LCC

S

C

Iiou(btsW“cmqrpb

IpuaCpa(ppp

ahpiqpIDci

oioblbhbrndbeat

L

Ttntoifdpmatwtesiirspna

gr

*vA

OCepf

Journal of the American College of Cardiology Vol. 52, No. 25, 2008© 2008 by the American College of Cardiology Foundation ISSN 0735-1097/08/$34.00Published by Elsevier Inc. doi:10.1016/j.jacc.2008.09.017

EDITORIAL COMMENT

imitations ofomputed Tomographyoronary Angiography*

teven E. Nissen, MD, MACC

leveland, Ohio

t seems so simple and elegant. Fifty years after thentroduction of coronary angiography, advances in technol-gy allow imaging of the coronary arteries noninvasivelysing multidetector computed tomography (CT) scanners1). Within a few short years, these imaging systems haveegun appearing everywhere, first in hospitals and clinics,hen in individual doctors’ offices, offering the promise ofafe and painless detection of coronary obstructions (2).

eekend courses allow the members of our profession tolearn” this new technology and apply it routinely to patientare (3). The manufacturers of the equipment, all largeultinational providers of radiological imaging devices, are

uite pleased to show practitioners how they can rapidlyecoup their million-dollar investments (4,5). What couldossibly be wrong with this picture? Medical progress to theetterment of patients (and practitioners).

See page 2135

n this issue of the Journal, an interesting and importantaper by Meijboom et al. (6) provides data that should gives pause. When applied routinely in symptomatic patientst risk of coronary disease, in more than 50% of subjects,T angiography “detected” coronary obstructions that sim-ly were not there. Strikingly, of 98 patients in whom CTngiography diagnosed 3-vessel coronary artery diseaseCAD), only 19 actually were found to have it, a false-ositive rate of 81%. This high false-positive rate hasotentially serious implications, leading to unnecessary andotentially risky procedures that threaten to accelerate

Editorials published in the Journal of the American College of Cardiology reflect theiews of the authors and do not necessarily represent the views of JACC or themerican College of Cardiology.From the Cleveland Clinic Foundation, Cardiovascular Medicine, Cleveland,hio. Dr. Nissen has received research support to perform clinical trials through theleveland Clinic Coordinating Center for Clinical Research from Pfizer, AstraZen-

ca, Novartis, Sankyo, Takeda, Sanofi-Aventis, and Eli Lilly, and consults for many

tharmaceutical companies, but requires them to donate all honoraria or consultingees directly to charity so that he receives neither income nor a tax deduction.

lready-excessive health care costs. Now that our nationalealth care expenditures exceed 16% of the gross nationalroduct, we must ask ourselves critical questions about thentroduction of any costly new technology. Does it improveuality of care compared with existing methods? Does itrolong life or improve quality of life? Does it reduce costs?s it cost-effective? Are there important safety concerns?espite its rapid adoption, for all of these metrics, CT

oronary angiography has yet to show that it can deliver onts promise.

Routine application of CT angiography is largely basedn flawed assumptions about the nature of CAD. After thentroduction of angiography by Sones in 1957, a generationf practitioners came to believe that coronary disease couldest be defined by the presence or absence of obstructiveesions. This view was reinforced by the introduction ofypass surgery in 1968. Coronary disease was simple. If youad enough obstructions, you carried a high risk for mor-idity and mortality, and would benefit from surgicalevascularization. If there were no obstructions, CAD wasot present. Then, a decade later, Andreas Gruntzig intro-uced balloon angioplasty, and luminal obstructions coulde readily treated without surgery. Within a few years, anntirely new discipline, interventional cardiology, was cre-ted, rapidly growing into the dominant approach used toreat stable CAD.

imitations of Angiography

hen came disquieting research. Most myocardial infarc-ions did not occur at the sites of significant coronaryarrowing, and percutaneous treatment of coronary obstruc-ions in patients with stable angina did not reduce the riskf myocardial infarction or death (7,8). Whether performednvasively or via multidetector CT, angiography is con-ounded by the phenomenon of coronary remodeling, firstescribed in 1987 by Glagov et al. (9). The remodelingrocess is observed histologically as the outward displace-ent of the external vessel wall overlying the atheroma. The

dventitial enlargement opposes luminal encroachment,hereby concealing the presence of disease. Thus, the vesselall may contain a large atheroma, despite an angiogram

hat shows little or no luminal narrowing. Although remod-led lesions do not restrict blood flow, clinical studies havehown that these low-grade lesions represent the mostmportant source of acute coronary syndromes (10). Accord-ngly, an absence of luminal narrowing does not preclude aisk of plaque rupture, resulting in myocardial infarction orudden cardiac death. It must be acknowledged that aotential advantage of CT angiography is detection of theseonobstructive plaques, but the clinical utility of suchssessments remains unproven.

Beginning soon after the introduction of coronary an-iography, studies began to question the accuracy andeproducibility of angiography. Carefully performed inves-

igations demonstrated that visual interpretation of angiog-

romiesaoarl

F

AtrnwrotnevttaisFitmnUtocao

C

Cpqutcglstdrc

pcwdt

qm(pttusacrccobtp

lasdatfaet

C

WmcarmruCdadcCtpte

2146 Nissen JACC Vol. 52, No. 25, 2008CT Coronary Angiography Limitations December 16/23, 2008:2145–7

aphy showed clinically significant intraobserver and inter-bserver variability (11). Other investigators documentedajor differences between the apparent angiographic sever-

ty of lesions and the extent of disease at post-mortemxamination (12). Then, using functional testing, studieshowed a prominent discordance between the apparentngiographic severity of lesions and the physiological effectsf stenoses (13). We learned the difficult lesson, nowpparently forgotten, that the silhouette or luminogram is aelatively poor representation of coronary anatomy and aimited standard on which to base therapeutic decisions (14).

unctional Testing

ccordingly, cardiovascular medicine evolved to use func-ional testing to assess the likelihood that symptoms wereelated to impaired coronary blood flow. Because luminalarrowing alone did not accurately identify patients whoould benefit from revascularization, practitioners began

outinely evaluating symptomatic patients for the presencef physiological abnormalities associated with these obstruc-ions. This approach was logical, although its cost effective-ess has not been rigorously studied. If there was goodxercise capacity and no ischemia, angiography and inter-ention could be safely deferred. In some cases, functionalesting was performed after angiography, rather than before,o determine whether a lesion observed by angiographyctually represented a flow-limiting stenosis. Two compet-ng imaging technologies developed, first exercise nuclearcintigraphy, and subsequently, stress echocardiography.or patients who could not exercise, elegant pharmacolog-

cal approaches were developed to substitute for exerciseesting. Of equal importance, the exercise capacity andagnitude of the ischemic burden provided valuable prog-

ostic information that could be used to guide therapy.nfortunately, the rapid proliferation of CT angiography

hreatens to take us back to the overly simplistic approach ofur predecessors, once described as “our preoccupation withoronary luminology” (14). Just find the stenosis and treat itggressively, an approach sarcastically described as theculostenotic reflex (14).

T Angiography: A Step Backward

ompared with conventional angiography, the angiogramsroduced by multidetector CT scanners are much lower inuality and are suboptimal for diagnostic purposes. Tonderstand the deficiencies in CT angiography, it is impor-ant to review the physics of the technology used fororonary angiography. The resolution of modern angio-raphic equipment is surprisingly modest, only about 4 or 5ine pairs per millimeter. Prudence limits radiation expo-ure, resulting in an image flaw known as quantum statis-ical noise that can be reduced only by increasing radiationoses. Angiographic resolution is also compromised byapid coronary artery motion, which is reduced in impact for

onventional angiography by using pulse-mode radiogra- a

hy. The translational velocity of the right coronary arteryan reach 50 mm/s. A coronary artery moving at 50 mm/sill produce a motion blur of approximately 0.35 mmuring a typical 7-ms X-ray pulse width, an acceptableemporal resolution for high-quality angiographic imaging.

CT angiography represents a step backward in imageuality. The spatial resolution is only about 2 line pairs perillimeter, about one-half that of conventional angiography

15). The temporal resolution of CT angiography is veryoor, currently about 20-fold worse than a typical conven-ional angiogram (�150 ms) (15). To overcome the limi-ations in temporal resolution, electrocardiographic gating issed, but rarely represents a perfect solution and precludesuccessfully imaging in patients with arrhythmias such astrial fibrillation. Even more problematic, the presence oforonary calcium frequently obscures the underlying lumen,endering the imaging of specific stenoses unreliable. Ac-ordingly, CT angiography retains all of the limitations ofonventional angiography, including the absence of physi-logical information and poor correlation with histology,ut it is further constrained by poor image quality. Withhese limitations, it is not surprising that CT angiographyerformed poorly in the current study.To overcome quantum statistical noise and other quality

imitations, large radiation dosages are used, typically aver-ging �15 mSv, with specific organ systems receivingubstantially higher exposure levels, about 5 to 7 times theosage used by prudent practitioners in an optimal invasivengiogram (16). Recent studies suggest a small but impor-ant risk of inducing malignancy, ranging as high as 1 in 150or breast cancer in a younger woman (16). Practitionerslso must consider the likelihood that many patients willndure multiple CT scans during their lifetime, amplifyinghe radiation risk.

oronary Flow Reserve (CFR)

hy should we prefer functional testing to anatomicalethods such as CT angiography? In chronic ischemic

oronary disease, symptoms result principally from thebility of stenoses to blunt increases in blood flow inesponse to metabolic demands. This phenomenon is com-only called CFR (17). Animal studies show that CFR

emains normal (typically a 5- to 7-fold increase in flow)ntil the stenosis approaches 75%. Between 75% and 95%,FR decreases progressively. Accordingly, the angiographicifferences between moderate and severe lesions may be onlyfew tenths of a millimeter. Such differences are difficult toiscern given the limitations in resolution of the bestonventional angiography and impossible to resolve withT angiography. Other factors impair the correlation be-

ween angiographic lesion severity and CFR, including theresence of ventricular hypertrophy, the metabolic state ofhe myocardium, and microvascular disease. Thus, thepicardial stenosis represents only one factor responsible for

reduction in flow reserve in patients with clinical symp-

tpAtd

pTistpimdaicmaaatgpFao

RCEn

R

1

1

1

1

1

1

1

1

1

1

K

2147JACC Vol. 52, No. 25, 2008 NissenDecember 16/23, 2008:2145–7 CT Coronary Angiography Limitations

oms. A stenosis incapable of producing angina in oneatient may result in severe functional limitation in another.ccordingly, testing for the presence of ischemia, rather

han the presence of a stenosis, is essential for good clinicalecision making.Some ardent advocates will point out that CT angiogra-

hy and physiological testing are not mutually exclusive.his advocacy ignores the issue of cost-effectiveness. Test-

ng via multiple imaging modalities has shown neitheruperior effectiveness nor cost savings over current diagnos-ic paradigms. Proponents also emphasize the negativeredictive value of the test. However, we must consider themplications and costs of a false-positive study, which, in

any cases, will lead to additional costly testing or proce-ures. Therefore, we are now left with a conundrum. CTngiography is being deployed rapidly, and there exist noncentives for industry or physician advocates to seek thelinical trial data we need to effectively use this imagingodality. Accordingly, we need a restricted use of CT

ngiography until adequate clinical evidence becomesvailable showing the cost-effectiveness and safety of thispproach (18,19). We need real-world trials that use thisechnology, not just in carefully selected patients whoenerate optimal images, but in all comers, includingatients with a low, medium, and high risk of CAD.inally, we need outcome studies to show whether this newpproach really improves what we care most about: patientutcomes.

eprint requests and correspondence: Dr. Steven E. Nissen,leveland Clinic Foundation, Cardiovascular Medicine, 9500uclid Avenue, Desk F-15, Cleveland, Ohio 44195-0001. E-mail:issens@ccf.org.

EFERENCES

1. Hoffmann MHK, Shi H, Schmitz BL, et al. Noninvasive coronaryangiography with multislice computed tomography. JAMA 2005;293:2471–8.

2. The Evidence Gap. Weighing the Costs of a CT Scan’s Look Insidethe Heart. Available at: http://www.nytimes.com/2008/06/29/business/29scan.html?pagewanted�all. Accessed July 14, 2008.

3. Toshiba to Host Educational Series “CT Angiography Made Easy” at

ACC 2007. Available at: http://findarticles.com/p/articles/ c

mi_m0EIN/is_2007_March_21/ai_n27255350. Accessed July 14,2008.

4. CT Business. Reimbursement in the U.S. Available at: http://www.medical.siemens.com/siemens/en_US/rg_marcom_FBAs/files/DICpage/Reimb.pdf. Accessed July 14, 2008.

5. Links to Commercial Payer Policies. Available at: http://www.healthcare.philips.com/main/reimbursement/payers/links/index.html. Accessed July 14, 2008.

6. Meijboom WB, Meijs MFL, Schuijf JD, et al. Diagnostic accuracy of64-slice computed tomography coronary angiography: a prospectivemulticenter, multivendor study. J Am Coll Cardiol 2008;52:2135–44.

7. Little WC, Constantinescu M, Applegate RJ, et al. Can arteriographypredict the site of a subsequent myocardial infarction in patients withmild-to-moderate coronary artery disease? Circulation 1988;78:1157–66.

8. Parisi AF, Folland ED, Hartigan P, for the Veterans Affairs ACMEInvestigators. A comparison of angioplasty with medical therapy in thetreatment of single-vessel coronary artery disease. N Engl J Med1992;326:10–6.

9. Glagov S, Weisenberg E, Zarins CK, Stankunavicius R, Kolettis GJ.Compensatory enlargement of human coronary arteries. N Engl J Med1987;316:1371–5.

0. Schoenhagen P, Ziada KM, Kapadia SR, Crowe TD, Nissen SE,Tuzcu EM. Extent and direction of arterial remodeling in stable versusunstable coronary syndromes: an intravascular ultrasound study. Cir-culation 2000;101:598–603.

1. Zir LM, Miller SW, Dinsmore RE, Gilbert JP, Harthorne JW.Interobserver variability in coronary angiography. Circulation 1976;53:627–32.

2. Arnestt EN, Isner JM, Redwood CR, et al. Coronary artery narrowingin coronary heart disease: comparison of cineangiographic and nec-ropsy findings. Ann Intern Med 1979;91:350–6.

3. White CW, Wright CB, Doty DB, et al. Does visual interpretation ofthe coronary arteriogram predict the physiologic importance of acoronary stenosis? N Engl J Med 1984;310:819–24.

4. Topol EJ, Nissen SE. Our preoccupation with coronary luminology.The dissociation between clinical and angiographic findings in isch-emic heart disease. Circulation 1995;92:2333–42.

5. Poon M. Technology insight: cardiac CT angiography. Nat Clin PractCardiovasc Med 2006;3:265–75.

6. Einstein AJ, Henzlova MJ, Rajagopalan S. Estimating risk of cancerassociated with radiation exposure from 64-slice computed tomogra-phy coronary angiography. JAMA 2007;298:317–23.

7. Gould KL, Lipscomb K, Hamilton GW. Physiologic basis for assess-ing critical coronary stenosis: instantaneous flow response and regionaldistribution during coronary hyperemia as measures of coronary flowreserve. Am J Cardiol 1974;33:87–93.

8. Lord SJ, Irwig L, Simes RJ. When is measuring sensitivity andspecificity sufficient to evaluate a diagnostic test, and when do we needrandomized trials? Ann Intern Med 20066;144:850–5.

9. Schünemann HJ, Oxman AD, Brozek J, et al., for the GRADEWorking Group. Grading quality of evidence and strength of recom-mendations for diagnostic tests and strategies. BMJ 2008;336:1106–10.

ey Words: coronary artery disease y computed tomography y

oronary angiography y prospective y multicenter.