Author + information
- John A. Rumberger, PhD, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. John A. Rumberger, Cardiac Imaging and Lipid Management, Princeton Longevity Center, 136 Main Street, Forrestal Village, Princeton, New Jersey 08540
In the 1960s hit television show The Outer Limits, the following (paraphrased) introduction was meant to focus our minds on what we see and then make an interpretation:
There is nothing wrong with your television set … . We are controlling the transmission. We will control the horizontal. We will control the vertical. We can roll the image, make it flutter. We can change the focus to a soft blur or sharpen it to crystal clarity … . You are about to experience the awe and mystery which reaches from the inner mind to—The Outer Limits.
Cardiac computed tomography (CT) is certainly a subject of “awe and mystery” but most certainly not new. Physicians and scientists have been looking at the heart using CT since the 1970s, from the first scanners actually designed specifically to look at the heart such as the Mayo Clinic's single and unique dynamic spatial reconstructor (DSR), to the commercial development of electron beam computed tomography (EBT), to the current range of multidetector computed tomography (MDCT) spiral scanners. However, there is no question that cardiac CT has been vastly and wonderfully improved and “sharpened to crystal clarity” from its inception to the current state-of-the-art 64-slice and beyond MDCT.
Although EBT was faster in image acquisition and provided less radiation to the patient than almost any other scanner developed from 1995 to 2010, the drawbacks of EBT were its limited spatial resolution along with the fickleness of a camel and the reliability of an old TV. Much has been written about the issues of radiation from MDCT, especially for imaging of the heart, because overlapping images were necessary to account for gating with the electrocardiogram. Never in my 30 years of doing and reporting on all sorts of cardiac imaging has an industry responded to such a challenge and affront and solved many of these radiation problems through improved scanners, more detectors, and clever protocols to now allow diagnostic cardiac CT to be done with radiation doses that currently range from equivalency with a standard mammogram to less than that of standard diagnostic invasive cardiac catheterization, not to mention as little as one-tenth that of modern nuclear stress testing (the latter being the 2 best-known complementary methods to define the extent, locale, and severity of coronary artery plaque disease).
Regardless of the challenges for applications of EBT and MDCT to the care of patients, the path of the scientific information regarding coronary atherosclerotic plaque development, diagnosis, and prognosis has been established through many published studies. However, these publications have generally focused on 1 aspect of the spectrum of data derived, mostly the coronary artery calcium (CAC) score (Agatston/volume/mass) for noncontrast studies and the number of >50% stenosis on contrast-enhanced cardiac computed tomography angiography (CTA) studies. No matter how you look at these data, what is being assessed is the severity of focal or global atherosclerotic plaque severity, each providing clear prognostic data.
However, here is the rub: what is the potential answer in terms of prognosis derived from the diagnostic cardiac CT to look at multiple measures of plaque severity, for example, calcified plaque, noncalcified plaque, mixed plaque (both calcified and noncalcified), and severity of focal stenoses in the largest group seen by the practicing cardiologist—the outpatient referral? Additionally, it is vastly different, mainly because of pre-test likelihood that prior data derived from hospitalized and emergency department patients may not be reflective of what might be found in general cardiology practice when faced with a moderate- to high-risk patient referred as an ambulatory outpatient with traditional risk factors with or without some variable nonacute symptoms.
The paper by Hou et al. (1) from Fu Wai Hosptial, Beijing, China, published in this issue of iJACC, goes a long way to addressing these prior gaps in our knowledge of the varied power of cardiac CT to define atherosclerotic plaque locale and severity and the long-term prognostic implications of such findings in ambulatory outpatients.
The current study had complete follow-up of 4,425 patients referred by their cardiologists for cardiac CTA for suspected coronary artery disease (CAD). Most (69%) were symptomatic, ranging from nonanginal chest pain and atypical and typical anginal chest pains, and the remaining had multiple cardiovascular risk factors. The study was done at a single large, referral medical center, and no patients were included who had known prior heart disease or had nondiagnostic images. Additionally, no patient who had revascularization within 60 days of the cardiac CTA (assuming the result would have influence on the cardiologist's decision to proceed to revascularization) was included in the 4,425 subjects.
For each cardiac CTA, the patients had an initial noncontrast CT to define CAC score followed by a 64-slice contrast examination using beta-blockers as necessary to keep the heart rate slow. In each subject, CAC score using previously established guidelines (2) was defined, as well as cardiac CTA-defined atherosclerotic plaque type (calcified, noncalcified, and mixed) and cardiac CTA coronary artery stenosis severity (no plaque observed [i.e., a normal study], nonobstructive [i.e., <50% stenosis], obstructive single-vessel disease, obstructive double-vessel disease, obstructive triple-vessel disease, and obstructive left main disease).
Median follow-up was 1,081 days (first quartile, 960 days; third quartile, 1,192 days). Major cardiovascular adverse events (MACE) were defined as a composite of verified cardiac death, verified nonfatal myocardial infarction, or verified coronary revascularization.
The major findings were not, frankly, unexpected (“disease” always trumps “risk”) but are otherwise quite thought provoking, as cumulative probability of 3-year MACE increased: 1) across strata for CAC score; 2) across strata for cardiac CTA plaque type; and 3) across strata for cardiac CTA stenosis severity.
Furthermore, the area under the receiver-operating characteristic curve (ROC) for predicting 3-year MACE was as follows (for reference, ROC area = 0.5, equivalent of a coin flip; ROC area = 1.0, a perfect test): 1) 0.71 for conventional risk factors; 2) 0.82 for conventional risk factors plus CAC score (p < 0.001 versus risk factors alone); and 3) 0.93 for conventional risk factors plus CAC score plus cardiac CTA results (p < 0.001 versus risk factors plus CAC score).
Kaplan-Meier survival curves for increasing CAC score, increasing cardiac CTA stenosis severity, and advancing of plaque severity (calcified, noncalcified, and mixed) all clearly showed increased likelihood of MACE during follow-up.
A CAC score of zero seems to be relatively consistently sacrosanct and continues to confirm a very low yearly risk of <0.1% per year; likewise a CAC score >400 continued to confirm a moderately high yearly risk. Although some investigators argue that CAC scanning should not be part of a cardiac CTA examination, as it adds to the radiation dose to the patient, most of my colleagues and I who have developed the CAC applications over these many years would argue (and are supported by the research in the current paper) that achieving a zero CAC score in an outpatient (the incidence of noncalcified plaque was 4% in the current investigation) does not justify proceeding to the higher radiation dose cardiac CTA. Additionally, and most agree on this point as well, that a CAC score >1,000 reduces the ability to make a diagnostic cardiac CTA and is consistent already with advanced CAD, and thus does not justify performing a higher radiation dose cardiac CTA.
The value of the Framingham Risk Score in population studies is beyond question, but application to the care of a single person in a given population stratified by Framingham Risk Score is simply less precise—resulting in the over-treatment of lipids and other risk factors in patients for whom true CAD is a remote issue versus under-treatment of lipids and other risk factors in patients for whom true CAD is a more immediate concern. However, application of individual disease (and not risk) markers such as CAC, type of plaque, and stenosis severity further defines and focuses the treatment and goals of treatment of modifiable risk factors in a more precise manner in a given patient. Although it remains to be proved that using cardiac CT disease stratification over conventional risk factor stratification would ultimately save health care expenditure for CAD, a simple analogy would be treating you with hypertensive drugs because you have a risk for hypertension (family history, overweight, high salt intake, and so forth) or treating you with hypertensive drugs because you have the disease of hypertension (taking your blood pressure as a measurement of its severity).
It is like changing the focus and stage height on the diagnostic microscope to scan a few or many areas of what is on the slide and then taking this information over to the telescope and focusing on the long-term prognosis. We need to focus on both the forest and the trees.
What next? Regional myocardial blood flow and reserve using old methods (3)? New methods of nonprovocative functional flow reserve? Metabolic imaging using cardiac CTA to define the vulnerable patient among similar trees in a forest? There is always need to improve our predictions on the basis of current facts—but these facts should be subject to microscopic inspection of disease, and not simply risk for disease, and then applied with clarity to look forward with the prognostic telescope to render better short-term and long-term predictions, thus guiding medical and surgical interventions. However, in closing, all of these and other data are highly significant on the basis of predictor models—but just the same, when applying them to an individual patient, I am always reminded of a mentor who said this to me: “A p value is no substitute for using your brain.”
Dr. Rumberger has reported that he has no relationships relevant to the contents of this paper to disclose.
↵⁎ Editorials published in JACC: Cardiovascular Imaging reflect the views of the authors and do not necessarily represent the views of JACC: Cardiovascular Imaging or the American College of Cardiology.
- American College of Cardiology Foundation
- Hou Z.-h.,
- Lu B.,
- Gao Y.,
- et al.
- Bell R.M.,
- Lerman L.,
- Rumberger J.A.