[MUSIC PLAYING]

THURA T. HARFI: Thank you all for coming. My talk today is going to be about coronary artery calcification. And hopefully you will find the talk instructive and educational.

So I have no disclosure, and I'm going to just mention a few of the outlines of the talk. So we'll discuss about the pathogenesis of coronary artery disease-- excuse me, coronary artery calcification. Sure. And we'll mention how we measure coronary artery calcification. We'll talk about Agatston score, how exactly it's measured.

We'll talk a lot about the prognostic role of coronary artery calcification in prediction of cardiovascular outcomes. And we'll focus on the role of coronary artery calcification in the 2013 guidelines and in the most recent one, 2018, and what has changed. And we will talk about some ways where we can communicate coronary artery calcification to the patient in a way that our patient can understand us. And we'll talk a little bit at the end about the role of coronary artery calcification that's detected outside the realm of cardiac CT.

So let's start with two clinical cases that I hope that will explain the points of coronary artery calcification. This is a 49-year-old woman who has really no past medical history, quite healthy. The only reason she set up this appointment with the primary cardiologist is really to talk about her risk. Her brother just had a PCI to his right coronary artery with 99% stenosis at the age of 49 a month prior to her appointment, and her father had a PCI at the age of 60.

She has two other sisters who are completely healthy but importantly had high cholesterol. She's mildly obese, and her blood pressure is normal. Her lipid panel is reasonable, slightly elevated LDL. She's quite asymptomatic, exercise sporadically one to two times a week on her home treadmill.

So the question is, what do you tell her? Do you just tell her, don't worry about it? Just go and live your life? Do you counsel her about healthy lifestyle and diet? Do you start her on statin therapy off the bat given her scary family history?

Or do you order exercise EKG stress test or any kind of stress test? Or do you order coronary artery calcium scan? Or do you order coronary CTA because Dr. Harfi's rounding? Or you tell her that she should not have booked the appointment because you're too busy and the division is trying to shorten the period of time?

[LAUGHTER]

I'll go with the last. So the other case is similar profile, but a man a little bit older, 60 years of Nepali origin. No family history, though. Same BMI. Blood pressure normal. Cholesterol panel is quite unremarkable too. LDL is slightly elevated. HDL is quite good for him.

He is more an avid exercise. He exercises more regularly, three or four times a week, and doesn't have any symptoms. But he tells you off the bat he doesn't like taking medications. So the same option-- provide assurance, healthy lifestyle, calcium score, exercise test, order exercise EKG, echo, nuclear, stress MRI, stress CT, stress PET. You choose it, and those go on and on. Or just refer him to the Cardiology Prevention Clinic because you don't know what to do?

So we'll come back to those cases later. So very quickly about the pathogenesis of coronary calcification in early form. So it's similar to the process of any atherosclerosis-- start by the lipoprotein invading the subintimal layer, depositing there, and then that will stimulate the monocyte from the circulation to invade the subintimal layer.

Once they are there, they start ingesting the lipoprotein. They form into macrophage and foam cells that creates an inflammation. Inflammation release cytokines. The cytokines will stimulate the metamorphism of monocyte and the macrophage and the pericyte into what looks like osteoblast-like cells.

And those osteoblasts will start depositing calcium crystal and forming microcalcification. Those microcalcifications get more, and they coalesce, and they become macrocalcification.

As seen in this depiction here, this is a process where it's going to continue with inflammation where more and more plaque is being built up, more and more calcium built up. And that's associated with more fat, more inflammation, thin cap. And this is what we call thin-cap fibrous plaque. Here, the inflammation is controlled calcium buildup, but then the thick fibers there is formed in more unstable plaque.

The exact pathogenesis on the molecule level still has not been figured out. There's different hypotheses. None of them are really predominant yet. But this is the major theme for it.

So what's Agatston score? Agatston score is just one of the ways and almost predominantly one of the ways to quantify the presence of coronary artery calcification. So this came after the Arthur Agatston. He's still with us. In 1990, he published this paper describing his method.

And what he did is he did a cardiac CT, gated without contrast, and did 20 slices for the heart, each slice of 3 millimeter thickness. And with each slice, he looked around here. And this is the aorta, and this is the left main. This is the LAD.

And wherever he find white spot-- white spot is the calcium-- he basically measured the area and measured how bright that white spot. And the brightness, in CT world, is measured by Hounsfield unit, 0 being the water. The neutral area is negative, and anything above is positive score.

So he said anything above 130, he will give it a weight of factor 1. Anything 200 to 300, factor 2. Anything 300 and 400, 3 and 4. Anything more than 400, you use a factor of 4. So he'd get the area. And depending on the size, you're gonna multiply that area by a factor depending on how dense that spot is. And he did that for all the 20 slides for a bunch of patients and showed that it correlates very good with the presence of coronary artery calcification and the burden.

And since then, this has been highly reproducible, highly inter-observer and intra-observer. So it really took fire and has been the standard method of quantifying coronary artery calcification. Now, I have to mention this means gated and no contrast. So gating means you just connect the CT scan to an EKG so that you take the picture within only certain periods within the cardiac cycle to reduce the motion artifact.

So how to perform calcium score? So it's quite easy, very quick, very really non-labor-intensive. Patient lies in the CT scanner. Need EKG monitoring. So they have to put EKG stickers on the patient.

The patient needs to be able to cooperate and breath-hold for probably three to six seconds. I don't think it needs six seconds. That's probably for the older scanner. The total time is probably 10 to 15 minutes, and the radiation dose is 0.5 to 2 millisieverts, which is equivalent to about one mammogram or two mammograms depending on the CT scan machine.

It costs about $100 to $200 US if the patient pays out of pocket. I think in our institution, people who don't get covered by their insurance, they pay around $100. You don't need IV line. There is no need for beta blockers. We don't slow the patient, no IV contrast, no oral contrast. There is no need for even nitroglycerin, and there is no need even for a patient to be fasting.

And this is what you get when you do coronary artery calcification. You can see on the left of the slide. This is the coronary artery, left main, LAD seen here. There's no evidence of any white spots, no evidence of coronary calcification. Here, there's a clear spot in the proximal LAD. And here, there's severe calcification.

Now, if I tell you, can you tell me how much here, how much there? Hard to say, right? A little bit here, a lot here. But this is very subjective. And people who do cardiac CT, it's very intuitive, easy. But those who are not dealing with cardiac CT on a daily basis, quantification is important, at least semi-qualitative.

This is a scan where it's not really calcium score. It's just non-contrast CT of the chest. And you can see it's very easy to spot the coronary calcium on any scan.

So what? So what does that mean? Maybe it's just part of aging. Maybe it's part of having wrinkled skin. Well, it's not really exactly like that. Coronary artery calcification strongly predicts outcomes independent of risk factor.

So in this study, what they did, they actually looked at the level of coronary artery calcification and correlated it with the risk factor and depending on the outcomes. So they saw, if you have zero risk factor but your coronary artery calcification above 100 or above 400, your risk is almost more than somebody who has calcium score of 0 with three risk factors.

And you can see that the slope goes sharp from calcium 0, 0 to 100, 100 to 400, and more than 400. Almost all the people who have calcium score more than 400 have very elevated risk regardless of their risk factors.

So that study was a little bit older study. And since then, there has been many studies all correlating and confirming the finding that coronary artery calcification improved risk of prediction on top of traditional risk factors. And all those studies have finally been able to find their way into the 2013 guidelines, but not really strongly.

And when 2013 guidelines for cholesterol management came, there was a lot of controversy, a lot of debate. It changed the whole way we treat hypercholesterolemia and proposed a lot of changes. And it's actually stood the test of time, and now it's become really part of our daily life.

So just a little bit to refresh everybody's memories, the 2013 guideline dedicated statin group. So everybody who has prior coronary artery disease or atherosclerosis disease needs statin. Anybody who has familiar hyperlipidemia with an LDL of more than 119 needs statin. Diabetic people with a little bit elevated LDL needs statins.

Everybody else, they fall into the umbrella of primary prevention. And from their primary prevention, that's where most of the change is. They use a new equation called the pooled cohort equation. Before that, according to the Adult Treatment Panel III, we were using Framingham score. Now it's the pooled cohort equation. It's just different equation, different statistics, supposedly more powerful, and they account for the race.

So once you do the risk score, according to those people, you divide your population into mainly three categories. People who have 10-year risk of less than 5, they label as low risk. 5 to 7.5, borderline. And more than 7.5 is the high risk.

So anybody more than 7.5 needs statin. Anybody less than 5, they don't need a statin. They need lifestyle education. And borderline, kind of in between-- you can use statin, but it's reasonable.

And that led really to a lot of debate in the news, in the press. And they said, oh, this will almost exclusively put most of the population with a statin, and you better put statin in the water. But the truth is they didn't say start statin. They say start the discussion.

Talk to the patients. See what they prefer. Are they really into primary prevention, or are they into not taking any pills? Start this discussion and start the conversation, and see if the patient would be a good candidate for a statin.

So that can get looked over, but the bottom line is even with their new calculator, they end up really overestimating risk. So the equation supposedly incorrectly did some changes. The equation accounts not only for coronary event, but actually for cardiovascular event, including stroke and stroke death. So that was one advantage.

As I said, it included the race. The prior cohort, Framingham, was predominantly Caucasian. The other thing the guideline said, it introduced the concept of statin intensity-- low, moderate, or high. That was not there. And before that, we were targeting LDL levels.

And according to 2013, they dropped completely the LDL levels, and you just put the patient on high intensity and forget about it. So that was kind of unsatisfying to a lot of the patients and the physician are the same. And high means any statin that has the ability to reduce LDL of less than 50%. Moderate-intensity statin are the statin that are able to reduce between 30% and 50% from the baseline level. And low, less than 30%.

And the other major change, they actually changed the cutoff for high risk. So it used to be anything more than 20% is very high risk, and 10 to 20 is borderline. Now anything between 5% and 20% is intermediate risk, and statin can be reasonable.

And they introduced the concept of risk discussion, which is a new concept. It was not there, and basically talking to the patient about the risk and benefit and side effect and patient preferences.

Now, where is calcium score here? There is no mention. Even the data was there, but there was no mention. Well, it has a little bit of spot in the whole guideline. And it said that if the patient is eligible and you talk to the patient but decision is not clear, then you can use one of those four things to further reclassify or help inform the risk discussion.

And those things are either family history of premature coronary artery disease; if the patient has abnormal calcium score of more than 300 or above the 75th percentile; or abnormal ABI, ankle-brachial index; or if they have high C-reactive protein. And they give all of those class IIb, including the coronary calcification.

And you can see the cutoff for calcium score to trigger therapy is pretty high. 300 is a very high risk, and above 75th percentile is also very high. And so people who were really enthusiastic about calcium score, they thought that didn't get enough deserve it needed, and the cutoff were really high, and the strength of endorsement was quite weak.

However, since then, data kept accumulating. There are multiple, multiple cohort studies, big population similar to the Framingham studies, all studying the calcium score, most notably the MESA study, Multi-Ethnic Study of Atherosclerosis. The Heinz-Nixdorf Recall study-- this is American. This is German. And this one is a Rotterdam study, and the Framingham Record observational study.

All of those studies have thousands and thousands of patients-- healthy, asymptomatic. All of them have initial coronary artery calcium score at the beginning. So basically, they followed those people, but they have the addition of not only about their profile; they actually have a baseline coronary artery calcium score at the beginning. And then they followed them for 10 years or even more to see what's the risk factor and the prediction and the role of coronary artery calcium.

The MESA study has been very powerful and very-- because it's very well done. It has different races, and it's designed like that. And it's multiple locations in the United States. And now it has completed more than 12 years of follow-up.

So more data coming in. Now calcium score not only actually predicts independent of risk factor, actually works even within different age groups. You can see here in this study from the MESA that regardless of your age, really the predominant factor of risk is really how much calcium you have. If you are 45 and you have calcium score more than 100, you are way higher risk than somebody who is 75 or even 84 with a calcium score of 0.

What that means is really, once coronary calcium is known, chronological age becomes irrelevant. It's really your risk of having is depending on how much calcium you have, not how old you are. And that's the concept in chronological age and arterial age started.

Now, one might think, OK, well, the older you are, the more calcium you get. That's true, as you can see from this graph. The blue line is basically how prevalent is a 0 calcium score in the population. You can see the older the population gets, the less likely they have zero calcium.

But if you look at the age 60, half of the population of MESA have 0 calcium score. If you look at age 70, about a third. So it's not rare for elderly people to have very low risk, and vice versa, of course. Calcium score of more than 100 goes up with age, and so on. But really, a significant portion of the population-- half of them have 0 calcium score at 60, and a third by the age of 70.

Then they show, OK, calcium score even works better even regardless of your LDL. Regardless of what your LDL level is, the higher calcium score, the more risk. Almost LDL level becomes irrelevant if you account for the calcium score. And that also applies for the non-HDL. And see patients who have subject-- it should say "subject." Subjects who have calcium score of more than 100, they all have super elevated risk, around 26%, regardless of their LDL level.

So OK, but calcium score is a scan. That's fine, and you get radiation. It's additional thing. It's expensive. It costs money. Why don't we just go with the risk scores? It's cheap. It's available. It's just a few questions, yes or no.

That's fine, and nobody is abandoning risk score. But the problem, when you only use risk score, is that you end up with a lot of misclassification. So in this study, what they did, very intelligent, is they get the MESA cohort and they looked into the guideline 2013. And they divided the population into three groups-- those who are, according to 2013 guidelines, statin would be recommended; those who would be, according to the guideline, statin would say you don't give a statin-- they don't need it-- and those who would be borderline. Borderline means 5% to 7% 10-year risk.

And they found that if you go to those people where the guidelines say give a statin, 41% have calcium score of 0 and their risk is 5% in 10 years, and that's considered very low. And you need to treat 64 to prevent one outcome.

And if you look at those borderline cases, which means, yeah, you can use statin. It's reasonable, 5% to 7%-- around 57% of them have 0 calcium score and their risk is 1.5 in 10 years to have stroke, MI, or dying from stroke or MI. And if you need to treat almost 220 patients or subjects to prevent one outcome. Very cost-ineffective if you go only with the risk scores.

And without surprise, people who already don't need a statin, majority of them don't need calcium score, don't have calcium, and their number needed to treat is very high and the risk is very low. But on the flip side, those where statin said don't give them statin, 21% of them, one in five, have a really elevated calcium score, and 5% have very high elevated, more than 100. And their risk is about 10%.

And you go back to the other side of the coin, is, how about those people who have statin recommended? You get about 30% have really elevated calcium score, and this is really where most of the benefit is, people who have very elevated risk and they have calcium score. And you need to treat 22 to prevent one outcome, and that's a reasonable number. And their risk, of course, is elevated.

So what the bottom line is, even within the categories according to the guideline, statin recommendations, calcium score really performs very highly and can discriminate between low risk and high risk. And even further, if you look even within categories within the high-risk group, people who are above 7 and 220-- so this line reflects the 7.5 risk cutoff. And you see that this is the people 7.5 to 10, 10 to 15, 15 to 20, and above a 20.

Anybody who has calcium score-- the gray bar here is more than 100-- easily crosses the bar regardless of their risk. Even if they are barely borderline, they're touching almost 7.5. Anybody who is 0, they never touched the bar, only if they are really high risk already.

So with that, really, the paper ended up with putting really-- what ended up really translating into 2018 guidelines is that you're really going to change management or decision only in the intermediate group. So this is the group who are at low risk. You see calcium score is 0. The risk is 1.3. Calcium score is not 0. The risk is low.

So that means that using calcium score in the very low-risk patient, yes, you will discriminate, but not enough to change decision. So people who have calcium score of 0, they're already low. People who don't have calcium score of 0, they have high risk, but still not enough to trigger a statin. So you're not going to generally change management here.

The same thing applies to people who are high risk. If they are already more than 20%, you use calcium. Those who have no calcium, yes, they have lower risk, but not as low that they don't need a statin. So you really don't change management here.

But the bulk of our population in the primary prevention really falls in the middle, falls in those intermediate risk scores between 5 and 7 and 20. And in those populations, really, you change 50%. 50% of those patient will have 0 calcium score, and they would not need a statin because their risk would be really low.

So this analysis was reproduced exactly using the European cohort, the Heinz-Nixdorf study. So in the German cohort, they actually applied the same concept using the 2013 guidelines and the European guidelines. And they showed the same thing. Calcium score still reclassifies within each group whether using the American guideline or using the European guidelines.

And again here, even within German population, about 7.2% of the population who don't need a statin, they would be actually elevated risk. And those who will need a statin according to the guideline, about 20% will have calcium score of 0. So over-prescribing, over-indication, that was the dominant thing about if you only use the 2013 guidelines. And I think we all know that if your age is 65, you're almost guaranteed your risk would cross 7.5 if you only use the calculator.

And then in the statistical world when they added calcium score, calcium score significantly improved the C-statistic, the discriminator power, discrimination power of the model. However, all the other factors that the guideline endorsed-- family history, ankle-brachial index, or CRP did not. They did not improve. So they thought that calcium score has more power to it, and it should need really a more prominent recommendation.

So by that time, again with all this data accumulating, a lot of the controversy about the guidelines, societies had to really get under the pressure to say what's their position about the calcium score. And there was a lot of controversy. Nobody knows is it appropriate. Insurance's still not paying for it. And people are using variable practice patterns.

So the ACC guidelines say it's class IIb. The European guideline again says it's also class IIb to use it. The 2017 expert consensus from Society of Cardiac CT, they said that it's actually quite appropriate and they designed it that you should use it in the 5% to 20% intermediate risk group and not in the low or high risk or even in people who have low risk who have family history. But they were kind of the outlier. The US Preventive Task Force said that there is insufficient data to support its use.

So on that, then, basically the Society of Cardiac CT issued their recommendation and in more detailed way, outlining really what to do, how to interpret it. So they said any calcium score of 0 is low risk. Don't give statin.

1 to 900, mildly increased risk. Consider moderate-intensity statin, and use high if it's above 75th percentile. 100 to 300 is moderately elevated risk. Really consider statin with aspirin. And if it's more than 300, go full force, aspirin and statin.

And they have pulled all the data and the literature at that time, and they ended up early seven registries about 84,000 patients. All those are asymptomatic. And they correlated that the median annual event rate for the different levels of calcium score.

So calcium score of 0 becomes a category. 1 to 100 is another category, and 100 to 300 is a category, and more than 300. So those are the four major categories from calcium score. And you can see that the risk really goes into linear with the more calcium buildup you have.

So then there is this paper that came in 2018 just last year, and said that really state-of-art review about the calcium score. And they actually endorsed this approach even though it was not in the guideline. And they said just-- those are very well known names in the primary prevention and the role of calcium score.

And they outlined what really ended up into the next-generation iteration of the guidelines. They say if your risk less than 5, don't give a statin and calcium score is not effective. If your risk more than 20, give a statin, don't consider calcium score because it's not going to change your recommendation.

If your risk is intermediate-- 5 to 7.5, they called borderline. 7.5 to 20 is intermediate-- then calcium score has a role and can significantly up-classify or down-classify patient, about 50%. So half of the population, you will really make a difference if you use calcium score in their recommendations.

And what really tipped the edge was really when the full 10-year follow-up from the MESA was ready and mature and they have all the data to actual more than 10 years' follow-up. So about 7,000 patients in the MESA. They were already recruited. And half of the population have calcium score at the beginning. Half of them don't. And they have a median follow-up of 11 years.

And they found that calcium score was really associated strongly in integrated fashion with the 10-year risk, and that was independent of risk factor, age, gender, or ethnicity. And so here, this graphic can show you the effect of race on the risk score. The four lines indicate different categories of calcium risk-- 0, 0 to 100, 100 to 300, and more than 300. And the same applies to all four pictures. You look white, Chinese, black, and Hispanic.

You can see the people who have higher calcium score, more than 300, are already at elevated risk regardless of their race. And the people who have more than 100, they already cross the 7.5 threshold for statin therapy in any way regardless of their race. And even the presence of any calcium, even if you have a score of 10, that also will make you cross the 7.5 risk except if you were Chinese. Chinese have usually lower risk in general compared to other ethnicities.

So if you have 100 score calcium, 100, you definitely cross that 7.5 and you are definitely at high risk even if you have a little bit of calcium, not even necessary above 100. Below 100, your risk still elevated, and statin might be worth it.

Now, the other side, which is equally important, is look at the 0 calcium score. 0 calcium has such a low risk in 10 years in all the races regardless of your race for more than 12 years' follow-up. And the risk is always 7.5 and even lower.

The same pattern was shown even regardless of the age group. And I think what's most striking, if you at the elderly people, even within people who are more than 75 or 85, their risk, if they have calcium score of 0, is below 7.5. So the power of calcium score to reclassify risk holds valid whether your gender or your race or even your age. And calcium score of 0 is quite a negative predictive marker.

All right, so finally, we get to the new guidelines which was issued last year. And they follow the same theme of the 2013 guidelines. They say if you have familiar hyperlipidemia, LDL more than 90, statin is indicated. If you are diabetic, statin is indicated. And then they use the same risk calculator. They didn't change the risk calculator. They didn't change the outcome-- still stroke, MI, or death from stroke or MI.

And they adopted those four levels of group-- low risk, borderline, intermediate, and high risk. And they follow the same thing. If you are above 7.5, statin is indicated. High risk, statin indicated. Borderline risk, they say start the risk discussion. And the risk discussion is basically talk to the patient, what is their preference, risk factor, and drug interaction.

Now, the major change is that calcium has really had a much bigger role in the guideline. They say whenever you talk about risk discussion in the intermediate group and it's unclear what to do-- the patient is not sure-- calcium score is a class IIa indication. And if it's 0, you don't need to start statin. If it's above 100, definitely consider statin therapy. If it's 0 to 100, then consider statin, especially if the patient above the age of 55.

And the other thing that they introduced is the risk enhancer. So there are a lot of risk factors that we know that they are really strong predictive event, but they are not enough to be included in the risk calculator. Every time we include them, we don't really change the model to become stronger. So they are not worth it to be included in the general population. And those they call the risk enhancer.

And I think it's worth mentioning them one by one. So family history of family coronary disease, if your persistently elevated LDL of more than 160, chronic kidney disease, metabolic syndrome, inflammatory condition like rheumatoid arthritis, lupus, or even condition associated with pregnancy like preeclampsia or premature menopause, or even anybody who's South Asian. I think we all probably have encountered some of our Indian-origin patients who really have aggressive coronary disease, and they keep forming plaque and plaque and needing a lot of intervention.

And additionally, they also identify some laboratory markers that are really high indication of increased risk. Those are, for example, increased level of CRP, lipoprotein little a, ApoB, or ankle-brachial index. So those are risk enhancer. What that means is that you use those risk markers if somebody was low risk or borderline but they have additional risk markers, that needs to be considered. That might push you toward prescribing statin than holding it.

And the guideline went further to identify really special group of patients whose calcium scores really can be effective and helpful. Those are the hesitant people, people who are reluctant who are not sure about their risk. They don't want to take a prescription. And they look at the number. They say, that's mine, but I'm very healthy. I'm good.

That will be very important to show them whether they-- if they don't have calcium, that's great news for everybody. If they do have coronary calcification, maybe that's really a good reason for them to be on the statin, and maybe it's worth it.

The other group is people who tried statin but now they had some side effect or perceived side effect and you're convinced that are high risk. You want to try them more, but they are hesitant, and you want to be sure that it's really worth the trouble.

So then you order the calcium score to further reclassify, and maybe they really don't need it. Maybe calcium score will be 0. And then if it's high, then they will be more motivated to try and work with you about getting them on a good regimen.

And other thing thing is older people, older people who are healthy. They are completely healthy, but, because of their age, they were assigned higher risk category just because how the calculator works. So those can be really benefited from calcium score, and middle-aged adults who are on the borderline risk score who have some risk enhancer who are on the borderline and you want to make sure that they are fully protected as much as possible.

So we're going to switch a little bit gears here, and we're going to talk about a special population that we usually don't regularly encounter or address in the guidelines in this way. So the young adult. So the young adult between 30 and 50, the guidelines-- all those cohorts that we mentioned we really recruit people above 45 or 50.

And the reason, because they don't want to spend a lot of time monitoring young people for 30 or 40 years to wait for an event. So it's cost-effective approach when you do cohort studies. But in this calcium score consortium, they were able to pull 22,000 patients, and they actually have follow-up for 12 years.

And they found that actually the presence of calcium in the young is still a risk marker, and it increases the risk. And it's not rare. If you look here in the people who are between 30 and 40, at least one in five have calcium score above 0, and a tiny bit of them have even above 100. In people who are 40 to 50, about a quarter of them have some coronary calcification, and even about 5% have calcium score more than 100.

In total, people who are younger than 50, a third of them have coronary artery calcification, and 7% of them have calcium score more than 100. And the presence of coronary calcification increases the coronary heart death mortality by five times and cardiovascular mortality by three times. So it's not benign phenomena if you have somebody who is young who, for some reason, is known to have coronary calcification. Then that needs to be addressed or considered and treated if possible.

So the other special group I always think about is diabetic. So we always label diabetic as high risk of the guard, and we give them a statin. But the truth, they are not homogeneous. Not all diabetics are created equal.

If you look here, within the MESA study, they looked at the people who have diabetes, have metabolic syndrome, or neither. Of course, the risk increases if you have diabetes, but not homogeneous, not in the same pattern. If you look at people who have coronary of zero, their risk is very low compared to people who have more than 100 or even 400.

And this graph shows that the survival rate is greatly-- it's very well for people who have calcium score of 0 regardless of the duration of diabetes, more than 10 years or less than 10 years. However, the problem is even though calcium scores will lower the risk, if you are diabetic and have no calcium score, that will lower your risk, but not enough where statin is not needed.

And if you look at this bar where this is the group who are diabetic, the risk is almost 7.5. It's not as equal as 15 for people who are 0 to 100 or even 30% or be more than 400, but still high enough where a statin might be beneficial. But it's to help inform the decision.

And here, they looked further into what kind of subgroup of diabetes that who might be in such a low risk that you really don't need statin therapy. They found that people who calcium score of 0 and A1c less than 7, they're really into the low category. Or people who are non-insulin user, their risk is about 6.9, 7%. And people who are less than 10 or more than 10 years of diabetes, they're almost hovering around the borderline, 7.5.

So the bottom line is I would not stop prescribing statin for diabetic, but if I have a calcium score, I will think of them differently. Maybe less intense statin if somebody is calcium score of 0.

Now, how about family history? So even the family history is endorsed in the guideline as a risk enhancer. And the guidelines say that even if you have calcium score of 0 and you have family history or a smoker, you don't have to hold a statin. But really, the data doesn't support that. In this study, the positive family history or negative with a calcium score of 0 is still very low risk and really doesn't warrant statin therapy.

The last subgroup I want to talk about is people who have very high elevated familiar hyperlipidemia. Do they all need statin? Everybody is super high risk? The truth is no. Probably all my colleagues in prevention have experience that firsthand, is not everybody with hyperlipidemia would develop a coronary artery disease.

And the calcium score of 0 is very prevalent. About more than 37% of those patients have calcium score of 0, and their outcome is very, very low. So there is more into this group. There's other factors. As we know, atherosclerosis is a very complex disease. It's not all about high lipid.

And this finding was reproduced in another subgroup study where they studied only 200 patients who have genetically proven familiar hyperlipidemia. And they followed them for about 10 years. And they found that people who have calcium scores, nobody has experienced a mass. When people have calcium, any calcification experience very high risk of cardiovascular event. So even within the hyperlipidemia, calcium score can play a role and can differentiate between high risk and low risk.

So the bottom line is they looked into calcium score in regard to other negative risk markers, including low CRP, homocysteine, normal BNP, all the good things that we think about portrayed low risk. And they found calcium score is actually the one that has the highest NRI, net reclassification index, which is a statistical measure of telling you how much you actually reclassify people from low risk to high risk, from high risk to low risk. And they found 13%, 13.7. In the world of NRI, this is a really high number. Compared to all the other negative risk markers, none of them even hit the 5% mark.

So then there has been more attempts to try to see if we can get more of the calcium score. Can we use calcium score to guide us in terms of aspirin therapy? We all have issues with aspirin. Now we stopped using it for primary prevention. And now our patients ask us, and we don't really have a straightforward answer to them.

So in this paper, they showed that, really, calcium score can help you guide that strategy. So they found the number needed to treat and number needed to harm. So the red line is really number needed to harm from aspirin because of GI bleed and major bleeding. And the number needed to treat to prevent an event is really in the blue and yellow the red bar. And this is according to your risk score.

So the bottom line is if your calcium score of more than 100, the number needed to treat is way below the number needed to harm, which means that it's really reasonable to put patients who have elevated risk more than calcium score on aspirin, but not people who have 0 calcium score of 0, 0 or even a little bit, 1 to 100. Because those are the people where the number needed to treat is way higher than the number needed to harm.

For example here, people who have calcium score of 0, you need to treat 2,000 patients to protect an outcome while you need only to treat about 400 to cause a harm. So that's not yet endorsed in society on the preventive community, but it's still something I think about when I see my patients in my clinic.

And similarly, in this very interesting analysis by Bill McEvoy who was a co-fellow with me in Hopkins, he tried to reproduce the same thing with blood pressure targets with the new guidelines with much lower target. He did this interesting analysis where he looked at the number needed to treat to actually reduce an event, whether heart failure or cardiovascular event, in association with your risk score.

So for example, if you look at this group, patients who are-- their blood pressure between 140 and 160 and their atherosclerotic risk disease less than 15%, the number needed to treat to push their blood pressure to 120 to get benefit really varies from 36 if they have 0 calcium score, 15 if they have calcium score of 0 to 100, or 5 if they have calcium score of more than 100. And the same thing applied to all other combinations of groups.

So the number needed to treat becomes so high if you really have not much of risk, 0 calcium score, and becomes much more cost-effective if you have calcium score of more than 100 or even any calcium score. I think it really shows how addressing and picking up those people who are high risk and intensifying the therapy to them, whether by aspirin, statin, or even blood pressure, can really be a very effective approach.

So when we get the calcium score, you grade, and then you want to talk to the patient. It's hard to translate what does that mean. You have calcium, and that probably doesn't translate at all to the patient. Luckily, we have three ways we can communicate, and I want to share them with you today. So we can use the MESA score, or we can refer to reference values, or we can talk about the arterial age. So let's talk about each one very quickly.

So the MESA score, this is another risk calculator similar to the pooled cohort equation, similar to the one that we have in our iPad, except that it's derived from the MESA and validated from the German population, and by another group as well. And basically, it's available online, available as an app in iPhone.

And what you do, you basically use the same number-- the gender, the age-- but it has two field additional that are not present in the pooled cohort, the coronary calcification and the family history. And at the end, you get a score here, spit out. So in this example that I proposed-- so this is a hypothetical example-- is a 55-year-old man with healthy but some lipid panel abnormalities. His risk score will be 6.5 according to the MESA calculator.

Now, if you have a 0 calcium score, this 6.5 would be 2.8. And then you'll say, the risk is too low. It's not worth it. The risk-benefit doesn't really justify using statin. But if this individual has calcium score of 50, the risk would be 8, 10, 15, according to the calcium score going higher. So that's how I use it with my patients. I actually have the computer open and plug the number in front of them and tell them how the number changes.

So this is one way you can tell them that even though the general calculator scores says you are borderline, but actually, if we account for your calcium score, you're really low risk. So maybe we should hold on therapy.

The other method to communicate the calcium score which I find very interesting is arterial age. Arterial age is basically you're telling the patient, like, even though your chronological age is 50, but your arterial age is like 80. So in our example, the 55-year-old gentleman, if his calcium score is 50, his arterial age is really 68. And that means that we tell the patient, hey, even though you're 55, but your arteries at the age of about 68 or 65.

And that can be scary because if your calcium score is-- if this guy's calcium score is 150, you'll tell him, like, your arteries as old as 75. So you will need to be careful about whom to talk to and not to upset somebody important.

And this was really adopted and inspired by the lung cancer research. They found that smokers who were told that was their lung age, they are more likely to quit. If you see a 50-year-old woman who say, like, you are 50 but your lungs are the age of 80, I think that catches attention of patients, and similarly with the calcium score.

So the third method which probably most of you have seen, because that's the method we use in our lab report frequently, is the percentile. So we not only report calcium score of 70. Well, what does that mean? We provide a reference range. Let's say calcium score of 50, but that's really 75th percentile for your age and race and gender.

And this gentleman, our example, if the calcium score is 150, then it will be 85 percentile. If he had calcium of 600, it will be 97th percentile. That doesn't mean that you're the 97th percentile of your classmates where you graduate, but that's actually the other way around. That's bad.

So the power of zero. So with all this research coming with that whole calcium score of 0 is such a strong negative, but the concept of the power of zero came about, and mainly for a guy whose name is Nasir Khurram. He first threw that term and really catched interest.

So the power of zero is the power of calcium score of 0. And in this study, they looked even at the elderly. So they looked about 6,000 in the bioimage participants. That's our group of people who are elderly, all of them above the age of, I think, 60 or 65. The mean age was 69.