To the Editor: In their article on the Coronary Artery EvaluationUsing 64-Row Multidetector Computed Tomography Angiography (CORE64) trial (Nov. 27 issue),1 Miller et al. rather miss the pointof contemporary computed tomographic (CT) angiography, as doRedberg and Walsh in their accompanying Perspective article.2CT angiography will never replace catheter angiography in thesame manner that CT pulmonary angiography has usurped its invasivecounterpart as the standard. Even the use of CT scanners witha steadily rising number of detectors is unlikely to fully overcomethe limits on spatial resolution imposed by cardiac motion.
With this in mind, the sensitivity and specificity of CT angiographyand its negative predictive value of 83% compare more than favorablywith other noninvasive approaches. The role and key benefitsof CT angiography — not in the future but now —lie in the avoidance of invasive angiography and its inherentrisk of major cardiovascular complications in patients witha low-to-moderate risk of requiring revascularization. Arguingagainst the use of CT angiography on the basis of cost is illogical.If CT angiography is used judiciously with follow-up catheterangiography as needed, it should reduce rather than increasecosts in part by removing the medical and legal costs of thecomplications associated with catheter angiography.
Andrew C. McCulloch, M.B., Ch.B. Stirling Royal Infirmary Stirling FK8 2AU, United Kingdom andrew.mcculloch{at}nhs.net
References
Miller JM, Rochitte CE, Dewey M, et al. Diagnostic performance of coronary angiography by 64-row CT. N Engl J Med 2008;359:2324-2336. [Free Full Text]
Redberg RF, Walsh J. Pay now, benefits may follow -- the case of cardiac computed tomographic angiography. N Engl J Med 2008;359:2309-2311. [Free Full Text]
To the Editor: The comparative evaluation of 64-row CT withselective contrast angiography has a severe shortcoming in thatit uses a two-dimensional measurement of vessel diameter, expressedas percent narrowing, which is known to have a very low degreeof precision. In addition, such a measure cannot account forthe wide variation in local vascular changes that determinethe degree of obstruction of flow.1,2 Unfortunately, Milleret al. are not the first investigators to be misled by "ourpreoccupation with coronary lumenology," in the words of someleading cardiologists.3 The authors used an arbitrary stenosisthreshold of 50% as an indicator of obstruction and did notaccount for the clinical outcome of such a cutoff. They completelydisregarded the ability of contrast angiography to identifya delay in regional flow and the existence of relevant collateralsin relation to myocardial injury.4 Although I agree with theconclusion that multidetector CT angiography cannot replaceconventional coronary angiography at present, this conclusionis based on multiple considerations that were not covered inthis study.
Sven Paulin, M.D., Ph.D. Beth Israel Deaconess Medical Center Boston, MA 02215
References
Robbins SL, Bentov J. The kinetics of viscous flow in a model vessel: effect of stenoses of varying size, shape, and length. Lab Invest 1967;16:864-874. [Web of Science][Medline]
Paulin S. Assessing the severity of coronary lesions with angiography. N Engl J Med 1987;316:1405-1407. [Web of Science][Medline]
Topol EJ, Nissen SE. Our preoccupation with coronary lumenology: the dissociation between clinical and angiographic findings in ischemic heart disease. Circulation 1995;92:2333-2342. [Free Full Text]
Paulin S. Functional alterations in the coronary circulation as mirrored in the angiogram. Cardiovasc Intervent Radiol 1982;5:177-185. [CrossRef][Web of Science][Medline]
To the Editor: Miller et al. have compared the accuracy of 64-sliceCT angiography with conventional angiography for the identificationof coronary artery disease. Of the 405 patients who underwentscreening, 89 (22%) were excluded from the study because theyhad a coronary calcium score of more than 600. An additional,undisclosed proportion of patients were excluded because theyhad a serum creatinine level of more than 1.5 mg per deciliter,a creatinine clearance of less than 60 ml per minute, a historyof previous cardiac surgery or recent intervention, atrial fibrillation,heart failure, or a body-mass index (the weight in kilogramsdivided by the square of the height in meters) of more than40, among other criteria. These criteria characterize an additionalnontrivial proportion of patients with chest pain who couldbe evaluated for possible coronary disease. In contrast, noneof these criteria are necessarily exclusionary for other noninvasiveapproaches (e.g., myocardial perfusion scintigraphy) for theassessment of hemodynamically significant coronary obstruction.The potentially large proportion of patients who were deemedto be ineligible for coronary assessment on CT angiography promptsthe question posed in the Discussion section of the article:What is the role of this test in the diagnostic algorithm forpatients with suspected coronary artery disease?
Stephen K. Gerard, M.D., Ph.D. Seton Medical Center Daly City, CA 94015
To the Editor: Miller et al. conclude that in symptomatic patientswith suspected coronary artery disease, CT angiography doesnot have a sufficient positive or negative predictive valueto replace conventional angiography. Two issues merit highlightingin this context. First, whatever its accuracy, CT angiographyis unlikely to replace conventional angiography for patientsat intermediate or high risk for coronary artery disease becausethe prospective use of CT angiography in this population wouldexpose a large proportion of patients to the cumulative risks,radiation, and costs associated with undergoing both procedures.Furthermore, previous studies have suggested that CT angiographyhas the best diagnostic performance in low-risk patients, asevidenced by high positive likelihood ratios and low negativelikelihood ratios.1 Second, as noted by the authors, positiveand negative predictive values depend on the prevalence of disease,and it may therefore be more appropriate to present likelihoodratios to describe diagnostic-test performance (positive likelihoodratio, 8.5; negative likelihood ratio, 0.2).
Ganesan Karthikeyan, M.D., D.M. McMaster University Hamilton, ON L8L 2X2, Canada karthik2010{at}gmail.com
References
Meijboom WB, van Mieghem CA, Mollet NR, et al. 64-Slice computed tomography coronary angiography in patients with high, intermediate, or low pretest probability of significant coronary artery disease. J Am Coll Cardiol 2007;50:1469-1475. [Free Full Text]
To the Editor: Redberg and Walsh make several unsubstantiatedand misleading claims in their Perspective article. CT angiographyis a diagnostic test, not a therapeutic intervention, and thuscannot have a direct effect on clinical outcomes, as drugs ordevices can.
I fully support clinical trials examining the potential benefitthat CT angiography may have for patients. Luckily for Redbergand Walsh, such studies have been performed, and more are inthe pipeline. These trials have so far shown that the use ofCT angiography has a prognostic value,1 reduces costs,2,3,4and reduces the length of stay in overcrowded emergency rooms.3As for Redberg and Walsh's statement that CT angiography "bombardspatients with radiation," it should be noted that the dose ofradiation with prospective gating is now at half the level ofthat with diagnostic angiography.5 In addition, newer CT scanningtechniques have been devised that maintain full chest angiographyat less than 1 mSv of total radiation dose.
It undoubtedly makes sense to scrutinize new technologies anddrugs carefully with the public's interest at heart. But itis impractical and inappropriate to assume a lack of benefitfor all diagnostic techniques unless large, randomized, outcome-basedstudies demonstrate direct improvement in outcomes for patientsfrom the outset.
Gabriel Vorobiof, M.D. Yale University School of Medicine New Haven, CT 06520
References
Min JK, Shaw LJ, Devereux RB, et al. Prognostic value of multidetector coronary computed tomographic angiography for prediction of all-cause mortality. J Am Coll Cardiol 2007;50:1161-1170. [Free Full Text]
Min JK, Shaw LJ, Berman DS, Gilmore A, Kang N. Costs and clinical outcomes in individuals without known coronary artery disease undergoing coronary computed tomographic angiography from an analysis of Medicare category III transaction codes. Am J Cardiol 2008;102:672-678. [CrossRef][Web of Science][Medline]
Ladapo JA, Hoffman U, Bamberg F, et al. Cost-effectiveness of coronary MDCT in the triage of patients with acute chest pain. AJR Am J Roentgenol 2008;191:455-463. [Free Full Text]
Goldstein JA, Gallagher MJ, O'Neill WW, Ross MA, O'Neil BJ, Raff GL. A randomized controlled trial of multi-slice coronary computed tomography for evaluation of acute chest pain. J Am Coll Cardiol 2007;49:863-871. [Free Full Text]
Earls JP, Schrack EC. Prospectively gated low-dose CCTA: 24 months experience in more than 2,000 clinical cases. Int J Cardiovasc Imaging 2008 November 25 (Epub ahead of print).
To the Editor: Redberg and Walsh have written one of the mostimportant articles of the year. They clearly point out the needfor evaluation of new technology if we are to have the resourcesto revamp our broken health care system and to cover the manypeople who are uninsured. I would add that this evaluation shouldbe paid for by the government and performed by researchers whoare not beholden to the manufacturers of the technology in question.I would also suggest that the same restrictions apply to drugsthat should be tested against current therapies and not a placebo.Only technologies and drugs that are clearly superior and cost-effectiveshould be approved for use and insurance coverage. We, as physiciansand citizens, should demand nothing less than evidence-basedand cost-effective procedures and treatments for our patients.
William M. Fogarty, Jr., M.D. Saint Louis University School of Medicine St. Louis, MO 63119 wmfogartyjr{at}sbcglobal.net
Drs. Lima and Miller reply: We agree with McCulloch that 64-rowCT angiography is an excellent option for selected symptomaticpatients with suspected coronary artery disease.1 Our studywas not designed to test multidetector CT angiography as a replacementfor conventional angiography but, rather, to compare its diagnosticaccuracy with that of conventional angiography, the standardby which noninvasive tests have been established. The test'spatient-based area under the curve (AUC) (0.93; 95% confidenceinterval [CI], 0.90 to 0.96) supports the clinical use of CTangiography in selected symptomatic patients and suggests thatsuch use could reduce unnecessary procedures with their associatedrisks and costs. Our conclusions aimed to clarify that althoughCT angiography and the conventional method both use angiography,the role of conventional angiography extends beyond strict diagnosisby guiding revascularization procedures.
Paulin challenges the use of stenosis as measured on conventionalangiography as the reference standard for defining obstructivecoronary artery disease. We agree on the shortcomings of measurementof vessel diameter by both conventional and CT angiography,given that anatomical measures are inherently limited in assessingthe physiological severity of coronary stenosis. Conversely,conventional angiography continues to be the clinical standardfor assessing lesions and determining approaches to revascularization.Although there are increasing data supporting adjunctive physiologicalmeasurements,2 conventional angiography that is used in isolationremains the most widely available diagnostic test and the standardagainst which noninvasive tests are compared.
We agree with Gerard that the exclusion criteria we used inour study reduced the population of patients eligible for CTangiography. This fact does not limit the value of CT angiographyto the much larger group of patients who do meet the inclusioncriteria. Indeed, all approaches that are used for the noninvasiveassessment of coronary artery disease have important benefitsand limitations, and alternative diagnostic approaches offeran opportunity to select the best test for each patient. Moreover,the evolution of CT angiography continues at such a rapid pacethat the broadening of the population eligible to undergo suchtesting appears to be practically inevitable.
Karthikeyan notes the dependence of predictive values on diseaseprevalence. This highlights our choice of the AUC as the mostappropriate index of the performance of CT angiography.3 Previousstudies have not assessed the severity of stenosis on continuousscales and thus have allowed only dichotomous thresholds tobe validly applied (predictive values and likelihood ratios).By contrast, we obtained continuous measurements and showedthat the robust performance of CT angiography was actually independentof cutoffs above the 50% threshold for stenosis, as shown inFigure 2 of the article. Moreover, in the identification ofpatients with at least one stenosis of 50% or more, CT angiographyhad a positive likelihood ratio of 8.4 (95% CI, 5.0 to 14.1)and a negative likelihood ratio of 0.16 (95% CI, 0.10 to 0.28),which reflects its accuracy. Finally, the prevalence of 56%for coronary artery disease among patients in our study extendsdiagnostic performance to patients who are at intermediate riskfor disease.
In summary, we believe that the value of 64-row CT angiographylies in providing the clinician with enough certainty to assistin medical decisions by classifying patients on the basis ofwhether they may or may not need further invasive procedures.
João A.C. Lima, M.D. Julie M. Miller, M.D. Johns Hopkins Hospital Baltimore, MD 21287 jlima{at}jhmi.edu
for the CORE 64 Investigators
References
Meijboom WB, Meijs MFL, Schuijf JD, et al. Diagnostic accuracy of 64-slice computed tomography coronary angiography: a prospective, multicenter, multivendor study. J Am Coll Cardiol 2008;52:2135-2144. [Free Full Text]
Tonino PA, De Bruyne B, Pijls NH, et al. Fractional flow reserve versus angiography for guiding percutaneous coronary intervention. N Engl J Med 2009;360:213-224. [Free Full Text]
Pepe MS. The statistical evaluation of medical tests for classification and prediction. Oxford, England: Oxford University Press, 2003.
Drs. Walsh and Redberg reply: Vorobiof states that because CTangiography is a diagnostic test, it cannot have a direct effecton clinical outcomes. Diagnostic tests are valuable only iftheir use leads to a treatment that results in improved clinicaloutcomes. To date, CT angiography has not been shown to leadto improved clinical outcomes. Vorobiof goes on to state thatfour trials have shown a potential benefit of CT angiography.However, of these trials, only one was a clinical trial; twowere observational studies, and one was a microsimulation modelto evaluate cost-effectiveness.
In the sole clinical trial cited by Vorobiof, Goldstein andcolleagues compared CT angiography with nuclear stress testingfor the evaluation of acute chest pain in 197 patients who presentedto the emergency department. Although CT angiography eitherruled out or identified coronary artery disease in 75% of thepatients, the remaining 25% required stress testing to evaluatelesions of intermediate severity or nondiagnostic scans, whichthe authors acknowledge as an important limitation of the technology.1
One of the two observational studies cited by Vorobiof showedthat CT angiography, not surprisingly, predicted an increasedrisk of death among patients with chest pain.2 In the secondstudy, a retrospective analysis of a claims database showedthat expenditures were lower for patients who underwent CT angiographythan for those who underwent single-photon-emission computedtomography, with no difference in adverse cardiovascular outcomes.3Although these results are interesting, they do not show thatCT angiography improves clinical outcomes.
The goal of the microsimulation study was to evaluate the cost-effectivenessof CT angiography.4 However, in order to determine cost-effectiveness,actual effectiveness must be known. Since CT angiography hasnot been shown to be effective in improving outcomes for patients,cost-effectiveness is difficult to determine.
Finally, although prospectively gated, low-dose CT angiographymay reduce the dose of radiation, even a lower radiation riskmay not be acceptable if the technology does not have a provenbenefit.
We agree with McCulloch that the test characteristics of CTangiography compare favorably with those of other technologiesand that "arguing against the use of CT angiography on the basisof cost" does not make sense. The concern with CT angiographyis not whether the cost is warranted but, rather, whether atechnology without a proven benefit and with a potential riskfrom radiation and possible complications from the pursuit ofincidental findings should be covered by insurance, regardlessof how much it costs.
Finally, Fogarty emphasizes the importance of having unbiasedresearchers perform critical evaluations of new technologiesand the importance of comparing new technologies with the standardof care. We agree with him entirely.
Judith Walsh, M.D., M.P.H. Rita F. Redberg, M.D. University of California, San Francisco San Francisco, CA 94143
References
Goldstein JA, Gallagher MJ, O'Neill WW, Ross MA, O'Neil BJ, Raff GL. A randomized controlled trial of multi-slice coronary computed tomography for evaluation of acute chest pain. J Am Coll Cardiol 2007;49:863-871. [Free Full Text]
Min JK, Shaw LJ, Devereux RB, et al. Prognostic value of multidetector coronary computed tomographic angiography for prediction of all-cause mortality. J Am Coll Cardiol 2007;50:1161-1170. [Free Full Text]
Min JK, Shaw LJ, Berman DS, Gilmore A, Kang N. Costs and clinical outcomes in individuals without known coronary artery disease undergoing coronary computed tomographic angiography from an analysis of Medicare category III transaction codes. Am J Cardiol 2008;102:672-678. [CrossRef][Web of Science][Medline]
Ladapo JA, Hoffmann U, Bamberg F, et al. Cost-effectiveness of coronary MDCT in the triage of patients with acute chest pain. AJR Am J Roentgenol 2008;191:455-463. [Free Full Text]
Previous
