Upcoming events

Usefulness of coronary atheroma burden to predict cardiovascular events in patients presenting with acute coronary syndromes (from the PROSPECT study)

Selected in The American Journal of Cardiology by D. Milasinovic



Shan P, Mintz GS, McPherson JA, De Bruyne B, Farhat NZ, Marso SP, Serruys PW, Stone GW, Maehara A


Am J Cardiol. 2015 Dec 1;116(11):1672-7


December 2015


Read the abstract

My Comment

What is known

Coronary atheroma burden, as measured by the intravascular ultrasound (IVUS)-derived percent atheroma volume (PAV), was shown to be predictive of major cardiovascular adverse events (MACE). However, most of the previously studied patients had stable angina and the occurrence of MACE was not related to the imaged coronary tree segment.

A prospective natural-history study of coronary atherosclerosis (PROSPECT trial) that included a three-vessel IVUS after successful percutaneous coronary intervention (PCI) in patients with acute coronary syndrome (ACS), showed that the occurrence of MACE during 3-year follow-up was equally related to both initially treated and untreated lesions (12.9% vs. 11.6%, respectively).

Research question

This secondary, patient-level analysis of the PROSPECT trial investigated the impact of total baseline PAV on the rate of 3-year nonculprit lesion-related MACE.


  • 660 ACS patients were included (30% STEMI, 66.2% NSTEMI, 3.8% unstable angina), with 3229 IVUS-defined nonculprit lesions that were untreated based on angiographic findings.
  • Nonculprit lesion was IVUS-defined as plaque burden (PB) ≥40% on ≥3 consecutive frames.
  • External elastic membrane (EEM) and lumen areas were determined at ≈0.4mm intervals.
  • Overall atheroma burden was calculated by summing all nonculprit lesions on a patient level: PAV = (∑ (EEM area – lumen area) / ∑ (EEM area) x 100.
  • MACE was defined as cardiac death, cardiac arrest, myocardial infarction (MI) and/or rehospitalization for unstable/progressive angina, unrelated to the initially treated lesions.

Major findings

  • Cumulative 3-year nonculprit lesion-related MACE was significantly less frequent in the lowest patient tertile (PAV <47.7%) compared to both intermediate (47.7%≤ PAV <50.9%) and high (PAV ≥50.9%) tertiles (6.3% vs. 14.7% vs. 15.1%, respectively; p=0.009).
  • This difference in MACE was driven mainly by a higher rate of progressive angina in patients in high (12.2%) and intermediate (10.7%) versus low (3.1%) baseline PAV tertiles (p=0.003).
  • Multivariable regression found insulin-dependent diabetes mellitus, presence of ≥1 thin-cap fibroatheroma by virtual histology (VH-TCFA) and PAV to be independent MACE predictors.
  • The combined presence of at least two of the three high-risk plaque characteristics (minimal lumen area (MLA) ≤4 mm2, VH-TCFA and/or PB ≥70%) was more frequent in  patients in intermediate and high, compared to low PAV tertile (p<0.001 for every combination).
  • By multivariable linear regression, male gender, age and prior cardiac intervention were independently associated with an increase in baseline PAV.

My comment

This study reaffirmed the association between baseline atheroma burden and MACE, with the added value of specifically relating MACE to angiographically mild, previously untreated coronary lesions. Higher degree of baseline atheroma burden was associated mainly with more rehospitalizations due to progressive angina and subsequent revascularizations. Although the low overall rates of hard clinical endpoints, such as death or MI, in this and also previous studies, preclude a definitive conclusion regarding the clinical relevance of total baseline atheroma burden as expressed by PAV, the following two discussion points seem warranted:

  1. In ACS patients, the discrepancy between angiographic findings of mild nonculprit stenoses and IVUS-measured high degree of total atheroma burden may be associated with incomplete revascularization and consequently worse outcomes;   
  2. The mechanisms behind the observed association of PAV and MACE need to be further elucidated and may include high total baseline PAV being a marker of disease progression and/or related to the presence of high-risk plaque characteristics. 


  • Alexander Kharlamov 21 Dec 2015

    To my personal understanding the title of the paper doesn't properly reflect the content. Definitely, PB (plaque burden) matters when below 40% and above 70%. First, I would remind to other colleagues that there are two essentially different parameters to measure with IVUS: TAV (total atheroma volume) and PAV (per cent atheroma volume). This is not the same, and by the plaque burden authors mean exceptionally PAV. Actually they merely clarify some tendencies we already knew from the first report of the PROSPECT study. First, they enrolled patients with PB&gt;40% which means we are talking about patients with progressive atherosclerosis and potentially about vulnerable population. It was already known that PB&gt;70% is an independent predictor of nonculprit MACE. This is not a surprise, and in that case we mean truly advanced lesions. For me personally a question is does PB really matter when it's below 40%. From the statistical point of view extrapolating it to the pathology of atherosclerosis the PB of 40-70% must be crucial for awareness of the physicians in sense of any acute CV events, but this is truly tough to predict something just with a per cent of PB. Another question if we see that the PB is decreasing during any lipid-lowering strategy, or intervention can in that case PB predict anything?!

  • Dejan Milasinovic 11 Jan 2016

    Thank you for this interesting comment. There seem to be at least three distinct issues that are being addressed. First, what is the IVUS-derived measure of atheroma burden that is more closely associated with clinical outcomes, PAV or TAV? Whereas total atheroma volume (TAV) expresses absolute plaque volume and is based on the summation of plaque areas [∑(EEMarea - Lumenarea)], the percent atheroma volume (PAV) takes into account arterial wall remodeling and expresses the proportion of the summed EEM areas occupied by plaque [∑(EEM area – lumen area)/∑(EEM area) x100]. A previous study showed that baseline PAV was more consistently associated with clinical outcomes, as compared with TAV (Nicholls SJ et al. J Am Coll Cardiol. 2010 May 25;55(21):2399-407). Although no comparative analysis of the predictive capacities of TAV vs. PAV was presented in the above reviewed PROSPECT substudy, the potential of PAV to predict MACE was confirmed. Second, what could be the value of PAV if the definition of a nonculprit lesion (PB ≥40%) already presupposed higher risk of MACE? The study showed that given baseline PB ≥40% in all patients, those with low PAV had significantly lower risk of nonculprit lesion-related MACE, compared to both high and intermediate PAV tertiles. This finding seems to suggest that patient-specific PAV, which approximates the combined plaque burden of all IVUS-imaged lesions and thus takes into account the diffuse nature of coronary artery disease, may provide added value to PB of a focal nonculprit lesion alone. Third, what would be the predictive value of PAV if IVUS-measured atheroma burden were responsive to lipid-lowering medication? A previous study suggested that baseline PAV remains an independent MACE predictor, even in the presence of an effective lipid-lowering therapy (Puri R et al. Eur Heart J. 2013 Nov;34(41):3182-90). In summary, it seems that in ACS patients, IVUS-derived measures of the overall atheroma burden may offer more insight into the diffuse nature of coronary atherosclerosis and provide added predictive value to angiographic assessment of focal nonculprit stenosis alone. How is it done in your hospital? Please feel free to comment on any instructive cases of IVUS-based assessment of nonculprit lesion progression in ASC patients.