Association between intercellular adhesion molecule-1 to depression and blood-brain barrier penetration in cerebellar vascular disease

Abstract

BACKGROUND

Cerebral small vessel disease (CSVD) is a prevalent cerebrovascular disease in clinical practice that is often associated with macrovascular disease. A clear understanding of the underlying causes of CSVD remains elusive.

AIM

To explore the association between intercellular adhesion molecule-1 (ICAM-1) and blood-brain barrier (BBB) penetration in CSVD.

METHODS

This study included patients admitted to Fuyang People’s Hospital and Fuyang Community (Anhui, China) between December 2021 and March 2022. The study population comprised 142 patients, including 80 in the CSVD group and 62 in the control group. Depression was present in 53 out of 80 patients with CSVD. Multisequence magnetic resonance imaging (MRI) and dynamic contrast-enhanced MRI were applied in patients to determine the brain volume, cortical thickness, and cortical area of each brain region. Moreover, neuropsychological tests including the Hamilton depression scale, mini-mental state examination, and Montreal cognitive assessment basic scores were performed.

RESULTS

The multivariable analysis showed that age [P = 0.011; odds ratio (OR) = 0.930, 95% confidence interval (CI): 0.880-0.983] and ICAM-1 levels (P = 0.023; OR = 1.007, 95%CI: 1.001-1.013) were associated with CSVD. Two regions of interest (ROIs; ROI3 and ROI4) in the white matter showed significant (both P < 0.001; 95%CI: 0.419-0.837 and 0.366-0.878) differences between the two groups, whereas only ROI1 in the gray matter showed significant difference (P = 0.046; 95%CI: 0.007-0.680) between the two groups. ICAM-1 was significantly correlated (all P < 0.05) with cortical thickness in multiple brain regions in the CSVD group.

CONCLUSION

This study revealed that ICAM-1 levels were independently associated with CSVD. ICAM-1 may be associated with cortical thickness in the brain, predominantly in the white matter, and a significant increase in BBB permeability, proposing the involvement of ICAM-1 in BBB destruction.

Key Words: Cerebral small vessel disease; Intercellular adhesion molecule-1; Blood-brain barrier penetration; Cortical thickness; White matter

Core Tip: Cerebral small vessel disease (CSVD) is a prevalent cerebrovascular disease in clinical practice that is often associated with macrovascular disease. A clear understanding of the underlying causes of CSVD remains elusive. Several studies have reported the relationship between intercellular adhesion molecule-1 (ICAM-1) and blood-brain barrier (BBB) penetration in neurological disorders. Thus, this study examined the association between ICAM-1 and BBB penetration in CSVD.

INTRODUCTION

Cerebral small vessel disease (CSVD) is a cluster of clinical, imaging, and pathological symptoms resulting from multiple factors that affect the cerebral arterioles, capillaries, venules, and veins[1,2]. It is a prevalent cerebrovascular disease in clinical practice, is often associated with macrovascular disease, and accounts for 20%-30% of all ischemic stroke and cerebral hemorrhage[3]. Given its high incidence, CSVD is thought to play a key role in stroke, dementia, and aging[3,4]. Therefore, it should be considered a major public health problem, particularly among aging populations. Despite extensive research, a clear understanding of the causes of CSVD remains elusive[5]. Numerous studies have identified inflammation as an etiological factor in CSVD development and progression[6]. A community-based cohort study also demonstrated a correlation between inflammatory biomarkers in CSVDs and magnetic resonance imaging (MRI)[7].

Dysregulation of cell adhesion molecules (CAMs), which coordinate leukocyte trafficking, may affect blood-brain barrier (BBB) permeability and promote inflammation and immune-mediated transport to the central nervous system (CNS), thereby connecting peripheral and neuroinflammation in severe psychiatric disorders[8,9]. Intercellular adhesion molecule-1 (ICAM-1) is a protein found on different cell surfaces, including endothelial, epithelial, and white blood cells. White blood cell adhesion and migration are critical processes in immune response and inflammation, both involving ICAM. In addition, ICAM-1 is associated with cardiovascular disease, cancer, and autoimmune disorders. Several studies have reported the relationship between ICAM-1 and BBB penetration in neurological disorders[10,11]. ICAM-1 and BBB penetration correlate with Alzheimer’s disease[12]. Integrin receptors on leukocytes bind to CAMs and promote their migration across the BBB[13,14]. Although the BBB does not deteriorate in severe mental illness as in neurologic diseases and neurodegenerative disorders, postmortem transcriptomic studies have shown increased permeability in schizophrenia, revealing the dysregulated expression of tight junction-related genes in the prefrontal cortex[15,16] and BBB transcripts enriched in brain endothelial cells such as ICAM1[13] as well as aberrant immunoreactivity of claudin-5 in brain sections from the parietal lobe[17]. A recent study also revealed a correlation between ICAM-1 expression and CSVD development[18]. Endothelial dysfunction and subsequent BBB damage are key pathophysiological mechanisms in CSVD. Dynamic contrast-enhancement MRI (DCE-MRI) is the preferred imaging technique for evaluating BBB leakage in CSVD[11]. It can assess BBB damage in specific areas and quantify permeability changes.

Although the aforementioned studies discussed the relationships between ICAM-1 and CSVD, more studies are required to elucidate the precise mechanisms through which inflammatory processes contribute to the pathology of CSVD.

At present, the pathogenesis of CSVD remains unclear, no standard treatment or plan has been established, and its diagnosis mainly depends on brain MRI; thus, early diagnostic markers are needed. Accordingly, this study aimed to examine the association between ICAM-1 and BBB penetration in CSVD. The results of this study could identify surrogate markers that could help screen for patients with radiological CSVD before the onset of symptoms. Indeed, blood biomarkers are easier to assess than markers observed via MRI. Early disease detection may allow physicians to develop a tailored, individualized treatment plan for each patient with CSVD. This approach can also help avoid unnecessary treatments and lower healthcare costs, ultimately improving patients’ quality of life.

MATERIALS AND METHODS

Study design and participants

This study included patients admitted to Fuyang People’s Hospital and Fuyang Community (Anhui, China) between December 2021 and March 2022. The inclusion criteria were as follows: (1) Patients who met the diagnostic criteria for cerebrovascular disease and were confirmed by head MRI[19]; (2) Age > 50 years; (3) Lack of apparent subjective neurological symptoms; (4) Normal cognitive function with normal mini-mental state examination (MMSE) and Montreal cognitive assessment basic (MoCA-B) scale scores; (5) No history of brain trauma; (6) No long-term use of antibiotics, antiplatelet drugs, and statins; and (7) All underwent multisequence MRI. The exclusion criteria were as follows: (1) Presence of intracranial macrovascular lesions; (2) Abnormal liver and kidney function; (3) History of a brain tumor or brain trauma, CNS infection, thyroid function abnormality, vitamin B1/B12 deficiency, alcoholic brain injury, syphilis, malnutrition, or other diseases that may cause brain injury; (4) Cognitive impairment, pconsciousness impairment, visual-auditory impairment, aphasia, and inability to cooperate with scale testing; (5) Vascular infarction or severe hemorrhagic stroke or patients with white matter hyperintensity (WMH) and leukodystrophy, CNS inflammatory demyelinating disease and vasculitis on MRI, or the presence of white matter lesions from other causes; (6) Current infectious disease/autoimmune disease; (7) Susceptibility weighted imaging (SWI) sequence indicated severe micro bleeding lesions; (8) Intake of nonsteroidal anti-inflammatory drugs or corticosteroids; (9) Presence of malignant tumor; and (10) Any condition or treatment that could affect the study outcomes according to the investigators.

The study was approved by the Ethics Committee of Fuyang People’s Hospital, and all participants signed an informed consent form for participation.

In addition, individuals who did not present with such symptoms and remained undiagnosed with CSVD were included in the control group.

Diagnostic criteria for CSVD

At least one CSVD marker (lacunar infarction, high signal in the cerebral white matter, microhemorrhage, and perivascular gap) must be present on MRI. MRI findings are as follows: (1) Lacunar infarction is characterized by a well-defined round, oval, or lacunar lesion, 3-15 mm in diameter, with isosignal to the cerebrospinal fluid. T2-fluid attenuation inversion recovery (FLAIR) sequences show cerebrospinal fluid-like low signals in the center and high signals around it, and T1 and T2 sequences usually show cerebrospinal fluid-like signal signals; (2) High cerebral in the white matter is characterized by abnormal signal in the cerebral white matter area, high signal on T2 or T2-FLAIR, micro-equal or low signal on T1, and lack of cavity in the lesion, which is classified into deep white matter high signal and paraventricular white matter high signal, and early fusion of deep white matter high signal or irregular paraventricular white matter high signal extending to the deep white matter is positive for cerebral white matter high signal; (3) The perivascular gap is characterized by a basal ganglia region, enlarged perivascular gap, direct < 3 mm, isosignal with cerebrospinal fluid, linear (when imaging plane is parallel to vessel alignment) or round-like (when perpendicular to vessel alignment), and positive if the number of perivascular gaps in unilateral basal ganglia region is > 10; and (4) Microhemorrhages are small round or ovoid, well-defined, homogeneous foci of signal loss, 2-5 mm in diameter, surrounded by brain parenchyma, and shown on (SWI) sequences; one or more microhemorrhages are positive.

Procedures

All participants underwent cognitive assessment using the MMSE and MoCA-B scales during the routine course of their clinical workup. The normal ranges of MMSE are > 17, > 20, and > 24 points for those who are illiterate, completed primary school, and completed junior high school or higher, respectively. Normal MoCA-B scores were > 19, > 22, and > 24 for those who were illiterate or had primary school education, secondary school education, and university education, respectively[20].

All patients also underwent multisequence MRI and DCE-MRI for clinical indications. Conventional axial and sagittal T1-weighted imaging [repetition time (TR) 7.9 ms, echo time (TE) 3.5 ms, layer thickness 1 mm, layer spacing 0 mm, 140 layers], T2WI (TR 3000 ms, TE 90 ms, section thickness of 5 mm, section gap of 0 mm, 24 layers), and DCE-MRI (TR 3.5 ms, TE 1.64 ms, section thickness of 3 mm, section gap of 0 mm, 36 layers) were performed using an Ingenia CX 3.0-T MRI system (Philips, Best, The Netherlands). Four physicians, including two imaging specialists and two neurovascular experts, independently evaluated the imaging results. If discrepancies occurred, the four physicians engaged in collaborative discussions to reach a diagnosis under the guidance of a senior imaging specialist. T2-weighted FLAIR WMH was quantified using the Fazekas scale[21].

Venous blood (3 mL) was collected in the fasting state and left at room temperature for 10-20 minutes. The blood samples were centrifuged at 3000 rpm for 15 minutes, and the supernatant was collected. Serum was analyzed using the human ICAM-1 enzyme-linked immunosorbent assay kit according to the manufacturer’s instructions.

T1-weighted images were preprocessed using SPM8 in Matlab R2012b. Image normalization, segmentation, modulation, and smoothing were performed according to standard procedures. The multiple regression approach in SPM8 was used to analyze the correlation between ICAM-1 expression and brain volume. A significance level of P < 0.001 and identified clumps > 10 pixels were defined as regions of interest (ROIs). Professional imaging physicians analyzed the ROIs on DCE-MRI using the Philips Intellispace Portal (v 7.0.5.40155) on a workstation with a Philips Ingenia CX scanner to obtain average Ktrans values for the ROIs.

FreeSurfer was used to process the structural magnetic resonance images of all patients. “Recon-all” processing was used with default parameters to derive the brain volume, cortical thickness, and cortical area of each brain region.

Data collection

Demographic and clinical data, including sex, age, education, personal history, medical history, and medication history, were extracted from patient charts.

Statistical analysis

Statistical analyses were performed using SPSS 24.0 (IBM, Armonk, NY, United States). Continuous variables are presented as means ± SD and were analyzed using the independent samples t-test. Categorical variables were presented as n (%) and analyzed using the χ² test. Logistic regression analysis was used to identify factors independently associated with CSVD. Voxel-based morphometry (VBM) analysis was performed using SPM8 on Matlab R2012b. Multiple regression analysis was performed to identify associations between ROIs and ICAM-1.

RESULTS

This study included 142 patients. Their average age was 60.36 ± 6.83 years. Sixty-one were males. Compared with the control group, the CSVD group was older (61.69 ± 7.13 vs 58.71 ± 6.10 years, P = 0.008). Sex, hypertension, diabetes, smoking, and drinking were comparable between the two groups (all P > 0.05; Table 1). Depression and cognitive assessment results by Hamilton depression scale (HAMD), MMSE, and MoCA-B were significantly different (all P < 0.001) between the CSVD and control groups.

Table 1 Clinical characteristics of the patients.

Characteristics	All (n = 142)	Control (n = 62)	CSVD (n = 80)	P value
Age (years)	60.39 ± 6.84	58.71 ± 6.10	61.69 ± 7.13	0.008
Male, n (%)	61 (43.0)	28 (45.2)	33 (41.3)	0.641
Hypertension, n (%)	63 (44.4)	23 (37.1)	40 (50.0)	0.125
Diabetes, n (%)	19 (13.4)	5 (8.1)	14 (17.5)	0.101
Smoking, n (%)	40 (28.2)	16 (25.8)	24 (30.0)	0.582
Drinking, n (%)	56 (39.4)	23 (37.1)	33 (41.3)	0.615
ICAM-1 (ng/mL)	258.11 ± 62.78	272.85 ± 63.90	246.69 ± 59.83	0.013
VCAM-1 (ng/mL)	285.77 ± 59.99	286.11 ± 55.08	285.51 ± 63.88	0.953
HAMD	16.01 ± 6.50	11.39 ± 4.03	19.60 ± 5.74	< 0.001
MMSE	26.36 ± 3.14	27.35 ± 2.72	25.59 ± 3.24	0.001
MoCA-B	25.51 ± 2.67	26.37 ± 2.21	24.85 ± 2.81	< 0.001
CRP (mg/L)	2.65 ± 4.03	2.43 ± 3.97	2.81± 4.09	0.578

CSVD: Cerebral small vessel disease; ICAM-1: Intercellular adhesion molecule-1; VCAM-1: Vascular cell adhesion molecule 1; HAMD: Hamilton depression scale; MMSE: Mini-mental state examination; MoCA-B: Montreal cognitive assessment basic; CRP: C-reactive protein.

Depression was present in 53 of 80 patients with CSVD and absent in 27. The HAMD, MMSE, and MoCA-B scores of the two groups significantly differed (all P < 0.05, Table 2).

Table 2 Correlation between depression and cognitive functioning among cerebral small vessel disease patients.

Characteristics	Depression (n = 53)	Non depression (n = 27)	P value
Age (years)	62.28 ± 7.83	60.52 ± 5.47	0.298
Male, n (%)	21 (39.6)	12 (44.4)	0.679
Hypertension, n (%)	28 (52.8)	12 (44.4)	0.478
Diabetes, n (%)	10 (18.9)	4 (14.8)	0.652
Smoking, n (%)	15 (28.3)	9 (33.3)	0.642
Drinking, n (%)	22 (41.5)	11 (40.7)	0.947
ICAM-1 (ng/mL)	244.14 ± 60.41	251.69 ± 59.50	0.597
VCAM-1 (ng/mL)	286.35 ± 63.82	283.87 ± 65.19	0.871
HAMD	22.92 ± 3.30	13.07 ± 3.41	< 0.001
MMSE	24.62 ± 2.96	27.52 ± 2.83	< 0.001
MoCA-B	24.26 ± 2.77	26.04 ± 2.47	0.006
CRP (mg/L)	3.09 ± 4.61	2.26 ± 2.82	0.393

ICAM-1: Intercellular adhesion molecule-1; VCAM-1: Vascular cell adhesion molecule 1; HAMD: Hamilton depression scale; MMSE: Mini-mental state examination; MoCA-B: Montreal cognitive assessment basic; CRP: C-reactive protein.

The multivariable analysis showed that age [P = 0.011; odds ratio (OR) = 0.930, 95% confidence interval (95%CI) = 0.880-0.983] and ICAM-1 levels (P = 0.023; OR = 1.007, 95%CI: 1.001-1.013) were independently associated with CSVD (Table 3).

Table 3 Multivariable analysis on cerebral small vessel disease.

Characteristics	B	SE	Wald	P value	OR	95%CI
						Lower	Upper
Sex	0.653	0.487	1.793	0.181	1.921	0.739	4.993
Age	-0.072	0.028	6.545	0.011	0.930	0.880	0.983
Hypertension	-0.309	0.384	0.647	0.421	0.734	0.346	1.559
Diabetes	-0.476	0.579	0.678	0.410	0.621	0.200	1.930
Smoking	-0.500	0.558	0.802	0.370	0.606	0.203	1.812
Drinking	-0.002	0.471	0.000	0.996	0.998	0.396	2.510
ICAM-1	0.007	0.003	5.141	0.023	1.007	1.001	1.013

OR: Odds ratio; CI: Confidence interval; ICAM-1: Intercellular adhesion molecule-1.

Using the VBM method, brain regions significantly correlated with ICAM-1 levels were identified. ROI1 and ROI2 from the gray matter and ROI3 and ROI4 from the white matter were selected for analysis. The average Ktrans value of each ROI was tested (Figure 1). The two ROIs in the white matter showed significant intergroup differences (both P < 0.001; 95%CI: 0.419-0.837 and 0.366-0.878), whereas only one ROI in the gray matter (P = 0.04; 95%CI: 0.007-0.680) showed significant difference (Table 4).

Figure 1 Color and grayscale images of the regions of interest and their average Ktrans values. ROI: Region of interest.

Table 4 Comparison of the mean Ktrans values of the region of interest regions between the cerebral small vessel disease group and the control group.

Regions	CSVD group	Control group	Significance	Mean difference	95%CI
					Lower	Upper
ROI1	2.509 ± 1.082	2.166 ± 0.897	0.046	0.343	0.007	0.680
ROI2	2.698 ± 0.987	2.695 ± 1.287	0.988	0.003	-0.374	0.380
ROI3	1.529 ± 0.748	0.901 ± 0.51	< 0.001	0.628	0.419	0.837
ROI4	1.571 ± 0.888	0.949 ± 0.654	< 0.001	0.622	0.366	0.878

CI: Confidence interval; CSVD: Cerebral small vessel disease; ROI: Region of interest.

The DKTAtlas portioning scheme results showed that ICAM-1 was significantly correlated (all P < 0.05) with cortical thickness in multiple brain regions in the CSVD group, whereas only limited correlations were observed in the control group (Figure 2 and Table 5).

Figure 2 Brain regions (blue area) where the cortical thickness was significantly correlated with intercellular adhesion molecule-1 in the control and cerebral small vessel disease groups. A: Regions where cortical thickness correlated with intercellular adhesion molecule-1 (ICAM-1) in healthy control group (independent sample t-test); B: Regions where cortical thickness correlated with ICAM-1 in cerebral small vessel disease group (independent sample t-test).

Table 5 Brain regions where cortical thickness correlates with intercellular adhesion molecule-1.

Group	Region	Correlation	Significance
CSVD	Right lingual	0.292	0.009
	Right middle temporal	0.310	0.005
	Right supramarginal	0.329	0.003
	Left entorhinal	0.230	0.040
	Left inferior parietal	0.240	0.032
	Left inferior temporal	0.247	0.027
	Left medial orbitofrontal	0.278	0.013
	Left middle temporal	0.230	0.040
	Left pars opercularis	0.337	0.002
	Left pars orbitalis	0.363	0.001
	Left pars triangularis	0.331	0.003
	Left rostral anterior cingulate	0.256	0.022
	Left rostral middle frontal	0.241	0.031
	Left superior temporal	0.226	0.044
	Left supramarginal	0.331	0.003
	Left insula	0.256	0.022
	Left mean thickness	0.235	0.036
Control	Right middle temporal	0.329	0.009
	Left entorhinal	-0.325	0.010
	Left pars orbitalis	0.254	0.046

CSVD: Cerebral small vessel disease.

DISCUSSION

Age and ICAM-1 levels were independently associated with CSVD. ROIs in the white matter were significantly different between the two groups, whereas only one ROI in the gray matter region was different between the two groups. Thus, serum ICAM-1 Levels can be used to screen for early CSVD. The four main clinical manifestations of CSVD are stroke, cognitive impairment, recurrent lacunar infarction, and white matter lesions[22,23]. Acute phase lesions increase the risk of lacunar infarction and cerebral parenchymal hemorrhage. Chronic phase lesions cause persistent brain tissue damage, leading to cognitive impairment, abnormal physical function, and emotional and personality disorders[24-27]. CSVD is common among individuals aged > 60 years, and its incidence is close to 100% in individuals aged > 90 years[27,28]. The multivariable analysis showed that age was independently associated with CSVD.

ICAM-1, as a regulator of the inflammatory response to vascular endothelial injury, can induce the release of toxic substances, causing vascular endothelial damage, which in turn leads to the adhesion and aggregation of inflammatory cells and microvasculature obstruction, ultimately leading to CSVD[29]. ICAM-1 expression in endothelial and epithelial cells is associated with the induction of inflammatory responses[3]. Another study reported that an increase in ICAM-1 also enhances the levels of interleukin (IL)-6 and causes asymptomatic cerebral infarction and white matter lesions[30,31]. Moreover, IL-6 and other proinflammatory cytokines aggravate vascular endothelial cell damage and BBB destruction[7,32]. ICAM-1 is predominantly expressed in endothelial cells, is overexpressed in many pathological conditions, and has functions similar to those of chemokines. In patients with CSVD, ICAM-1 can induce leukocyte aggregation and infiltration from the circulation to the inflammation site and increase the permeability of endothelial cells, resulting in pathological manifestations such as BBB damage, insufficient cerebral perfusion, and increases in the total cranial MRI load in patients with CSVD[33,34].

In addition, by analyzing the BBB permeability of ICAM-1 and associated brain regions, this study quantitatively assessed the correlation between ICAM-1 and BBB damage. Previous studies have shown that BBB penetration in the CNS is more common in patients with CSVD than in healthy patients[23,35,36]. Correlation analyses between ICAM-1 and cortical thickness were performed to determine whether BBB damage can cause changes in cortical thickness. The results showed that ICAM-1 displayed significant intergroup differences and was possibly associated with the pathogenesis of CSVD. Brain regions associated with ICAM-1, particularly the white matter, showed a significant increase in BBB permeability, suggesting the involvement of ICAM-1 in BBB damage. Moreover, inflammatory response is a crucial factor leading to BBB destruction in the early stages of CSVD, and white matter damage occurs earlier. ICAM-1 is expressed on endothelial cells and binds leukocytes through lymphocyte function-associated antigen 1, which mediates leukocyte adhesion to epithelial cells and promotes transendothelial migration[37-39]. Vascular permeability is tightly regulated by adhesion molecules between endothelial cells, which are activated at inflammation sites and perilymphatic endothelial cells, leading to increased expression of adhesion molecules during leukocyte infiltration. When leukocyte infiltration is initiated, these adhesion molecules activate endothelial cell signaling, which then alters the shape of the endothelium and opens pathways for leukocyte migration[40,41]. Blood vessels are pathways for blood flow within the body, including arteries, veins, and capillaries[42]. Capillaries comprise approximately 85% of the total cerebral vascular length and are the main sites of the BBB. The blood vessels in the CNS are continuous and non-enclosed and strictly regulate the movement of cells, molecules, and ions between CNS tissues and the bloodstream[43]. The strict confinement of the BBB precisely regulates internal homeostasis in the nervous system, which protects the brain’s neural tissues from invasion by diseases, injuries, inflammation, toxins, and pathogens and safeguards the normal function of neurons[44]. Therefore, normal BBB function is essential for the CNS.

The BBB is located between the brain parenchyma and blood vessels and is a dynamic interface between the brain parenchyma and blood, which is a composite organization composed of endothelial cells, basement membranes, pericytes, and their tight junctions[45]. Maintaining the structural and functional integrity of the BBB is essential for controlling the chemical composition of the extracellular fluid in the brain, which helps ensure normal synaptic information processing and neuronal connectivity. When the structural integrity of the BBB is disrupted, dysfunction is manifested as increased vascular permeability, and toxic cytokines, immune molecules and cells, microorganisms, and viruses in the bloodstream enter the brain, causing various ionic imbalances, signaling homeostasis, and activation of immune cells in the brain, resulting in various inflammatory reactions, which can lead to different neurodegenerative changes and sequelae[27,46]. Thus, ICAM-1 Levels predict BBB permeability and may serve as a biomarker for the early detection of CSVD. Nevertheless, this study had a cross-sectional design, and causal relationships could not be determined. Moreover, previous studies have highlighted the early development of cerebral microangiopathy with high BBB leakage in individuals with diabetes before the emergence of CSVD[47,48]; however, it was not observed in the present study. The conflicting results could be attributed to the study population and patient characteristics. CSVD affects several cognitive domains in asymptomatic individuals and in those with stroke or dementia[1,49]. In this study, the CSVD group had lower MMSE and MoCA-B scores than the control group.

However, this study has limitations. First, this study followed a cross-sectional design, causal relationships could not be determined, and selection bias was possible. Second, the patients were limited to those with asymptomatic CSVD given that the study aimed to examine ICAM-1 in early-onset CSVD and explore its utility as a surrogate marker of MRI-detected CSVD before symptom onset. ICAM-1 changes in the CSVD course were not detected. Third, diffusion tensor imaging can more effectively clarify the damage to white matter fiber bundles, which was not investigated in this study. In addition, DCE-MRI is time-consuming and mainly depends on patient cooperation, and patient movements during the procedure may compromise the accuracy of the results. Thus, a well-designed, large-sample longitudinal study is needed. Fourth, age was independently associated with CSVD, and older individuals typically experience chronic sterile low-grade inflammation, which were attributed to multiple cellular and biological mechanisms. In older individuals, inflammation predicts the risk of various age-related chronic noncommunicable diseases, including age-related CSVD. The close relationship between inflammation and CSVD presents the potential for inflammation-related therapies. However, few studies have focused on inflammation-related therapies and their effects. This gap is an important area of unmet medical need and warrants further translational research.

CONCLUSION

This study found that age and ICAM-1 levels were independently associated with CSVD. ICAM-1 may cause cortical thickness in the brain, predominantly in the white matter, and a significant increase in BBB permeability, demonstrating the involvement of ICAM-1 in BBB damage. The potential clinical utility of ICAM-1 as a CSVD biomarker is important for its early detection and thus the prevention of its progression.