2083
Identification of Carotid Non-Hemorrhagic Lipid-Rich Necrotic Core by Magnetization-Prepared Rapid Acquisition Gradient-Echo Imaging: Validation by Contrast-Enhanced T1W Imaging1Center for Biomedical Imaging Research, Department of Biomedical Engineering, Tsinghua University School of Medicine, Beijing, China, 2Department of Radiology, University of Washington, Seattle, WA, United States, 3Beijing Tiantan Hospital, Capital Medical University, Beijing, China, 4China National Clinical Research Center for Neurological Diseases, Beijing, China
Synopsis
Lipid-rich necrotic core (LRNC) plays a key role in the vulnerability of atherosclerotic plaques and is associated with ischemic cerebrovascular events. Previously, LRNC was identified using contrast-enhanced T1W (CE-T1W) imaging which needs administration of contrast agent. This study determined the capability of MP-RAGE imaging in identification of non-hemorrhagic LRNC (NH-LRNC) validated by CE-T1W. We found that moderate to good agreements between MP-RAGE and CE-T1W imaging in identification and quantification of NH-LRNC. MP-RAGE showed smaller bias in measuring NH-LRNC than T2W imaging. Our results suggest that MP-RAGE might be a better non-CE imaging technique for assessing NH-LRNC in carotid arteries.
Introduction
Lipid-rich necrotic core (LRNC) plays a key role in the vulnerability of atherosclerotic plaques. It has been demonstrated that the presence and size of LRNC are significantly associated with ischemic cerebrovascular events 1. Therefore, characterization of carotid LRNC prior to plaque rupture is important. Multi-contrast magnetic resonance (MR) vessel wall imaging techniques have been largely utilized to detect LRNC in carotid arteries 2-4. In particular, contrast-enhanced T1W (CE-T1W) imaging is considered as the “gold standard” of vivo imaging technique for characterizing LRNC which is better than T2W imaging 3. Most recently, 3D magnetization-prepared rapid acquisition gradient-echo (MP-RAGE) imaging has been demonstrated to have the potential in identification of carotid non-hemorrhagic LRNC (NH-LRNC) 5. This study aims to determine the capability of 3D MP-RAGE imaging in evaluating carotid artery NH-LRNC validated by CE-T1W imaging.Methods
Study sample: Fifty-one asymptomatic patients (mean age, 61.7 ± 11.7 years; 40 males) with carotid atherosclerotic plaques on ultrasound were included in this study. All recruited subjects underwent carotid multicontrast MR vessel wall imaging. MR imaging: Carotid MRI was performed on a 3.0T MR scanner (Achieva TX, Philips Healthcare, The Netherlands) with a dedicated 4-channel carotid coil (Machnet BV, Roden, Netherlands). The MR imaging protocol included TOF, T1W, T2W, CE-T1W, and MPRAGE imaging sequences. The imaging parameters are detailed (Table 1): TOF: turbo field echo (TFE); TR/TE 20/5 ms; flip angle 20°; T1W: TSE; TR/TE 800/10 ms; flip angle 90°; T2W: TSE; TR/TE 4800/50 ms; flip angle 90°; CE-T1W: TSE; TR/TE 800/10 ms; flip angle 90°; and MPRAGE: TFE; TR/TE 9/5.5 ms; flip angle 15°. The in-plane FOV and spatial resolution was 140×140 mm2 and 0.6×0.6 mm2 for all sequences. Image analysis: The presence or absence and the size of carotid LRNC were evaluated by two reviewers with >5 years’ experience in neurovascular MRI with an interval of one month and consensus using 3D CASCADE software respectively on the following 3 different MR sequence combinations: 1) TOF, T1W, and T2W; 2) TOF, T1W, and MP-RAGE; and 3) TOF, T1W, and CE-T1W. NH-LRNC was defined as components with iso-intense on TOF and T1W images and hypointense on T2W, MP-RAGE (66% signal drop compared with muscle 5) and CE-T1W images using published criteria 2-3. Statistical analysis: The agreement of MP-RAGE and T2W imaging in identification of NH-LRNC with CE-T1W imaging was evaluated using Cohen’s kappa analysis. The area of NH-LRNC was compared between MP-RAGE and CE-T1W imaging and between T2W and CE-T1W imaging using paired t test. The intraclass correlation coefficient (ICC) and Bland-Altman were utilized to assess the agreement between MP-RAGE and T2W and CE-T1W imaging in measuring NH-LRNC when present. A p value <0.05 was considered as statistically significant.Results
In total, 1894 slices from 51 subjects were included in this study. Of 1894 slices, 582 (30.7%), 448 (23.7%), 551 (29.1%) slices were found to have NH-LRNC on MP-RAGE, T2W, and CE-T1W images, respectively. Moderate agreement was found between MP-RAGE and CE-T1W imaging (kappa=0.52) and between T2W and CE-T1W imaging (kappa=0.59) in identification of NH-LRNC (Table. 1). For NH-LRNC detected by both MP-RAGE and CE-T1W imaging, no significant difference was found in the area of NH-LRNC between MP-RAGE and CE-T1W imaging (6.1 ± 6.1 mm2 vs. 6.0 ± 6.6 mm2, p=0.628). In contrast, the area of NH-LRNC measured by T2W imaging was significantly smaller than that measured by CE-T1W imaging (5.5 ± 5.9 mm2 vs. 6.1 ± 6.6 mm2, p=0.023). Good agreement can be observed in quantification of NH-LRNC between MP-RAGE and CE-T1W imaging (ICC=0.774, 95% CI 0.723-0.815) and between T2W and CE-T1W imaging (ICC=0.791, 95% CI 0.742-0.831). Bland-Altman analysis revealed that the bias of measurement of NH-LRNC by T2W was greater than that measured by MP-RAGE as compared with CE-T1W imaging (Fig. 1). An example that MP-RAGE successfully identified carotid LRNC was shown in Fig. 2.Discussion and Conclusions
This study determined the capability of MP-RAGE imaging in identification of NH-LRNC compared with T2W imaging validated by CE-T1W imaging. Moderate to good agreements were found in identification and quantification of NH-LRNC between MP-RAGE and CE-T1W imaging. MP-RAGE had smaller bias in measuring the size of NH-LRNC compared to T2W imaging. In addition, we found that MP-RAGE detected more NH-LRNCs compared with T2W and CE-T1W imaging, suggesting that MP-RAGE might be a more sensitive imaging techniques for lipid-rich atherosclerotic plaques. In conclusion, MP-RAGE sequence might be a better non-contrast enhanced imaging technique than T2W imaging for assessing non-hemorrhagic lipid-rich necrotic core in carotid arteries.Acknowledgements
This work was supported by grants of Beijing Municipal Science and Technology Project (D171100003017003) and National Natural Science Foundation of China (81271536; 81361120402).References
1. Zavodni AE, Wasserman BA, McClelland RL, et al. Carotid artery plaque morphology and composition in relation to incident cardiovascular events: the Multi-Ethnic Study of Atherosclerosis (MESA). Radiology. 2014;271:381-389.
2. Trivedi RA, U-King-Im JM, Graves MJ, et al. MRI-derived measurements of fibrous-cap and lipid-core thickness: The potential for identifying vulnerable carotid plaques in vivo. Neuroradiology. 2004;46:738-743.
3. Cai J, Hatsukami TS, Ferguson MS, et al. In vivo quantitative measurement of intact fibrous cap and lipid-rich necrotic core size in atherosclerotic carotid plaque: Comparison of high-resolution, contrast-enhanced magnetic resonance imaging and histology. Circulation. 2005;112:3437-3444.
4. Takaya N, Cai J, Ferguson MS, et al. Intra- and interreader reproducibility of magnetic resonance imaging for quantifying the lipid-rich necrotic core is improved with gadolinium contrast enhancement. J Magn Reson Imaging. 2006;24:203-210.
5. Qiao H, Li F, Xu D, et al. Identification of carotid lipid-rich necrotic core and calcification by 3D magnetization-prepared rapid acquisition gradient-echo imaging. Magn Reson Imaging. 2018;53:71-76.
Figures
