Purpose: Recent evidence shows that blood biomarkers, in isolation, are largely ineffective in risk stratification of carotid atherosclerotic plaque (CS) symptomatic transformation. Stenosis degree is a poor marker of symptom risk while plaque morphology may play a role. Intravascular ultrasound (IVUS) provides high-resolution (axial ≤0.12mm for 20MHz transducer) imaging, but conventional phenotypic virtual histology (VH-IVUS) plaque classification poor between-center/-observer reproducibility limits any wider applicability. Moreover, conventional VH-IVUS, addressing total content of eg necrotic core (NC), does not discriminate focal vs. dissociated NC that is relevant to rupture risk. We investigated whether quantitative measures of key plaque components implicated in rupture risk are related to levels of several circulating biomarkers.
Methods: We developed a novel software-based algorithm for detailed, fully-quantitative VH-IVUS analysis of key plaque components known for their role in plaque rupture/thrombosis. In 21 plaques we validated inter-transducer (2 transducers) and inter-observer (3 observers) reproducibility of qVH-IVUS analysis including minimal fibrous cap (FC) thickness, and peak confluent NC area, thickness and arc. Next we employed our qVH-IVUS algorithm to evaluate CS lesions in 252 consecutive patients (age 47–83, 63.4% men, h/o CS-attributable symptoms in 50.3%) presenting for potential CS revascularization. Finally, in the first 200 subjects we determined the levels of a panel of biomarkers and preformed regression analysis in search for quantitative CS morphology/biomarker associations.
Results: qVH-IVUS revealed significant differences in minimal FC thickness (0.41±0.04 v 0.34±0.05 v 0.16±0.02 v 0.19±0.03mm), peak confl NC area (3.0±0.2 v 2.5±0.3 v 4.4±0.4 v 3.4±0.5mm2), arc (87.1±6 v 67.2±6 v 121.6±9 v 94.0±9deg) and thickness (0.88±0.04 v 1.07±0.07 v 1.34±0.06 v 1.16±0.11 mm); data for asympt CS in absence of contralat symptoms, asympt CS in presence of contralat symptoms, recently symptomatic and remotely symptomatic CS, p<0.001 for all.
While hsCRP was not correlated with min FC (r=-0.24, p=0.74) or NC area (r=0.05, p=0.48), TIMP correlated with min FC (r=0.34, p=0.001). Modest though highly significant correlations were identified between Lp-PLA2 and confl NC area (r=0.3, p=0.0001), HDL and confl NC thckn (r=-0.21, p=0.002). Fibrinogen level correlated with % plaque fibrotic content (r=0.19, p=0.008).
Conclusion: These findings provide novel insights into circulating biomarker/quantitative plaque morphology associations that may be relevant to plaque biology and risk.