Introduction

Intracranial atherosclerosis is a prevalent cause of ischemic stroke worldwide, especially in the Asian population [1]. Both luminal stenosis and vulnerable plaques contribute to ischemic events. Intraplaque hemorrhage (IPH), as a marker of vulnerable plaques, has been established as an important factor for ischemic stroke. The presence of IPH is also associated with atherosclerosis progression and stroke recurrence [2,3,4,5]. One study suggested that IPH may also lead to cerebral blood flow reduction [6]; thus, the contribution of IPH to stroke prevention needs to be strongly considered. Identifying the modifiable risk factors associated with IPH is critically important for preventing atherosclerosis progression and stroke recurrence.

Both systemic and focal factors play a role in the development of vulnerable plaques. Some scholars have proposed that systemic factors are important in plaque instability after reviewing the morphology of several thousand carotid artery plaques with digital subtraction angiography and analyzing their association with clinical events [7]. Since a proper method is lacking to noninvasively assess intracranial vessel wall atherosclerotic plaques, the determinants of IPH have rarely been reported. In 2012, the Rotterdam study investigated the association of systemic blood pressure parameters and carotid artery IPH measured by high-resolution magnetic resonance imaging (HR-MRI). Their results suggested that pulse pressure (PP) is a critical risk factor for IPH [8] and this association is independent of other cardiovascular disease and blood pressure components. Another small sample size population-based study demonstrated that a lower diastolic blood pressure (DBP) is related to IPH in the carotid artery [9]. In addition, data about the associations between blood pressure parameters and aortic arch atherosclerosis indicated that systolic blood pressure (SBP), as well as its variability, is correlated with aortic arch complex plaques in severe aortic stenosis [10]. From the results above, we can conclude that systemic blood pressure parameters are associated with atherosclerotic plaque characteristics, but the results are rather confusing. Furthermore, the few recent studies about blood pressure and IPH focus on the carotid artery in the Western population, and the association between blood pressure parameters and intracranial IPH in Asian populations has not been investigated.

The purpose of this study was to thoroughly explore the associations between blood pressure parameters (SBP, DBP, PP, and their variability) and IPH in ischemic stroke patients with intracranial atherosclerosis. The results of this study will provide some insight into the pathophysiology of high-risk plaques and give more guidance for blood pressure management in this population.

Methods

Subjects

Ischemic stroke patients with symptomatic intracranial atherosclerosis admitted to a single comprehensive stroke center in 2017–2018 were retrospectively screened. The inclusion criteria for this study were as follows: (1) ischemic stroke with more than one atherosclerosis risk factor; (2) age ≥ 18 years old; (3) stroke caused by intracranial atherosclerosis with or without significant vessel stenosis; and (4) MRI performed within 4 weeks of symptom onset. Subjects were excluded if they had (1) MRI contradictions or inadequate image quality; the image quality (IQ) was evaluated by a 4-point scale (1 = poor; 2 = adequate; 3 = good; 4 = excellent) [11], and those with images scored as IQ = 1 were excluded, (2) significant extracranial artery stenosis (≥50) or cardiac embolisms, or (3) nonatherosclerotic intracranial artery disease including dissection, Moyamoya disease, vasculitis, or other undetermined causes. In addition, patients who underwent endovascular procedures before vessel wall imaging were also excluded because of the possible effects on the vessel wall lesions.

This study was conducted within the framework of the 1964 ethical guidelines of the Helsinki Declaration. Informed consent was waived by the local ethics committee due to the retrospective analysis.

Blood pressure measurement

Daily morning blood pressure indicators, including SBP and DBP, were obtained from the standard medical records. The measurement protocol is briefly described here. Trained nurses usually measured the brachial blood pressure of stroke patients twice daily using a corrected blood pressure monitor (Omron Corporation, Kyoto, Japan). For each measurement, the performance could be completed in a seated or flat position after full rest, usually on the nonparalyzed side, and was repeated two times. In the case of a significant difference between the two sides, the higher reading would be recorded. For this analysis, blood pressure measurements from the first three days and the seventh day after admission were collected. PP was acquired from SBP and DBP, and the individual blood pressure variability parameters were calculated from the four measurements, including the standard deviation (SD) and coefficient of variation (CV).

Hypertension was diagnosed based on the Eighth Joint National Committee (JNC8) and Chinese hypertension guidelines [12, 13]. Subjects with mean SBP ≥ 140 mmHg, DBP ≥ 90 mmHg or self-reported use of any antihypertensive drugs during the past 2 weeks were defined as having hypertension.

Covariate assessment

The basic characteristics and atherosclerosis risk factors were extracted from the medical records. Diabetes was considered if patients had fasting glucose ≥ 7 or nonfasting glucose ≥ 11.1 mmol/L or were on antidiabetic medication. Hyperlipidemia was determined based on the adult treatment panel III guidelines [14, 15] (total cholesterol ≥ 240 mg/dL, low-density lipoprotein cholesterol ≥ 160 mg/dL, or high-density lipoprotein cholesterol < 40 mg/dL) or for those who had been on lipid-lowering medication. Smoking and drinking histories were also obtained from the medical history.

MRI protocol

All brain and vessel imaging was performed on a 3.0 T MRI with an 8-channel head coil (GE Discovery MR750). The vessel wall imaging protocol included 3D time-of-flight magnetic resonance angiography (TOF-MRA) and 3D T1-weighted imaging (T1WI) followed by routine brain MRI. The detailed imaging parameters were as follows: 3D TOF MRA obtained in the axial plane, repetition time (TR)/echo time (TE): 22/2.5 ms, flip angle 20°, field-of-view (FOV) 22 cm × 18 cm, and spatial resolution: 0.6 × 1.0 × 1.2 mm3. 3D T1WI (Cube): TR/TE: 800/16 ms, flip angle 90°, FOV 23 cm × 18.4 cm, and spatial resolution 0.7 × 0.6 × 0.6 mm3. 3D TOF-MRA and T1WI were used to visualize stenosis and IPH.

Image analysis

Two experts blinded to the clinical information assessed the vessel wall images independently using the GE-Extended Workstation. In case of inconsistencies, a third senior expert was involved to settle any disagreements. The infarct and culprit vessels were reviewed by an independent vascular neurologist.

The axial slices of the vessel wall underwent multiplanar reconstruction perpendicular to the arterial centerline. The atherosclerotic plaque characteristics, including maximum wall thickness (max WT), luminal stenosis degree, and IPH, were assessed in all intracranial vascular beds, including the bilateral A1 segment of the anterior cerebral artery, M1-2 segments of the middle cerebral artery, terminal segment of the internal carotid artery, P1 segment of the posterior cerebral artery, V4 segment of the vertebral artery and basilar artery.

Atherosclerotic plaques were considered in patients with eccentric wall thicknesses with or without significant luminal stenosis. The max WT and stenosis degree were acquired in the culprit vessel. IPH was defined as T1-hyperintensity plaque with signal intensity higher than that of the surrounding brain parenchyma or muscle [16]. For this analysis, only IPH located in the culprit vessel was counted. Luminal stenosis was measured according to the Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) criteria [17] and divided into 4 grades in this analysis (0 for <30%; 1 for 30–49%; 2 for 50–69%; and 3 for ≥70%). Multiple plaques were defined as more than three plaques in all detected vascular beds of each patient. To ensure the accuracy of the image analysis, the interrater reproducibility in discriminating IPH and the max WT measurements was also tested in ten randomized patients.

Statistical analysis

The variables are presented as the mean ± SD or number (%) depending on the variable type. Student’s t test or χ2 test was used for comparisons between groups. Cohen’s kappa and the intraclass correlation (ICC) tests were performed to assess the interrater reliability in discriminating IPH and the max WT measurements. Univariate and multivariate logistic regression analyses were utilized to demonstrate the odds ratio (OR) and corresponding 95% confidence interval (CI) of different blood pressure indicators in identifying atherosclerotic plaque characteristics. To analyze the association between blood pressure parameters and atherosclerotic plaque burden, the degree of stenosis and number of plaques were taken as dichotomous variables in the regression analysis. All p values < 0.05 were considered statistically significant. All statistical analyses were performed by using SPSS 22.0 (IBM, New York, USA).

Results

Characteristics of the subjects with and without IPH

A total of 306 patients were included in this study; the mean age was 66 ± 11 years old, and 69.2% of the subjects were male. The consistency test for image interpretation between the two readers was fairly good, with kappa = 1.0 (p < 0.01) for diagnosing IPH and an ICC = 0.94 (95% CI: 0.85–0.97) for determining the max WT measurements. In general, a larger atherosclerotic plaque burden could be observed in this cohort. A total of 59.8% of the subjects presented with ≥50% luminal stenosis, and 71.2% of the patients had more than 3 atherosclerotic plaques in the detected vascular beds.

Of all subjects included, 124 subjects (40.5%) were considered IPH positive. Compared with subjects without IPH, the subjects with IPH were more likely to be male (p < 0.05). The other atherosclerosis risk factors were well balanced between the two groups (Table 1). Regarding the vessel wall plaque characteristics, the IPH (+) group had a larger max WT, a higher proportion of multiple plaques and more severe stenosis than the IPH (−) group (all p < 0.05).

Table 1 Characteristics of the subjects with and without intraplaque hemorrhage
Full size table

In terms of the blood pressure parameters, the mean SBP was much higher in the IPH (+) group than in the IPH (−) group (mean 142.8 vs. 138.8 mmHg, p < 0.05). None of the other blood pressure indexes were different between the two groups, including blood pressure variability (Table 2).

Table 2 Blood pressure parameters of the subjects with and without intraplaque hemorrhage
Full size table

Associations between blood pressure parameters and atherosclerotic plaque burden

The degree of luminal stenosis was divided into two groups (0–1 as the mild group [<50% stenosis] and 2–3 as the severe group [≥50% stenosis]). The associations of blood pressure indicators with atherosclerotic plaque burden were assessed (Table 3). The results suggested that both SBP and PP were correlated with multiple plaques and severe luminal stenosis in the univariate and multivariate regression analyses, with OR = 1.071, 95% CI: (1.044–1.098) and OR = 1.039, 95% CI: (1.019–1.060) per 1 mmHg increase in SBP in the multivariate adjusted regression analysis, and OR = 1.058, 95% CI: (1.027–1.089) and OR = 1.044, 95% CI: (1.019–1.070) for PP, respectively. We did not find any associations between DBP or blood pressure variability and atherosclerotic plaque burden.

Table 3 Regression analysis of blood pressure parameters and atherosclerosis plaque burden
Full size table

Associations between blood pressure parameters and IPH

A further investigation of the associations between blood pressure parameters and IPH (Table 4) demonstrated that SBP was associated with IPH in the univariate analysis, with OR = 1.020, 95% CI: (1.003–1.036), as well as in the multivariate analysis (adjusted for age, sex, and cardiovascular risk factors), with OR = 1.021, 95% CI: (1.003–1.038). However, when atherosclerotic plaque burden was corrected in model 3, this positive relationship disappeared (OR = 1.014, 95% CI: (0.996–1.032)). No relationships between the other blood pressure parameters and IPH were found.

Table 4 Regression analysis of blood pressure parameters and intraplaque hemorrhage
Full size table

Prevalence of atherosclerotic plaque burden and IPH according to the SBP categories

Finally, we investigated the prevalence of multiple plaques, severe luminal stenosis, and IPH according to different SBP levels. Figure 1 shows the increased frequency of multiple plaques and severe stenosis with increased quartiles of SBP (Pearson χ2 test: p < 0.01, p for trend <0.01) but not IPH (p > 0.05).

Fig. 1
figure 1

Atherosclerotic plaque characteristics according to systolic quartile. Pearson χ2 test p < 0.01 for the presence of multiple plaques and severe stenosis; p = 0.26 for the presence of IPH in four categorical groups

Full size image

Discussion

To the best of our knowledge, this is the first study to thoroughly investigate the associations between systemic blood pressure parameters and intracranial vulnerable plaques in subjects with symptomatic intracranial atherosclerosis. Our study revealed some novel findings: (1) a high prevalence of IPH can be found in patients with advanced atherosclerosis; (2) SBP and PP are independently associated with intracranial atherosclerotic plaque burden; (3) SBP is associated with IPH after adjusting for cardiovascular risk factors but not after further correcting for plaque burden; and (4) no associations between blood pressure variability and intracranial atherosclerotic plaque burden or IPH were found in this study.

Vessel wall plaque characteristics in stroke patients with intracranial atherosclerosis

As a surrogate marker of high-risk stroke, IPH is not uncommon in ischemic stroke with intracranial atherosclerosis. Previous studies on intracranial atherosclerosis have mainly focused on luminal stenosis, and data about vessel wall vulnerable plaques in intracranial arteries are limited. Compared with the nonselected stroke population of a few recent intracranial vessel wall plaque studies [18, 19], our population had a much larger atherosclerotic plaque burden, showed more extensive vascular bed lesions and had more severe stenosis. Both atherosclerotic plaque pathology and vessel wall imaging studies have suggested that IPH is associated with advanced atherosclerosis [20, 21], and subjects with advanced atherosclerosis have a higher prevalence of IPH, which can explain the results in our study.

Association of blood pressure parameters with atherosclerotic plaque burden and IPH

Hypertension has been established as a risk factor for atherosclerosis and cardiovascular events [22], while the associations between blood pressure parameters and atherosclerotic plaque characteristics in different vascular beds remain inconclusive. Data from the Rotterdam study indicated that both SBP and PP were associated with carotid artery IPH after correcting for age, sex, cardiovascular risk factors, and carotid vessel wall thickness. When further adjusted for other blood pressure components, only PP was associated with IPH in the carotid artery [8]. Another study regarding cardiovascular risk factors and carotid IPH illustrated that lower DBP was correlated with carotid artery IPH [9]. This inconsistency could possibly be explained by the population included, sample size, blood pressure levels, and adjusted risk factors. Thus, we should explain this association with caution. Our results demonstrated positive associations between SBP, PP, and intracranial atherosclerosis and a correlation between SBP and IPH in the multivariate adjusted models. However, after adjusting for plaque burden, the positive association between SBP and IPH disappeared. Thus, atherosclerosis burden could be a confounder between SBP and IPH, and the positive association between SBP and IPH may possibly be mediated by atherosclerosis. In addition, this controversial result could also be explained by the fact that the effect of systemic blood pressure on plaque vulnerability differs between the proximal and distal cerebral arteries. Compared with the extracranial arteries, the intracranial arteries have thinner media and adventitia and fewer elastic medial fibers; moreover, the tortuosity and hemodynamic changes in the intracranial arteries are also different from those of the extracranial arteries. We assumed that plaque vulnerability in distal vascular beds may be less affected by systematic blood pressure. More prospective studies about blood pressure with vessel wall imaging in different vascular beds need to be conducted to validate these results.

Association of blood pressure variability with atherosclerotic plaque burden and IPH

Several studies have suggested that short-term or long-term blood pressure variability is associated with atherosclerosis and its progression, as well as with cardiovascular events [23, 24]. One study investigated the correlation between circadian blood pressure changes and carotid atherosclerosis progression in 286 subjects over 55 years old and found that daytime SBP variability could be used as a predictor of early carotid atherosclerosis progression [25]. A similar study also indicated that a nondipping blood pressure pattern was associated with carotid atherosclerosis in a middle-aged population [26]. Regarding long-term blood pressure variability, a prospective study about high-risk cardiovascular disease also established that visit-to-visit blood pressure variability (SBP variability) can provide prognostic information about major adverse cardiovascular events [27].

Plaque rupture is an important cause of cardiovascular disease. Of all the contributory factors to plaque rupture, fluid–structure interactions are receiving increasing attention. Blood pressure variability could significantly affect the blood flow velocity and wall shear stress [23]. Plaque focal wall shear stress has been recognized as a critical factor contributing to atherosclerosis progression and vulnerable plaque rupture [28]. There are limited studies on blood pressure variability and plaque vulnerability, especially in patients with symptomatic intracranial atherosclerosis. Thus, our study is hypothesis-generating, and we assumed that blood pressure variability could possibly be associated with IPH. Unfortunately, we did not find any associations between blood pressure variability and IPH in this study or any relationships between blood pressure variability and atherosclerotic plaque burden. In fact, the negative association between blood pressure variability and plaque vulnerability could also be observed in coronary arteries with stable angina [29]. Some speculations were proposed here. First, the effect of blood pressure variability on atherosclerosis in different vascular beds is different. A previous study on blood pressure variability and atherosclerosis mainly focused on the aortic arch and carotid artery, and no intracranial atherosclerosis was involved. Moreover, unlike that in the aortic arch or other large vessels, dynamic cerebral autoregulation in the cerebral artery could provide a protective mechanism to keep the blood flow stable in cases of drastic systemic blood pressure changes [30, 31]. Second, the blood pressure variability used in our study is also different from that in previous studies. Some studies used daily blood pressure variability, while others used visit-to-visit blood pressure variability; in contrast, we used day-to-day blood pressure variability during a short period of time, which might also have an effect on the result. Patients in hospitals usually have more frequent blood pressure monitoring than those in other settings and consequently receive better blood pressure management, which can also underestimate the blood pressure variability to some degree.

Strengths and limitations

The key strength of this study is that we presented a general description of vessel wall atherosclerotic plaque burden and the prevalence of IPH in subjects with intracranial atherosclerosis from a HR-MRI prospective and comprehensively investigated the associations between blood pressure parameters and intracranial vessel wall vulnerable plaques. The association between blood pressure variability and intracranial vulnerable plaques has never been reported before.

There are indeed some limitations to this study. First, the cross-sectional analysis failed to conclude the cause–effect relationship between the blood pressure indicators and atherosclerotic plaque characteristics. The dynamic blood pressure parameters before stroke or plaque rupture may be much more significant. Second, we focused on the associations between blood pressure parameters and atherosclerotic plaque burden and IPH, and no vessel wall remodeling or plaque enhancement was analyzed. Finally, the day-to-day blood pressure variability in the acute period of stroke was used in this analysis, and as we know, subjects with acute stroke often present with elevated blood pressure or significant fluctuations, which can also affect the results to some degree. Further studies including daily blood pressure variability or visit-to-visit blood pressure variability are needed.

Implications

This study has some implications. Atherosclerosis is a systemic pathophysiological process with extensive vascular lesions. Despite the common risk factors, dynamic atherosclerosis progression in different vascular beds is not fully understood. The contributory roles of systemic blood pressure on different vascular beds are different; perhaps the focal vascular biological factors are much more important in distal vessels than in proximal vessels. Furthermore, our study can also provide some guidance on the use of antihypertensive medication for atherosclerotic plaque burden. Blood pressure, as a treatable risk factor for atherosclerosis, should be strongly considered, especially with SBP management.

Conclusion

Of all the blood pressure parameters included, SBP is associated with intracranial atherosclerotic plaque burden and IPH, and its association with IPH could possibly be mediated by atherosclerosis. The associations between blood pressure variability and vessel wall plaque characteristics need further study.