Example: bankruptcy

Aortic Size Assessment by Noncontrast Cardiac …

Aortic Size Assessment by Noncontrast Cardiac Computed Tomography: Normal Limits by Age, Gender, and Body Surface Area …




Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Text of Aortic Size Assessment by Noncontrast Cardiac …

Aortic Size Assessment by Noncontrast CardiacComputed Tomography: Normal Limits by Age,Gender, and Body Surface AreaArik Wolak, MD,* Heidi Gransar, MS,* Louise E. J. Thomson, MB CHB,*John D. Friedman, MD, FACC,* Rory Hachamovitch, MD, FACC,* Ariel Gutstein, MD,*Leslee J. Shaw, PHD, FACC,* Donna Polk, MD, MPH,* Nathan D. Wong, PHD, Rola Saouaf, MD,* Sean W. Hayes, MD,* Alan Rozanski, MD, FACC,*Piotr J. Slomka, PHD,* Guido Germano, PHD, FACC,* Daniel S. Berman, MD, FACC*Los Angeles and Irvine, CaliforniaOBJECTIVESTo determine normal limits for ascending and descending thoracic aorta diameters in alarge population of asymptomatic, low-risk adult of aortic size is possible from gated noncontrast computed tomography(CT) scans obtained for coronary calcium measurements. However, normal limits for aortic size by thesestudies have yet to be 4,039 adult patients undergoing coronary artery calcium (CAC) scanning, systematicmeasurements of the ascending and descending thoracic aorta diameters were made at the level of thepulmonary artery bifurcation. Multiple linear regression analysis was used to detect risk factorsindependently associated with ascending and descending thoracic aorta diameter and exclude subjectswith these parameters from the final analysis. The final analysis groups for ascending and descendingthoracic aorta included 2,952 and 1,931 subjects, respectively. Subjects were then regrouped by gender,age, and body surface area (BSA) for ascending and descending aorta, separately, and for each group,the mean, standard deviation, and upper normal limit were calculated for aortic diameter as well as forthe calculated cross-sectional aortic area. Also, linear regression models were used to create BSA versusaortic diameter nomograms by age groups, and a formula for calculating predicted aortic size by age,gender, and BSA was , BSA, gender, and hypertension were directly associated with thoracic aortadimensions. Additionally, diabetes was associated with ascending aorta diameter, and smoking wasassociated with descending aorta diameter. The mean diameters for the final analysis group were 33 4 mm for the ascending and 24 3 mm for the descending thoracic aorta, respectively. Thecorresponding upper limits of normal diameters were 41 and 30 mm, limits of ascending and descending aortic dimensions by noncontrast gatedcardiac CT have been defined by age, gender, and BSA in a large, low-risk population of subjects undergoingCAC scanning. (J Am Coll Cardiol Img 2008;1:200 9) 2008 by the American College of CardiologyFoundationFrom the *Departments of Imaging (Division of Nuclear Medicine), Department of Medicine (Division of Cardiology), andCSMC Burns & Allen Research Institute, Cedars-Sinai Medical Center, and the Department of Medicine, University ofCalifornia at Los Angeles, School of Medicine, Los Angeles, California; and the Heart Disease Prevention Program, Divisionof Cardiology, University of California, Irvine, California. This study was supported in part by a grant from The EisnerFoundation, Los Angeles, California. Drs. Wolak and Gutstein are fellows of Save a Heart Foundation, Los Angeles,California, and American Physicians Fellowship, Boston, Massachusetts. H. William Strauss, MD, acted as Guest Editor forthis received September 7, 2007; revised manuscript received November 14, 2007, accepted November 21, : CARDIOVASCULAR IMAGINGVOL. 1, NO. 2, 2008 2008 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATIONISSN 1936-878X/08/$ BY ELSEVIER From: on 01/29/2013Thoracic aortic aneurysm is a common, po-tentially lethal, but treatable disease, particu-larly if detected before dissection or assessment of aortic size is a keycomponent in this detection and in guiding thera-peutic decisions. Multiple imaging modalities areavailable for assessing the thoracic aorta, includingX-ray angiography, transesophageal echocardiogra-phy (TEE), computed tomography (CT), and mag-netic resonance imaging (MRI). Although all ofthese modalities have diagnostic value, CT hasevolved to be the mainstay of evaluation owing to itsaccuracy and reproducibility, as well as its speed,simplicity, and true 3-dimensional distinguish the normal from the enlargedaorta, it is necessary to standardize the values of normal aortic dimensions. Due to fundamentaldifferences in the imaging techniques, normal limitsof thoracic aorta dimensions for CT are has been published regarding normal limitsfor thoracic aortic dimensions with CT (1 3).Moreover, to our knowledge, no publications todate have reported these limits with gated, noncon-trast chest CT studies the studies commonly usedfor coronary artery calcium (CAC) , our aim was to determine normal limitsfor ascending and descending thoracic aorta diam-eters in a large population of asymptomatic, low-risk adult subjects undergoing CAC examined the noncontrast gated CT findings of4,387 patients, age 26 to 92 years, free of knownclinical coronary heart disease (CHD), who under-went CAC scanning during the period from July2004 to March 2007 at Cedars-Sinai MedicalCenter. Subjects were self-referred (n 44, 1%),referred by their physician (n 3,308, 75%), orrecruited as part of ongoing research (EISNER[Early Identification of Subclinical Atherosclerosisusing NoninvasivE Imaging Research]-1, -2, and-3) protocols (n 1,035, 24%). An additional 547patients with incomplete data were excluded ( ,missing height, missing weight, and/or aorta diam-eter not measured). The study was approved by theCedars-Sinai Medical Center Institutional regarding the presenceof categorical cardiac risk factors was collected inevery patient through written questionnaires. Sys-temic arterial hypertension was defined as a docu-mented history of high blood pressure. Currentsmoking or history of smoking was defined aspositive smoking status. Hypercholesterolemia wasdefined according to the National Cholesterol Ed-ucation Program (4) guidelines: low-density li-poprotein (LDL) 100 mg/dl and Framinghamrisk score 20%, 2 or more risk factors and LDL 130 mg/dl, or LDL 160 mg/dl. All lipid mea-surements were within 1 week of the CT were classified as having diabetes if theycarried an established diagnosis of diabetes mellitusmade by a physician and/or were receiving treat-ment with insulin or oral hypoglycemic agents, or iftheir measured fasting glucose was 126 history of coronary artery disease (CAD)was defined as a CAD event occurring in a first-degree relative (men age 65 years and women age 55 years). Weight and height were alsoobtained and body mass index (BMI) andbody surface area (BSA) were calculatedusing the Mosteller (5) method. Amongthe 4,387 subjects, 348 subjects with miss-ing clinical data (56, 111, 269, and 277cases of missing information about smok-ing, diabetes mellitus, dyslipidemia, andhypertension, respectively) were imaging CT studies wereperformed on 1 of 4 CT scanners: 2electron beam computed tomography(EBCT) scanners (GE Imatron, C-150 ore-Speed, GE Medical Systems, San Fran-cisco, California), a 16-slice multidetectorcomputed tomography (MDCT) scanner(Brilliance CT scanner, Philips, Cleve-land, Ohio), and a dual source computedtomography (DSCT) scanner (SomatomDefinition, Siemens, New York, NewYork). For EBCT, we used a protocol of 3-mmslice thickness, 50- or 100-ms exposure time, 130kVp, 630-mA tube current, either a 300- or350-mm field of view for reconstruction, and asharp reconstruction kernel. For MDCT, the pro-tocol was slice thickness, 140 kVp,168-mA tube current, 250-ms exposure time, and a350-mm field of view. For DSCT, the protocol was3-mm slice thickness, 140 kVp, 150-mA tubecurrent, 200-ms exposure time, and a 350-mm fieldof view. Prospective electrocardiogram triggering atthe heart rate-dependent percentage of the R-Rinterval was used for all scanners. The CT imageswere acquired in the craniocaudal direction fromimmediately inferior to the aortic arch to the levelof the diaphragm in a single breath-hold for allABBREVIATIONSAND ACRONYMSBMI body mass indexBSA body surface areaCAC coronary artery calciumCAD coronary artery diseaseCHD coronary heart diseaseCT computed tomographyDSCT dual source computedtomographyEBCT electron beamcomputed tomographyLDL low-density lipoproteinMDCT multidetectorcomputed tomographyMRI magnetic resonanceimagingTEE transesophagealechocardiographyJACC: CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT201Downloaded From: on 01/29/2013examinations. For interpretation, the images weretransferred to a dedicated workstation (NetraMD,ScImage, Los Altos, California). Systematic mea-surements of the outer aortic wall perpendicular tothe axis of rotation of the aorta (Fig. 1) in the axialplane at the lower level of the pulmonary arterybifurcation in both the ascending and descendingaorta were made by a licensed radiology technicianexperienced in cardiac CT. Measurements in excessof cm were verified by an imaging cardiologist( , , , or ). Ascendingand descending aorta cross-sectional areas werecalculated from the measured diameter and indexedto continuous variables wereassessed for normality using the Shapiro-Franciatest and assessed visually by inspection of histo-grams and standardized normal probability (P-P)plots. Continuous variables were compared usingthettest for 2 groups and categorical variables werecompared using the Pearson chi-square multiple linear regression analysis, wesought to detect parameters that are associated withascending and descending thoracic aorta size and toexclude patients with these parameters from thefinal analysis groups; therefore, the final analysisgroup would include only patients without param-eters that influence the aorta size and can be calledessentially normalized. We employed a 2-stagemodel. A main effects model, which included po-tentially clinically important predictors ( , age,gender, smoking, family history of CHD, diabetes,dyslipidemia, and hypertension), was developed instage I and a main effects plus interactions model instage II. Separate models were developed for as-cending and descending aorta. Although eitherBMI or BSA formulas can be used for body size,BSA was chosen as the adjusting body size variablefor all subsequent analyses. This is because BSAwas previously found to have a greater associationwith thoracic aortic diameter than BMI does (6,7),and BSA was the body size variable that enteredinto selection models most , regression diagnostics (statisticalmethods used for quality assurance of regressionmodels) (8), including checking the assumptions oflinear regression, analysis of residuals, checking formulticollinearity, model specification, and modelvalidation using a jackknifing procedure, were usedto assess the stage I models. The most influentialoutliers or high leverage subjects were thus identi-fied and then excluded as part of the the stage I models, all 2 2 interactionswere tested for significance at the relaxed level, both singly and against their maineffects. Additionally, the significant interactionterms were tested against each other using stepwisemodel selection to produce the stage II models. Allpotential interactions were investigated. Regressiondiagnostics were done on stage II models, as de-scribed for stage I. None of the predictors wereshown to be collinear, except when interactionterms were introduced. Standardized regression co-efficients were used to assess the relative contribu-tion of the predictors in both stage I and II on these analyses, we were able to identifysingle predictors or interactions that had bothsignificant (p ) and strong influence on aorticsize (nonstandardized beta coefficient in ab-solute value, meaning either mm or ). Subjects with influential predictors or mani-festing high leverage on the model s diagnosticswere later excluded from the final analysis the final analysis, we grouped the subjects bygender, age, and BSA for the ascending and de-scending aorta, separately. We calculated the meanaortic diameter, standard deviation, and upper nor-mal limit (mean 2 standard deviations) for eachgroup. We also used linear regression models tocreate BSA versus aortic diameter nomogramsbased on the former stratification method. BecauseFigure Aorta MeasurementsTransaxial slice at the lower level of the pulmonary artery bifurcation from acomputed tomography coronary artery calcium scan in a patient with nor-mal aortic dimensions showing the method for deriving the ascending anddescending aortic dimensions. For interpretation, the images were trans-ferred to a dedicated workstation (NetraMD, ScImage, Los Altos, California).Thewhite arrowsrepresent outer wall thoracic ascending aortic diameterand thoracic descending aortic diameter measurements perpendicular to theaxis of rotation of the : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT202Downloaded From: on 01/29/2013previous reports used aortic diameter and areainconsistently, we also calculated aortic area. Fi-nally, we created a formula for calculation of pre-dicted aortic size by age, gender, and BSA. Theupper limit of normal was chosen as 2 standarddeviations above the data were analyzed using Stata version 8(StataCorp, College Station, Texas), Analyze Itversion (Analyse-it Software, Leeds, England),and SPSS version 12 (SPSS Inc., Chicago, Illinois).RESULTSTable 1shows the clinical characteristics of thepopulation. Male patients were significantly differ-ent from female patients in all parameters otherthan prevalence of diabetes and 2shows the stage I models of potentialclinical predictors of aortic diameter for the ascend-ing and descending aorta, separately. Both modelsfound that age, gender, BSA, diabetes, and hyper-tension were significant predictors of aortic diame-ter. Smoking was an independent predictor only ofdescending aortic II models (significant stage I variables interactions) are presented inTable 3. For both theascending and descending aorta, age, BSA, diabe-tes, hypertension, and an interaction between ageand male gender were significant predictors ofaortic diameter. Interactions between BSA andsmoking, BSA and hypertension, and hypertensionand dyslipidemia were found to be exclusively asso-ciated with descending aortic the ascending aorta, hypertension and diabe-tes mellitus were found in stage II models to besignificant and influential parameters; therefore,1,225 subjects with hypertension and/or diabeteswere excluded from thefinal analysis. For the de-Table Characteristics of the Initial Study PopulationOverall(n 4,039)Female Patients(n 1,529)Male Patients(n 2,510)pValueAge (yrs) (m2) (kg/m2) ,626 ( )645 ( )981 ( ) history of CAD1,443 ( )608 ( )835 ( ) ( )82 ( )110 ( ) ,444 ( )477 ( )967 ( ) ,083 ( )396 ( )687 ( ) aorta (mm) aorta (mm) body mass index; BSA body surface area; CAD coronary artery I Models: Potentially Clinically Important Predictors of Aortic DiameterAscending Aortan 4,039Descending Aortan 4,035Adjusted R2 R2 (Standardized)StandardErrorpValueUnstand ardizedCoefficient(Standardized) ( ) ( ) ( ) ( ) * ( ) ( ) ( ) ( ) history of CAD ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) *Per as in Table : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT203Downloaded From: on 01/29/2013scending aorta, hypertension and smoking were sig-nificant and influential parameters; therefore, 2,286patients with hypertension and/or smoking were ex-cluded from the final analysis. We additionally ex-cluded 193 and 55 cases of outliers for the ascendingaorta and descending aorta, respectively, after deter-mining that they were influential and had high lever-age and, thus, were likely to be abnormal. The finalanalysis groups included 2,952 subjects (1,147 femaleand 1,805 male subjects) and 1,931 subjects (736female and 1,195 male subjects) for the ascending anddescending aorta, mean aortic diameters for the final analysisgroup in the combined genders were 33 4mmand 24 3 mm for ascending and descendingthoracic aorta, respectively. The corresponding up-Table II Models: Stage I Models InteractionsAscending AortaDescending Aorta(n 4,038)(n 4,034)Adjusted R2 R2 (Standardized)StandardErrorpValueUnstand ardizedCoefficient(Standardized)* ( ) ( ) gender ( ) ( ) * ( ) ( ) ( ) ( ) history of CAD ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) , ( ) ( ) , smoking ( ) , BSA ( ) , dyslipidemia ( ) *Per increments; past or current as in Table and Descending Aortic Diameters by Gender, Age, and BSAAge (yrs)BSA (m2)Ascending (mm)* (n 2,952)Descending (mm)* (n 1,931)Female Patients(n 1,147)Male Patients(n 1,805)Female Patients(n 736)Male Patients(n 1,195) 45 , ( ) , ( ) , ( ) , NA ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , NA ( ) , ( ) , NA ( ) , ( ) 45 54 , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( )55 64 , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) 65 , ( ) , ( ) , ( ) , NA ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , NA ( ) , ( ) , ( ) , ( )*Values are expressed as mean 1 SD, upper limit (range). In cases when there were fewer than 6 patients in a group, SD was not calculated and the upper limit of normal was not as in Table : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT204Downloaded From: on 01/29/2013per limits of normal were 41 and 30 mm. The meanascending and descending thoracic aortic diametersfor females were mm and , respectively, and the corresponding upperlimits of normal were and The meanascending and descending thoracic aortic diametersfor males were mm and mm,respectively, and the corresponding upper limits ofnormal were and 4details aortic diameter parameters strat-ified by gender, age, and BSA. In the final ascend-ing aorta analysis group there were 787 (27%), 106(4%), and 1 ( ) subjects with an ascendingthoracic aortic diameter greater than 35, 40, and 50mm, respectively. Nomograms for aortic diameter,ascending and descending, for gender and agegroup by BSA are shown inFigure 2. The formulafor predicting ascending aortic diameter is ADiameter (mm)BSA m2<45 y/oy = + 2 (mm)BSA m255-64 y/oy = + 2 (mm)BSA m245-54 y/oy = + 2 (mm)BSA m2>65 y/oy = + 2 (mm)BSA m2<45 y/oy = + 2 (mm)BSA m255-64 y/oy = + 2 (mm)BSA m245-54 y/oy = + 2 (mm)BSA m2>65 y/oy = + 2 Ascending and Descending Thoracic Aortic Diameter by Gender, Age, and BSA(A)Nomogram of normal value of thoracic ascending aorta female patients.(B)Nomogram of normal value of thoracic ascending aorta male on next : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT205Downloaded From: on 01/29 age (years) (if male) age(if male) BSA; the formula for predictingdescending aortic diameter is age(years) (if male) age (if male) 5shows the ascending anddescending aortic cross-sectional area measure-ments by BSA, gender, and of thoracic aortic size is important indetecting aortic aneurysm. The present study dem-onstrates that measurements can be made by gated,noncontrast CT scans obtained for assessment (mm)BSA m2<45 y/oy = + 2 (mm)BSA m255-64 y/oy = + 2 (mm)BSA m245-54 y/oy = + 2 (mm)BSA m2>65 y/oy = + 2 (mm)BSA m2<45 y/oy = + 2 (mm)BSA m255-64 y/oy = + 2 (mm)BSA m245-54 y/oy = + 2 (mm)BSA m2>65 y/oy = + 2 Ascending and Descending Thoracic Aortic Diameter by Gender, Age, and BSA(C)Nomogram of normal value of thoracic descending aorta female patients.(D)Nomogram of normal value of thoracic descendingaorta male patients. Thegold linesrepresent the estimated mean thoracic ascending aorta diameter. Theorange linesrepresent theestimated 95% normal confidence upper and lower limits. BSA body surface area; y/o years of : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT206Downloaded From: on 01/29/2013Relationship to previous mean aorticdiameter and cross-sectional area found in thecurrent study are comparable to the dimensionspreviously reported by CT, MRI, and echocardiog-raphy studies (Table 6)(1 3,6,7,9 12). Our find-ings that aortic dimension relates directly to age(2,3,6,7,9) and BSA confirm previous results (6,7).Interestingly, we also confirmed that the associationbetween BSA and aortic diameter was stronger thanthat between BMI and aortic diameter (6,7).Regarding the relationship between thoracic aor-tic dimensions and gender, previous works havereported that male gender is associated with a largeraortic diameter (3,6). In our study, however, malegender was a significant predictor only when inter-acting with age, such that older men have, onaverage, larger aorta than women of a similar age,but the difference is smaller for younger men andwomen (Table 3).In addition to hypertension, smoking was found tobe an independent predictor only of the descendingaortic diameter. It has been suggested that the etiologyof thoracic aneurysms differs between the ascendingand the descending segments and that the pathogen-esis of aneurysms in the descending thoracic aorta maymore closely resemble that of the abdominal aortaaneurysms than that of ascending thoracic aneurysms(10). Because smoking has been shown to be the riskfactor most strongly associated with abdominal aorticaneurysms, it is not surprising that the current studyfinds that smoking is also associated with largerdescending thoracic aorta diameters (11).The similarity of the findings of Guthaner et al.(1) and the current study suggests that at the level ofthe pulmonary artery, there is reasonable resem-blance between gated noncontrast CT and non-gated contrast CT measurements. Though the cur-rent study was not designed to address this issue,this observation, if further proven by additionalstudies, might suggest that the normal limits pro-posed in the current work can be used for contrastnongated studies too. Gated aorta studies, however,can provide additional information in the form oftemporal diameter changes during the cardiac and Descending Aortic Cross-Sectional Area for BSA by Gender and Age (cm2/m2)*Age (yrs)Ascending Aorta (n 2,952)Descending Aorta (n 1,931)Female Patients(n 1,147)Male Patients(n 1,805)Female Patients(n 736)Male Patients(n 1,195) , ( ) , ( ) , ( ) , ( )45 , ( ) , ( ) , ( ) , ( )55 , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( ) , ( )*Values are expressed as mean 1 SD, upper limit (range).Table of All Previous Data Regarding Normal Limits for Thoracic Aortic Dimensions Using Any ModalityModalityAuthor (Ref. #)YearAgeRange (yrs)Anatomical Landmark ofthe Ascending AortaAscending AortaDimensions Diameteror Area/m2(N)Descending AortaDimensions Diameteror Area/m2(N)CommentsMRIKersting-Sommerhoff et al. (9)1987 YoungadultsPulmonary artery level30 4 mm (20)24 4 mm (20)Adopted also forCT ( 10)Mohiaddin et el. (11) 1990 10 60Pulmonary artery cm2/m2(70) cm2/m2(70)EchoRoman et al. (7)1989 20 74Proximal ascending aorta30 4 cm (68 malepatients), 27 4(67 female patients)TEEDrexler et al. (12)1990 19 302 cm above the aortic cm2/m2(25) cm2/m2(25) TEEAgmon et al.(6)2003Pulmonary artery level33 4 (373)26 3 (373)TEECTGuthaner et al. (1)1979Pulmonary artery level32 5 mm (15)25 4 mm (15)Aronberg et al. (2)1984 21 61Caudal to the aortic arch35 mm (102)26 mm (102)Hager et al. (3)2002 17 89Caudal to the aortic arc at maximal size 31 4 mm (70)25 4 (70)Current study2008 26 75Pulmonary artery level33 4 mm, (2,952)24 3 mm, (1,930)Data are provided as reported in the original manuscript (some of the manuscripts provide only partial data). Diameter is given in millimeters and cross-sectional area in square computed tomography; Echo echocardiography; MRI magnetic resonance imaging; TEE transesophageal : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT207Downloaded From: on 01/29/2013The current study was not designed to provideprognostic data. However, a previous study (12) hasshown that thoracic aortic dissection and rupture isrelated to aortic diameter and that the annualcomplication rate for ascending thoracic aorta di-ameter below 35 mm (the mean in the current studyis 33 mm) and below 40 mm (the upper limit ofnormal in the current study was 41 mm) was 0%and , respectively. Following the above-mentioned criteria, we found that only 4% of thepatients in the final analysis group had more of annual risk for the current study, we used CT data from alow-risk population in a selective manner. Identifica-tion of clinical parameters that were associated withincreasing aortic dimensions allowed exclusion of pa-tients with these parameters from the final analysisgroup. A final analysis group that more closely repre-sents a normal population was defined, as opposed tothe original unselected asymptomatic population. Inthe future, when it is impractical to examine a largenumber of healthy subjects of various ages and bothgenders, this statistical approach to selecting a normal patient group may be appropriate to determine normallimits for anatomic or physiologic describe herein an approach toarrive at a normal sample from a nearly normal unselected group of patients. The major limitation ofsuch an approach, in our opinion, is that if the riskfactor profile of the initial unselected group is signif-icantly biased away from normal, no amount offine-tuning can make the sample normative and,therefore, fewer initial risk factors enable a better normalization process. The risk factor profile of ourinitial group is given inTable 1. When we comparethe current study group with the American HeartAssociation heart disease and stroke statistics 2007 update data (13) and with the MESA (Multi-Ethnic Study of Atherosclerosis) population (14),which included a cohort of patients with no CAD anda low-risk factor profile, we see that the current studygroup has a lower risk profile. Therefore, we believethat the initial group of the study is suitable for thewhittling down process and the final group representsa truly normal other limitations should also be ad-dressed. The dataset comprises a patient populationfrom a single center. Although our initial unselectedgroup consisted of a fairly large number of patients,owing to the whittling down and stratificationprocess, in a few subgroups, we were not able toprovide an upper normal limit because of lowsubject count. It is also noted that measurement ofaortic diameters by noncontrast CT in the level ofthe pulmonary artery may not be representing thetrue short axis, and to measure the true short axis,contrast-enhanced study may be required. Lastly,because different scanners were employed duringthe course of this study, there could be somevariability due to the calibration of the measure-ment itself. However, given the strict acceptancecriteria applied to our CT scanners, this differencefor aortic measurements is likely to be of aortic size is possible from CT scansobtained for CAC measurements. Given the impor-tance of detecting thoracic aortic aneurysm, consider-ation should be given to including these measure-ments as part of the clinical report of the CAC nomograms provided by our study could allowsuch a report to include a statement regarding therelationship of a patient s measurements to the ex-pected upper normal for a given patient (15 17).Reprint requests and correspondence:Dr. Daniel S. Ber-man, Director, Cardiac Imaging, Cedars-Sinai MedicalCenter, 8700 Beverly Boulevard, Los Angeles, California90048. E-mail: Guthaner DF, Wexler L, Harell demonstration of cardiac struc-tures. AJR Am J Roentgenol 1979;133:75 Aronberg DJ, Glazer HS, MadsenK, Sagel thoracic aorticdiameters by computed Comput Assist Tomogr 1984;8:247 Hager A, Kaemmerer H, Rapp-Bernhardt U, et al. Diameters of thethoracic aorta throughout life as mea-sured with helical computed tomogra-phy. J Thorac Cardiovasc Surg 2002;123:1060 Executive summary of the third reportof the National Cholesterol EducationProgram (NCEP) expert panel on de-tection, evaluation, and treatment ofhigh blood cholesterol in adults (adulttreatment panel III). JAMA 2001;285:2486 Mosteller RD. Simplified calculationof body-surface area. N Engl J Med1987;317 Agmon Y, Khandheria BK, Meiss-ner I, et al. Is aortic dilatation anatherosclerosis-related process? Clini-cal, laboratory, and transesophagealechocardiographic correlates of tho-racic aortic dimensions in the popula-tion with implications for thoracicaortic aneurysm formation. J Am CollCardiol 2003;42:1076 Roman MJ, Devereux RB, Kramer-FoxR, O Loughlin J. Two-dimensionalechocardiographic aortic root dimen-sions in normal children and J Cardiol 1989;64:507 : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT208Downloaded From: on 01/29/20138. Neter J, Kutner MH, Wasserman W,Nachtsheim CJ, editors. Applied LinearStatistical Models. 4th edition. NewYork, NY: McGraw-Hill/Irwin, Kersting-Sommerhoff BA, SechtemUP, Schiller NB, Lipton MJ, HigginsCB. MR imaging of the thoracic aortain Marfan patients. J Comput AssistTomogr 1987;11:633 Webb WR, Higgins CB. ThoracicImaging: Pulmonary and Cardiovas-cular Radiology. Philadelphia, PA:Lippincott, Williams, and Wilkins, Mohiaddin RH, Schoser K, AmanumaM, Burman ED, Longmore DB. MRimaging of age-related dimensionalchanges of thoracic aorta. J ComputAssist Tomogr 1990;14:748 Drexler M, Erbel R, Muller U, Witt-lich N, Mohr-Kahaly S, Meyer of intracardiac dimen-sions and structures in normal youngadult subjects by transesophagealechocardiography. Am J Cardiol1990;65:1491 Isselbacher EM. Thoracic and ab-dominal aortic aneurysms. Circulation2005;111:816 Lederle FA, Johnson GR, Wilson SE,et al. Prevalence and associations ofabdominal aortic aneurysm detectedthrough screening. Ann Intern Med1997;126:441 Elefteriades JA. Natural history ofthoracic aortic aneurysms: indicationsfor surgery, and surgical versus non-surgical risks. Ann Thorac Surg 2002;74:S1877 80; discussion S1892 Rosamond W, Flegal K, Friday G, etal. Heart disease and stroke statis-tics 2007 update: a report from theAmerican Heart Association statisticscommittee and stroke statistics sub-committee. Circulation 2007;115:e69 McClelland RL, Chung H, Detrano R,Post W, Kronmal RA. Distribution ofcoronary artery calcium by race, gender,and age: results from the Multi-EthnicStudy of Atherosclerosis (MESA). Cir-culation 2006;113:30 : CARDIOVASCULAR IMAGING, VOL. 1, NO. 2, 2008MARCH 2008:200 9Wolaket Aorta Measurements by Noncontrast Cardiac CT209Downloaded From: on 01/29/2013

Related search results