Coronary Calcium Scanning: Role in Preventive Cardiology

Hello, my name is Dr. Nathan Wong, professor and director of the Heart Disease Prevention Program at the University of California, Irvine. I'm here to talk to you today about the role of coronary calcium screening in preventive cardiology. I've been involved with research on coronary calcium and cardiovascular disease now for 30 years, so this topic is quite dear to my heart. These are my disclosures, most of which are not related to this presentation. What we're going to be covering in this lecture is first a little bit about the background and pathophysiology of coronary calcium and the rationale for coronary calcium screening. Then we'll discuss the relation of coronary calcium to future cardiovascular disease events and mortality. We'll learn about the benefits of coronary calcium screening and then finally, we'll end with discussing current guidelines for coronary calcium screening and it's recently heightened role in choosing which patients may be most suitable for treatment, according to the 2018 cholesterol management guidelines. Let's begin by just discussing the background and rationale for coronary calcium screening. As most of you are aware, there was a dramatic increase in deaths due to heart disease over much of the past century. And then this, of course, has declined substantially over the last 30 years or so. However, coronary heart disease still remains the leading cause of death in the United States. Certainly an important challenge is the fact that many people who die suddenly of coronary disease will not have this adequately detected initially and will not have had, for example, a positive exercise stress tests. Many of these people have only mild non flow limiting coronary plaque that is not sufficient to result in a positive exercise stress test so, thus receive a clean bill of health yet oftentimes can drop dead of a myocardial infarction or sudden coronary death with rather modest amounts of disease present. We see from these data involving pulling 4 major studies that about two thirds of myocardial infarctions are actually caused by low grade stenosis of less than 50%. So it's crucial that we find better ways to detect coronary atherosclerosis before it has become clinically significant. If we look at the diagnostic sensitivity of different test, we can see that many of the tests we traditionally use to diagnose coronary artery disease are not going to be positive unless significant disease has become present. And many of these tests are quite a bit more expensive than coronary calcium screening. Invasive modalities of course, can detect disease earlier, but they are not normally indicated for people who are asymptomatic. What are criteria required for a good screening test? Well, first of all, the test needs to provide an accurate determination of the likelihood that an asymptomatic person has a condition, in this case, atherosclerosis. The test should be reproducible and reliable. So if we repeat the test, say 5 minutes from now in the same person, that we get pretty much the same result. Importantly, the test should be able to detect individuals where early intervention, such as with preventive pharmacologic medicines, can have a beneficial impact and very importantly, perhaps most importantly, should provide incremental value to risk predicted by office based risk assessment. So the idea here is that the new tests, if if we're going to spend whatever it it costs to do the tests and and if it's more than that of office based risk factor screening such as such as doing doing blood pressure test and lipid test and assessing family history and cigarette smoking, for example, that the new test really needs to be able to show incremental statistical prediction for cardiovascular events beyond the traditional office space measures. There are tests for just about every part of the body, as one can see, ranging from looking at carotid plaque in the carotid arteries using carotid intima media thickness. Also, one can do MRI. And of course, we have coronary calcium that we're going to talk about. And then in other extremities of the body, for example, one can measure vascular compliance by radial tonometry and even microvascular reactivity by fingertip tonometry. However coronary artery calcium evaluation by CT is really the only standardized noninvasive tool to assess the extent of coronary arterosclerosis. We know that coronary calcium invariably indicates the presence of atherosclerosis, even though atherosclerotic lesions do not always contain coronary calcium. And while coronary calcium generally occurs in the later stages of plaque development, we know that calcium deposition can occur fairly early in life. We've we've seen, for example, coronary calcium in adolescents with familial hypercholesterolemia and even in lesions that are not necessarily that advanced. Coronary calcium, the amount of coronary calcium generally correlates with the overall amount of atherosclerotic plaque, and hence it's highly sensitive for angiographic disease. The tip of the iceberg concept is an excellent way to kind of illustrate the relation of coronary plaque to overall atherosclerotic plaque. The idea being here that the amount that is above the ocean is the coronary calcium, and the larger that amount, the larger the amount below the surface of the ocean, which represents the rest of the atherosclerotic plaque. However, there is great variability around this relationship. So some people may have a large amount of calcified plaque and smaller amounts of non calcified plaque, whereas the opposite can also be true. But in general, the more the amount of calcified plaque, the more the amount of overall atherosclerotic plaque. Atherosclerotic calcification is an active process but needs to occur when other aspects of atherosclerosis are also present. There's evidence that osteopontin and its messenger RNA involved in bone mineralization have been identified in calcific lesions. And alkaline phosphatase is central to early calcium deposition. Also, vascular smooth muscle cells produce matrix vesicles which regulate mineralization in the vascular intima and media. And we know that atherosclerosis, atherosclerotic calcification occurs mainly in the intima and medial classification mainly contributes to vascular stiffness and is common in those of advanced age and with diabetes and chronic kidney disease. Data from correlation studies with coronary angiography show a very high level of sensitivity for for obstruction, meaning that in in persons who have obstructive coronary arteries from the angiogram that that virtually 100% will have at least some calcification. Also the negative predictive power is more than 99%. So in persons with a zero calcium score, there's virtually virtual certainty that there is no obstructive angiographic disease. So it has been felt that the coronary calcium scan could potentially be used as a filter prior to angiography to help weed out doing unnecessary angiograms that would turn out to be negative. This is the basics of the coronary calcium scan and protocol, which generally involves triggering of the the ECG to do the scan at 60 to 70% of the R R interval and generally 3 mm contiguous scans going from the carina to the apex. However, there is some variability sometimes in the actual, um, uh, in the actual distance between the scans or even the R R interval that is used. This is an example of the, the, the what the software shows with regard to the calcium score obtained from each of the major coronary arteries that are assessed, which are generally the left main, left anterior descending, right coronary artery and the left circumflex artery. And one can see that a total score for each artery is provided and then the sum of the scores represents the total score that is generally reported on the patient report and some software also calculates other measures such as the total volume. This is an example of the calcium scan report that a patient might might get from their scan and centers. That basically describes where where the calcium is and how significant it is and what clinical recommendations may be appropriate for further tests and or follow up. This is an example of significant coronary artery calcium and the left main and left anterior descendant artery. We know that the dosage of coronary calcium scan is is quite low similar to that of a mammogram, and a long distance air flight. And about one third that of annual background radiation. Let's talk now about the relation of coronary calcium to cardiovascular events and mortality. We know when we see coronary calcium, this means that atherosclerosis is present in that particular vessel. And we also know that higher levels of coronary calcium do correlate with higher risk for cardiovascular and coronary heart disease events. Also a zero calcium calcium score, which is quite common, actually suggests a very low probability of obstructive disease and less than a 1% chance of a heart attack and stroke over the next five years. So let's review some of the data that supports these points. Importantly, we see from these data that we published a number of years ago from the multi-ethnic study of atherosclerosis or MESA that you see in the third, the third bar, the middle bar in each group is the prevalence of a positive coronary calcium score. And one can see that that ranges in men age 45 to 54 from about 39% up to about 82% in those aged 75 to 84. And and correspondingly, in women, they start out a lot lower at only 10% for those age 45 to 54. But end up similarly at around 80% by age of 75 to 84. Note that the prevalence that the presence of coronary calcium is closely associated with the likelihood of having evidence of atherosclerosis in other vascular beds, which was the purpose of the study. So for example, abdominal aortic calcium or AAC is actually present at most age groups in both men and women to a greater extent than is coronary calcium. So you can see that in in men it's also around 38% in ages 45 to 54, but ends up at around 95% in those who are age 75 to 84. In women, you don't see the gender difference that you see in coronary calcification being quite a bit lower in women. But you see that the prevalence of abdominal aortic calcium is almost as high as that in men going from about 29% in ages 45 to 54 to 94% in those who are age 75 to 84. We also see increase in prevalence with age for carotid intima media thickness, as well as a positive ankle-brachial index of less than 0.9. These also increase with age, but are less common than having a coronary or abdominal aortic calcification. Basically by the age of 65, 90% or more of both men and women have evidence of atherosclerosis in at least one of these four vascular territories. So basically atherosclerosis as essentially ubiquitous by the seventh decade of life. These are some initial data that showed the relationship between a calcium score and cardiovascular events from one of the early studies which we published now nearly 20 years ago. And this showed, for example, people with a calcium score of greater than 271 to have nearly a ninefold greater increase in risk of cardiovascular events compared to those with scores of 1 to 15. Other much larger studies, such as this one from Leslie Shaw and colleagues corroborated these results demonstrating a significant relation of higher calcium scores with worsening survival in both men and in women. So clearly one can see poor survival, especially for those with calcium scores of 400 or higher, and especially when the calcium score is greater than a thousand. Also, Budoff and colleagues showed a a significant relation of greater levels of coronary calcium with cumulative survival over a longer follow up of ex extending beyond 10 years. So here one can see for example in people with calcium scores of more than a thousand that greater than 25% of these people will be dead after about 10 years. The MESA study was the seminal longitudinal prospective study to confirm these earlier studies and showed showed also higher levels of coronary calcium to be directly related to the incidence of coronary heart disease events. In this subanalysis of the MESA cohort we also demonstrated this relationship to hold in persons with diabetes and with metabolic syndrome. In fact, in persons with diabetes you can see a ten fold risk in coronary heart disease event rates going from a calcium score of zero to a calcium score of 400 or higher. So this really proves a point that diabetes is a very heterogeneous condition and is not necessarily a coronary heart disease risk equivalent. Basically, persons with either diabetes or metabolic syndrome need to have a calcium score of 100 or higher before they reach that approximate 2% per year risk that has been equated with a cardiovascular disease risk equivalent. These data tend to support guidelines released several years ago recommending calcium scoring for cardiovascular risk assessment in asymptomatic adults aged 40 and over who have diabetes. The MESA study was also instrumental in showing that coronary calcium improves the C statistic a measure of risk prediction beyond the Framingham Risk score more than any of the other measures tested here in this paper that included carotid intima media thickness, flow mediated dilatation, c reactive protein, family history or ankle-brachial index. While most of these did provide statistically significant improvement in the C statistic, the magnitude of the improvement was far less than that seen from adding coronary calcium to Framingham Risk Score, where the C statistic improved from 0.6 to 2.78. Also, the Rotterdam study around the same time showed that the calcium score in that study was a more important predictor of risk than any of these other novel biomarkers are subclinical modalities. What does progression of coronary artery calcium mean? Well, we know that if you have coronary calcium at baseline, you're more likely to progress over the subsequent years. This is an example of a patient whose coronary calcium in the right coronary artery increased going from 1993 to 1995 to 1997. However, it's the baseline calcium score that is is far more powerful in predicting future cardiovascular events. And we also do not have proven interventions that can demonstrate effects on reducing the amount of coronary calcium. So hence, guidelines have not supported repeating the calcium scan for measuring progression of coronary calcium. However, we will return to this issue in a later slide. These are also data we published from MESA showing that yes, the greater the rate of progression, we do see a higher incidence of coronary heart disease, but the baseline calcium score is by far more important in predicting who's going to get an event. What are the benefits of coronary calcium? We have quite a bit of data that's accumulated that show that that by visualizing your calcium scan that seems to motivate better adherence to therapy. So in effect, is a picture worth a thousand words? These data we publish now more than 20 years ago shows us point in and from surveying about 700 men and women who had a calcium scan. And what we found was that people who had a positive calcium score and a higher to score, they were more likely to start new use of aspirin, cholesterol lowering medicine, see their doctor, lose weight, decrease in dietary fat. But also they worried more. So we know that potentially risk reducing behaviors may be reinforced by knowledge of a positive scan and this was even independent of their coronary risk factor status. The PACC study done by Alan Taylor and colleagues also showed greater incidence of statin use, as well as aspirin use over a six year period in people who had a positive coronary calcium scan. And more recently, the Eisner randomized controlled trial where more than 2000 middle aged adults with risk factors were randomized between having a scan versus no scan. What was shown was that people who had a scan and in particular if their calcium scan was greater than 400, they were more likely to show reductions in LDL cholesterol, systolic blood pressure, start exercising, take lipid lowering or blood pressure lowering therapy or aspirin, as well as adhere to their lipid lowering drugs. The scan group had no change in the Framingham risk score compared to an increase in the Framingham risk score and the people who were not scanned. Importantly, what was noted was there was no increase in downstream tests and costs from having a coronary calcium scan, which was a concern that were raised by some of the critics earlier on. So this slide shows there was actually lower cost, both procedure cost and medication cost in the scan group compared to the no scan group. However, what remains to be determined is whether coronary calcium scan will ultimately improve clinical outcomes down the line. So the ROBINSCA trial being done in Europe is going to be the definitive trial and is currently enrolling some 39,000 participants from the Netherlands and will randomize them into having either the European risk score calculated. Coronary calcium done or no screening. Those with a higher risk level from the European score algorithm will get risk factor modification and if they have a high calcium scan of 100 or higher, they'll get statin and ACE inhibitor treatment according to the Dutch guidelines. And these people will be followed for five years for fatal and non-fatal coronary heart disease. So it will be important to show whether coronary calcium scanning is able to result in improved medical treatment that then would presumably result in better clinical outcomes down the line compared to traditional risk factor based screening, which is basically what the score risk algorithm involves. How about for people who had a scan with a coronary calcium score of zero? What are some of the lessons that can be learned from this power of zero? Well, we know from data published just a few years ago that even with multiple risk factors, regardless of the number of risk factors, one has, if the calcium score is zero, these individuals have a very low event rate. So perhaps it's the amount of calcification that is a stronger determinant of risk than is the number of risk factors. In addition, we can see even older people who generally have a much higher risk of developing coronary heart disease. These events in younger people that their risk remains low if their calcium score is zero. So again, perhaps it's the amount of subclinical disease from coronary calcium that's a more important indicator of risk than someone's age. Similarly, we see this as LDL cholesterol, even with the high LDL cholesterol, one sees the zero calcium score associated with very low risk of cardiovascular outcomes. So clearly, we see here again that it's a calcium score that matters rather than the LDL cholesterol. The concept of a long term warranty against having cardiovascular events in people with the zero calcium score has been promulgated by several investigators. These data here show that over a period of 15 years that having a zero calcium score was a stronger predictor of not dying and was associated with the lowest mortality rate than even people who had a Framingham Risk score of less than 10% or no risk factors for that matter. These these groups had actually higher mortality rates. So finally, what are the guidelines for coronary calcium screening and its role in identifying patients for treatment according to the 2018 cholesterol management guidelines? Well, in 2013, the workgroup noted that assessing coronary artery calcium is likely to be the most useful of the current approaches to improving risk assessment among individuals found to be at intermediate risk after formal risk assessment. It's been noted that imaging has at least three virtues. One of these is that once subclinical atherosclerosis is detected, the intensity of drug therapy might be able to be adjusted, depending on the amount of plaque burden present. Several papers, such as this one by Michael Blaha and colleagues, have noted that the absence of CAC predicts excellent survival with 10 year event rates of only about 1%. And it was suggested here that a finding of a zero CAC might be used as a rationale to emphasize lifestyle therapies rather than pharmacotherapy, and to forgo repeated imaging studies. Remember that the ASCVD risk estimator its updated version of risk estimator plus is recommended as a first step in preventive cardiology to assess a person's risk of atherosclerotic cardiovascular disease over the next 10 years. Because it's from this risk that we consider the intensity of treatment. This using the risk score is intended to promote a patient provider risk discussion about the best ways to reduce risk and is not to be used as an automatic determination as to who should be on a statin. Our cholesterol guidelines recently updated in 2018 point out that in people age 40 to 75, we should be doing this risk score and persons can be divided into four risk categories. Those who are at low risk. We focus on lifestyle measures to reduce risk factors. Those people at high risk greater than or equal to 20%, we initiate a statin to reduce LDL by at least 50%. The people who are in the borderline to intermediate risk level is where we use a list of potential risk enhancers shown on the left here, such as a family history of premature Ascvd, persistently elevated LDL, chronic kidney disease, metabolic syndrome, certain conditions specific to women, inflammatory conditions, ethnicity, or certain lipid biomarkers or other measures that can help inform the treatment decision. It is then recommended that if the risk decision is uncertain, as noted on the bottom of this slide, to consider measuring coronary calcium in selected individuals and a zero calcium score would be rationale to down stratify the person's risk and perhaps withholding or delaying statin therapy unless diabetes, family history of premature CHD, or cigarette smoking are present, whereas a calcium score of 1 to 99 would favor statin therapy and a calcium score of 100 or higher or greater than or equal to the 75th percentile would be a more definite indication for initiating statin therapy. So this kind of demonstrates here we use the clinician patient discussion and consider risk enhancing factors. But if uncertainty still remains consider a calcium score to further revise the decision based on the results of the calcium scan. And again, people with the zero calcium score are shown to have been shown to be below the threshold for statin benefit. And we might consider avoiding or postponing drug therapy unless the person has diabetes or is a heavy smoker or strong family history of ASCVD. And clearly if one is has a calcium score of 100 or higher or the 75th percentile or higher, they are above the threshold for statin benefit and therefore a statin would definitely be recommended. It's in the borderline area 1 to 99 or below the 75th percentile that we could consider statin therapy now or maybe postpone it. And and this is where one could consider repeating the calcium scan in five years to see if the result has changed because some of these people will progress more rapidly than others. And this may further inform the treatment decision. There is also a MESA risk score that has coronary calcium as one of factors, so this may be more useful than risk scores that relies solely on standard risk factors to estimate a person's risk. These are the data from the paper from Nasir and colleagues in the MESA study that shows that if you have a zero calcium score and your 10 year ASCVD risk is under 20% as shown in the blue bars, that these people are below that 7.5 threshold for statin benefit. However, if you're greater than 20%, even if you have a calcium score of zero, you would still benefit from statin therapy. And this further emphasizes the point here. We do know that if you have an intermediate level of calcification that this will in cases of risk from 10 to 20% stratify a persons risk so that they they may benefit from statin therapy. And of course clinical judgment is recommended. And finally, if the calcium score is greater than 100, regardless of the 10 year risk, one can be shown to be in the nets that statin benefit range. This recent study also nicely shows that in people who have calcium scores of of more than 100 that statin users had a better outcome over a period of about 12 years than nonstatin users. So this demonstrates where statin therapy may be most useful, indicating that it may be useful, most useful for reducing risk in persons who already have a moderate to significant amount of atherosclerosis present based on a calcium score of greater than 100. So finally, we can kind of summarize by indicating that the future of cardiovascular risk assessment should be predictive and optimized in identifying people at increased risk of CVD. It should be preventive increasing the focus on delaying or preventing the onset of cardiovascular disease and preemptive, meaning that we want to be able to apply preventive strategies long before clinically apparent disease. And finally should be personalized. We need to use accurate markers based on our understanding of improved cardiovascular pathophysiology to tailor preventive strategies to individual needs. The recommended approach to cardiovascular risk assessment can be summarized in this slide. As previously indicated we begin with the 10 year atherosclerotic cardiovascular disease risk score. We consider factors such as family history and certain other risk enhancing factors such as metabolic syndrome and various biomarkers. And then to further refine the treatment decision if needed in selected patients, we can then consider the use of coronary calcium screening, as we have discussed and has been now incorporated into the latest 2018 cholesterol management guidelines. So with that, I would like to thank you for your attention.