Abstract
Introduction
Carotid and femoral plaque burden is a recognized biomarker of cardiovascular disease risk. A new electronic-sweep 3-dimensional (3D)–matrix transducer method can improve the functionality and image quality of vascular ultrasound atherosclerosis imaging. This study aimed to validate this method for plaque volume measurement in early and intermediate–advanced plaques in the carotid and femoral territories.
Methods
Plaque volumes were measured ex vivo in pig carotid and femoral artery specimens by 3-dimensional vascular ultrasound (3DVUS) using a 3D-matrix (electronic-sweep) transducer and its associated 3D plaque quantification software, and were compared with gold-standard histology. To test the clinical feasibility and accuracy of the 3D-matrix transducer, an experiment was conducted in intermediate–high risk individuals with carotid and femoral atherosclerosis. The results were compared with those obtained using the previously validated mechanical-sweep 3D transducer and established 2-dimensional (2D)–based plaque quantification software.
Results
In the ex vivo study, we assessed 19 atherosclerotic plaques (plaque volume, 0.76-56.30 μL), finding strong agreement between measurements with the 3D-matrix transducer and the histological gold-standard (intraclass correlation coefficient [ICC]: 0.992; [95% CI: 0.978-0.997]). In the clinical analysis of 20 patients (mean age 74.6 ± 4.45 years; 40% men), we found 64 (36 carotid and 28 femoral) of 80 scanned territories with atherosclerosis (measured atherosclerotic volume, 10-859 μL). There was strong agreement between measurements made from electronic-sweep and mechanical-sweep 3DVUS transducers (ICC: 0.997 [95% CI: 0.995-0.998]). Agreement was also high between plaque volumes estimated by the 2D and 3D plaque quantification software applications (ICC: 0.999 [95% CI: 0.998-0.999]). Analysis time was significantly shorter with the 3D plaque quantification software than with the 2D multislice approach with a mean time reduction of 46%.
Conclusions
3DVUS using new matrix transducer technology, together with improved 3D plaque quantification software, simplifies the accurate volume measurement of early (small) and intermediate–advanced plaques located in carotid and femoral arteries.
Introduction
Imaging-based biomarkers of subclinical atherosclerosis improve individual cardiovascular risk prediction compared with conventional risk scales based on clinical parameters.1,2 The most studied biomarker is the computed tomography–based coronary artery calcium score (CACS); nevertheless, ultrasound (US) is free of ionizing radiation and can detect plaques in large peripheral arteries from very early stages before calcification.3 The European Society of Cardiology guidelines now recommend US assessment of carotid and/or femoral atherosclerosis burden for cardiovascular risk evaluation,4,5 reflecting the ability of atherosclerosis burden measured by 3-dimensional vascular ultrasound (3DVUS) to predict individual cardiovascular risk,6 almost matching the prognostic performance of CACS.7,8 This good performance is likely because 3DVUS provides a more comprehensive evaluation of overall atherosclerosis burden, avoiding the drawbacks of 2DVUS, together with good reproducibility of plaque measurements.9
Several 3-dimensional (3D) approaches have been developed to simplify and standardize 3D image acquisition. The VL13-5 volumetric-linear array probe uses the “mechanical-sweep” method and generates accurate and reproducible measurements of carotid and femoral atherosclerosis burden from early to more advanced disease stages9,10 regardless of plaque size.11,12 However, the functionality of this approach is limited by the large probe footprint, which hinders examination of angulated surfaces or small fields of view. A new commercially available 3D vascular probe (XL14-3) based on matrix technology performs an “electronic-sweep” that improves image quality for the study of atherosclerosis, and the transducer’s smaller footprint makes it easier to manipulate during the US exploration. The 3D-matrix probe is supported by dedicated semiautomatic software that allows 3D analysis of the explored arterial segment. A previous report confirmed excellent interscan reproducibility for carotid atherosclerosis assessment13; however, the accuracy of the 3D-matrix probe for plaque burden quantification has not been established. More importantly, the new 3D-matrix transducer has not been tested previously for its ability to detect and quantify early atherosclerosis (plaques smaller than 69 μL), a cornerstone of primary prevention strategies, that is underestimated by old 3D methods because of technical limitations.10,14,15 In this study, we present the first validation of the electronic-sweep 3D-matrix transducer for accurate plaque volume quantification ex vivo in a pig model of atherosclerosis, with a focus on small plaques. In addition, we compared the ability of the 3D-matrix transducer to detect and quantify carotid and femoral plaques in patients with intermediate–advanced atherosclerosis with that of the previously validated mechanical-sweep VL13-5 volumetric-linear array transducer.
Discussion
Accurate assessment of cardiovascular disease risk would require precise quantification of individual subclinical atherosclerosis burden. In the present study, we demonstrate that 3DVUS with the novel XL14-3 3D-matrix probe and CM2020 software accurately quantifies plaque volume in a shorter time and shows high precision for the evaluation of early atherosclerotic plaques, a challenging scenario where risk stratification is of greatest value. In addition, electronic-sweep 3DVUS with the XL14-3 transducer accurately quantifies plaque burden in the carotid and femoral arteries, a recommended image-based biomarker for cardiovascular disease risk assessment in clinical guidelines.4,5
Due to poor axial resolution, first-generation 3D volume acquisition methods with external mechanical sweep do not accurately detect and quantify early (small) plaques,15 even for plaque volumes up to 69 μL.14 New methods for the assessment of subclinical atherosclerosis cannot be clinically validated without invasive procedures, the exposure of ostensibly healthy individuals to radiation (with computed tomography), or long scan times (with magnetic resonance). Animal models of atherosclerosis provide a helpful alternative, especially those that closely resemble the human disease. Previous studies support the use of artificial phantoms resembling plaques for validation purposes,12,15,19 prompting us to design a set of realistic phantoms using diseased porcine arterial specimens. This ex vivo analysis showed that plaque volume quantified by electronic-sweep XL14-3 probe closely matched that measured by gold-standard histology. A key feature of this analysis is that fresh specimens, although evaluated ex vivo, maintain their ultrasound features,20 thus allowing a realistic and detailed assessment of plaques that explores possible quantification biases arising from differences in plaque composition or shape. However, modern histological techniques of fixation and processing have been linked to changes in plaque features and slightly shrinkage-related changes of larger atherosclerotic plaques.21 This may explain the small proportional bias in plaque volumes obtained by electronic XL14-3 transducer ex vivo compared with histology. Although histological shrinkage cannot be excluded, the XL14-3 probe and histology estimates of plaque volume showed close agreement.
Advances in 3D probes have not only improved image quality and resolution, but have also simplified and standardized image acquisition, making procedures easier to implement in clinical practice. The XL14-3 probe is compact and has a smaller footprint than high-resolution mechanical-sweep 3D probes. This resulted in a slightly smaller acquired field-of-view (FOV) in electronic acquisitions with the XL14-3 probe than that expected for mechanical acquisitions with the VL13-5 probe. The overall excellent agreement in volume measurements between the 2 methods with standardized acquisition protocols may reflect the fact that the theoretical ≈6-cm FOV in the mechanical method translates into an actual variable 3- to 5-cm FOV available for plaque burden analysis, depending on artery depth and tortuosity. This is because the mechanical method uses a fan-like sweep that generates a pyramid-shaped 3D data set, producing a smaller FOV for superficial arteries than for deep arteries, as well as decreased image quality at greater depths and at the lateral image borders (Figure 2). Despite its smaller FOV, the small size of the XL14-3 probe facilitated scanning maneuvers in narrow and angulated areas, allowing more accurate diagnosis of challenging atherosclerotic lesions located in difficult (deep and tortuous) arteries and patient anatomies (short or profound neck or inguinal areas).
We used a standardized protocol centered at the carotid bulb or femoral bifurcation; this is a single-region protocol that acquires a selected region or vessel landmark with the 3D probe. This protocol has been shown to be highly reproducible9,13 and to facilitate the monitoring of changes in serial plaque burden evaluations,3 suggesting that 3DVUS methods would be especially appropriate for studies examining plaque progression (mechanistic studies, clinical trials of drug therapies, or lifestyle interventions). More importantly, limiting 3D studies to the evaluation of carotid bulb alone has been shown to reliably predict events,19 thus simplifying the assessment of atherosclerosis burden for cardiovascular risk estimation. On the other hand, several lines of evidence support the value of imaging the femoral arteries: 1) in young to middle-age individuals, femoral territory is more frequently affected than are the carotid territory or coronary arteries by CACS9; 2) femoral atherosclerosis shows a stronger association than carotid plaques with positive CACS, a surrogate marker of cardiovascular events,22 and femoral burden is strongly associated with the presence of significant coronary artery disease23,24; 3) atherosclerosis progression, also a surrogate marker of events, is more frequently detected in peripheral arteries by US than in the coronary arteries by CACS in middle-aged individuals3; and 4) more importantly, some prospective outcome-based studies have shown that the presence25,26 and the extent27 of both carotid and femoral plaques are associated with clinical cardiovascular events, independently of risk factors. In this regard, our study results will strengthen the case for multiterritorial assessment of plaque burden, because the procedure is equally feasible for the femoral and carotid territories. However, further research is needed to determine the added value of multiterritorial plaque burden assessment by VUS28 and whether the decision to assess both territories or only the carotids should be guided by patient age or disease stage.
The study of atherosclerosis by US has historically been hindered by limitations derived from the inherent characteristics of US technique. The most important of these limitations are as follows: 1) the poor detection of low-echogenic plaques and juxtaluminal black areas (JBAs; those with an echo-density similar to blood and a thin fibrous cap below the resolution of US); and 2) posterior acoustic shadowing from severe plaque calcification. The methodology validated in this study still has these limitations. To improve the study of low-echogenic plaques, 3D-power Doppler and 3D contrast-enhanced US technology is currently under development. These future methods will undoubtedly be of value for the detection of JBAs, and more importantly, they will allow exploration of new directions in the assessment of atherosclerosis-based risk markers by US (ie, the measurement of low-echogenic plaque volume and the combined study of 3D plaque burden and plaque characterization). Nevertheless, notwithstanding the inherent limitations of US, plaque burden measurement with current US methods predicts clinical cardiovascular events and has prognostic value. Further studies should seek to determine the actual effect of the lack of detection of JBAs with current 3DVUS technology for cardiovascular risk assessment. Regarding plaque calcification, in our study cohort, we found no significant limitation in plaque burden assessment caused by posterior acoustic shadowing, indicating that the new 3DVUS method can be used with confidence from early to mid-advanced disease stages. Our results align with those from the HRP study demonstrating the feasibility of carotid 3VDUS in asymptomatic intermediate- to high-risk 65- to 80-year-old individuals.7,8 However, previous studies of 3DVUS reported high drop-out rates caused by severe plaque calcification; these rates varied depending on the clinical context, ranging from 23% in patients undergoing revascularization of peripheral artery disease12 to 33% in patients with recent ischemic stroke.11 These findings suggest that 3DVUS performs worse in very-high risk or symptomatic individuals in whom severe plaque calcification co-occurs with stenotic lesions of complex morphology. Nevertheless, there is currently no consensus recommendation identifying which patients are suitable for plaque burden analysis by 3DVUS, and large prognostic studies will be needed to define the feasibility and prognostic value of 3DVUS in each age stratum and risk category.
The new CM2020 3D analysis software increased analysis speed while maintaining high intraobserver accuracy and reproducibility. This is mainly because CM2020 uses custom-made 3D algorithms that facilitate segmentation and simplify the manual correction of contours, thus significantly reducing analysis time, whereas VPQ is based on slice-by-slice 2D segmentation of the vessel. Also, 3D segmentation produces a 3D bifurcated model of the vessel that allows simultaneous analysis of the proximal segments of the 2 carotid and femoral branches, an approach not possible with VPQ. Altogether, these features make 3D US a simple, user-friendly, and radiation-free method with true potential to become a population screening tool for bedside cardiovascular risk assessment.
ARYAN'S BLOG 