New cardiovascular imaging: the who, what, when and why

MDCT was the topic at this recent round table, with discussion about when and how it should be used, changes it may bring, and further research needed.

Cardiology Today convened this round table in November during the American Heart Association Scientific Sessions 2005 in Dallas. Chief Medical Editor Carl J. Pepine, MD, moderated the discussion. Part one of the round table is presented here. Part two will be published in the March issue of Cardiology Today.

Moderator
Carl J. Pepine, MD Eminent Scholar, Professor and Chief, Division of Cardiovascular Medicine, University of Florida, Gainesville; Chief Medical Editor of Cardiology Today.
Daniel Berman, MD Director, Nuclear Cardiology/Cardiac Imaging at Cedars Sinai Medical Center; Professor at University of California, Los Angeles School of Medicine.	W. Greg Hundley, MD Associate Professor of Radiology and Medicine (Cardiology) at Wake Forest University Health Sciences in Winston Salem; Medical Director, Cardiovascular Magnetic Resonance.
Mario Garcia, MD Director, Cardiovascular Imaging, Department of Cardiovascular Medicine at The Cleveland Clinic.	João A. C. Lima, MD Director of Cardiovascular Imaging in Cardiology; Associate Professor of Medicine and Radiology; Johns Hopkins University.
James A. Hill, MD Professor of Medicine and Pharmacology, Co-Director of CTMR for CV Medicine at the University of Florida, Gainesville.	Norbert Wilke, MD Associate Professor of Radiology and Medicine, University of Florida; Chief of Cardiovascular MR and CT at University of Florida Health Science Center, Jacksonville.

CARL J. PEPINE, MD: Cardiovascular imaging is a rapidly growing area that appears to be moving well beyond the image specialists. The first topic we want to address today pertains to multidetector computed tomography and the rapid explosion in 64-slice MDCT. Is MDCT replacing other forms of cardiovascular imaging?

JOÃO A. C. LIMA, MD: MDCT is a technology that has grown extremely fast in the last few years because it gives spatial resolution in the main axis of the body. The slice thickness with which CT can take pictures of the heart is very thin and has allowed us to look at coronaries. It is that power that brought MDCT to the front burner of CV imaging.

The addition of 64 slices, the ability to go from four to eight to 16 and now 64, has made it possible for all these slices to be acquired in one revolution. The stacking of these slices becomes very simple and much less prone to error and that’s where the power of the technique came from: thin slices and obtaining many of them at the same time.

DANIEL BERMAN, MD: CT coronary angiography is a practical clinical tool that is already making an impact. We started the Society of Cardiovascular Computed Tomography [SCCT] eight months ago and in the last month we have added 500 new members. We have 1,600 members that have joined in less than a year.

The image quality of CT is routinely excellent, providing you take all necessary precautions. It will have an application in certain groups of patients.

Where does it fit in? There is agreement that people who have chest pain, particularly younger patients with chest pain that doesn’t sound like angina and who previously would have been given a stress test, a stress echo or stress nuclear, might now get a cardiac CT. The CT coronary angiogram results, in some cases, might lead to the cath lab. In many cases, however, it will rule out any kind of atherosclerosis of the coronary tree. When there is doubt, there are still other forms of stress testing that can be used. That’s the low-intermediate likelihood of coronary disease in symptomatic patients.

In asymptomatic patients, I don’t think CT is the right test. I think coronary calcium scanning or carotid IMT scanning – some kind of plaque imaging – would be the right test for the asymptomatic patient. It has a role, and in that role it will probably compete with the stress-testing modalities.

PEPINE: Dr. Wilke, has 64-slice totally replaced 16-slice? Is there any current role for 16- or 32-slice?

NORBERT WILKE, MD: The 64-slice MDCT offers much improved spatial resolution compared with the 16-slice scanner. Many papers have been published using the 16-slice scanners, however, and they produce very nice pictures of the coronaries showing very nice lumenograms. The greatest benefit of the 64-slice is the ability to look at noncalcified plaque. It will become the primary scanner to look at noncalcified and calcified plaque and the lumen of the coronary vessel in one study.

PEPINE: Dr. Garcia, what about this concept of visualizing the wall and the plaque? How will the current version of CT perform relative to MRI, echo, IVUS and high-resolution echo?

MARIO GARCIA, MD: CT is probably the most robust technology right now for noninvasive visualization of the coronary anatomy. Although MRI can do that, the potential to apply this to a large population with a relatively low cost favors CT. Also, there is greater spatial resolution with CT.

I think that ultrasound will remain a primary tool for evaluation of left ventricular function and sequela of coronary disease such as wall motion abnormalities and left ventricular dyssynchrony. The coronary anatomy is far away from being resolved by coronary ultrasound.

PEPINE: I would put forward the notion that MR and CT are probably the two best methods for assessing LV anatomy, atrial anatomy and function.

So if we accept that, then let’s move to the coronaries. In the coronaries, you’re going to pacify the lumen with some kind of a contrast agent and you also have interest in the wall of the artery. So for the lumen, where are we? Would you say that 64-slice is clearly superior to 16? Is it equivalent to contrast MR or selective coronary angiography or do we not know the answers?

W. GREG HUNDLEY, MD: Right now, spatial resolution is much higher with CT for the lumen of the coronaries. MRI is a different technology in that it uses nonionizing radiation. That may be useful if we want to follow some aspect of coronary disease over time. We might look at the wall of the coronaries or measure coronary artery blood flow, a functional consequence.

Another advantage of MRI is that it can appreciate the functional aspect of the disease process in addition to the anatomy. The spatial resolution is not as high as CT currently, but the functional consequence is actually quite high. MRI may have a use in patients with advanced disease or in patients with a clinical disease process who should be followed over time.

As our patients age, that group of patients becomes a significant percentage. MRI will have a role because of the nonionizing radiation allowing us to study patients repeatedly over time and also assess the functional consequence.

JAMES HILL, MD: All of the noninvasive studies that we’ve done — exercise testing, nuclear studies, dobutamine stress echo — are looking at the effects of disease and using those effects to predict coronary disease or disease risk.

CT is really the opposite. You are looking at the presence of CAD whether it is the lumen or whether it is the wall or the combination of the two. The challenge will be taking that anatomic information and trying to understand the clinical and functional significance.

BERMAN: Here is a scenario for us to consider: A 76-year-old man in good health has a coronary calcium scan showing that he’s in the 32nd percentile, with a score of 126. He has moderate LDL elevation, is placed on a statin and his LDL is lowered to about 70. He is also assigned aspirin.

A couple of years later he is sent for a CT coronary angiogram. The CT coronary angiogram shows a mid-LAD lesion that looks like a significant stenosis right in the area where there was calcification over a three-year period of time, but there is no change in calcification. The patient is then referred to the cardiac cath lab.

On the way, I say to the referring physician: ‘Why don’t you first consider a noninvasive test and see if there’s any ischemia associated with this calcification?’ Nine minutes of exercise, no chest discomfort, no ST-segment depression, no abnormality on the nuclear test. It is all completely normal. The referring physician says, ‘Well, we’ll take that into account when we go the cath lab.’

So we go into the cath lab, and the patient has what looks like a moderate stenosis on the CT angiogram. We do IVUS, and we see that there’s a large, noncalcified component of this plaque and the patient gets a drug-eluting stent.

Now, this just happened last week and the patient is fine. He feels great now that he has a stent, but he felt great before. He had no symptoms.

We define the anatomy now, but we have no concept of what really is happening. What is the activity of the disease based on what we see just on the anatomic lesion? If it’s a critical stenosis in a proximal location I can understand saying he should go right to the cath lab. With a moderate lesion, however, we have to be careful or we’ll open up a Pandora’s box and nobody will be able to pay for all the stents that we put in.

PEPINE: As an anecdote, those of us who grew up in the cath lab know this: If you want to intervene on a lesion, do IVUS because it is generally looks worse than on angiography. If you don’t want to intervene on a lesion, you do flow reserve because most of the time with intermediate lesions it will be >2.5.

LIMA: The power of MDCT is the ability to look at the vessel wall. If we had data, for example on Dr. Berman’s case, that aggressive lipid-lowering therapy would reduce that lesion 15% in a certain period of time, and that could have stopped this stent. Without that information, what you see is an ugly plaque and then it becomes difficult to stop the train of interventional momentum. I don’t know if that’s bad.

PEPINE: I had a patient a couple of weeks ago sent by his general practitioner who had his CT in his hand. It said: ‘carotid calcification indicative of atherosclerosis.’ So, the patient wants to know what that means. Dr. Berman, since you brought up the calcium notion, what would you tell him?

BERMAN: I wouldn’t argue with the report; I think it’s correct. Part of the problem is that those reports don’t give any indication about what needs to be done. It leads to questions in either the family practitioner’s mind or in the patient’s mind. We have to get the report directly and be careful about how we use that information.

I think that calcification in the arteries means atherosclerosis, and I’m on the aggressive side. When I see atherosclerosis, I believe in aggressive medical treatment but not angiographic procedures unless there’s some evidence of functional significance. Now the calcium doesn’t tell us anything about stenosis. You can have extensive positive remodeling of a vessel and have no encroachment on the lumen at all, so I don’t think we should jump from calcification to an invasive procedure.

We learned a lesson several years ago with calcium. The organized cardiology world was worried that if we do calcium scanning on everybody’s arteries, patients would end up in the cath lab unnecessarily. So everyone was aware of the problem and decided we’re not going to allow that to happen. The societies involved in making recommendations about calcium said not to go straight to the cath lab. If over a certain score, consider a functional test to see if there’s ischemia. I think we’re in the same ballpark now with CT coronary angiogram where we have to put out caution statements about going from stenosis to cath unless it’s really convincing.

WILKE: Clearly 64-slice MDCT is probably going to change the way we look at coronary disease. People are excited because it clearly affords us not only a lumenogram but also a look at the inside and outside of the vessel wall. It also allows us to look at calcification. Now why is this important?

Patients at higher risk for a plaque rupture are those who have mild coronary disease; these are not the ones with significant lesions. We have the capability to identify these early, mild disease stages that may carry high risk for patients. We do not have the data yet, but we now have the noninvasive tool to assess total plaque burden as it is related to calcified and, most importantly, to noncalcified plaque. Then maybe we have different strategies for treatment in those patients, whether it’s conservative or other interventions.

There are no data yet, but we are interested in early detection of the disease in a way we’ve never had before. This is very different from nuclear and intravascular ultrasound, which is an invasive procedure. Even so, the learning process about the disease process of nonstenotic plaques is coming from earlier studies with intravascular ultrasound. There could be a lot of potential research in the future using this CT imaging in combination with different drug treatments whether it’s antiinflammatory treatment of noncalcified plaque or nonstenotic plaque. We are talking about noninvasively studying the outcome of patients with lesions that are not signficant but may carry a high risk in particular patients.

PEPINE: Dr. Wilke, when you do your reports, do you routinely report plaque volume or some measure of plaque?

WILKE: Yes, we report whether there is a significant lesion or not, and the amount of calcification and the amount of noncalcified plaque. Even so, no recommendation can be derived yet from that noncalcified data. This hasn’t been implemented in any ACC or AHA guidelines, as we do not have any published confirmative data yet. However, there are several studies – such as the MESA [Multi-Ethnic Study of Atherosclerosis], looking into that, and there are anecdotal studies performed in Europe. We’ll see in a couple of years what the outcomes are.

PEPINE: Are you concerned about over utilization of MDCT?

GARCIA: There is potential for overutilization and over-referral for invasive procedures. It’s not on untested ground. It’s not an issue of limitation of the technologies.

I’m sure that Dr. Berman will not send every patient with a nuclear scan defect for a cardiac catheterization procedure, but there are users with less experience who will do that.

One important reality is that the consumer drives the use of this technology. The patients demand to have this done because they understand what they see on an image much more than other imaging modalities that they have been shown in the past. The potential power of behavioral modification to demonstrate to the patient that they need long-term secondary prevention could be an interesting consequence of the use of CT.

Indeed, compliance with primary prevention in the patient population is very low. I believe there is significant potential to increase compliance by pairing imaging data in support of anatomical findings.

HILL: That may be a double-edged sword. You have the opportunity for the patient to see their disease and to make modifications, but at the same time the patient sees the disease and gets the same attachment to these pictures as we do. You see an angiogram in the cath lab, and it looks so bad that you just have to do something about it even though it has no functional significance based on a functional study and long-term outcome data are lacking.

The danger here is that coronary calcification in the past did not have a wide market penetration in a sense that it was relatively isolated and was not advertised after its initial introduction. This new imaging technique, on the other hand, will have many more cardiologists involved; there will be much different referral bias, and a larger patient population to potentially screen. It really offers the opportunity to dramatically increase catheterization volume and increase interventional volume, perhaps beyond what’s appropriate.

HUNDLEY: Standards for reporting need to be developed so that information gathered at one institution or at a physican’s office can be amalgamated and followed for many purposes: outcomes research, financial impact, assurance that it’s used properly.

BERMAN: If we screen people with noninvasive coronary angiography, we may open the doors to tremendous numbers of unnecessary procedures. Screening should not directly lead to catheterization. For the patient who has some symptoms, it is clear. You may send them for nuclear scans if their symptoms don’t sound cardiac. The nuclear scans are normal 80% or maybe 90% of the time. And where are you?

You don’t know the source, you don’t know if this patient has enough atherosclerosis that they should be treated aggressively even if it’s not what is causing their symptoms. In this kind of patient, use of CTA allows you, in a substantial proportion of patients, to say that their coronary arteries don’t have stenosis and that they don’t have sufficient atherosclerosis to be causing a big problem. You may come up with the source of the chest discomfort from seeing something else that’s on the CT angiogram.

PEPINE: Some would argue that the functional capacity that you obtained from the traditional stress test before doing CT imaging of the coronary arteries provided a measure of prognosis that we understand and is well validated. While the image might eventually get there, we really don’t know the answer yet.

BERMAN: I’m talking about symptoms. I’m saying that when a patient has chest pain and they are worried about dropping dead, you can exercise them and say they have a good prognosis because of the 90% sensitivity of these tests. There is a possibility, although it is rare, of having a normal scan with nuclear even though the patient has severe disease with balance reduction and flow. You really haven’t completely answered the question, and I think you can do that effectively with a CTA.

HILL: So you’re talking about using it as a screening tool?

BERMAN: Yes, in symptomatic patients.

WILKE: All data published so far are from patients with symptoms or atypical chest pain, in particular the group with immediate risk and TIMI score of 3 to 4 or less. These are the patients who will eventually benefit from a CTA. The last published studies using 64-slice CTA again have shown that the negative predictive value for ruling out coronary disease is more than 97%. So this is an excellent test, particularly in the intermediate risk group in patients with chest pain to rule out coronary artery disease.

As we know from TACTICS [Treatment with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy], there is no clear benefit from conservative or invasive tests in the intermediate-risk group (TIMI score 3). That’s the group where CTA potentially can provide more information and where all the other tests were often not conclusive. In the low-risk group in certain clinical situations and in the symptomatic patients it would be an interesting case to offer more cardiac CTA. For the high-risk group of CAD there is no role for any of those tests. Interventional catheterization is the standard of care for that group.

PEPINE: If you had a patient with symptoms suggesting ACS (new onset rest angina), who had no positive biochemical markers in the emergency department or in your office and you did a 64-slice CT showing no severe stenosis, would you all agree that that would be enough evidence to dismiss the patient and go on with conservative management?

HUNDLEY: I don’t think we know yet. We need to do that study because it’s a question begging to be answered. I believe we are spending around $20 billion a year in the United States providing care for acute chest pain syndromes in emergency rooms. That cost includes observing the patient over a day, getting the functional tests, going on to intervention.

It would be useful to test with the potential power of CT or MR whether a fast anatomic image or a fast anatomic coupled with a fast functional image would make things move faster. Could that snapshot provide us that same piece of information as the 36-hour observation and further testing? If it would, that would be useful not only for patient management and health care, but would also save lots of health care dollars.

BERMAN: I agree. Several of us on the board of SCCT have undertaken a study with exactly that question as our first research project. We’re doing a randomized trial comparing usual care, which will include a nuclear arm, with going straight to CT. I think the results will show that you have to be nearly normal to have the patient go home. We don’t have to see a severe stenosis to say you’re not going home. With a mild to severe stenosis we will say wait, we’ve got to do something more. But if it’s less than 25% stenosis and a low calcium score, we’re going to test whether we can send the patient home.

WILKE: It’s correct to say that there are no definitive studies yet published in the use of cardiac CTA in chest pain patients coming from the ED. There are several institutions that have studies going on, including the University of Florida, which have several hundred patients in a database. We see that this is a test with a very high negative predictive value for ruling out disease in patients with intermediate risk and also when used in combination with an MR rest-stress perfusion test. If you can provide and build the infrastructure in your institution, there are clear advantages over current diagnostic tests, particularly in cost savings and time the patients have to stay in the hospital for ruling out CAD, which is still the majority (70%) of patients with atypical chest pain.

PEPINE: This very high negative predictive value that you refer to: Is that established in a high-risk population such as a patient coming to the ER with chest pain, or is that established in a more general population where the risk is low?

WILKE: Most studies published so far are in preselected patient groups and not usually in all-comers. They do not refer to the broader spectrum of patients we see. Other groups are well studied and those data could be considered to justify guideline changes and proper reimbursements now.

BERMAN: Actually, there is a large single-center study in press, but the results have not been released. We all look forward to hearing those results. It’s from a high-risk group in an emergency department from Beaumont Hospital. The patients who have been studied are people on the way to cath lab. There have been a lot of invasive coronary angiography to noninvasive coronary angiography tests and the patients are not just the asymptomatic people walking around on the street. These are people who have been referred to coronary angiography. Within that group, a complete normal study is associated with a miniscule, more like 1%, of frequency in a technically adequate study of a significant coronary lesion.

PEPINE: Dr. Lima, where do you stand on this?

LIMA: The images are very seductive. One has to keep that in mind because what is missing are data on prognosis. We have a lot of data comparing angiography with invasive angiography in patients who are on the way to the cath lab so they had some symptoms that got them referred to an invasive angiogram. That’s where the bulk of the data come from. I think the dilemma is that we are validating this tool against invasive angiography with the background goal of one day using it for screening. The greatest power is to look at the vessel wall. We eventually want to develop a tool that we can actually use on the patient who has a lot of risk factors when we think we need to intervene pharmacologically and medically to prevent progression of disease to nonfatal MI.

I’m very interested in preventing sudden death in people who have CAD. But, in order to move from that stage — from the symptomatic patient in whom we have data to the other populations — radiation remains a concern.

To use it for that function, one would have to reduce radiation by a large factor. Right now the radiation is acceptable for someone who has symptoms but is not acceptable for a patient who does not have symptoms.

We need data on prognosis and we need some technological development to reduce radiation.

PEPINE: So you are encouraged but want more data. Dr. Hill?

HILL: I feel the same way. The concern I have is taking it from the high-risk patient with symptoms to the relatively high-risk patient who is asymptomatic – such as a 55-year-old man with high LDL, family history and so on – and applying it to that group of patients.

You have an early detection bias in any kind of study that you would do, so you clearly have to define the patient population you are concerned about and do all the studies, not just with symptomatic patients but also with patients who will eventually be moving into the realm of having these studies done.

PEPINE: If the major goal is to prevent sudden death in a largely asymptomatic population, you really do have to take this tool to the wide-open, free-living people.

GARCIA: When you look at all the studies that have been published so far, these included patients who were selected to go to cardiac catheterization presumably because they had clinical criteria that made catheterization warranted. In fact only 30% to 40% of these patients ended up having significant disease.

We have been stricken by very similar observations in our practice in a multicenter trial that we have just completed. We know that over 75% to 80% of these patients have a positive or equivocal stress test result. So the practice of performing normal coronary angiography is significantly prevalent.

Recently while in Canada, I looked at the Ontario statistics on rates of normal coronary angiography and was very surprised that their prevalence rates of normal coronary angiography are 25% to 30%. This, then, is not a problem of our practice in the United States but is really an issue of not having significant diagnostic accuracy or confidence with the results of a stress test.

Part two of this round table will be published in March.