Acute Asthma Responsiveness and B2 Adrenergic Receptors Polymorphisms

The most recently available statistics indicate that approximately 22.2 million people in
the United States have asthma. Of these, about 55% experienced at least one attack in the
year before the survey resulting in 1.9 million visits to emergency departments and 497,000
hospitalizations. The natural history of acute asthma is one of resolution and 70 to 80% of
people rapidly clear their airway obstruction following treatment with 7.5 to 10.0 mg of
nebulized albuterol. However, there is a subset of patients with recurrent severe episodes
that are poorly responsive to this treatment and require admission to hospital for
resolution. Although these patients represent a minority of asthmatics, they have persistent
disease that accounts for the vast majority of expenditures for urgent care. Yet virtually
nothing is known about them. In the present application, we wish to test the hypothesis that
these individuals do poorly with albuterol because of genetic polymorphisms of their beta 2
adrenergic receptors (β2AR) that either influence severity and/or reduce responsiveness to
short acting bronchodilators.

One possibility of great clinical significance is that the albuterol non-responders have
genetically determined differences in β2AR that down regulate effectiveness. Various
polymorphisms have been described in this receptor with the greatest attention being devoted
to single nucleotide polymorphisms (SNPs) substitutions at amino acid positions 16 (arginine
to glycine, Arg 16 Gly) and 27 (glutamine to glutamic acid, Gln 27 Glu). The Gly 16 receptor
exhibits enhanced down regulation in vitro after exposure to agonist while the Arg 16 allele
is more resistant. The Gly 16 allele is strongly associated with asthma destabilizations as
manifested by nocturnal symptom and an increased risk of severity. In contrast, Arg/Arg
homozygosity at position 16 has been associated with poorer outcomes in prospective studies
of regular use of short-acting β2 agonists. This phenomenon was not seen in asthmatics that
used albuterol on an as needed basis or those who were homozygous for glycine. Since over 90
% of the subjects in our studies, and those of others, on acute asthma take this drug
routinely and the vast majority are treated with it emergently, the presence of the Arg/Arg
allele could lead to poorer responses. It has also been suggested that this polymorphism may
also produce similar effects with long-acting β2 agonists. Again, since a large percentage
of asthmatic patients chronically use long acting sympathomimetics, this factor may have a
contributory effect. Alternatively, β2AR pharmacogenetics may be totally unimportant. Two
studies have implied that neither single genotypes at Codon 16, nor haplotypes in the β2AR
have any effect on the acute bronchodilator response to either albuterol or salbutamol.
However, since these trials were performed in stable patients, they may not be
representative of the non-responsive group described above.

In the current study we propose to build upon previous methods we have used in the care of
acute asthma. Clinical assessments and treatment regimens are standardized using a care path
as in previous studies. Demographics, height, weight, race, symptoms, signs, past medical
histories, and routine medication use for the previous month prior to the index visit are
recorded on an intake sheet. We routinely obtain information on disease duration, frequency
of attacks, hospitalizations, ICU admissions, intubations, and a list of the general and
specific triggers for acute exacerbations. We also record asthma medications including oral
and inhaled steroids, short and long-acting beta-adrenergic agonists, antileukotriene drugs,
anticholinergics and methylxanthines. We include in our histories all of the concurrent
medications including eye drops (beta-blockers and prostaglandins), cardiac and
anti-hypertensive drugs, (beta-adrenergic active agents and ACE inhibitors. We ask people
how they are taking their asthma drugs, but have no way of knowing if they truly are using
them as prescribed. If their medications are obtained at our hospital it may be possible to
determine how many prescriptions are filled, but we still will not know use. We also record
the use of illicit drugs such as cocaine, heroine, and OxyContin, etc. Body mass index (BMI)
is calculated as weight in Kg divided by height in meters squared. Obesity is classified as
a BMI ≥ 30 Kg/M2. Arterial oxygen saturation is determined by pulse oximetry on room air and
supplemental oxygen is administered as necessary for values less than 90%. Before treatment,
the best of three peak expiratory flow rates (PEFR) is taken as representing the patient's
initial state of flow limitation. The PEFR data is recorded in absolute terms as well as a
percentage of predicted normal. As in previous studies, if a patient's airway obstruction
prevented him or her from achieving the minimum level on the peak flow meter, a value of 10
is arbitrarily assigned to avoid dividing by zero when the percentage improvement following
albuterol is calculated.

The subjects receive either 2.5 mg of nebulized albuterol every 20 min for three doses or
two doses of 5.0 mg of albuterol 20 min apart. Albuterol non-responsiveness is defined as a
failure of the PEFR in an acutely ill asthmatic to exceed 40% of predicted following ≥7.5 mg
of albuterol (2.5 mg albuterol aerosols q.20 min x3). After completion of each schedule,
PEFR is repeated and the patients reexamined. Admission and discharge decisions are made
according to published predetermined criteria. Patients are considered ready to be sent home
if they are asymptomatic, free of accessory muscle use, have absent or diminished wheezing,
and have achieved a peak flow of 60% of predicted. Those not meeting these requirements are
given further treatment with adrenergic and anticholinergic bronchodilators and
glucocorticoids at the discretion of the ED physician and reassessed hourly. If they
subsequently meet the discharge criteria, they are released. If not, they are admitted to
hospital. The false-positive admission rate with these algorithms has been found to be less
than 1% and the 24-hour relapse rate less than 2%.

Only individuals in whom the index ED visit was for the treatment of an acute asthma attack
are studied. Patients with histories suggestive of congestive heart failure, pneumonia,
chronic bronchitis, or emphysema are excluded.

Venous blood will be drawn from 1250 subjects. It is expected that 25% will have obstruction
that is poorly responsive to albuterol and require admission to hospital for resolution. The
other 75% will have obstruction that quickly resolves with albuterol and who are discharged
home. The DNA will be isolated and assayed blindly by the Center for Human Genetics
Research, Vanderbilt Medical Center for identification of β2 AR SNPs at positions 16 and 27
and receptor haplotypes. Single-nucleotide polymorphisms will be genotyped by use of TaqMan
single-nucleotide polymorphism genotyping assays (Applied Biosystems, Foster City, CA, USA).
Eleven polymorphisms will be selected for genotyping from within the ADRB2 coding region and
up to about 5 kb 5' Codon 1 (rs11958940, rs17778257, rs2895795, rs2053044, rs12654778,
rs11959427, rs1042711, rs1042713, rs1042714, rs1800888, and rs1042718). Quality-control
measures include a genotype call rate of 95% and no significant departures from
Hardy-Weinberg Equilibrium within each racial group (exact p value > 0.01). Haplotypes will
be predicted with publicly available software (SNPHNP-a program for estimating frequencies
of large haplotypes of single-nucleotide polymorphisms [version 1.3]). Haplotype pairs 2/2,
2/4, 2/6, 4/4, 4/6 and 6/6 will be evaluated.

The primary outcome variable will be the genetic patterns (SNPs and haplotypes) in each
population. We wish to evaluate SNP patterns and albuterol responsiveness using a case
control study. Between-group comparisons will be made by analysis of covariance. The
associations between baseline PEFR and the percent improvement following albuterol will be
stratified for SNP pattern and will be assessed by analysis of covariance. Single
polymorphisms may be insufficient and it has been suggested that haplotypes containing
multiple polymorphisms may be more informative. Drysdale and coworkers identified 13
polymorphisms and organized them into 12 haplotype groups with varying frequencies between
ethnic groups. The four most common were haplotypes 1, 2, 4 and 6. Haplotype 2, which
contains Gly16 homozygotes, was associated with the greatest FEV1, reversibility while
haplotype 4, which contains Arg16 homozygotes, was the least. Five haplotype pairs were
common in asthmatics. Homozygotes, haplotypes 2/2 had the greatest impact on FEV1 and
protein expression while haplotype pair 4/4 showed the least effect in both categories.
Since no isolated SNP may have any predictive utility, that is why we are measuring the
haplotypes. The haplotype combination 4/4 produces the smallest impact and may be
operational in the non-responders. The haplotype pair 4/6 produces the largest improvement
with albuterol, and, therefore, may be operational in the responsive group.

We will estimate haplotype frequencies across the pool of study patients. Using logistic
regression, we will calculate a global score statistic to rest for the overall association
between haplotypes and responsiveness to albuterol. Next, haplotype specific scores and
corresponding p-values will be calculated. Similar statistics will be calculated adjusting
for baseline covariates. These analyses will be carried out using R functions in
haplo.stats.

We will evaluate the associations of haplotypes 1-6 with response/non-response to adrenergic
agonists using a case-control study. Using the data in Table 1 and assuming that the cases
make up 25% of the study subjects, and using the function haplo.power.cc. We require a
sample size of 1243 patients to be able to detect at least a 40% increase in risk of
non-responsive to adrenergic agonists in patients with haplotypes containing Arg16Arg
compared to haplotypes without Arg16Arg with 80% power at 0.05 level of significance. Under
similar assumptions, we require a total sample size of 865 to be able to detect a 50%
increase in risk of treatment failure in patients with haplotypes containing Arg16Arg
compared to patients without. The number of subjects represents the total required to
achieve sufficient power given a 75:25 treatment success-failure ratio.

The results of this study will demonstrate whether Gly16Arg or Gln27Glu genotypes and/or
β2AR haplotypes are associated with the response to albuterol in patients who experience
acute episodes of asthma and whether genotype pattern varies between different clinical
phenotypes. They will also provide insights into whether such polymorphisms determine the
severity of the episodes as measured by initial pulmonary function and symptoms. In this
investigation, both positive and negative results would be extremely important. Positive
findings would suggest that we can begin to make sense out of therapeutic responsiveness at
a very basic level. They also suggest that we need to begin to search for alternative
treatments for a significant segment of the asthmatic population when they become acutely
ill. Equally important it means that our efforts on asthma control need no longer be
diffusely applied, but now can be focused on prophylaxis in affected individuals. Negative
findings will conclusively demonstrate that genetically determined changes in B2AR cannot
readily be translated into clinical responses and that we need to look for other factors
that determine severity and resource utilization in the beta2 unresponsive population.

Inclusion Criteria:

- Acute asthma

Exclusion Criteria:

- Any other condition