Author + information
- Ilan Gottlieb, MD,
- Jeffery Brinker, MD,
- João A.C. Lima, MD and
- Carlos E. Rochitte, MD* ()
- ↵*Heart Institute (InCor), University of São Paulo Medical School, Av. Dr. Enéas de Carvalho Aguiar, 44, Andar AB - Setor de Ressonância Magnética, São Paulo, SP 05403-000, Brazil
We thank the authors for their letters and for the interest in our study and their thoughtful remarks. We would like to add some comments.
As Drs. Correia and Blaha noted, the predictive values found in our paper (1) have slightly different meanings than commonly utilized in other trials (2). This is so because we chose to take a different perspective. Our aim was to determine if a calcium score (CS) of 0 (positive scan for us) could predict the absence of obstructive coronary artery disease (CAD), whereas the other trials they refer to examined the question of whether the presence of calcium increases the likelihood of chest pain being related to significant stenosis (2). We thank both for the opportunity to further clarify this issue.
Using our approach of calling a zero CS a positive scan, the positive predictive value refers to the ability of zero calcium to rule out obstructive CAD. This is in fact the same message of a negative predictive value using the “conventional” approach. This predictive value was low in our study: 68%. Accordingly, the sensitivity of zero calcium to detect the absence of disease (i.e., to rule out obstructive CAD) was also low at 45%. As Dr. Blaha points out, when our results are interpreted from this perspective, they are clearly consistent with previously published studies (1,2).
We feel that our approach more accurately tests the utility of the CS when applied for this specific purpose, i.e., to rule out obstructive disease in symptomatic patients with suspected CAD to allow for discharge from the emergency department or to direct outpatient investigation to other causes of chest pain.
We agree with Drs. McEvoy, Timmis, and Blaha that high-quality research has been performed in determining the epidemiologic value of coronary calcium as a marker of atherosclerosis-related adverse events in asymptomatic individuals, and we thank them for stressing once again that our study did not investigate this patient population. Our study documents the limitations of coronary calcification in symptomatic individuals suspected of having obstructive CAD. In fact, it quantifies something that experienced clinical cardiologists already know and have incorporated in their clinical practice (i.e., noncalcified plaque can rupture and cause a myocardial infarction) and this phenomenon is not that uncommon, particularly among patients usually considered to be at low risk for coronary disease (e.g., women and younger individuals) (3). This was again confirmed in vivo in our study in which 20% of the totally occluded vessels were free from calcification (1).
Referring to the ACCURACY (Assessment by Coronary Computed Tomographic Angiography of Individuals Undergoing Invasive Coronary Angiography) trial (4), Dr. Budoff states that it “demonstrated CAC sensitivity of 94% and specificity 42% for >50% stenosis by quantitative coronary angiography” and continues by stating that “Gottlieb et al. present the opposite results (sensitivity 45% and specificity 91%), calling into question study design, equipment, or CAC methodology, not validity of CAC testing.” In fact, their results are very similar to ours, just expressed differently. As noted above, the ACCURACY CS sensitivity of 94% for the presence of stenosis matches our (CORE64 [Coronary Evaluation Using Multi-Detector Spiral Computed Tomography Angiography Using 64 Detectors]) CS specificity of 91% for the absence of stenosis, whether their specificity of 42% for the presence of stenosis matches our sensitivity of 45% for the absence of stenosis.
Regarding our CS methodology, we followed standard imaging parameters and requirements recommended by the American College of Cardiology/American Heart Association guidelines (5). Although we recognize that multidetector computed tomography (MDCT) scanners have different performance parameters as compared with electron beam computed tomography (EBCT), in clinical practice, CS is more often measured with MDCT than EBCT due to the former's much better performance in coronary angiography.
One could be tempted to generalize our findings to all subgroups of patients, mixing symptomatic and asymptomatic patients as being the same. This is a grave mistake. Dr. Budoff states that our study trumps more than a 1,000 studies and that prognosis of zero CS has been assessed in over 100,000 patients, but he regrettably misses the fact that the vast majority of the published CS literature refers to asymptomaticpatients.
Dr. Budoff questions exclusion of patients with CS >600 in our study. This group would, by definition, be irrelevant to our paper, the main point of which was to demonstrate the prevalence of significant disease in symptomatic patients having no coronary calcium.
While we take the opportunity to thank Dr. Rita Redberg for the time and effort in appraising our work and the comments on the strength of our study, we agree with Dr. Timmis, Blaha, and McEvoy that the associated editorial to our paper took a broader view than our data warrants. We acknowledge the role CS has for risk stratification in selected asymptomatic populations as well as in epidemiologic studies of atherosclerotic disease.
Searching for surrogate evidence of stenosis, as is the case with CS, makes the performance of the test rely heavily on the prevalence of obstructive CAD and other biological factors in the population it is being applied to, rendering CS unsuitable for ruling out obstructive CAD in symptomatic patients. In summary, we believe our work reflects the application of what we know in pathophysiology to clinical medicine and supports the results of previous studies indicating that symptomatic patients with suspected CAD should not be discharged from the emergency department based solely on the results of coronary calcium scores assessed by unenhanced CT.
We thank the colleagues who have expressed interest in our paper and provide a forum for further discussion of the analytic and statistical methods used in our paper.
- American College of Cardiology Foundation
- Gottlieb I.,
- Miller J.M.,
- Arbab-Zadeh A.,
- et al.
- Sarwar A.,
- Shaw L.J.,
- Shapiro M.D.,
- et al.
- Stary H.C.,
- Chandler A.B.,
- Dinsmore R.E.,
- et al.
- Budoff M.J.,
- Dowe D.,
- Jollis J.G.,
- et al.
- Budoff M.J.,
- Achenbach S.,
- Blumenthal R.S.,
- et al.