Use this URL to cite or link to this record in EThOS:
Title: The clinical impact of intravascular ultrasound derived virtual histology on percutaneous coronary intervention
Author: Murray, S. W.
ISNI:       0000 0004 6059 2909
Awarding Body: University of Liverpool
Current Institution: University of Liverpool
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
It has been shown that early interventional treatment of patients with high risk acute coronary syndromes (ACS) has a favourable effect on mortality. It is also known that coronary plaque rupture and atherothrombosis creates a mileu of necrotic and thrombotic material, which is difficult to treat. Moreover, angiographic assessment of coronary artery disease is highly flawed; 2-dimensional luminal silhouettes are not ideal templates to guide very important interventions. Intravascular ultrasound (IVUS) is the gold standard imaging tool able to delineate vessel dimensions, plaque burden, length and now with virtual histology (VH) - plaque composition. Despite optimal medical care and urgent revascularisation, 12-20% of ACS patients will suffer a further major adverse cardiac event (MACE) at 30 months. Our aim was to evaluate the angiographic treatment of high risk ACS patients by performing IVUS-VH pre and post-intervention, with the operators blinded to the images. Our hypotheses for this work were as follows: 1. Significant compositional and structural differences exist between culprit, non-culprit and stable plaques when analysed by Virtual Histology. 2. The histologically most unstable plaque does not occur at the site of maximum angiographic stenosis (in culprit lesions). 3. Angiographically guided stent length selection and positioning is flawed, leaving unstable plaque behind in the reference segments. Following recruitment, 135 lesions split into: 70 ACS culprit; 20 ACS non-culprit and 35 Stable lesions underwent analysis for inter-observer and intra-observer variability. We were able to show good standard markers of correlation but a large repeatability co-efficent, for some outputs from the analysis. This has raised questions with regard to the ability of the technique to detect differences between plaque types. In relation to our main hypothesis, we have been able to show structural and compositional volume differences between “active” ACS plaques (n=70) and “stable” angina plaques (n=35). We have used the most important of these to generate a plaque risk score based upon ROC statistics and logistic regression. The most important discriminators were: Remodelling index at the minimum lumen area; Plaque Burden; Presence of VH-TCFA; minimum lumen area (MLA) < 4mm2 and necrotic core to dense calcium ratio (NC/DC). The subsequent risk model was tested on an independent, blinded cohort of plaques from the Thoraxcentre, Rotterdam (n=50). This confirmed good discriminatory power for the equation (AUC – 0.71). Within this hypothesis we also explored the differences in individual areas of plaque. At two separate sites (MLA and MAX NC) in each lesion type (ACS and stable) n=210, we showed that the MAX NC site lies proximal to the MLA in most cases and contains more positively remodelled plaque disease with less calcification. This is important as positively remodeled plaque disease is often not visible on a plain coronary angiogram when treating ACS lesions. Finally, as a follow on from thie previous chapter, we examined the treatment of ACS lesions by blinded IVUS examination. The operator completed their stent procedure with only angiographic guidance. We were able to show in 56 ACS lesions that systematic errors of judgement occur related to sizing of the vessel, the choice of stent sizes and the subsequent stent deployment. 36%, 40% and 65.5% of stents met three separate standard criteria for good stent deployment. Moreover, between 5-40% of stents had some form of significant abnormality. With regard to the sizing of stents to the vessel, the mean reference vessel size was 10.58mm2 (±2.51) yet the minimum stent area achieved was only 6.79mm2 (±2.43). If the stent that was chosen had been symmetrically deployed to its nominal size (e.g 2.5; 3.0; 3.5) then the stent area achievable should have been 8.87mm2 (±2.68). This has allowed us to calculate for the first time an estimated “under deployment area”. This was 2.08mm2 (±1.87).
Supervisor: Stables, R. H. ; Palmer, N. D. Sponsor: Not available
Qualification Name: Thesis (M.D.) Qualification Level: Doctoral