Residual Inflammatory Risk and Intracranial Atherosclerosis Plaque Vulnerability: Insights From High-Resolution Magnetic Resonance Imaging
Article information
Abstract
Background and Purpose
This study aimed to investigate the association between residual inflammatory risk (RIR) and vulnerable plaques using high-resolution magnetic resonance imaging (HRMRI) in symptomatic intracranial atherosclerotic stenosis (ICAS).
Methods
This retrospective study included 70%–99% symptomatic ICAS patients hospitalized from January 2016 to December 2022. Patients were classified into four groups based on high-sensitivity C-reactive protein (hs-CRP) and low-density lipoprotein cholesterol (LDL-C): residual cholesterol inflammatory risk (RCIR, hs-CRP ≥3 mg/L and LDL-C ≥2.6 mmol/L), RIR (hs-CRP ≥3 mg/L and LDL-C <2.6 mmol/L), residual cholesterol risk (RCR, hs-CRP <3 mg/L and LDL-C ≥2.6 mmol/L), and no residual risk (NRR, hs-CRP <3 mg/L and LDL-C <2.6 mmol/L). Vulnerable plaque features on HRMRI included positive remodeling, diffuse distribution, intraplaque hemorrhage, and strong enhancement.
Results
Among 336 included patients, 21, 60, 58, and 197 were assigned to the RCIR, RIR, RCR, and NRR groups, respectively. Patients with RCIR (adjusted odds ratio [aOR], 3.606; 95% confidence interval [CI], 1.346–9.662; P=0.011) and RIR (aOR, 3.361; 95% CI, 1.774–6.368, P<0.001) had higher risks of strong enhancement than those with NRR. Additionally, patients with RCIR (aOR, 2.965; 95% CI, 1.060–8.297; P=0.038) were more likely to have intraplaque hemorrhage compared with those with NRR. In the sensitivity analysis, RCR (aOR, 2.595; 95% CI, 1.201–5.608; P=0.015) exhibited an additional correlation with an increased risk of intraplaque hemorrhage.
Conclusion
In patients with symptomatic ICAS, RIR is associated with a higher risk of intraplaque hemorrhage and strong enhancement, indicating an increased vulnerability to atherosclerotic plaques.
Introduction
Intracranial atherosclerotic stenosis (ICAS) is the predominant cause of global ischemic stroke, carrying a high risk of stroke recurrence despite optimal medical treatment, with a recurrence rate exceeding 20% [1-4]. The presence of vulnerable plaques, mainly characterized by plaque rupture, can potentially escalate subsequent recurrent events in symptomatic ICAS, a process that may be mediated and regulated by inflammation [5-9]. The residual inflammatory risk (RIR), which considers the effects of both high-sensitivity C-reactive protein (hs-CRP) and low-density lipoprotein cholesterol (LDL-C) on atherosclerosis progression, has recently been extensively investigated in the cardiovascular field as a novel inflammatory marker [10,11]. Preliminary exploration of RIR in patients with acute ischemic stroke and transient ischemic attack (TIA), based on data from the Third China National Stroke Registry (CNSR-III), indicates that RIR can predict stroke recurrence, especially among those with large-artery atherosclerosis [12,13]. The potential role of vulnerable atherosclerosis plaque in this strong relationship between RIR and stroke recurrence remains understudied.
High-resolution magnetic resonance imaging (HRMRI) can help visualize the morphology and composition of intracranial plaque and assist in the evaluation of plaque characteristics [14,15]. Intraplaque hemorrhage, strong plaque enhancement, positive remodeling, and diffuse distribution have been identified as features of vulnerable plaques on HRMRI [16-19]. A recent study demonstrated that elevated hs-CRP levels were significantly associated with plaque enhancement on HRMRI in patients with symptomatic ICAS [20]. However, there are currently limited studies concerning the relationship between residual cholesterol inflammatory risk (RCIR), RIR, and residual cholesterol risk (RCR), and vulnerable plaques in symptomatic ICAS. We investigated the association between residual risk and vulnerable intracranial atherosclerotic plaques in patients with symptomatic high-grade ICAS using HRMRI.
Methods
Study design and population
This retrospective study was conducted in a high-volume stroke center and consecutively enrolled patients with 70%–99% symptomatic ICAS hospitalized from January 2016 to December 2022, which was approved by the Institutional Ethics Committee of Beijing Tiantan Hospital (approval number: KY2021-075-02). Written informed consent was obtained from all the patients or their guardians. Patients who underwent HRMRI of the culprit plaque and concurrent laboratory testing, including hs-CRP and LDL-C, were screened. Baseline data (including sex, age, National Institutes of Health Stroke Scale [NIHSS] score, comorbidities, qualifying events, and culprit plaque location) were collected by trained professionals.
Patients who met the following criteria were included: (1) Age ≥18 years. (2) 70%–99% stenosis of the intracranial artery (including intracranial segments of internal carotid artery, middle cerebral artery, intracranial segment of vertebral artery, and basilar artery). The stenotic degree of the culprit artery was measured using the Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) method with computed tomography angiography or digital subtraction angiography [21]. (3) Ischemic stroke or TIA within 3 months. Ischemic stroke was defined as a new focal neurological deficit lasting over 24 hours or less than 24 hours with a new infarction on imaging. TIA was defined as the acute onset of a focal neurological deficit lasting less than 24 hours without a new infarction on imaging. (4) One or more risk factors contributing to atherosclerosis, including hypertension, hyperlipidemia, diabetes mellitus, and tobacco use. The exclusion criteria were as follows: (1) non-atherosclerotic intracranial stenosis, such as moyamoya disease, vasculitis, dissection, or fibromuscular dysplasia; (2) cardioembolism; (3) concurrent intracranial aneurysm, tumor, or arteriovenous malformation; (4) stroke caused by in-stent restenosis in the culprit artery; (5) previous endovascular or surgical procedure on the culprit artery; and (6) concurrent pneumonia, urinary tract infection, and autoimmune disease.
Laboratory analysis
The hs-CRP, LDL-C, and high-density lipoprotein cholesterol (HDLC) levels in all enrolled patients were measured during hospitalization. hs-CRP levels were assessed using a Roche Modular P800 analyzer (Roche, Basel, Switzerland). LDL-C and HDL-C levels were determined using a fully automated biochemical analyzer (model 008AS, Hitachi, Tokyo, Japan). Fasting antecubital venous blood samples were collected and analyzed by professionals who were blinded to the study design.
Based on previous reports [12,22], patients with hs-CRP ≥3 mg/L and LDL-C ≥2.6 mmol/L were categorized into the RCIR group, those with hs-CRP ≥3 mg/L and LDL-C <2.6 mmol/L into RIR group, those with hs-CRP <3 mg/L and LDL-C ≥2.6 mmol/L into RCR group, and those with hs-CRP <3 mg/L and LDL-C <2.6 mmol/L into no residual risk (NRR) group.
HRMRI acquisition
HRMRI was performed using a 3T GE Discovery MR 750 (GE Healthcare, Waukesha, WI, USA), 3T Siemens TrioTim (Siemens Healthcare, Erlangen, Germany), or 3T Philips Ingenia CX (Philips Healthcare, Best, the Netherlands) MRI scanner. The imaging protocols included three-dimensional (3D) time-of-flight MR angiography, 3D T1-weighted imaging (3D T1WI), magnetization-prepared rapid acquisition with gradient-echo (MPRAGE) or simultaneous noncontrast angiography and intraplaque hemorrhage (SNAP), and contrast-enhanced T1WI. The intracranial culprit artery was covered by sequences obtained in either the coronal or axial views. Additional details on the sequence parameters are provided in Supplementary Table 1. Enhanced T1WI (3D CUBE T1 [GE Healthcare], 3D SPACE T1 [Sampling Perfection with Application-Optimized Contrasts using Different Flip Angle Evolutions; Siemens], and 3D VISTA T1 [Volume ISotropic Turbo spin echo Acquisition; Philips]) was performed 5 minutes after the administration of gadopentetate dimeglumine at a dose of 0.1 mmol/kg, using the same parameters as those of the precontrast T1WI.
Imaging analysis and measurements
The images were processed with RadiAnt DICOM (Digital Imaging and Communications in Medicine) Viewer software (Medixant, Poznan, Poland). HRMRI images were reviewed by two neurologists (Y.Y. and R.C.) who were blinded to the clinical data and laboratory results. The images were categorized based on their qualities into poor, adequate, and good, with exclusion criteria applied to poor-quality images exhibiting severe motion artifacts or low signal intensity-to-noise ratio. To assess the inter-observer reproducibility of the same scanner, the two neurologists (Y.Y. and R.C.) independently reviewed the images of 40 randomly selected patients. One month after the initial review, the images of 40 randomly selected patients were reanalyzed by the same neurologist (Y.Y.) to assess intra-observer reproducibility.
Cross-sectional areas of the maximal lumen narrowing site and reference site of the culprit artery were measured. The remodeling index was calculated by dividing the vessel area at the site of maximal lumen narrowing by the vessel area at the reference site. The reference site was selected using the WASID method [21]. Positive remodeling was defined as a remodeling index >1.05, intermediate remodeling as 0.95–1.05, and negative remodeling as less than 0.95. The degree of stenosis was calculated using the WASID method by measuring the lumen diameters at the stenosis and reference sites [23]. The angle of stenosis was calculated as the angle of intersection between the proximal and distal segments of the stenosis lesion [24].
The distribution patterns were discerned in the narrowest slices. Plaques distributed in at least four quadrants of the lumen perimeter were defined as diffusely distributed, whereas those distributed in at most three quadrants were classified as non-diffusely distributed. Intraplaque hemorrhage was defined as a hyperintense signal intensity exceeding 150% of that of the adjacent gray matter on T1WI and MPRAGE (GE and Siemens), or SNAP (Philips) [25].
Plaque enhancement was categorized into three grades: nonenhancement, moderate enhancement, and strong enhancement. Moderate enhancement was defined as inferior to that of the pituitary infundibulum, whereas strong enhancement was equivalent to or surpassing that of the pituitary infundibulum [25].
The culprit intracranial plaques were identified by two experienced neurologists according to clinical symptoms and neurovascular imaging. Based on previous studies [16-19], culprit plaques with positive remodeling, diffuse distribution, strong plaque enhancement, or intraplaque hemorrhage (Figure 1) were regarded as vulnerable plaques. The corresponding diagrams of each type of vulnerable plaque on HRMRI are also displayed in our previous study [17].
Schematic diagram of study enrollment, categorization, and intracranial vulnerable plaque. LDL-C, low-density lipoprotein cholesterol; hs-CRP, highsensitivity C-reactive protein; RCIR, residual cholesterol inflammatory risk; RIR, residual inflammatory risk; RCR, residual cholesterol risk; NRR, no residual risk; HRMRI, high-resolution magnetic resonance imaging.
Statistical analysis
All enrolled patients were divided into four groups: “RIR,” “RCR,” “RCIR,” or “NRR” group. The Shapiro–Wilk test was used to test the distribution normality of continuous variables. Normal distribution variables were presented as mean±standard deviation and compared among groups using one-way ANOVA. Non-normally distributed variables were presented as median (interquartile range) and compared using the Kruskal–Wallis H test. The differences in the categorical variables among the four groups were tested with χ2 or Fisher’s exact tests. Intra-observer and inter-observer variabilities in identifying the vulnerable plaque were evaluated using the Cohen’s κ statistic. The association between residual risk and vulnerable intracranial plaque was analyzed using multiple logistic regression analysis. Variables with P<0.10 in univariate analysis, as well as age, sex, and potential factors associated with vulnerable plaques, were adjusted for in the multivariate logistic regression model. For sensitivity analysis, a predefined LDL-C threshold of 1.8 mmol/L was applied to assess the robustness of the results. The unadjusted and adjusted odds ratios (aOR) and 95% confidence intervals (CI) were calculated. Statistical analyses were performed using R (version 4.2.0; R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a two-tailed P-value <0.05.
Results
Baseline characteristics
A total of 336 eligible patients were included in this study after screening 771 potential participants with symptomatic ICAS who underwent HRMRI of the culprit artery (Figure 2). The mean age of the entire cohort was 56.60±10.38 years, and 78.6% (264/336) of them were male. Among them, 21, 60, 58, and 197 patients were classified into the RCIR, RIR, RCR, and NRR groups, respectively. For baseline characteristics, the proportion of stroke as a qualifying event (RCR [87.9%]>RIR [78.3%]>RCIR [76.2%]> NRR [61.4%]; P<0.001) was significantly different among the four groups. The HDL-C level in the RCR, RCIR, and NRR groups was higher than that in the RIR group (1.10 vs. 1.05 vs. 1.00 vs. 0.86; P<0.001). The time from symptom onset to laboratory testing in the NRR group (18 days) was longer than that in the RIR (10.5 days), RCR (4 days), and RCIR (4 days) groups (P<0.001). The proportion of males, hypertension, diabetes, and smoking; length of stenosis; degree of stenosis; angle of stenosis; and statin use did not differ significantly among the four groups (Table 1).
Flowchart of patient enrollment. ICAS, intracranial atherosclerotic stenosis; HRMRI, high-resolution magnetic resonance imaging; LDL-C, low-density lipoprotein cholesterol; hs-CRP, high-sensitivity C-reactive protein; RCIR, residual cholesterol inflammatory risk; RIR, residual inflammatory risk; RCR, residual cholesterol risk; NRR, no residual risk.
Comparison of baseline characteristics between RCIR, RIR, RCR, and NRR groups
Intra-observer and inter-observer reproducibilities
The intra-observer reproducibility for identifying diffuse distribution (κ, 0.837; 95% CI, 0.663–1.000), intraplaque hemorrhage (κ, 0.811; 95% CI, 0.607–1.000), and stroke enhancement (κ, 0.881; 95% CI, 0.720–1.000) was excellent, and the intra-observer reproducibility for identifying positive remodeling was substantial (κ, 0.776; 95% CI, 0.568–0.984). The results of the inter-observer reproducibilities were similar (κ, 0.760–0.846) (Supplementary Table 2).
Plaque characteristics
The plaque characteristics of the four groups are shown in Table 2. The remodeling indices among the four groups were not significantly different (0.84 vs. 0.85 vs. 0.78 vs. 0.86; P=0.207). There was also no significant difference in the proportion of positive remodeling (23.8% vs. 26.7% vs. 19.0% vs. 28.9%; P=0.499) or diffuse distribution (38.1% vs. 38.3% vs. 34.5% vs. 37.6%; P=0.972) among the four groups. However, there were significant differences among the four groups in the proportion of intraplaque hemorrhage (RCIR [42.9%]>RIR [33.3%]>RCR [22.4%] >NRR [19.3%]; P=0.024) and strong enhancement (RCIR [61.9%] >RIR [60.0%]>NRR [30.5%]>RCR [27.6%]; P<0.001).
Comparison of plaque characteristics between RCIR, RIR, RCR, and NRR groups
Association of residual risk with intraplaque hemorrhage and strong enhancement
Age, sex, hyperlipidemia, qualifying events, NIHSS score, culprit plaque location, HDL-C level, time from symptom onset to laboratory testing, time from laboratory testing to HRMRI, degree of stenosis, angle of stenosis, and intensive statin use were included in the multivariate logistic regression model. The results showed that patients with RCIR (aOR, 2.965; 95% CI, 1.060–8.297; P=0.038) had a higher risk of intraplaque hemorrhage than those with NRR. In addition, patients with RCIR (aOR, 3.606; 95% CI, 1.346–9.662; P=0.011) and RIR (aOR, 3.361; 95% CI, 1.774–6.368, P<0.001) were more likely to have strong enhancement than those with NRR (Table 3).
Association of residual risk with intraplaque hemorrhage and strong enhancement
Sensitivity analysis
When using an LDL-C cutoff value of 1.8 mmol/L, RCIR (aOR, 5.482; 95% CI, 2.218–13.551; P<0.001) and RCR (aOR, 2.595; 95% CI, 1.201–5.608; P=0.015) were both significantly associated with intraplaque hemorrhage compared to NRR. The association of RCIR (aOR, 3.962; 95% CI, 1.844–8.512; P<0.001) and RIR (aOR, 3.160; 95% CI, 1.339–7.454; P=0.009) with strong plaque enhancement was similar to previous results (Table 4).
Association of residual risk with intraplaque hemorrhage and strong enhancement in the sensitivity analysis (cutoff level of LDL-C=1.8 mmol/L)
Discussion
The present study investigated the correlation between the RIR and intracranial plaque vulnerability and revealed that both RCIR and RIR were significantly associated with strong plaque enhancement in patients with ICAS who experienced ischemic stroke or TIA. Moreover, we observed that RCIR was independently associated with an increased probability of intraplaque hemorrhage. After lowering the cutoff value of LDL-C in defining RIR, the sensitivity analysis suggested that RCR had an additional correlation with an augmented risk of intraplaque hemorrhage.
Inflammation and dyslipidemia play pivotal roles in the development, progression, and eventual rupture of atherosclerotic plaques, demonstrating their collaborative influence on atherogenesis [26,27]. hs-CRP and LDL-C levels serve as indicators of the systemic inflammatory state and lipid burden, reflecting the extent of atherosclerosis. We concurrently evaluated hs-CRP and LDL-C levels to investigate the implications of vulnerable intracranial plaques. RCIR and RIR were significantly associated with strong plaque enhancement in patients with symptomatic ICAS, suggesting a more important role for inflammation in the process of plaque enhancement. A previous study involving 143 patients with symptomatic ICAS showed that elevated levels of hs-CRP were closely associated with strong plaque enhancement (OR, 7.497; 95% CI, 2.633–21.349; P<0.001) [20], which is consistent with our findings. The key discrepancy was that we explored the role of LDL-C in plaque enhancement. We found that RCIR may exhibit a stronger association with strong plaque enhancement than RIR (aOR, RCIR vs. RIR, 3.606 vs. 3.361), revealing a potential synergistic role of LDL-C in plaque enhancement. The additive effects of these factors are likely due to their distinct yet interconnected roles in atherosclerosis pathophysiology. Systemic inflammation driven by elevated hs-CRP levels can exacerbate LDL-C–mediated plaque progression, accelerating lipid accumulation and the expansion of the necrotic core. Concurrently, hs-CRP may directly bind to lipoproteins to form immune complexes that further stimulate local inflammation within the plaque and accelerate intravascular thrombosis [28].
Another important finding of our study is that RCIR indicates a higher risk of intraplaque hemorrhage in patients with symptomatic ICAS. Notably, RCR showed an additional relationship with intraplaque hemorrhage in the sensitivity analysis after employing an LDL-C cutoff value of 1.8 mmol/L. Our results are in line with a prior study utilizing a similar LDL-C cutoff value of 1.8 mmol/L, which reported that acute coronary syndrome patients with RCR (aOR, 7.95; 95% CI, 1.60–39.53, P=0.011) were more susceptible to plaque rupture, as determined by optical coherence tomography [29]. Intraplaque hemorrhage occurs when the rupture of plaque microvessels contributes to free cholesterol deposition, macrophage infiltration, erythrocyte membrane accumulation, and an increase in necrotic core size, which increases the risk of plaque instability [30]. These findings imply that dyslipidemia exerts a greater influence than inflammation on initiating and exacerbating intraplaque hemorrhage or plaque rupture during the progression of atherosclerosis.
Both intraplaque hemorrhage and strong enhancement of culprit plaque in patients with ischemic stroke imply plaque instability, potentially heightening the risk of neurological deterioration or future stroke events [6,31]. A meta-analysis demonstrated that intracranial plaque enhancement is strongly associated with ischemic stroke.6 Another meta-analysis of eight studies showed that the presence of carotid intraplaque hemorrhage on HRMRI serves as a robust predictor of cerebrovascular events during a median follow-up time of 19.6 months [31]. Our study indicates that RCIR is correlated with both intracranial hemorrhage and plaque enhancement, affirming the indispensable roles of inflammation and dyslipidemia in driving plaque progression and instability. Li et al. [12] found that RCIR could predict poor functional prognosis and stroke recurrence in acute ischemic stroke or TIA patients, particularly in those with a large-artery atherosclerosis subtype. We hypothesized that RCIR might increase the risk of stroke recurrence in patients with symptomatic ICAS by influencing the stability of the culprit plaque. However, this hypothesis requires further investigation. Mediation analysis may help explore the relationships among the three factors (RIR, vulnerable plaque, and stroke recurrence).
In recent years, with the growing understanding of RIR, the concept of a dual LDL-C and hs-CRP-lowering therapy has been proposed and applied to patients with cardiovascular diseases [32]. Ischemic stroke patients exhibiting RCIR, RIR, or RCR may be the target population for this dual LDL-C and hs-CRP therapy in the future. Although the results of the Colchicine in Patients With Acute Ischemic Stroke or Transient Ischemic Attack (CHANCE-3) trial recently showed that low-dose colchicine could not reduce the risk of subsequent stroke within 90 days, the potential role of other anti-inflammatory drugs, such as canakinumab, in the secondary prevention of cerebrovascular diseases remains worth investigating [33,34].
Our study has several limitations. First, it was a single-center study that exclusively included Asian populations, and potential selection bias may be inherent. Second, some baseline characteristics such as time from symptom onset to laboratory testing and time from laboratory testing to HRMRI, were unbalanced across groups, which may stem from the retrospective nature of this study. More rigorous prospective studies are required to validate these results. Third, HRMRI data originating from three different types of MRI scanners and the intra-scanner consistency of analysis of intracranial vulnerable plaques could not be assessed because of the retrospective nature of the study. However, inter- and intra-observer consistencies among scanners have been reported to be excellent in previous studies [25].
Conclusions
In patients with symptomatic ICAS, RIR is associated with a higher risk of intraplaque hemorrhage and strong enhancement, indicating increased vulnerability to atherosclerotic plaques. RIR is expected to be a novel inflammatory marker for screening high-risk patients with vulnerable intracranial plaques.
Supplementary materials
Supplementary materials related to this article can be found online at https://doi.org/10.5853/jos.2024.03251.
Parameters of multiple sequences on GE, Siemens, and Philips MR scanners
Intra-observer and inter-observer reproducibilities
Notes
Funding statement
This work was supported by the National Natural Science Foundation of China (Contract grant number: 82171894 to N.M.), the National Natural Science Foundation of China (Contract grant number: 82151309, 82327803 to X.L.), and the China Postdoctoral Science Foundation (Contract grant number: 2023M742437 to Z.H.).
Conflicts of interest
The authors have no financial conflicts of interest.
Author contribution
Conceptualization: YY, NM. Study design: YY, LY, NM. Methodology: YY, RC, NM. Data collection: RC, XH, XS, ZH, ML, JY. Investigation: ZM, RW, YW, XL. Statistical analysis: YP, YY. Writing—original draft: YY. Writing—review & editing: LY, NM. Funding acquisition: NM, XL, ZH. Approval of final manuscript: all authors.
Acknowledgments
We thank all patients and healthcare providers who participated in this study. We gratefully acknowledge Servier Medical Art image bank (https://smart.servier.com/) for providing the graphic elements used in the schematic diagram (Figure 1).