OUP user menu

Beyond greyscale IVUS assessment of progression/regression: it should be simple, but it's not

Gary S. Mintz
DOI: http://dx.doi.org/10.1093/ehjci/jet291 378-379 First published online: 21 January 2014

In this issue of European Heart Journal — Cardiovascular Imaging, Puri et al. report the serial virtual histology (VH)–intravascular ultrasound (IVUS) data from SATURN (Study of Coronary Atheroma by Intravascular Ultrasound [IVUS]: Effect of Rosuvastatin versus AtorvastatiN) that compared the efficacy of two potent statins, each prescribed at the highest doses. A pre-specified exploratory analysis combined subsets of patients from both randomization groups to assess serial changes in VH-IVUS data.1

There are limitations to greyscale IVUS analysis of atherosclerosis progression/regression. (i) As anti-atherosclerotic therapies have improved, the differences between therapies—especially, the effects on changes in absolute or percent atheroma volume—have almost disappeared. (ii) Baseline atheroma burden is a stronger determinant of clinical events than progression/regression.2,3 (iii) The clinical impact of anti-atherosclerotic therapies on events, especially the hard endpoints of death and myocardial infarction, is greater than the impact on atheroma progression/regression, suggesting that protective effects may result from compositional changes rather than mere reduction in atheroma burden.

Promises of second-generation intravascular imaging techniques—VH-IVUS (the technique used in the current study), integrated backscatter (IB)-IVUS, iMAP (two other ultrasound techniques incorporating radiofrequency analysis), optical coherence tomography (OCT, the high-resolution light analogue of IVUS), near-infrared spectroscopy (NIRS), and endothelial shear stress (ESS)—include the ability to detect a vulnerable plaque, predict clinical events, and better discriminate among anti-atherosclerotic therapies. However, all of these techniques have limitations. For one, while all of these techniques have been validated in vitro, in vivo compositional analysis of the same patient may differ among various techniques.

  1. VH-IVUS, IB-IVUS, and iMAP can assess tissue composition in addition to detecting atheroma burden. However, they are confounded by thrombus and do not have the resolution to measure fibrous cap thickness, the ECG-gating necessary with VH-IVUS makes precise co-registration of serial studies difficult, and the changes in plaque composition (most of which have been with VH-IVUS) have been inconsistent with the current study of Puri et al.1 showing a decrease in fibrofatty plaque versus a meta-analysis of other studies showing a decrease in the necrotic core.4 Nevertheless, VH-IVUS has been shown to predict events in three prospective registries: PROSPECT (Providing Regional Observations to Study Predictors of Events in the Coronary Tree), VIVA (VH-IVUS in Vulnerable Atherosclerosis), and the ATHEROREMO (European Collaborative Project on Inflammation and Vascular Wall Remodeling in Atherosclerosis)-IVUS substudy.57

  2. OCT can measure fibrous cap thickness; it can also detect, but not quantify lipid and macrophage infiltration of the fibrous cap, especially changes in lipid and macrophage infiltration; and it cannot assess atheroma burden. In one study, statin use was shown to increase fibrous cap thickness;8 and in another study, an OCT-detected thin-cap fibroatheroma (TCFA) and microchannels were predictors of angiographic progression.9

  3. NIRS can identify and quantify lipid-rich plaque. Aggressive statin therapy (compared with less aggressive statin therapy) reduced the amount of NIRS-detectable lipid-rich plaque in the YELLOW (reduction in yellow plaque by aggressive lipid-lowering therapy) study.10 The ability of NIRS to predict events is being tested in PROSPECT-II.

  4. In the PREDICTION (Prediction of Progression of Coronary Artery Disease and Clinical Outcome Using Vascular Profiling of Shear Stress and Wall Morphology) study, an increase in the plaque area (primary end point) was predicted by a large baseline plaque burden; and a decrease in the lumen area (secondary end point) was predicted by both a large baseline plaque burden and low ESS.11

Therefore, none of these techniques is perfect; no one technique provides comprehensive atherosclerosis assessment; and each has technical, methodological, and/or practical limitations. The logical conclusion is either to (i) perform multimodality imaging in a single clinical study that, while possible, can be a real challenge in large numbers of patients for clinical trialists and their interventional colleagues, core laboratories, and ethics committees, or (ii) develop combination devices, a technical feat that has only recently been realized in a combined NIRS/IVUS catheter.

However, there are reasons why even a multimodality imaging approach might be frustrating in attempting to correlate anti-atherosclerotic therapy with lesion stabilization and lesion stabilization with better clinical outcomes. Atherosclerotic lesions that cause events are either TCFAs, the most common type of vulnerable plaque, erosions, or rarely calcified nodules. TCFAs are focal manifestations of coronary atherosclerosis. Contrary to a New York Times editorial of 28 March 2004 that stated ‘[Patients] may have hundreds of vulnerable plaques elsewhere that are more apt to burst and trigger a heart attack than are the more stable plaques in the narrow section,’ coronary arteries of high-risk patients contain only a few vulnerable plaques;12,13 this impacts the likelihood that an arbitrarily selected and imaged segment of one moderately diseased coronary artery will contain even a single vulnerable plaque. Even pre-specified three-vessel imaging detected only half of the plaques that caused events in PROSPECT.5 Erosions have a different underlying plaque composition compared with lipid-rich TCFAs and are difficult to identify in vivo. TCFAs, especially in stable patients, can ‘heal’ spontaneously, whereas those in unstable patients are more likely to remain unstable or to develop spontaneously during follow-up.14,15 When patients are entered into clinical trials, are treated with modern medical therapy, and are monitored closely, the event rate is typically lower than anticipated, events are more often revascularization than the hard events of death or myocardial infarction, and the relationship between the segment of artery imaged and a particular revascularization event may be difficult to adjudicate.2,57,11 Finally, the optimal duration of serial studies is far from clear ranging from 8 weeks in the YELLOW to 2 years in the SATURN; and with some techniques, there is limited data regarding variability as well as limited data on which to base sample size calculations.

While assessing compositional changes in atherosclerosis may seem like the next logical step in progression/regression research, clinical investigators and their corporate sponsors face a true dilemma in selecting the optimal imaging modality, the primary endpoint, the sample size, and the duration of the study to best detect and quantify disease progression/regression as well as assess plaque composition, vulnerability, and stability, especially as a surrogate for clinical events.