• Research Review
• Journal Review

Authors: K Johnson, D Dowe

Reference: AJR 2010; 194:1235-1243 (open access)
http://www.ajronline.org/cgi/content/abstract/194/5/1235

Reviewer: Robert Goldberg, PhD, Contributing editor, ProCor; Professor of Medicine and Epidemiology, Division of Cardiovascular Medicine, University of Massachusetts Medical School, Worcester, Massachusetts, USA

Reviewer comments: Given the magnitude of coronary heart disease (CHD) in the US and throughout the world, and the impact of CHD on quality of life, functional status, and long-term prognosis, the need for a cost-effective and efficient screening program for identifying adult men and women at increased risk for CHD has been recognized for many years.

While a number of coronary risk schemas have been developed over the years to assist in the identification of men and women at increased risk for CHD in whom either lifestyle or treatment regimens could be directed, the Framingham risk score has been the most commonly accepted algorithm used for predicting the risk of CHD in American men and women. While this risk model has been shown to have good discriminatory utility, it does not include information on BMI, family history of CHD, markers of inflammation, or the results of various cardiac testing modalities, including both invasive and noninvasive test results, that might further aid in the detection of at risk individuals. This predictive instrument has been shown to be useful in clinical practice but certainly can be improved upon, like all predictive models, especially if it enhances an individual's efforts to modify their risk factors for CHD through the use of different preventive and/or treatment strategies.

Given this backdrop, the results of the present study suggest that the presence of coronary calcium as visualized on coronary angiography can be used to detect individuals at increased risk for coronary atherosclerosis. Indeed, the presence of coronary calcium added value to the commonly accepted risk factors for CHD in predicting extent of underlying atherosclerosis. This modality cannot, however, be recommended for population-wide screening at the present, however, until its effect on various CHD-related outcomes, as well as risks associated with the procedure and accompanying costs, can be more systematically evaluated. While the costs of diagnostic testing modalities for the identification of coronary artery calcium need to be reduced and shown to be cost-effective as a population-based screening approach, the present results suggest that if the Framingham risk algorithm is used as the first step in a screening program for detecting coronary atherosclerosis, and patients with a 10 year risk score for developing CHD of at least 10% are included, nearly two thirds of women and one quarter of men with considerable atherosclerosis will be missed by reliance on the Framingham risk score. In contrast, the visualization of any coronary calcium on diagnostic testing was shown to be highly sensitive for detecting the presence of moderate to more advanced degrees of coronary atherosclerosis.

Despite the present findings, the Framingham risk score should clearly not be abandoned for the evaluation of adult men and women considered to be at varying risk for CHD. This is a relatively simple inexpensive test, which combined with careful questioning of the patient and collection of additional data on BMI, family history of premature CHD, and the results of various lab tests can be used to identify patients at moderate or high risk for CHD.

Moreover, the role of other novel risk factors, such as markers of inflammation or other indicators of underlying atherosclerosis (e.g., intima-media thickness) need to be considered as possible risk predictors of extent of CHD or coronary artery stenosis.

Irrespective of the screening approach utilized to identify at risk men and women, given the ongoing burden of CHD in the US, and the emerging epidemics of obesity and diabetes, it would appear prudent for most, if not all, adult American men and women to adhere to a heart healthy diet, exercise on a regular basis, and maintain their weight on as optimal level as possible. These are tried and proven approaches to lowering one's risk for CHD as well as likely enhancing one's quality of life. The results of the present study certainly do not rule out the utility of screening for the presence of coronary calcium in individuals deemed to be at moderate and higher risk for CHD as well as the use of several easily asked questions that comprise the Framingham risk score. This risk score and careful questioning by one's physician remain as important gatekeepers in an effective screening program for coronary atherosclerosis and before more expensive tests, with varying sensitivity and specificity and associated patient apprehension and anxiety, are undertaken.

Purpose of study: To compare the sensitivity of the Framingham risk score with the presence of coronary artery calcium, based on the results of CT angiography, in detecting moderate or more severe degrees of underlying coronary atherosclerosis.

Location of study: New Haven, CT

Study design: Cross-sectional

Results: The study population consisted of persons who underwent coronary CT angiography at a private outpatient radiology practice. These patients completed a survey questionnaire, had their coronary risk factors assessed so that the Framingham risk score could be calculated and patients classified into various risk strata (e.g., low, intermediate, high risk), and subsequently underwent imaging of their coronary arteries. Each of the imaging studies was interpreted by a radiologist and two methods were utilized for estimating the extent of plaque burden and coronary artery blockage detectable on coronary CT angiography; one of these methods quantified the amount of plaque burden in a coronary segment as being either none, mild, moderate, or heavy while the second method utilized the Duke prognostic index to quantify degree of coronary artery stenosis. The study sample consisted of 2123 individuals who underwent cardiac imaging studies. Approximately two thirds of the sample were men (n=1416), the average age of this screened population was early 50 years, more than 90% were white, and 44% of men and 15% of women were considered to be at intermediate or high risk for CHD based on the Framingham risk score. The majority of the study sample was asymptomatic, their average BMI was approximately 29, slightly more than one half had hypertension, and approximately 10% had diabetes.

Overall, 30% of the men and 20% of the women had moderate or greater plaque burden and 11% of the men and 7% of the women had moderate or greater extent of coronary stenosis (Duke prognostic score greater than 3). In contrast, approximately one third of the men and nearly one half of the women failed to have any detectable atherosclerotic plaque.

In identifying a segment plaque score greater than 4, which was determined to indicate the presence of at least moderate atherosclerotic disease burden, the presence of any detectable calcium was highly sensitive in both women (97%) and in men (98%). The sensitivity of calcium for detecting moderate or greater stenosis burden was 97% in men and 92% in women. On the other hand, a Framingham risk score greater than 10% was considered to be 74% sensitive in men and only 36% sensitive in women for detecting who was at risk for having more severe coronary stenosis.

The negative likelihood ratio for the presence of calcium for detecting moderate or greater plaque burden was approximately 0.4 for men and women with only slightly higher negative likelihood ratios of calcium for detecting moderate or greater coronary artery stenosis. On the other hand, these likelihood ratios were considerably higher in men and women with a Framingham risk score greater than 10%.