Respiratory Medicine (2009) 103, 98e103

ava i lab le a t www.sc iencedi rec t .com

journa l homepage : www.e lsev ie r . com/ loca te / rmed

CT scanning-based phenotypes vary with ADRB2polymorphisms in chronic obstructive pulmonarydisease

Woo Jin Kim a,o, Yeon-Mok Oh b,o, Joohon Sung c, Young Kyung Lee d,Joon Beom Seo e, NamKug Kim e, Tae-Hyung Kim f, Jin Won Huh g,Ji-Hyun Lee h, Eun-Kyung Kim h, Jin Hwa Lee i, Sang-Min Lee j,Sangyeub Lee k, Seong Yong Lim l, Tae Rim Shin m, Ho Il Yoon n,Sung-Youn Kwon n, Sang Do Lee b,*

a Department of Internal Medicine, College of Medicine, Kangwon National University, Chuncheon, South Koreab Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine and Clinical Research Center forChronic Obstructive Airway Diseases, Asan Medical Center, University of Ulsan College of Medicine,Seoul, South Koreac Department of Cancer Prevention and Epidemiology, National Cancer Center, South Koread Department of Radiology, Bundang CHA Hospital, University of Pocheon Jungmoon College of Medicine, Seongnam,South Koreae Department of Radiology, Asan Medical Center, University of Ulsan College of Medicine, Research Institute of Radiology,Seoul, South Koreaf Division of Pulmonology, Department of Internal Medicine, Hanyang University Guri Hospital, Hanyang University Collegeof Medicine, Guri, South Koreag Department of Internal Medicine, Ilsan Paik Hospital, Inje University, Goyang, South Koreah Department of Internal Medicine, Bundang CHA Hospital, University of Pocheon Jungmoon College of Medicine,Seongnam, South Koreai Department of Internal Medicine, Ewha Womans University Mokdong Hospital, College of Medicine, Ewha WomansUniversity, Seoul, South Koreaj Division of Pulmonary and Critical Care Medicine, Department of Internal Medicine, Seoul National University College ofMedicine, Clinical Research Institute, Seoul National University Hospital, Lung Institute, Medical Research Center, SeoulNational University College of Medicine, Seoul, South Koreak Division of Respiratory and Critical Care Medicine, Department of Internal Medicine, College of Medicine, KoreaUniversity Anam Hospital, Seoul, South Koreal Division of Pulmonary and Critical Care Medicine, Department of Medicine, Kangbuk Samsung Hospital, SungkyunkwanUniversity School of Medicine, Seoul, South Koream Department of Internal Medicine, Kangnam Sacred Heart Hospital, Hallym University College of Medicine,Seoul, South Korea

* Corresponding author. Tel.: þ82 2 3010 3140; fax: þ82 2 3010 6968.E-mail address: sdlee@amc.seoul.kr (S.D. Lee).

o Woo Jin Kim and Yeon-Mok Oh contributed equally to this article.

0954-6111/$ - see front matter ª 2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.rmed.2008.07.025

CT phenotypes vary with the ADRB2 polymorphism 99

n Respiratory Center, Seoul National University Bundang Hospital, Department of Internal Medicine, Seoul NationalUniversity College of Medicine, Seongnam, South Korea

Received 5 March 2008; accepted 25 July 2008Available online 11 September 2008

KEYWORDSCOPD;Polymorphism;Computed tomography

Summary

Background: Chronic obstructive pulmonary disease (COPD) is a heterogeneous disease that ischaracterized by varying degrees of involvement of airway and lung parenchyma. Althoughcigarette smoke is the major risk factor for COPD, the principal determining factors of involve-ment of the airway or lung parenchyma have not been clearly defined. Genetic variability inCOPD patients might influence the varying degrees of involvement of airway and parenchyma.We therefore studied whether airway and parenchyma involvement might be associated withthe ADRB2 genotype, which has been reported to be associated with COPD susceptibility andthe bronchodilator response.Methods: One hundred and eleven COPD subjects, whose post-bronchodilator FEV1/FVC valueswere less than 0.7, and who had histories of smoking exceeding 10 pack-years, were prospec-tively recruited from pulmonology clinics of 11 hospitals in Seoul, Korea. The degrees ofinvolvement of airway and parenchyma were evaluated by volumetric computed tomography(CT) scans. In-house software automatically calculated luminal areas, airway wall areas,percentages of wall areas in segmental bronchi, emphysema indices, and mean lung densitiesin the whole lung parenchyma. The ADRB2 genotypes at codon 16 were determined for allpatients.Results: Gly16 was associated with lumen diameter, luminal area, and percentage of wall areain patients with COPD (p Z 0.02), whereas neither wall area nor wall thickness differed withADRB2 genotype. Neither emphysema index nor mean lung density was associated with ADRB2genotype.Conclusion: Gly16 variant in ADRB2 gene was associated with airway wall phenotypesmeasured using CT scanning in COPD patients.ª 2008 Elsevier Ltd. All rights reserved.

Introduction

Chronic obstructive pulmonary disease (COPD) is defined asairflow limitation measured by spirometry, and is a hetero-geneous disease characterized by varying degrees ofinvolvement of airway and lung parenchyma.1 Chronicinflammation and structural changes are the major patho-logic features of COPD.1 Although some subjects showemphysema as the predominant feature,2 remodeling inairways is seen to various extents in COPD patients.3

Although cigarette smoking is the most important riskfactor, the mechanisms of persistent inflammation, thecauses of structural damage, and the factors that modifythe involvement of airway or parenchyma, have not beenclearly elucidated.

Computed tomography (CT) imaging is a useful tool in theevaluation of morphological changes in COPD patients andcan also be helpful in evaluating heterogeneous COPDphenotypes.4 CT phenotypes including airway wall thick-ening, and emphysema severity, have been related todifferent clinical characteristics and varying treatmentresponsiveness in COPD,5 and such phenotypes may be usefulresearch biomarkers.6 Genetic variability in COPD patientsmight explain persistent inflammation and the varyingdegrees of involvement of airway and parenchyma. To date,

there have been several reports regarding the association ofCT phenotypes and particular genotypes.7e9 The phenotypesassociated with particular genes are mainly the severity ofemphysema and emphysema distribution. Recent tech-nology has made it possible to evaluate airway thick-ness.10,11 Airway phenotypes can therefore be evaluated toexamine whether there might be an association betweensuch phenotypes and particular genotypes. A previous reportsuggested that gender may influence airway lumen andairway thickness.12 However, little research on the geneticbasis of airway thickness has been conducted.

ADRB2 is one of the most studied genes for COPDsusceptibility and work on this gene has yielded contra-dictory results.13e15 Some data suggest that ADRB2 geno-type may be associated with bronchodilator responsivenessto both short-term and long-term use of b2-agonist inasthma treatment.16,17

Recently, we developed volumetric CT analysis tools forevaluation of emphysema and airway characteristics inCOPD patients.18 Quantified phenotypes of emphysema andairway using volumetric CT might be helpful in searching formore genetically homogeneous subgroups displayingparticular genetic influences.

We therefore investigated whether there might be anassociation between emphysema and quantitative CT

100 W.J. Kim et al.

phenotypes of the airway, and ADRB2 genotypes on theother, which could be associated with the pathogenesis ofairway disease in COPD patients.

Methods

Subjects and informed consent

The subjects for this study were extracted from a patientcohort entitled ‘‘The Korean Obstructive Lung Disease(KOLD) Cohort’’, which consists of patients with chronicobstructive pulmonary disease (COPD) or asthma. The KOLDCohort was designed primarily to develop a systematicdiagnostic model and an integrative prognostic factor ofobstructive lung diseases. For the KOLD Cohort, thepatients with chronic respiratory symptoms as well asairflow limitation or bronchial hyperresponsiveness havebeen and will be recruited in the pulmonary clinics of 11hospitals in South Korea from June 2005 to October 2012.For this study, a kind of interim analysis of the KOLD Cohortstudy, a total of 145 patients were recruited from June 2005to December 2006, and complete CT scanning data andother clinical information, such as blood analysis and dataon pulmonary function were obtained.

In this study, the 118 COPD subjects extracted from theKOLD Cohort of 145 patients met all of the followingcriteria. They had post-bronchodilator FEV1/FVC values of<0.7 and had more than 10 pack-years of smoking history aswell as no or minimal abnormalities on chest radiographs.Among 118 COPD subjects, only 111 patients were analyzedfor this study after the exclusion of patients whose CT scanswere not performed adequately.

Our Institutional Review Board approved the analyses ofthe clinical and imaging data. Individual informed writtenconsent was obtained from all patients.

Measurements of airway and lung parenchymausing CT scans

Volumetric CT scans were performed on all patients using16-channel, multidetector, CT machines of three manu-facturers. These included the Somatom Sensation 16(Siemens Medical Solutions, Forchheim, Germany), the GELightspeed Ultra (General Electric Healthcare, Milwaukee,WI), and the Philips Brilliance 16 (Philips Medical Systems,Best, Netherlands). Patients were scanned during sus-pended full inspiration and expiration in the supine posi-tion. CT parameters used in the different CT scanners wereas follows: 16� 0.75 mm collimation, 100 eff. mAs, 140 kVp(Somatom Sensation 16); 16� 0.625 mm, 300 mAs, 140 kVp,Pitch 0.938, 0.5 s/rot (GE Lightspeed); and 16� 0.75 mm,133 mAs, 140 kVp, pitch 1, 0.75 s/rot (Philips 16). Theacquired data were reconstructed using a standard algo-rithm with 0.625e0.8 mm thickness and 0.625e0.8 mmincrement. The CT machines were calibrated every weekwith an AAPM standard phantom. The image data werestored in the Digital Imaging and Communications in Medi-cine (DICOM) format; this is the international standard forinterconnecting medical imaging devices on standardnetworks.

Using in-house software, images of the whole lung wereextracted automatically, and the attenuation coefficient ofeach pixel was measured and calculated. The cutoff levelbetween normal lung density and a low-attenuation areawas defined as �950 HU. From the CT data, the volumefraction of the lung below �950 HU (V950) and the meanlung density (MLD) were calculated automatically.Measurements of the airway dimensions was performednear the origins of segmental bronchi (RB1, LB1þ 2)selected by a radiologist who was blind to clinical results,using in-house software. The software automaticallydetects the airway lumen, and the inner and outer bound-aries of the airway wall, using a full-width-half-maximum(FWHM) method. The FWHM method is typical of objective,quantitative approaches to automatic airway measurementmethods.19e21 The software was validated using polyacryltubes with variable inner diameters and wall thick-nesses.22,23 In each segmental bronchus, the airwaydimensions, including the wall area (WA), lumen area (LA),and wall area percentage (WA %), were measured. The wallarea percentage was defined as WA % Z WA/(WAþ LA)� 100. The mean value of each segmental bron-chus was used for statistical analysis. All of these analyseswere performed by one of the authors (YL) who was blind tosubjects’ background data.

Genotyping

ADRB2 genotyping was performed on all patients. GenomicDNA was prepared from blood for genotype analysis.Genotypes of codon 16 (rs1042713) were determined usingpolymerase chain reaction and restriction fragment lengthpolymorphism (RFLP) techniques as previously described.16

Restriction digests were electrophoresed on 4% (w/v)agarose gels and visualized using ethidium bromide.

Statistics

Analysis of variance (ANOVA) was used to analyze thebaseline characteristics to determine any differencesamong genotypes. Statistical analyses of CT parameterswere used to determine any differences between geno-types. HardyeWeinberg equilibrium was tested by the chi-square method. Phenotypes tested included inspiratoryand expiratory emphysema indices, inspiratory and expi-ratory mean lung densities, lumen area, lumen diameter,wall thickness, wall area, and wall area percentage.Inspiratory emphysema indices and expiratory emphysemaindices were log transformed to approximate normaldistributions. The associations between CT parameters andADBR2 genotype were examined under the additive modelassuming that an addition of each risk allele increases asthe same amount of risk. For example, the model assumesthat having two Gly alleles doubles the risk in comparisonto having one Gly allele. The analysis was adjusted forage, gender, smoking status, body mass index (BMI), andbaseline FEV1. A random effect model was used wherehospital-influenced characteristics with CT phenotypeswere modeled as random errors24 using Proc Mixedprocedure of PC-SAS for Windows version 9.2 (SAS Insti-tute, Cary, NC).

Table 2 Airway wall indices using CT scanning accordingto ADRB2 genotype.

Genotype Arg/Arg Arg/Gly Gly/Gly p value

Lumenarea(mm2)

11.1� 2.9 10.7� 2.6 9.5� 3.0 0.02

Lumendiameter(mm)

3.63� 0.48 3.58� 0.46 3.34� 0.55 0.02

Airwaythickness(mm)

1.23� 0.12 1.21� 0.12 1.24� 0.10 0.59

Wall area(mm2)

17.8� 5.4 17.7� 5.2 16.4� 6.0 0.22

Wall area (%) 63.6� 4.2 63.7� 4.2 66.4� 5.4 0.02

Mean values (�SEM) for data are shown. Associations of gly16polymorphism and CT parameters were estimated assumingadditive genetic model (adjusting age, sex, smoking, FEV1, BMI,and random effects of recruited hospitals).

CT phenotypes vary with the ADRB2 polymorphism 101

Results

Characteristics of subjects

One hundred and eleven patients (108 men and 3 women)were analyzed. Their mean (�SD) age, FEV1, and smokinghistory were 65.4� 7.4 years (range 47e81 years),1.48� 0.57 L, and 45.9� 25.0 pack-years, respectively.

Amongst the 111 COPD patients, 40 patients were of theArg/Arg genotype, 46 patients were Arg/Gly, and 25patients were Gly/Gly. The prevalence of Arg16 poly-morphisms did not significantly deviate from the HardyeWeinberg equilibrium. The ages, FEV1 values, and smokinghistories were not significantly different amongst the threegroups of patients (Table 1).

Airway and lung parenchymal measurements usingCT scanning according to genotype

Lumen area was significantly smaller in patients with Gly16(p Z 0.0225). Lumen diameter was significantly shorter(p Z 0.0208) and wall area percentage was significantlyhigher in the presence of the Gly allele (p Z 0.0205)(Table 2). Airway wall thickness and wall area were notassociated with ADRB2 genotypes.

Inspiratory emphysema index, inspiratory mean lungdensity, expiratory emphysema index, and expiratory meanlung density did not vary with ADRB2 codon 16 genotype(Table 3).

Discussion

In this study, lumen area, lumen diameter and wall areapercentage of airway were associated with the ADRB2genotype whereas emphysema phenotypes were not. Ourfinding is the first report suggesting that genetic variabilitymay influence airway phenotypes in COPD patients.

Our finding that COPD patients with the Gly16 allele inthe ADRB2 showed smaller lumen areas and larger wall areapercentages could be explained in several ways.

Firstly, the Gly16 allele in the ADRB2 might increasesusceptibility to COPD. A report in Chinese claimed thatGly16 increased susceptibility to COPD.15 Although theclaim was not supported by other reports of inconsistent

Table 1 Baseline characteristics of 111 subjects withCOPD by ADRB2 genotypes.

Genotype Arg/Arg Arg/Gly Gly/Gly p Value

Number ofsubjects(male/female)

40 (39/1) 46 (45/1) 25(24/1)

Age (years) 65.3� 6.6 65.0� 8.5 66.3� 6.5 0.76Baseline

FEV1 (L)1.41� 0.58 1.53� 0.53 1.45� 0.56 0.56

Smoking(pack-years)

46.4� 25.5 50.0� 27.2 38.3� 17.8 0.17

Data are presented as mean (�SD), unless otherwise noted.Analysis of variance (ANOVA) was used to calculate p values.

findings,13,25 as race could have influenced associationresults, our finding suggesting smaller lumen areas in Gly16alleles is compatible with the work showing the risk ofCOPD in Chinese subjects with Gly16 alleles.

Second, the Gly16 allele in ADRB2 might be related toa greater degree of airway obstruction. A previous reportdemonstrated that Gly16 was associated with betterresponse to the long-term use of a short acting b2-agonist,17,26 leading others to hypothesize that a greaterobstruction at baseline leads to better response to bron-chodilator treatment in patients with the Gly16 genotype.27

Our findings also support the hypothesis that moreobstruction is seen in patients with the Gly16 allele.

Lastly, Gly16 allele in the ADRB2 might be associatedwith the chronic bronchitis subtype. Previous reportrevealed that COPD patients with chronic bronchitissymptoms showed higher wall area percentage and thick-ness to diameter than COPD patients without chronicbronchitis symptoms.28 In our study, ADRB2 genotype wasassociated with wall area percentage but not with wallthickness nor with wall area. This result suggests an asso-ciation with chronic bronchitis. However, we cannot ruleout the possibility that this genotype is related to lumendilation but not to airway wall phenotype. This warrantsfurther research.

Our finding that genetic variability in the ADRB2 mayinfluence airway phenotypes in COPD patients could raisean important issue in clinical practice. Since airwayphenotypes are important in COPD, factors influencingthese phenotypes can be a target for drug development.

There are some limitations in this study. First, accordingto a recent report results may differ with the use of varioustypes of CT scanners, and with the radiation doses used.29

To minimize the variation of the result, we modified theimaging protocols with similar reconstruction methods,resolution, and radiation dose. However, usage of differentmachines from different vendors with different imagingprotocols may increase the measurement errors, resultingin weakening the correlation with the CT phenotypes.

Table 3 Emphysema indices according to ADRB2 genotype.

Genotype Arg/Arg Arg/Gly Gly/Gly p Value

Inspiratory emphysemaindex (%)

24.4� 16.2 21.1� 16.0 17.1� 12.7 0.38

Inspiratory meanlung density (HU)

�887.4� 24.3 �882.6� 29.4 �881.4� 18.6 0.57

Expiratory emphysemaindex (%)

14.0� 14.1 11.1� 13.0 10.0� 11.9 0.75

Expiratory meanlung density (HU)

�839� 45.2 �834� 44.3 �840� 31.7 0.90

Mean values (�SEM) for data are shown. Associations of gly16 polymorphism and CT parameters were estimated assuming additivegenetic model (adjusting age, sex, smoking, FEV1, BMI, and random effects of recruited hospitals).

102 W.J. Kim et al.

Secondly, we could not measure small airways, butrather only large airways, so this study may only reflectlarge airways. According to one study, however, largeairway data parallel information from small airways.30

Thirdly, we analyzed only codon 16 in our ADRB2 geno-typing. It is suggested that ADRB216 is of functional impor-tance and this genotype has been extensively studied in bothclinical trials, and to define functional aspects of COPD.17,27

Although there may be unsuspected between-race differ-ences in the importance of this codon, and although explo-ration of other haplotypes might be informative,31 codon 16seems, at present, to be the most informative.

Lastly, most of our patients were male, and there maybe gender differences in CT phenotypes of COPD subjects.12

Our results may therefore not apply to female patients.In conclusion, ADRB2 gene polymorphism was associated

with airway wall phenotypes measured using CT scanning inCOPD patients. CT phenotype differences may be related todifferent subtypes that may be explained by geneticfactors. Future research might include replication in addi-tional cohorts and identification of functional relevance.

Conflict of interest

None declared.

Acknowledgment

This study was supported by a grant from the Korean Health21 R&D Project, Ministry of Health and Welfare, Republic ofKorea (A040153).

References

1. RabeKF,Hurd S,AnzuetoA,Barnes PJ,Buist SA,CalverleyP, etal.Global strategy for the diagnosis, management and prevention ofchronic obstructive pulmonary disease. Am J Respir Crit CareMed 2007;176:532e55.

2. Boschetto P, Miniati M, Miotto D, Braccioni F, De Rosa E,Bononi I, et al. Predominant emphysema phenotype in chronicobstructive pulmonary disease patients. Eur Respir J 2003;21:450e4.

3. James AL, Wenzel S. Clinical relevance of airway remodeling inairway diseases. Eur Respir J 2007;30:1420e41.

4. Muller NL, Coxson H. Chronic obstructive pulmonary disease 2:imaging the lungs in patients with chronic obstructive pulmo-nary disease. Thorax 2002;57:982e5.

5. Kitaguchi Y, Fujimoto K, Kubo K, Honda T. Characteristics ofCOPD phenotypes classified according to the findings of HRCT.Respir Med 2006;100:1742e52.

6. Schuster DR. The opportunities and challenges of developingimaging biomarkers to study lung function and disease. AmJ Respir Crit Care Med 2007;176:224e30.

7. DeMeo DL, Hersh CP, Hoffman EA, Litonjua AA, Lazarus R,Sparrow D, et al. Genetic determinants of emphysema distri-bution in the national emphysema treatment trial. Am J RespirCrit Care Med 2007;176:42e8.

8. Sakao S, Tatsumi K, Igari H, Watanabe R, Shino Y, Shirasawa H,et al. Association of tumor necrosis factor-a gene promoterpolymorphism with low attenuation areas on high-resolutionCT in patients with COPD. Chest 2002;122:416e20.

9. Ito I, Nagai S, Handa T, Muro S, Hirai T, Tsukino M, et al. Matrixmetalloproteinase-9 promoter polymorphism associated withupper lung dominant emphysema. Am J Respir Crit Care Med2005;172:1378e82.

10. de Jong PA, Muller NL, Pare PD, Coxson HO. Computed tomo-graphic imaging of the airways: relationship to structure andfunction. Eur Respir J 2005;26:140e52.

11. Hasegawa M, Nasuhara Y, Onodera Y, Makita H, Nagai K, Fuke S,et al. Airflow limitation and airway dimensions in chronicobstructive pulmonary disease. Am J Respir Crit Care Med2006;173:1309e15.

12. Martinez FJ, Curtis JL, Sciurba F, Mumford J, Giardino ND,Weinmann G, et al. Sex differences in severe pulmonaryemphysema. Am J Respir Crit Care Med 2007;176:243e52.

13. Matheson MC, Ellis JA, Raven J, Johns DP, Walters EH,Abramson MJ. b2-adrenergic receptor polymorphisms areassociated with asthma and COPD in adults. J Hum Genet 2006;51:943e51.

14. Brøgger J, Steen VM, Eiken HG, Gulsvik A, Bakke P. Geneticassociation between COPD and polymorphisms in TNF, ADRB2and EPHX1. Eur Respir J 2006;27:682e8.

15. Ho LI, Harn HJ, Chen CJ, Tsai NM. Polymorphism of theb2-adrenoceptor in COPD in Chinese subjects. Chest 2001;120:1493e9.

16. Martinez FD, Graves PE, Baldini M, Solomon S, Erickson R.Association between genetic polymorphisms of theb2-adrenoceptor and response to albuterol in children withor without a history of wheezing. J Clin Invest 1997;100:3184e8.

17. Israel E, Chinchilli VM, Ford JG, Boushey HA, Cherniack R,Craig JJ, et al. Use of regularly scheduled albuteroltreatment in asthma: genotype-stratified, randomized,placebo-controlled cross-over trial. Lancet 2004;364:1505e12.

CT phenotypes vary with the ADRB2 polymorphism 103

18. Lee YK, Oh YM, Lee JH, Kim EK, Lee JH, Kim N, et al. Quan-titative assessment of emphysema, air trapping, and airwaythickening on computed tomography. Lung 2008;186:157e65.

19. Amirav I, Kramer SS, Grunstein MM, Hoffman EA. Assessment ofmethacholine-induced airway constriction by ultrafast high-reso-lution computed tomography. J Appl Physiol 1993;75:2239e50.

20. Block M, Liu YH, Harris D, Robb RA, Ritman EL. Quantitativeanalysis of a vascular tree model with the dynamic spatialreconstructor. J Comput Assist Tomogr 1984;8:390e400.

21. Wood SA, Zerhouni EA, Hoford JD, Hoffman EA, Mitzner W.Measurement of three-dimensional lung tree structures byusing computed tomography. J Appl Physiol 1995;79:1687e97.

22. Kim N, Seo JB, Song KS, Chae EJ, Kang SH. Semi-automaticmeasurement of the airway dimension at computed tomog-raphy using the full-width-half-maximum method: study on themeasurement accuracy according to ct parameters and size ofthe airway. Korean J Radiol 2008;9:226e35.

23. Kim N, Seo JB, Song KS, Chae EJ, Kang SH. Semi-automaticmeasurement of the airway dimension at computed tomog-raphy using the full-with-half-maximum method: study on themeasurement accuracy according to orientation of an artificialairway. Korean J Radiol 2008;9:236e42.

24. Zeger SL, Liang KY, Albert PS. Models for longitudinal data:a generalized estimating equation approach. Biometrics 1988;44:1049e60.

25. Summerhill E, Leavitt SA, Gidley H, Parry R, Solway J, Ober C.b2-adrenergic receptor Arg16/Arg16 genotype is associatedwith reduced lung function but not with asthma in the Hut-terities. Am J Respir Crit Care Med 2000;162:599e602.

26. Hizawa N, Makita H, Nasuhara Y, Betsuyaku T, Itoh Y,Nagai K, et al. b2-adrenergic receptor genetic polymorphismsand short-term bronchodilator responses in patients withCOPD. Chest 2007;132:1482e92.

27. Liggett SB. Pharmacogenetics of beta-1 and beta-2-adrenergicreceptors. Pharmacology 2000;61:167e73.

28. Orlandi I, Moroni C, Camiciottoli G, Bartolucci M, Pistolesi M,Villari N, et al. Chronic obstructive pulmonary disease: thin-section CT measurement of airway wall thickness and lungattenuation. Radiology 2005;234:604e10.

29. Yuan R, Mayo JR, Hogg JC, Pare PD, McWilliams AM, Lam S,et al. The effects of radiation dose and CT manufacturer onmeasurements of lung densitometry. Chest 2007;132:617e23.

30. Nakano Y, Wong JC, de Jong OA, Buzatu L, Nagao T, Coxson HO,et al. The prediction of small airway dimensions using computedtomography. Am J Respir Crit Care Med 2005;171:142e6.

31. Silverman EK, Kwitatkowski DJ, Sylvia JS, Lazarus R,Drazen JM, Lange C, et al. Family-based association analysis ofb2-adrenergic receptor polymorphisms in the childhoodasthma management program. J Allergy Clin Immunol 2003;112:870e6.