Proposed Methods To Diagnose Vulnerable Plaque

Imaging modalities that have been used to study plaques within the current concept of vulnerable plaque (VP) are summarized in this section. Under the current concept of VP, most CVD events are caused by plaques at high risk of rupture. Therefore, studies of biomarkers predicting CVD events might be useful to identify the presence of "VP." We reviewed studies that examined the association between serum biomarkers and the occurrence of ACS events. Since these are not studies of specifically defined "VP" populations, their inclusion here is merely to suggest possible roles of these biomarkers.

The evaluation of a diagnostic test requires a reference standard. Given the lack of a reference standard for VP, we discuss some issues concerning such a definition. This discussion is followed by summaries of peripheral blood biomarkers and imaging modalities. We provide one table for blood biomarkers (Table 1) and another for imaging tests (Table 2). Appendix 1 and Appendix 2 provide detailed descriptions of each biomarker and imaging test, respectively.

Reference Standard Issues of Tests for Vulnerable Plaque

The current concept of VP focuses on the characterization of the structure and behavior of the plaque lesion. Given the location of the plaques and the need to accurately characterize intra-luminal plaque features, an invasive imaging technology will most likely be used as the reference standard for VP. It will be necessary to conduct natural history studies (as discussed earlier in Section 4) to compare these tests to determine the technology that most accurately predicts the occurrence of acute coronary syndrome (ACS) or acute ischemic neurological events.

Assuming the validity of the concept of VP, a peripheral blood marker could potentially be used for diagnostic purposes if it was unique to the VP features and if it was sufficiently elevated to be detected by a laboratory method before the plaque ruptures. Considering that most cellular markers are far from being specific because they usually come from more than one part of the human body, it would be difficult to distinguish whether the increased level of a marker was due to these plaques or to some other process such as inflammation unrelated to the coronary arteries.

Additionally, complex techniques might be required for the detection of certain markers. Furthermore, a plaque with vulnerable features must be identified reasonably quickly to allow for effective therapies to be instituted. Traditional risk assessment predicts long-term outcome, but does not provide information about whether a specific plaque will rupture in 1 month, 6 months, or 1 year. This raises the question of how the diagnostic threshold value of a marker would be established.

Ideally, to evaluate a peripheral blood marker to diagnose the condition, we would need a prospective study of subjects that would be followed for outcomes attributable to plaque rupture (ACS or sudden death). In this ideal study, baseline measurements of the marker with the properties already described, as well as the results of the imaging technique, should be obtained. The imaging technique would allow us to focus on certain segments of coronary arteries that may correspond to VPs and which would serve as a reference standard to test the proposed marker. After an outcome, new imaging studies (or autopsy) would confirm whether the event was due to the rupture of one of those segments that was initially characterized as VP. At the end of follow up, the data would be analyzed to determine whether there was a relationship between the baseline values of the cellular marker and the outcomes that were attributed to VPs.

Such a prospective cohort study would face several barriers. These include the need to use invasive techniques in an apparent healthy population, the difficulty of obtaining complete followup data on the study population, the lack of a reference standard for imaging techniques as well, and the difficulty of selecting one of the proposed markers to be tested since none meets all the requirements.

Studies of Inflammatory and Cellular Markers

Candidate markers have been evaluated by prospective studies for predicting clinical outcomes—ACS or cerebrovascular events, or death. Since they may identify singly or in combination only those patients at risk for those outcomes, their association with plaque vulnerability must be inferred.

The paradigm for peripheral blood biomarker detection for "VP" is predicated on the principle that cytokines and other biologic factors within the plaques can enter or leave the bloodstream, and be detected circulating in peripheral blood. Alternatively, "VP" may cause unique cell types, cytokines, or a combination of substances and cells to appear in the peripheral blood of affected patients (Schwartz 2003). Finally, certain types of patients, i.e., those with a systemic inflammatory state or those prone to producing inflammatory markers, might be more likely to form "VP."

Although unique identifying substances that would permit high sensitivity and specificity have not yet been characterized, several candidate markers have been proposed that may be used to estimate the "VP" risk in populations (Schwartz 2003). This non-invasive assessment of risk could be used to identify patients for future non-invasive or invasive imaging studies.

Table 1 summarizes inflammatory and other cellular markers that have been evaluated by prospective studies for predicting cardiovascular or cerebrovascular events. A narrative descriptive of these markers is included in Appendix 1.

Review of Studies That Used a Biomarker to Assess ACS Risk

Several studies described below investigated prospectively whether cellular and inflammatory markers are associated with progression of ACS.

Blankenberg et al. (2002) evaluated serum concentration of IL-18 and other markers of inflammation in 1,229 patients with CAD and found IL-18 to be a strong independent predictor of cardiovascular disease (CVD) death in patients with CAD; the authors suggested that IL-18mediated inflammation leads to accelerated vulnerability of atherosclerotic plaques.

Blankenberg et al. (2001) evaluated the effect of soluble adhesion molecules on the risk of future cardiovascular events among 1,246 patients with CAD with a mean 2.7 years of follow-up. They concluded that soluble adhesion molecules level of sVCAM-1, sICAM-1 and sEselectin were significantly related to future death from cardiovascular causes among patients with CAD.

Chakhtoura et al. (2000) evaluated 25 patients (17 with unstable angina and 8 with stable angina) and found that patients who had unstable complex lesions had a fivefold higher expression of the platelet activation epitope CD63 than patients with stable angina. In addition, patients with unstable angina had 15 percent more glycoprotein IIb/IIIa aggregation sites expressed on their platelet membrane. A direct relationship was observed between the morphology of ruptured plaque and platelet activation in patients with unstable angina.

Considering the studies available in the published literature, the hs-CRP measurement appears to be the strongest marker for future clinical events (i.e., MI, unstable angina, and stroke) due to arterial inflammation in both diseased and apparently healthy, asymptomatic patients (Morrow 2000; Koenig 1999; Tracy 1997; Reuben 2000; Biasucci, Liuzzo, and Fantuzzi 1999; Ridker 1998; Kuller 1996; Taaffe 2000; Danesh 2000; Lowe 2001). The CRP plasma level is also suggested as the best method of risk assessment in patients with either stable or unstable angina (Liuzzo 1994; Heeschen 2000; Biasucci, Liuzzo, and Grillo 1999; Sabatine 2002; Morrow 1998; Toss 1997) and for long term assessment after MI (Ridker, Rifai, Pfeffer, et al. 1998).

Furthermore, significant lipid elevation occurs in fewer than 50 percent of patients with ACS, MI, and unstable angina. For this reason, cholesterol and lipid measurement alone do not seem to be satisfactory markers for those at risk of sudden vascular events. One study (Ridker, Stampfer, and Rifai 2001) showed CRP and Total/HDL cholesterol ratio as the only cardiovascular risk indicators, using multivariate and age adjusted analysis. If CRP, IL-6 and ICAM-1 levels are added to lipid levels, risk assessment can be improved over lipids alone (Ridker, Stampfer, and Rifai 2001). Appendix 1 describes each marker and its potential association with clinical outcomes in detail.

In summary, although there is some indication that cellular markers might be correlated with ACS risk (i.e., they suggest the presence of VP), there is no evidence from studies specifically designed to evaluate whether such markers can accurately predict the presence of "VP."

Imaging Modalities

Development of clinically useful imaging techniques for identifying "VP" has been an active area of research in the last decade. Table 2 lists non-invasive and invasive imaging modalities which have been proposed and studied to assess these plaques. Appendix 2 describes each modality and its potential association with clinical outcomes in detail.

Summary of Methods to Diagnose Vulnerable Plaque

Despite the rapidly growing support for the VP concept among the research community, current clinical practice has not been substantially altered by this concept. This is largely because there is not yet a consensus on the operational definition of a VP, and there is no readily available imaging or laboratory test that would allow clinicians to diagnose the presence of a VP in routine clinical practice. Moreover, even if such a diagnostic test were available, too little is known about the natural history of "VP," so it is difficult to see how such a diagnostic test would inform clinical decisionmaking at this time. Further, there are substantial barriers to the development of accurate diagnostic methods of the plaques. One major barrier derives from the fact that the description of the plaques was based largely on histopathological descriptions of culprit lesions. This gold standard is only available after a plaque becomes a culprit plaque, and indeed only after the patient dies. Large-scale studies of the natural history of plaques are needed, to identify what characteristics on the various imaging modalities are most important in prospectively predicting plaque rupture. These studies have not been done to date.

Among the imaging techniques described above, a clear consensus has not emerged about which method should be the reference standard, nor has a unique identifying substance such as CRP been identified that would uniquely identify these plaques with any reasonable sensitivity or specificity. Therefore, it is our opinion that large scale, prospective studies employing serial imaging of the coronary arteries of asymptomatic high-risk patients, or of patients with stable angina, is required before an estimate of the risk of an acute coronary event can be made either in asymptomatic patients or in patients with stable angina with plaques of a particular morphology.

In summary, the clinical diagnosis of VP remains problematic. Currently, neither invasive nor non-invasive imaging technology can reliably identify VP prospectively, i.e., before rupture.

Return to Contents
Proceed to Next Section