Notes From a Cardiac AV Superuser: Wm. Guy Weigold, MD
For a cardiologist, Wm. Guy Weigold, MD, spends an unusual amount of time in front of a monitor. “I happen to be a cardiologist who has expertise in cardiac CT,” he explains. “I spend the majority of my time looking at images.” Weigold is director of the cardiac CT program at the MedStar Washington Hospital Center in Washington, DC, and he directs that institution’s cardiac CT core laboratory (part of a large cardiovascular-imaging laboratory that also offers cardiac MRI, echocardiography, intravascular ultrasound, and basic angiography under the auspices of the MedStar Research Institute). To keep from slipping over to the dark side, Weigold maintains a regular schedule of clinical service each week, either participating in rounds with colleagues or covering the cardiac ICU, for instance. “I try not to turn completely into a radiologist,” he says, only half in jest. Nonetheless, when it comes to the mysteries of atherosclerotic disease—understanding, for instance, what triggers an acute myocardial infarction—Weigold believes that CT imaging and advanced visualization software will provide at least some of the answers. Weigold has played an active role in cardiac-imaging research, participating in the SPARC trial,¹ a comparative-effectiveness study that looked at the impact of various diagnostic strategies (including coronary CT angiography) on downstream testing. In his capacity as CT core laboratory director, Weigold helps research organizations conduct the cardiac CT portions of their studies, using the suite of advanced visualization tools in Synapse® 3D (FUJIFILM Medical Systems USA, Inc) to probe and manipulate the image information. Research organizations often turn to a core laboratory for analysis of imaging data for consistency, expertise, and accuracy—but also to avoid confounding and bias by keeping analysis walled off from clinical evaluation of the patient, he explains. Rigorous Demands In fact, reading cardiac CT studies for research often demands more specific detail than would a straightforward clinical study. “A core laboratory will often engage in a more rigorous quantification of the imaging data and calculations, and a very thorough going over of the data,” he notes. “The power of the Fujifilm system is that it has a lot of capability and applications built within it.” On any given day, Weigold might need to do a large-scale analysis of the aorta and peripheral vasculature or a detailed analysis of the heart’s structure and function. The Synapse 3D tools make detailed analysis of the coronary arteries possible, from the detection of coronary calcification and quantification of coronary stenosis to drilling down to the analysis of plaque composition. “Research organizations are going to come to you with any variety of questions, depending on what their studies are, and they are going to ask for a variety of different types of analysis,” he explains. “One study may looking at aortic diameters, aortic aneurysms, and aortic dissections, and that’s a completely different set of analyses from those needed by someone interested in plaque characteristics of coronary atherosclerotic lesions. Having a panoply of applications is a required feature when you are looking for a robust system for a core laboratory, which is why we like it.” The Elusive Trophy As Weigold explains it, atherosclerotic lesions are histologically complex, presenting medicine with an enduring mystery: “The challenge in coronary atherosclerotic disease is to try to do something to address this problem of a disease entity that has a long subclinical course that can undergo an acute destabilization and produce a lethal affliction—which is an acute myocardial infarction. It is well known that people can harbor this disease for years without any symptoms whatsoever until the day of their infarction, yet we still know very little about what triggers that destabilization; frequently, the first manifestation of coronary atherosclerosis is an acute myocardial infarction,” he says. Ideally, physicians would detect this disease early in its subclinical phase and intervene in the disease process, Weigold says, but that has been stymied by multiple complexities: Many people walk around with this disease process and never have a myocardial infarction; others can walk around with extensive, severe disease that never causes an infarct, but might lead to bypass surgery. Still others have a very small burden of disease—only one small portion of the coronary vasculature is affected—but that happens to be the one that destabilizes and produces an acute myocardial infarction. “There’s a lot of interest in trying to predict the future—to identify people who will have a heart attack down the road—and do something about it,” Weigold says, “but that has turned out to be very difficult. The goal of an accurate prediction method coupled with a truly effective early-disease treatment is turning out to be an elusive trophy.” A Detection Gap Beginning with the ongoing Framingham Heart Study in 1948, medicine has produced a well-known set of risk factors that physicians can feed into an algorithm to predict the likelihood that a patient will have a heart attack. “The problem is that doesn’t work as well as we would like,” Weigold says. “There still are large numbers of people who should be at low risk, according to the risk-prediction algorithm, but who go on to have a heart attack. There is a detection gap where we are still missing people. This is where the hope for imaging has emerged.” Direct visualization of the coronary arteries has produced two ways of assessing risk. According to Weigold, the more accepted (given the available data) is the coronary-artery calcium score, but his interest is clearly in the second method: plaque analysis using coronary CT angiography. “The the burden of calcified disease seems to be a marker for risk; it actually may be the noncalcified cholesterol deposition that is the bad actor here,” he says. “Rupture of a cholesterol-rich, thin-capped fibroatheroma is the most common mechanism for acute coronary thrombosis. In some cases, the angiogram in an acute infarction demonstrates the crater left behind where there used to a plaque that was unroofed—that erupted and opened. That process is highly irritating to the blood flowing past it, which responds by clotting, producing the heart attack.” Digging Into the Tool Kit Using the Synapse 3D coronary-analysis tools, Weigold looks at the attenuation data within the plaque itself to separate the different gradations of plaque histology. Calcified lesions produce a very high attenuation threshold. He divides noncalcified lesions into two groups: fibrotic lesions with intermediate attenuation (which might be more stable lesions, with more collagen deposition) and others, with very low attenuation, that are believed to contain pools of lipids and cholesterol buildup. Histology and pathology studies indicate that the latter are the lesions that can destabilize and rupture and produce the inflammatory and thrombotic cascade, he says. “The software allows you great flexibility, in terms of being able to set multiple different thresholds and up to five different levels of attenuation, and to set the characteristics you would like to establish in order to discriminate between one type of plaque and another,” Weigold adds. Beyond the attenuation values, Weigold also makes use of the data to be derived from a 3D volume acquisition. “A simple example would be the volume of the left ventricle,” he says. “There’s clinical utility in knowing whether someone’s actual heart size is too big or too small, at a volume level expressed in milliliters. The same kind of thing can be done, at a smaller level, within the blood vessels—to look at plaques.” The Bad Actor Weigold describes some plaques as small, atherosclerotic pimples on the wall of the vessel, while others are very large, and might or might not narrow the lumen to obstruct blood flow. “As it turns out, from the pathology and histology studies, the lesions that seem to go on to produce myocardial infarction (perhaps contrary to intuition) are not those that are the most stenotic, flow-limiting lesions, but are actually those that start out as lesions that do not limit flow, but then go on to rupture and produce acute thrombosis,” he says. The vessel can expand to preserve its luminal diameter, accommodating the plaque. He explains, “It seems that the really high-risk lesions are these large-volume plaques, outwardly expanded and remodeled, that are producing minimal lumen stenosis.” He adds, “With CT and these kinds of analysis tools, we are way past where we ever were, in the past, with plain old angiography, looking at luminal stenosis.” The hope, he says, is that further studies and data collection will produce better understanding of the pathophysiology of the plaque and what triggers the myocardial infarction that happens today, even though the lesion was there yesterday, the day before, and a year before that. “That’s the goal,” he says. Cheryl Proval is editor of Radinformatics.com.