Virtual Histology Imaging in Acute Coronary Syndromes: Useful or Just a Research Tool?

ABSTRACT: Acute coronary syndromes (ACS) frequently cause considerable morbidity and mortality with a high risk of further events within the following year, despite the use of percutaneous coronary intervention (PCI). Numerous studies have described the concept of acute, partial or complete thrombotic occlusion of the coronary artery, which occurs at the site of a friable atherosclerotic plaque with a lipid-rich necrotic core and a ruptured overlying thin fibrous cap (“culprit lesion”). Moreover, this process appears independent of the severity of the underlying stenosis. Most of our knowledge about the morphological characteristics of culprit lesions has been obtained from necropsy studies of lesions at the extreme end of the ACS spectrum. The development of intravascular ultrasound (IVUS) with virtual histology (VH), using spectral analysis of the radiofrequency ultrasound backscatter signals to identify the components of atherosclerotic plaque, has allowed in-vivo delineation of the relative contributions of necrotic core and fibrous atheroma in unstable lesions. This evidence suggests that there may be variations in the morphology of plaques that rupture and promote thrombosis in ACS, rather than the traditionally accepted view that a common pathological mechanism is at play. This imaging modality, therefore, provides great potential for in-vivo information about the culprit lesion. In this review article, we explore the background and potential application of virtual histology to improve the invasive treatment of ACS patients.

J INVASIVE CARDIOL 2010;22:84–91

___________________________________________
Acute coronary syndromes (ACS) frequently cause considerable morbidity and mortality, with a high risk of further events within the following year despite the use of percutaneous coronary intervention (PCI).¹ Numerous studies have described the concept of acute, partial or complete thrombotic occlusion of the coronary artery which occurs at the site of a friable atherosclerotic plaque with a lipid-rich necrotic core and a ruptured overlying thin fibrous cap (“culprit lesion”).^2,3 Moreover, this process appears independently of the severity of the underlying stenosis.⁴ Most of our knowledge about the morphological characteristics of culprit lesions has been obtained from necropsy studies of lesions at the extreme end of the ACS spectrum.⁵ The development of intravascular ultrasound (IVUS) with virtual histology (VH), using spectral analysis of the radiofrequency ultrasound backscatter signals to identify the components of atherosclerotic plaque, has allowed in vivo delineation of the relative contributions of necrotic core and fibrous atheroma in unstable lesions.^7–9 This evidence suggests that there may be variations in the morphology of plaques that rupture and promote thrombosis in ACS, rather than the traditionally accepted view that a common pathological mechanism is at play. This imaging modality, therefore, provides great potential for in vivo information about the culprit lesion.⁶ The spectrum of coronary heart disease is heterogeneous. Most acute events arise from lesions that were originally non-flow-limiting. Precise lesion characteristics could influence interventional or medical treatment, but currently this is not evaluated. Clinically, ACS presentations vary from unstable angina to non-ST/ST-elevation myocardial infarction (NSTEMI, STEMI). Considerable variation exists in the degree of myocardial injury, assessed by troponin release and the extent of myocardial ischemia, as demonstrated by patient symptoms and electrocardiographic (ECG) changes. Moreover, the response of patients to conventional pharmacological therapy with respect to the resolution of myocardial ischemia is variable, ranging from a single episode of chest pain to prolonged, intractable symptoms. This variation is multifactorial and, in addition to clinical factors, includes extent of coronary disease at the culprit lesion site. However, contemporary medical treatment is not lesion- or patient-specific, largely due to the inability to accurately characterize the lesion. Current IVUS-VH data suggest significant variations exist in the lipid-rich necrotic core content of the culprit lesion.^3,5,7–9 In this review we aim to explain the role of IVUS-VH and its potential to add knowledge to the field of intervention on ACS lesions.

The Technology

Grayscale IVUS is a tomographic imaging tool that can visualize coronary atherosclerosis in vivo, elucidating plaque area, plaque distribution, lesion length and coronary remodeling. IVUS has demonstrated discrepancies between the extent of atherosclerosis seen by coronary angiography and the actual extent of atherosclerotic disease.¹⁰ Angiography or “lumenography” is able to determine only whether there is adequate space for blood to flow well, but it cannot assess the underlying vessel changes and “hidden” plaque burden. Quantitative assessment of this plaque composition within a lesion has not been possible with grayscale IVUS analysis, until now.¹¹ IVUS-VH uses advanced radiofrequency analysis of ultrasound backscatter signals to overcome the limitations of grayscale IVUS by providing a more detailed analysis of plaque morphology.12 In addition, IVUS-VH has the potential to provide in vivo patient-specific plaque analysis to determine the range of characteristics in relation to clinical factors and risk, rather than making assumptions from a highly selected autopsy population. Obviously, many different cell and tissue types are commonly found in atherosclerotic plaques. To simplify image interpretation and because of the fundamental resolution limitations of the underlying ultrasound signal, plaque components are grouped into four basic tissue types during IVUS-VH imaging. These components are displayed on the image as different color pixels. This technique is based on advanced radiofrequency analysis of reflected ultrasound signals in a frequency domain analysis and displays the reconstructed color-coded tissue map of plaque composition superimposed on cross-sectional images of the coronary artery obtained by grayscale IVUS.¹² Basic IVUS enables real-time, high-resolution tomographic visualization of the coronary arteries. Both lumen and vessel dimensions and the distribution of plaques can be analyzed. In addition, grayscale IVUS can be used to assess the presence of intraluminal thrombus and plaque rupture.15 Grayscale IVUS has demonstrated the multiplicity of plaque ruptures seen in patients with ACS.16–18 A recent study demonstrated that the number of vulnerable plaques with less than 75% luminal obstruction identified by IVUS had a positive correlation with future cardiovascular events.19 Of note, serial IVUS analysis of a small patient cohort showed that 50% of ruptured coronary plaques detected on the first ACS event had spontaneously healed at 22 months' follow up.²⁰ Grayscale IVUS imaging is, however, limited with regard to analysis of plaque composition. Both calcified and dense fibrotic tissues, such as those found in plaques, have strong echoreflections with lateral shadowing and are, therefore, not easy to differentiate. Currently, VH can better distinguish between areas with low echoreflections than can grayscale IVUS. Nevertheless, quantification of hypoechogenicity is important, as it has been related to adverse event rates.^21,22

Validity

There has been debate over the best way to validate IVUS-VH, as ex-vivo analysis removes blood attenuation, ECG gating and coronary contraction/motion. To assess validity in vivo, VH-IVUS backscatter data from 51 ex vivo left anterior descending coronary arteries were recorded and compared with histological interpretation of the same sites.²⁴ The overall predictive accuracies were 93.5% for fibrotic tissue, 94.1% for fibrofatty tissue, 95.8% for necrotic core and 96.7% for dense calcium. This demonstrated the potential of this imaging tool for analysis of plaque vulnerability. In the Carotid Artery Plaque Virtual Histology Evaluation (CAPITAL) study, there was a strong correlation between plaque characterization and that following true histological examination of the plaque following endarterectomy.³⁵ The predictive accuracy for TCFA was 99.4% and 96.1% for calcified TCFA. In vivo studies using IVUS-VH have shown that presumed vulnerable plaques (TCFAs according to IVUS criteria) occur more often in patients with ACS than in patients with stable angina.40 Patients who are stable at clinical presentation show more intimal thickening and plaque composition with considerably more fibrotic tissue than that typically seen in vulnerable plaques.

Tissue Types

Fibrous. Fibrous tissue is represented by dark green pixels. Histologically, this tissue is collagenous, with no lipid.^23,24 On grayscale IVUS, these tissues tend to be medium-bright regions. Fibrofatty. Fibrofatty tissue is denoted in VH by light green pixels. This tissue is loosely packed collagen, but it can have a cellular quality, with potential for foam cells to start invading.²⁴ There is usually no necrotic core and even cholesterol products are rare. If thrombus or plaque rupture are included as plaque during analysis, then they are displayed as fibrofatty plaque. Necrotic core. In VH the necrotic core is seen as red. This tissue is a mixture of soft, lipid-like dead cells, foam cells and trapped blood cells.²³ Most of any real structure is lost, and with some areas producing microcalcification as a byproduct (from the dead cells), this leads to a recipe for gross instability and rupture with friable areas next to sharp calcification. Dense calcium. White pixels represent dense calcium. These calcified regions can be lost during histology processing, but on plain grayscale IVUS, they act as extremely strong reflectors of signal and appear as bright white with a “red” shadow behind. There is ongoing debate among the IVUS-VH research community about how much information can safely be derived behind various types of calcium. Stent struts may also appear as calcium if included in plaque analysis.

Can IVUS-VH Improve Our Knowledge about Coronary Disease?

Plaque risk assessment. In-vivo plaque classification with VH-IVUS is based on a histopathological classification system developed by Virmani et al in 2000.^2,5 From this system, coronary lesions can be classified as adaptive intimal thickening, pathologic intimal thickening, fibroatheroma, fibrocalcific and TCFA plaques. For risk stratification or the likelihood of lesion rupture, it may be important to distinguish between the above-mentioned plaque types. This classification system represents the different stages in the development of atherosclerosis through to atherothrombosis. Adaptive and pathological intimal thickening (AIT)/(PIT). In AIT, the intimal tissue largely comprises smooth muscle cells within a collagen matrix. In PIT, the intimal tissue is comprised of mixed fibrous and fibrofatty plaque (lipid pools) with very little or no necrotic core. There can be microcalcification, but less than 10% of the plaque.²⁵Fibroatheroma (FA). Fibroatheromatous plaques have a true necrotic core greater than 10% of total plaque volume, and this contains cholesterol products. The fibrous cap that protects the coronary lumen from the necrotic core is significant. As the lesions progresses, the fibrous cap is assumed to diminish to a thickness less than 65 µm, which would increase vulnerability and allow classification as a TCFA. The limited resolution of IVUS-VH (approx 150 µm) means the cap can be measured only if greater than 150 µm thick. Fibrocalcific plaque. Fibrocalcific plaques are mainly fibrous plaques that are dense in calcium (> 10%) and some have a small amount of confluent necrotic core (28 TCFA: IVUS-VH definition of vulnerability. TCFAs have a confluent necrotic core (> 10%) in direct contact with the lumen (no evidence of a fibrous cap) and a minor amount of calcium ( 20%, no evidence of a fibrous cap, a calcium content > 10% with a speckly appearance and a positive remodeling index with a high plaque burden/luminal narrowing confer the greatest vulnerability to plaques. This appears to be in some way different from the commonly taught theory that nonsignificant plaques are the ones that are more likely to rupture.²⁹ It may simply be that nonsignificant plaques are more frequent along a given coronary segment.

Can IVUS-VH Influence Coronary Intervention?

As the majority of acute coronary events are triggered by plaque rupture,^17,18 defining the anatomic features that lead to plaque rupture should be of central importance to lesion imaging. Post-mortem analyses have shown that TCFA is probably the main precursor lesion for plaque rupture.² TCFAs are characterized by a large necrotic core separated from the coronary lumen by a thin fibrous cap. According to histological studies, the size of the necrotic core and the thickness of the fibrous cap have a critical influence on plaque stability. In addition, other characteristics of vulnerable lesions include localized expansive enlargement of the vessel wall (“positive remodeling”) and microcalcification within the lesion. The location of the lesion within the coronary tree, the length of the lesion and the narrowing of the artery relative to healthy reference lumen size are also important parameters for the evaluation of plaque vulnerability. Coronary dimensions and elements of plaque composition such as the presence and amount of necrotic core, the degree of calcification and coronary remodeling are all anatomic features visualized by IVUS and IVUS-VH, but not by traditional angiography. The reconstruction of VH-IVUS images in a longitudinal view enables a comprehensive analysis of the total length of the plaque and its complete orientation to the rest of the coronary tree.

Should We Consider Utilizing IVUS-VH in ACS?

The longitudinal IVUS-VH view of the target segment provides a wealth of information about the length of the plaque and its general composition. According to post-mortem data, plaque composition is a better predictor of ACS events than the degree of coronary stenosis.²⁹ The formation of intraluminal thrombus after plaque rupture or erosion plays a central role in the clinical course of ACS. Coronary thrombosis is a dynamic process and the culprit thrombus can assume different degrees of organization. With repeated thrombosis, the imaging of lesions becomes increasingly difficult. As the plaque ruptures, the lesion site is covered with thrombus and the underlying plaque and necrotic core are filled by intramural hemorrhage.³⁰ An intraluminal thrombus tail rich in red blood cells can form proximal to the site of plaque rupture, especially in the presence of slow blood flow. This thrombotic tissue can be organized like fibrotic or fibrofatty tissue, causing luminal narrowing further from that at the rupture site. Post-mortem studies have clearly shown that repeated plaque ruptures have a key role in plaque progression. Pathologic studies have detected complex plaques with previous ruptures concentrating at one site, suggesting that certain sites are chronically vulnerable.²⁶ Subclinical episodes of plaque disruption followed by healing is considered a mechanism by which plaque burden increases and leads to inward remodeling processes and finally luminal narrowing.²⁶ The origin or focus of the plaque rupture can be proximal or distal to the minimum lumen of the lesion.³¹ Hence, assessment of atherosclerotic lesions should include the whole length of the lesion to ensure detection of the site of plaque rupture, which might not necessarily be the site of the minimum lumen cross-sectional area. This phenomenon has been seen earlier in IVUS studies.²⁷ In ACS patients, despite early PCI, significant numbers of patients have a recurrent event, which raises the question of whether procedural factors influence the outcome. The mechanism of angioplasty in acute plaques is poorly understood.³⁶ In a recent first-in-man paper, Wei et al³⁶ reported that in 20 ACS patients undergoing predilatation, 35% of lumen enlargement was due to an increase in vessel area and 65% to a decrease in plaque area. They concluded that fibrous and fibro-fatty elements redistribute into the reference segments and one-third of necrotic tissue was “lost”, potentially supporting a theory of embolization. Angiographically guided stent placement is an imperfect but realistic compromise in the setting of “real-world” PCI. Many subtle forms of stent placement abnormalities exist, and unfortunately, these may contribute to long-term adverse events such as stent thrombosis and in-stent restenosis. The STLLR (Prospective Evaluation of the Impact of Stent Deployment Techniques on Clinical Outcomes of Patient Treated With the Cypher® Stent)³⁷ using quantitative coronary angiographic (QCA) evaluation of stent placement has shown that suboptimal placement (or “geographic miss”) is widespread and confers a three-fold increased chance of heart attack and twice the chance of a repeat procedure. The types of geographic miss range from longitudinal miss, with the stent being too small to cover the lesion, to axial miss, which means the stent is under- or oversized. The use of IVUS-VH before any intervention would allow further inspection and measurement of the vessel size, plaque length and inspection of side-branch anatomy and plaque constituents (e.g., significant calcification). This extra information could go a long way to improve strategies in PCI. Another area of interest is in the angiographic placement of stents directly into necrotic areas of plaque, which are known not to occur at the mimium luminal area (MLA) of the culprit plaque.³⁸ Stent struts embedded within necrotic tissue delay endothelialization and adversely affect tissue level homeostasis and thrombotic status.³⁹ This is a very important area where IVUS-VH could positively influence stent placement in ACS patients and would ensure not only that the MLA is covered, but also the more unstable necrotic shoulder of the distal/proximal plaque, which may prevent further late thrombotic sequelae. As post-mortem data have demonstrated, longitudinal analysis of the target lesion using IVUS-VH can show the site of plaque rupture, which is often proximal to the MLA site. Covering only the MLA site might not be sufficient to fully treat the problem. IVUS-VH could reduce the rate of future events related to reoccurring ruptures of an uncovered lesion. This concept is also being evaluated in ongoing clinical studies.

Clinical Studies Assessing and Utilizing

IVUS-VH in ACS Recent studies have confirmed that plaque composition has a nonuniform distribution along the coronary arteries.⁴² Depending on the distance from the coronary ostium, the proximal segments show a significantly larger necrotic core, but no change in other plaque components.⁴³ As the proportion of necrotic core is related to overall lesion vulnerability, these findings could explain the higher incidence of plaque ruptures in the proximal segments of the coronary tree. According to previous studies, IVUS-VH analysis could confirm the relation of outward (positive) and inward (negative) remodeling processes to plaque composition.⁴³ In a small in vivo study, plaque composition and morphology assessed using IVUS radiofrequency analysis were shown to relate to coronary remodeling, supporting a role of plaque composition in vessel remodeling.⁴⁴ The size of necrotic core is significantly larger in lesions with positive remodeling; fibrous plaque burden, however, has a significant inverse relationship with the remodeling index. Positive remodeling is, therefore, associated with presumed high-risk lesions such as fibroatheromata and TCFAs, while negative remodeling is associated with less-vulnerable lesion types such as those with pathologic intimal thickening and fibrotic plaques.⁴⁴ Recently IVUS-VH analyses were performed in the first 990 patients enrolled in the 3,000+ patient global IVUS-VH Registry to assess the impact of gender and age on in vivo plaque characterization. The 990 patients were divided into three age-group terciles ( 68 years) and again divided according to gender. Naturally, both women and men had an increase in plaque with increasing age, and at any age, men had more plaque than women. The percentages of dense calcium and necrotic core increased with increasing patient age and diabetic status. In both men and women, gender differences were lowest in the oldest tercile (> 68 years).⁵¹ In December 2008, a group from Korea found no-reflow/slow-flow in 12 of 57 patients (21%) undergoing PCI for ACS.⁵² Having performed pre-PCI IVUS-VH, they found that fibrofatty plaque volume across the entire lesion length correlated with slow flow, which appears to incriminate this type of plaque, or more likely thrombus, as the main culprit for reduction in thrombolysis in myocardial infarction (TIMI) flow following lesion treatment. This suggests that IVUS-VH in ACS can identify angiographically silent thrombus burden and potentially allow thrombus aspiration before balloon angioplasty or stent implantation. However, other authors looking at post thrombus-aspiration STEMI patients and stable angina patients, appear to suggest that the necrotic core volume is the only independent predictor of no-reflow.^53–55 It is well known from previously conducted meta-analyses and reviews that using IVUS-guided stent implantation has no differential effect on long-term death or nonfatal MI in comparison with angiographically guided stenting. However, there is a reduction in angiographic restenosis and target vessel revascularization (TVR) at 6 months with IVUS-guided stenting. Moreover, a very recent paper by Roy et al⁵⁶ using IVUS guidance to place DES showed reductions in both stent thrombosis and target lesion revascularization (TLR) when compared with an angiographically guided group. The authors believe that a growing evidence base is accumulating to support optimum lesion assessment and stent deployment during PCI procedures to ensure the clinical benefit is passed on to patients.

Discussion

Catheter-based invasive diagnosis of plaque composition with IVUS-VH can identify the features of different plaque types that are important predictors of plaque vulnerability. The in vivo predictive power and accuracy of IVUS-VH and the assumptions made from post-mortem data are currently being assessed in a clinical trial that aims to identify lesions at high risk of plaque rupture leading to either plaque progression (silent plaque progression) or a clinically symptomatic event. The current characteristics being examined as possible indicators of increased lesion vulnerability are the presence and extent of a confluent necrotic core, the absence of a fibrotic cap, the pattern of calcification, positive remodeling, the degree of luminal stenosis and the location of the interrogated lesion. Compared with other diagnostic tools, IVUS-VH provides improved in-vivo diagnostic accuracy of atherosclerotic plaques. As an in-vivo diagnostic tool with online, on-screen data display, it has the potential to bypass the most important bias of post-mortem data (that analyzed patients have already died from acute cardiac death). It therefore can accurately assess plaque progression on an individual basis. Of all currently available diagnostic tools, the pathology-based criteria for plaque vulnerability regarding plaque composition and plaque type can be most comprehensively assessed by grayscale IVUS and IVUS-VH. This evaluation enables more comprehensive risk assessment and stratification on an individual basis for secondary prevention. Grayscale IVUS has established some clear interventional applications such as the characterization and quantitative analysis of left-main disease and the imaging of in-stent restenosis.^9,48 The additional benefit of plaque composition analysis for these clinical indications is under investigation at present. IVUS has a role in the assessment of stent underexpansion and stent malapposition; however, it is not recommended for routine use during stent implantation. As discussed earlier, IVUS-VH guidance of coronary interventions could achieve complete coverage of virtual histologically defined high-risk lesions, in addition to treatment of the MLA, and, therefore, potentially reduce the risk of restenosis or progression of atherosclerosis in the reference segments. The efficacy of this application, in comparison with that of optimal medical therapy, is also yet to be determined. Until now, we have had no diagnostic tools and no evidence to support the treatment of non-flow-limiting lesions with a preventive strategy other than risk- factor modification. There is uncertainty regarding the restenosis risk after preventative stent treatment of a vulnerable lesion in comparison with the spontaneous rupture rate of high-risk lesions that have not been treated with stenting. The duration of the possible vulnerable stage is also unknown, as is the time of the progression or regression of coronary artery disease. Given that the vulnerability of high-risk lesions may be only temporary owing to the possibility of changes in plaque structure. Prospective, serial IVUS-VH studies are being performed to clarify the natural history of these vulnerable lesions (PROSPECT/ATLANTA trials). Our knowledge of the natural history of atherosclerosis, including lesion classification, relies on post-mortem histology data. As IVUS-VH enables in vivo identification of four different plaque characteristics and their location, this technique may enable a more accurate classification of lesions with regard to progression and regression. With time, we may even be able to determine with high accuracy which lesions should be treated immediately with intervention and which with long-term systemic medical therapy.

IVUS-VH Current Limitations

The limitations of this technique are described elsewhere.49 The most crucial aspect of any IVUS examination is a steady pullback. However, accurate border detection is critical when post-processing the images. The VH software classifies the entire plaque area, and poor detection can seriously influence composition results and plaque risk assessment. The axial resolution of IVUS-VH (150 µm) is too low to detect thin fibrous cap thickness, which is currently defined as 65 µm. This threshold for critical cap thickness has, however, been established for ruptured plaques, but not TCFAs. Histopathological studies apparently demonstrate a greater thickness for caps in TCFAs.⁵⁰ This is related to the dehydration produced when preparing specimens. As previously discussed and shown, thrombus detection could help localize both the extent and origin of the plaque rupture in patients with ACS. Thrombus as the primary surrogate for acute coronary thrombosis cannot yet be assigned a color classification, therefore we should focus harder as a group of researchers to spot thrombus on grayscale IVUS to exclude it from analyses. More attention should be paid to thrombus-associated plaque (TAP), as this effectively represents plaque that has been involved directly in an atherothrombotic episode. In a similar situation, plaque rupture is visible on grayscale IVUS, but is not traceable by the current software and has to be included within the plaque burden (increasing fibrofatty proportions). It is hoped by our research group that eventually the operator may be able to to trace plaque rupture or thrombus and color code/quantify it better. This would also allow greater focus on the true culprit plaque in these areas. There also remain concerns with regard to the true ability of the IVUS-VH classification tree algorithm to accurately predict what is behind areas of significant calcification within plaques. This is due to the loss of echo intensity in these areas. Work continues constantly on modifying and improving these systems to further aid accurate tissue diagnosis.

Conclusions

IVUS-VH is a promising imaging tool that can influence the critical decision-making process of PCI. It currently fits in as one piece in a complex puzzle of information related to the natural history and the outcome of ACS. It is hoped that data provided by ongoing IVUS-VH studies will prove influential in PCI guidance and establish a firm role for this technology to progress from a research tool to improving patient outcomes. Acknowledgement. We wish to thank Volcano Corporation for allowing us the use of various figures in this article.

_________________________________________ From the Liverpool Heart and Chest Hospital, Liverpool, United Kingdom. Manuscript submitted September 15 and accepted September 25, 2009. Disclosure: Dr. Palmer discloses that he has received honoraria from Volcano Corp. for IVUS training services. Address for correspondence: Scott Murray, MB ChB, MRCP, Liverpool Heart and Chest Hospital, 14 Thomas Drive, Broadgreen, Liverpool, United Kingdom L14 3PE. E-mail: scottmurray@doctors.org.uk