Improving Risk Stratification for Cardiovascular Disease
 |
SPECT & patient subgroups
Precisely defining short- and long-term cardiac risk is pivotal for guiding therapy in individual patients. SPECT MPI has been shown to be an important method to risk stratify patients for CAD, and it has proved beneficial in a variety of patient subgroups.For example, Al-Mallah et al evaluated more than 7,000 patients to better understand the link between kidney function and SPECT MPI results. Patients with a mean age of years were followed for a mean of 2.6 years (Circ Cardiovasc Imaging 2009;2:429-36). Researchers found that the risk of death increased with worsening kidney function; however, at each stage of impaired renal function, patients with abnormal SPECT MPI had a significantly increased hazard of adverse events. In addition, the magnitude of the total perfusion deficit and ischemia on MPI was associated with worse outcomes.
“For any degree of ischemia, the worse the renal failure was, the worse the cardiac event rate was,” says Daniel S. Berman, director of cardiac imaging at Cedars-Sinai Heart Institute in Los Angeles and a colleague of one of the study authors. “The low-and high-risk groups, based on abnormal nuclear scans, were greatly separated in terms of their risk as the renal dysfunction progressed.”
Berman notes that this patient population is not well suited for imaging with contrast-enhanced CT, given the propensity for dense calcification and the risk of kidney injury associated with iodinated contrast medium. Therefore, the use of nuclear imaging not only leads to improved risk stratification, compared with CT, but it also reduces the chance of acquiring suboptimal images and of inducing renal complications.
In another study, Hachamovitch et al followed 5,200 elderly patients (75 years or greater) for a median of 2.8 years who underwent either pharmacological or exercise stress imaging (Circulation 2009;120:2197-206) to determine SPECT’s prognostic discrimination. They noted that CAD in the elderly frequently presents with silent ischemia, atypical symptoms or nonspecific functional status deterioration, making it challenging to identify those at risk.
They found that both reversible and fixed defects on MPI were highly associated with cardiac death, and that ejection fraction and perfusion data added incrementally to each other, “further enhancing risk stratification.” They also noted that increasing ischemia was associated with increasing survival with early revascularization, whereas in the setting of little or no ischemia, medical therapy had improved outcomes.
Finally, Hachamovitch et al found that even in the elderly, risk increased “dramatically with increasing age and other clinical factors, thus marking these cohorts as heterogeneous and necessitating examination of multiple patient subsets to understand prognostic test performance.”
Calcium scoring
While an abnormal SPECT scan is prognostic for cardiac events, a normal SPECT scan is not necessarily an indication of the absence of disease. Chang et al found that as the coronary artery calcium (CAC) score, as measured by noncontrast CT, increased in asymptomatic people with normal perfusion scans, the risk of having a cardiac event also increased significantly, especially in those with severe CAC scores—greater than 400 (J Am Coll Cardiol 2009;54;1872-1882).Researchers, who followed more than 1,100 patients for a median of 6.9 years, found that approximately half of the 977 patients with a normal SPECT had a CAC score of at least moderate severity, signaling cardiac risk that would not otherwise have been predicted. In addition, an abnormal SPECT result increased with increasing CAC score, from less than 1 percent (CAC score less than 10) to 29 percent (CAC score greater than 400).
“The premise is not just to identify high-risk people with normal SPECT studies; it’s also to identify people with early disease so we can intervene and prevent cardiac events,” says senior author John J. Mahmarian, MD, a cardiologist at the Methodist DeBakey Heart and Vascular Center in Houston. “At this point, however, it is more speculation rather than a proven concept.”
There is supportive data, however, mostly from intravascular ultrasound (IVUS) studies that lipid-lowering therapy can reduce noncalcified, or vulnerable, plaque burden. “If someone has evidence of calcification, he or she most likely has five-times as much noncalcified plaque, which is more prone to rupture and cause an occlusion,” Mahmarian says.
Interestingly, Inoue et el used serial coronary CT angiography (CCTA) imaging to demonstrate noncalcified plaque regression in a small cohort of statin-treated patients (J Am Coll Cardiol Img 2010;3:691-698). They said the technique could be used for clinical trials designed to assess plaque progression and regression, and noted, “The use of statins even at a low dose may result in significant changes in plaque morphology.” In fact, Voros et al recently initiated the AFRICA (Atorvastatin plus Fenofibric acid in the Reduction of Intermediate Coronary Atherosclerosis) trial, a single-center, prospective, double-blind, randomized, placebo-controlled study intending to enroll nearly 400 patients (J Cardiovas CT 2010;4,164-172).
While IVUS has been the standard of care to measure plaque morphology, it is not appropriate in the primary prevention setting. “Therefore, the validation of CT as a noninvasive tool to potentially characterize plaque progression in primary prevention patients is an important aspect of this trial,” they wrote. Investigators will use the 320-multidetector Aquilion One CT scanner (Toshiba America Medical Systems).
While current ACC/ASNC (American College of Cardiology/American Society of Nuclear Cardiology) guidelines recommend SPECT imaging to assess for ischemia in asymptomatic subjects with a severe CAC score (over 400), Mahmarian’s standard protocol is to order a CAC test for patients with normal perfusion studies but multiple Framingham risk factors. He notes that the results of his group’s study and others support this protocol because approximately 20 percent of those with normal perfusion studies will have a CAC score of at least moderate severity that “cannot be predicted from the patient’s clinical profile.”
“If you have any calcium, you have coronary disease, and I use CAC scores to determine how aggressively I should treat patients with lipid-lowering therapies,” Mahmarian says.
Anand et al found similar results looking at about 500 asymptomatic diabetic patients (Eur Heart J 2006;27:713-721). A total of 127 patients who had CAC scores greater than 100 underwent SPECT MPI. In addition, a random sampling of 52 patients with scores less than 100 underwent perfusion imaging. Thirty percent of patients had perfusion abnormalities, the majority of which were reversible.
The amount of CAC was linearly correlated with fixed, mild, moderate and large reversible defects. An analysis showed that the CAC score was the only predictor of perfusion abnormality. At two years, no one with a CAC score less than 10 had a cardiac event, even in those with at least a 5 percent perfusion abnormality. The majority of events occurred in patients with scores above 400.
Researchers found that the CAC score predicted events more accurately than the Framingham risk score. Yet, they noted, “It is paradoxical that although established cardiovascular risk factors predict CAC, they are not predictive of abnormal myocardial perfusion, yet CAC predicts myocardial ischemia.”
American Diabetes Association guidelines recommend exercise stress test screening for CAD in diabetic patients with an abnormal resting ECG indicative of MI and two or more risk factors. However, Anand et al concluded, “Our study shows that high-risk but asymptomatic diabetic patients can be identified by a strategy of initial CAC imaging followed by selective MPI, which has the advantage of combining the high sensitivity of CAC imaging with the specificity of MPI for predicting angiographic stenosis.”
In a landmark move last November, new ACC/AHA (American Heart Association) guidelines for assessing cardiovascular risk in asymptomatic adults for the first time acknowledged the value of CAC scoring. However, there was not enough evidence to recommend serial CAC exams. In December, however, Budoff et al released a study showing the value of serial CAC tests to predict mortality (J Am Coll Cardiol Img 2010;3:1229-1236). Budoff concedes that his study of more than 4,500 subjects is not sufficient to update the guidelines just yet. However, if there are similar findings in a MESA (Multi-Ethnic Study of Atherosclerosis) cohort of more than 6,000 looking at serial CAC scanning, “that should be good enough for the guidelines to recommend repeat CAC scans.” The MESA data are expected within a year.
Cardiac CTA
While there have been many declamations about overutilization of diagnostic medical imaging, in general, imaging is suboptimally utilized for risk stratification, says Udo Hoffmann, MD, MPH, director of the cardiac MR, CT, and PET program at Massachusetts General Hospital in Boston. In a landmark study, researchers combed the CathPCI Registry of the National Cardiovascular Data Registry (NCDR), which is sponsored by the ACC and the Society for Cardiovascular Angiography and Interventions (SCAI), to determine the diagnostic yield for invasive catheter angiograms. They found that more than half of all referrals to the cath lab had no significant CAD (N Engl J Med 2010; 362:886-895).In the study, which included patients with no known CAD, Patel et al found that only 37 percent had obstructive CAD. They also noted the limited incremental value of a positive result on a noninvasive test (including any of a broad range of tests such as resting ECG, echocardiography, CT, or SPECT). Researchers were not able to determine which noninvasive test was used, but most were nuclear studies, says Hoffmann, who was not involved with the study.
Patel et al found that noninvasive testing was “paradoxically used more often in patients with a high Framingham risk score than in those with an intermediate or low risk score, a practice that is at odds with current guideline recommendations.” Hoffmann says these results provide a gateway for the use of CCTA to improve patient management. “CCTA has a very high negative predictive value for ruling out stenosis; it also can detect noncalcified plaque, which further helps to stratify risk; and radiation exposure for the exam can be as low as 1 to 3 mSv.”
The National Institutes of Health recently initiated the PROMISE trial, which will compare anatomic (CT) and functional (stress echo, nuclear and/or exercise treadmill) results to understand the value they bring to patient care. “This type of comparative-effectiveness research is rarely done in medical imaging,” Hoffmann says. Researchers hope to enroll 10,000 low- to intermediate-risk patients who present to the emergency department and analyze results for health and economic outcomes. “This is a new area of inquiry that will grow in the future and will provide the data necessary to justify reimbursement for cardiovascular imaging and techniques,” he says.
PET imaging
An important advance in last several years has been demonstrating the ability to measure myocardial blood flow reserve with rubidium-82 (Rb-82) PET, which “increases PET’s capability to identify patients at risk for developing cardiac events,” says Berman. Cardiac PET excels at identifying perfusion defects, particularly in specific patient populations that might be suboptimally imaged with SPECT. But rather than comparing one region of the heart with another, measuring absolute blood flow reserve allows each region to be compared to itself and to determine how much blood flow can increase in that region in a given vessel.Myocardial blood flow with 13N-ammonia and 15O-water as PET flow tracers has been validated; however, these tracers require an onsite cyclotron. Rb-82 does not rely on a cyclotron and already is commonly used to assess perfusion. Several studies have now documented the accuracy and reproducibility of Rb-82 to measure blood flow reserve.
Another development has been the preliminary data on flurpiridaz F18 (Lantheus Medical Imaging), a new PET perfusion agent. Berman, whose center is involved in the trial of the drug, says the agent is extracted “linearly across the whole range of blood flow, giving us the ability to see more minor perfusion defects than we could see previously.” It is being compared with technetium-99 sestamibi SPECT.
The agent has shown a higher specificity in the right coronary artery, suggesting fewer false positives compared with conventional SPECT MPI and, ultimately, fewer unnecessary trips to the cath lab. In addition, an F18-labeled perfusion agent does not need an onsite generator or cyclotron. Its long half-live (109 minutes) means it can be produced in regional cyclotron centers and shipped as a unit dose.
Biomarker debate
Biomarker screening is typically performed in low- and intermediate-risk patients and has revolved around when to start statin therapy, but there may be more to it, says James A. de Lemos, MD, of the University of Texas Southwestern Medical Center in Dallas. “Part of the personalized medicine approach is not just to identify who is at risk of heart disease, but also to look at different tests that predict different forms of heart disease. To date, we’ve focused only on predicting CAD and MI, and that is absolutely critical, but the epidemic now is heart failure, due to either thickening or weakening of heart muscle, and it is likely there will be different sets of factors that predict these complications,” he says.De Lemos and colleagues used a highly sensitive assay to detect minute levels of troponin T, undetectable with the conventional assay, in patients enrolled in the Dallas Heart Study (JAMA 2010;304:2503-2512). Higher levels of the biomarker—that would not normally be found—correlated with the prevalence of hypertension, diabetes, left ventricular (LV) mass, LV wall thickness and the proportion of individuals classified as having LV hypertrophy. Self-reported heart failure, CAD and cardiovascular disease were more frequent with higher levels of troponin.
The median follow up of the more than 6,000 patients was six years and troponin levels remained independently associated with all-cause mortality. De Lemos sees this marker as an opportunity to detect subclinical disease and intervene early or to stratify those already at high risk for more targeted therapy.
Regarding conventional biomarkers such as the Framingham risk score and C-reactive protein (CRP) to stratify CAD risk, De Lemos says they are imperfect, but cautions that newer candidates such as lipoprotein-associated phospholipase A2 must be rigorously tested before adopted into clinical practice. In a recent point/counterpoint essay, he and Enriquez argued that “traditional risk factors used alone or in combination do not predict risk well enough, leaving a significant number of ‘unpredicted’ cardiovascular events occurring in individuals classified as low risk by standard classification methods” (Prev Cardiol 2010;13[4]:152-159).
“Simply put,” they wrote, “the status quo is not sufficient and the need for novel risk assessment tools is substantial.” They see promise with markers that are more specific for cardiac disease such as B-type natriuretic peptide rather than nonspecific inflammatory markers like CRP.
Interestingly, Wilkins and Lloyd-Jones in the counterpoint said, “Existing biomarkers for clinical prediction used in isolation are ineffective” and are “unlikely to result in improved discrimination or substantial reclassification for population screening.”
They suggested that the “prevention of risk factors in the first place [such as hypertension, hyperlipidemia and smoking, among others] may be the best complementary, population-based method to reduce the burden of coronary heart disease.”
The debate on the various ways to better stratify cardiac risk will continue, and rightly so. Researchers are increasingly focused on ways to make risk stratification more personal. Genetic approaches are in their infancy, while the search for more specific blood biomarkers has heated up. CAC scoring is evolving into a highly effective means to personalize risk, while advances in SPECT, PET and CT will carry those modalities into the next era of improved risk stratification.