Carotid IMT and Plaque Presence Improve Prediction of Coronary Heart Disease Risk

April 09, 2010

April 9, 2010 (Houston, Texas) — Adding an assessment of plaque and carotid intima-media thickness (CIMT) to traditional risk factors significantly improves the prediction of coronary heart disease, according to the results of a new study [1]. Importantly, nearly one-quarter of patients were reclassified into another risk category, with approximately 40% of patients considered intermediate risk reclassified to the lower risk category.

"If you look at our traditional risk-stratification methods, they are far from perfect," lead investigator Dr Vijay Nambi (Baylor College of Medicine, Houston, TX) told heartwire . "The majority of heart attacks that happen in the United States happen in people who are low or intermediate risk. This is the basis on which C-reactive protein [CRP] was developed [as a risk marker] and the Reynolds risk score was developed. We're looking for additional biomarkers, imaging markers, to help us further identify individuals who are at risk for clinical events."

In an editorial accompanying the study, Drs James Stein and Heather Johnson (University of Wisconsin, Madison) note that the United States Preventive Services Task Force (USPSTF) recommends against measuring anatomic markers of atherosclerosis, including CIMT [2]. The recommendations are based on the lack of specific data regarding the independent predictive value of CIMT in patients at intermediate risk of coronary heart disease. Also, there are concerns about the ability of CIMT to reclassify these intermediate patients into lower or higher risk categories.

This new study "provides clear answers to several concerns expressed about CIMT imaging by the USPSTF, and their findings validate the recent consensus statement recommendations for appropriate patient selection for use of carotid cardiovascular disease risk prediction," according to Stein and Johnson. "This paper closes the discussion about the incremental value of carotid ultrasound for coronary heart disease risk prediction in patients at intermediate risk, thus opening the door for outcomes research studies that are required to determine if atherosclerosis imaging is as helpful as its proponents believe."

The results of the study and editorial are published in the April 13, 2010 issue of the Journal of the American College of Cardiology.

Adding Incremental Value to Traditional Risk Scores

In the present study, the researchers used data from the Atherosclerosis Risk in Communities (ARIC) trial to assess the incremental benefits of adding assessments of CIMT and plaque burden to traditional risk-factor screening for coronary heart disease. In total, the group included 13 145 individuals who were free of coronary heart disease or stroke and followed for an average of 15 years. Overall, there were 1812 clinical events, a majority of these being myocardial infarction or coronary death.

Several different models were used to assess coronary heart disease risk: traditional risk factors only, which includes age, sex, systolic blood pressure, antihypertensive-medication use, total cholesterol, high-density lipoprotein (HDL) cholesterol, diabetes mellitus, and smoking status; traditional risk factors plus sex-specific CIMT measurements; traditional risk factors plus the presence/absence of atherosclerotic plaque; and traditional risk factors plus CIMT plus the presence/absence of atherosclerotic plaque. The area-under-the-curve (AUC) for the 10-year risk of coronary heart disease was used to measure the ability of the models to predict future disease.

Overall, adding the presence of plaque plus CIMT information, individually or together, to traditional risk factors significantly improved the AUC in men and women. Adding CIMT data alone in women, however, did not significantly improve the AUC, while models that included the traditional risk factors and CIMT were not significantly improved with the addition of the presence of plaque in men.

Adjusted AUC for the Different Models

Model Overall Men Women
Traditional risk factors 0.742 0.674 0.759
Traditional risk factors plus CIMT 0.750 0.690 0.762
Traditional risk factors plus plaque 0.751 0.686 0.770
Traditional risk factors plus CIMT plus plaque 0.755 0.694 0.770

Using the data from the carotid ultrasound, the addition of plaque information and CIMT resulted in the reclassification of risk, particularly among individuals in the intermediate-risk category. For those with a 10-year risk of coronary heart disease of 5% to 10% and 10% to 20%, 37.5% and 38.3% were reclassified into another risk category. Overall, plaque presence was a more important marker than CIMT for reclassification in women.

The net reclassification index (NRI), a measure that examines the net effect of adding a biomarker to the risk-prediction model, and the clinical NRI, defined as the NRI in intermediate-risk patients only, were used to assess how correctly CIMT and the presence or absence of plaque reclassified predicted patient risk. Traditional risk factors plus CIMT and carotid plaque led to an NRI of 9.9% in the overall sample, or 8.9% in men and 9.8% in women. Adding either CIMT or carotid plaque led to a clinical NRI of 16.7% and 17.7%, respectively, and a clinical NRI of 21.7% when used together.

A Better Crystal Ball

Interestingly, the majority of reclassified patients moved from the intermediate-risk category to the lower-risk category. No patient from the low-risk group was reclassified as high risk, and nobody from the high-risk group was classified as low risk. Speaking with heartwire , Nambi said the issues these carotid ultrasound findings raise--whether or not clinicians should back off therapy in intermediate-risk patients reclassified as lower risk or ramp up therapy in patients bumped to a higher-risk category--needs to be studied in a randomized, clinical trial.

"Overall, the net effect of adding the imaging marker was good," he said. "In other words, it correctly identified people who had an event and moved them to a higher-risk group and also correctly identified people who did not have an event and moved them to a lower-risk group. But it doesn't mean that it identified everybody correctly. Some people that were dropped into a lower-risk group still ended up having events. Basically, what I would do is, if they moved into a higher-risk category, I would be more aggressive in treating them, but if they moved to a lower-risk category, I would not necessarily hold off treating them at this time point."

Dr Amit Khera (University of Texas Southwestern, Dallas), who was not part of the study but who spoke with heartwire about the results, said that carotid ultrasound fits into a group of "tie-breaker" tests that help clinicians refine patient risk. This is predicated on the fact that the current clinical risk-prediction algorithms are imperfect, particularly in younger individuals and women, leading to a search for a "better crystal ball." Regarding the 62% of intermediate-risk patients reclassified as lower risk, Khera said this finding is not unique to CIMT or this particular cohort.

"The irony is that our call to arms, what we're all saying, is that we're missing people and we need better tests," said Khera. "It turns out that while the tests certainly do pick up people that we're missing, they're also showing that we might be treating more people who might not need it. From a medical standpoint, that gives some doctors a little more angst, withdrawing therapy from an individual who is now low risk. They might be now moved to a lower-risk category, but lower risk means that some of patients are still going to have events. There is no answer to that right now, downgrading risk and when to pull back on therapy."

Khera noted, however, that most of these tie-breaker tests, including CRP and coronary artery calcium screening, are used in patients not currently on therapy. In primary-prevention cases, a 45-year-old patient might have borderline low-density lipoprotein cholesterol, around 120 mg/dL, and also have a mild family history of coronary disease. "These patients don't want to take a drug for the next several decades if they don't have to," Khera told heartwire . In these instances, plaque in the arteries or a higher IMT might be the push to start treatment with a statin.

In their editorial, Stein and Johnson note that while increments in AUC might seem small, they are similar to increases observed as the individual contributions of smoking status and systolic blood pressure to the Reynolds risk score for women and greater than the contributions of lipids, family history, and high-sensitivity CRP.