Multi-target Vulnerable Plaque Imaging

The search for the vulnerable plaque has been a prolonged, circuitous journey that despite considerable investment and resource expenditure has hitherto failed to alter clinical practice. The rationale remains enticing. We have known for some time that plaques at risk of rupture and precipitating myocardial infarction and stroke have certain characteristics that extend beyond simple stenosis severity. These include inflammation, microcalcification, angiogenesis, positive remodelling, a thin fibrous cap, a large necrotic core and plaque hemorrhage. Rapid recent advances now allow us to detect such features using both invasive and non-invasive imaging technology ¹. The hope has been that such technology might allow us to predict with accuracy the plaques and more importantly the patients at risk of rupture-related events. However recent data has been somewhat disappointing, at least at the plaque level. The PROSPECT trial used virtual histology intravascular ultrasound (VH-IVUS) to identify high-risk lesions (the VH-IVUS defined thin capped fibroatheroma), demonstrating that such plaques were a predictor of events and offered good negative predictive value. However out of a total of 595 such lesions only 26 caused an event and most of these were not myocardial infarctions ². Whilst this in part may reflect the limitations of VH-IVUS it is increasingly being appreciated that high-risk plaques are in fact relatively common and only rarely go on to cause a clinical event. Indeed the positive predictive value is currently so low that questions have been raised about whether treatment to individual plaques will ever be justified ³. Faced with this fundamental issue what then is the future of high-risk plaque imaging?

There is more hope at the patient level. Where present high-risk plaques rarely exist in isolation but tend to form at multiple sites across the vasculature in patients with an active inflammatory disease process. Whilst the majority of these plaques will heal without consequence, it is believed that patients with active disease and a tendency to forming unstable lesions will be at increased risk of a clinical plaque rupture event ⁴. Recent data supports this hypothesis. T1 weighted MR imaging can identify plaques with intraplaque hemorrhage or intraluminal thrombus. Patients with evidence of such plaques in the carotid vasculature had a 3-fold increase in cardiac events ⁵. Similarly patients with high-risk plaques in their coronary arteries detected using the same MR approach⁶ as well as CT plaque characterisation have also been found to have an adverse prognosis ⁷.

Whilst more encouraging there remains an urgent need to improve the field of vulnerable plaque imaging so that it might at last make a useful contribution to clinical practice. This will require extensive scientific endeavour. In particular improved understanding of what makes a plaque truly vulnerable, how the array of different high-risk features are related, which particular characteristics offers the best risk prediction and how their presence evolves over time.

The manuscript by Demeure et al. in the current issue of Circulation Cardiovascular Imaging attempts to address some of these key issues ⁸. The authors investigated both carotid inflammation and angiogenesis using FDG PET and contrast-enhanced ultrasound respectively, performed in patients undergoing carotid endarterectomy (11 were symptomatic and 19 asymptomatic). The contrast-enhanced ultrasound used a contrast agent that remains intravascular and so informs about new vessel formation rather than vessel leakiness. Imaging data were then compared with histological analysis of the excised carotid atheroma. This provided both validation of the imaging findings and an alternative assessment of the presence of inflammation and angiogenesis in the different plaque types. A good correlation between FDG uptake and macrophage burden was observed as previously reported. Similarly patients thought to have angiogenesis on ultrasound had a higher vessel density on histology than those that did not, although the binary nature of this assessment does represent a limitation of this technique.

The key finding of the study was that inflammation (assessed using both FDG PET and macrophage burden on histology) was increased in symptomatic versus asymptomatic patients despite similar plaque burden. This supports previous data. The same however was not true of angiogenesis, whether assessed by imaging or histology. This suggests that inflammation and angiogenesis may not be as closely related as previously thought and that angiogenesis may not be such a high-risk marker, at least in the absence of associated plaque haemorrhage.

The authors are therefore to be congratulated on advancing our knowledge in the field. Moreover they deserve credit for delivering such a trial. Performing complex imaging protocols in the increasingly narrow window between symptom development and endarterectomy is a major logistical challenge that should be recognized. Such endeavors are nevertheless worth pursuing, providing a rare opportunity to directly correlate imaging findings with histology and to compare symptomatic with asymptomatic plaques.

So where next for the field? Prospective studies are now required to assess whether high-risk plaque imaging and measures of disease activity can provide incremental prognostic information and whether this might be cost-effective. These should also address which patient populations would derive maximal benefit and whether combinations of different plaque characteristics add value. A recent study by Motoyama et al supported the latter notion demonstrating that coronary plaques with both positive remodeling and a large necrotic core on CT identified patients at increased future risk of cardiac events ⁷. Other PET/CT trials underway will develop this theme further and examine whether PET measures of coronary disease activity are complementary to CT plaque characterisation and associated measures of plaque burden (clinicaltrials.gov NCT02278211).

Ultimately atherosclerosis is a highly complex multi-faceted disease process in which risk prediction is challenging. Combined assessments of plaque burden, plaque characteristic and disease activity are likely to be required to accurately identify patients at imminent risk of myocardial infarction and stroke. Further innovative research of the kind presented by Demeure et al. will be required if this concept is going to be translated in to clinical reality.