfield AE, O'Byrne P, Barnes PJ, Ullman A; Formoterol and Corticosteroids Establishing Therapy (FACET) International Study Group. Effect of inhaled formoterol and budesonide on exacerbations of asthma. N Engl J Med 1997;337:1405–1411.
32. Chervinsky P, Vanas A, Bronsky EA, Dockhorn R, Noonan M, LaForce C, Pleskow W. Fluticasone propionate aerosol for the treatment of adults with mild to moderate asthma. J Allergy Clin Immunol 1994;94:676–683.
33. Haahtela T, Jarvinen M, Kava T, Kiviranta K, Koskinen S, Lehtonen K, Nikander K, Persson T, Reinikainen K, Selroos O, et al. Comparison of a beta 2-agonist, terbutaline, with an inhaled corticosteroid, budesonide, in newly detected asthma. N Engl J Med 1991;325:388–392.
34. Nelson JA, Strauss L, Skowronski M, Ciufo R, Novak R, McFadden ER. Effect of long-term salmeterol treatment on exercise- induced asthma. N Engl J Med 1998;339:141–146.
35. Green SA, Cole G, Jacinto M, Innis M, Liggett SB. A polymorphism of the human β2-adrenergic receptor within the fourth transmembrane domain alters ligand binding and functional properties of the receptor. J Biol Chem 1993;268:23116–23121.
36. Lee DKC, Currie GP, Hall IP, Lima JJ, Lipworth BJ. The arginine-16 beta2-adrenoceptor polymorphism predisposes to bronchoprotective subsensitivity in patients treated with formoterol and salmeterol. Br J Clin Pharmacol 2004;57:68–75.
37. Shore SA, Drazen JM. Beta-agonists and asthma: too much of a good thing? J Clin Invest 2003;112:495–497.
38. McGraw DW, Almoosa KF, Paul RJ, Kobilka BK, Liggett SB. Antithetic regulation by β-drenergic receptors of Gq receptor signaling via phospholipase C underlies the airway β-agonist paradox. J Clin Invest 2003;112:619–626.
39. Taylor DR, Epton MJ, Kennedy MA, Smith AD, Iles S, Miller AL, Littlejohn MD, Cowan JO, Hewitt T, Swanney MP, et al. Bronchodilator response in relation to β2-adrenoceptor haplotype in patients with asthma. Am J Respir Crit Care Med 2005;172:700–703.
40. Mann M, Chowdhury B, Sullivan E, Nicklas R, Anthracite R, Meyer R. Serious asthma exacerbations in asthmatics treated with high-dose formoterol. Chest 2003;124:1544.
41. Ellsworth DL, Coady SA, Chen W, Srinivasan SR, Elkasabany A, Gustat J, Boerwinkle E, Berenson GS. Influence of the b2-adrenergic receptor Arg16Gly polymorphism on longitudinal changes in obesity from childhood through young adulthood in a biracial cohort: the Bogalusa Heart Study. Int J Obes Relat Metab Disord 2002;26:928–937.
42. Choudhry S, Ung N, Avila PC, Ziv E, Nazario S, Casal J, Torres A, Gorman JD, Salari K, Rodriguez-Santana JR, et al; the Genetics of Asthma in Latino Americans Study. Pharmacogenetic differences in response to albuterol between Puerto Ricans and Mexicans with asthma. Am J Respir Crit Care Med 2005;171:563–570.
Correspondence and requests for reprints should be addressed to Elliot Israel, M.D., Brigham and Women's Hospital, 75 Francis Street, Boston, MA, 02115. E-mail:

Related

No related items
American Journal of Respiratory and Critical Care Medicine
173
5

Click to see any corrections or updates and to confirm this is the authentic version of record