Transient ischemic attack (TIA) is defined as sudden, transient neurological dysfunction resulting from focal cerebral ischemia.¹ The symptoms of TIA resemble those found early in a stroke and include sudden onset of weakness or numbness, typically on one side of the body, slurred speech, blindness, double vision, vertigo or loss of coordination. The etiology of TIA is heterogeneous. TIA can be caused by a variety of pathophysiological mechanisms.² Atherosclerosis leads to narrowing or blockage of the cerebral blood vessels. Embolus from the heart also cause TIA. Lastly, conditions such as hypotension or polycythemia can result in reduced cerebral blood flow and perfusion pressure. Precise classification of TIA patients based on their underlying pathophysiology is essential for their evaluation and management.

Blood viscosity (BV) is a measure of the thickness and stickiness of blood.³ BV is defined as the intrinsic resistance applied to blood flow and an essential predictor of endothelial shear stress. High BV can increase resistance against vessel walls, burden the heart, and reduce oxygen delivery to tissues, potentially causing a TIA.⁴ Studies have shown that BV is elevated in patients with stroke risk factors such as hypertension, diabetes, coronary artery disease (CAD), and tobacco use.⁵ Additionally, increased hematocrit (Hct) or fibrinogen concentration, key components of BV, are associated with a heightened stroke risk.^3,5,6

We previously reported that BV levels varied according to the potential cause of stroke within the stroke of undetermined etiology, negative evaluation (SUDn) group.⁷ SUDn due to possible atherothrombosis (AT) group exhibited a higher BV levels than the SUDn due to possible embolism (E) group. Endothelial damage, impaired blood flow, and hyperviscosity due to hemorheological alterations in intracranial atherosclerosis might be related to the higher BV in the SUDn-AT group. Regarding to TIA, little is known about the relationship between blood rheology and TIA onset. Previous study suggests that platelet count, white blood cell count, fibrinogen, plasma viscosity are not reliable predictors of a single TIA event.⁸ On the contrary, other research shows that patients with TIA have a hemorheologic deficit, resulting in an increase of viscous resistance.⁴ The hemorheologic changes exacerbate the negative effects of a vascular lesion, affecting blood perfusion and potentially leading to hypoperfusion and stroke.

Considering the unique mechanisms of stroke and TIA related to atherothrombosis and embolism, we hypothesized that BV levels might differ depending on the presence of cerebral atherosclerosis in TIA patients. In this context, this study investigate if there were any differences in BV levels according to the presence of cerebral artery stenosis in TIA patients.

This research was carried out at a single institution, employing a retrospective observational design to examine patients with TIA, from March 2017 to December 2021. For the study, consecutive patients aged over 20 who were admitted within 24 hours of symptoms onset were screened. TIA was defined as a brief episode of focal neurological dysfunction due to focal cerebral ischemia, which resolves within 24 hours.¹ Instances of transient visual disturbances linked to retinal ischemia, such as transient monocular blindness, were classified as TIA. Isolated transient symptoms like altered consciousness, dizziness, amnesia, confusion, vertigo, dysarthria, dysphagia, or diplopia were not classified as TIAs unless they were accompanied by other symptoms of brainstem ischemia.⁹ Patients who were clinically diagnosed with TIA and later found to have signs of cerebral infarction on magnetic resonance diffusion-weighted imaging (MR DWI) were considered to have a TIA. Exclusion criteria included: 1) patients whose baseline Hct level was either below 30% or above 50%, given the potential impact of Hct on BV, 2) patients who had taken antithrombotics within a 5-day period prior to the onset of TIA. 3) patients who had cerebral artery occlusion or highly suspicious of arterial dissection or hypercoagulable states.

During their hospital stay, all patients underwent inclusive examinations. Every patient had a brain computed tomography or MR imaging scan, along with an angiographic study. Factors such as demographics, medical history, and traditional vascular risk factors were assessed. Tests including 12-lead electrocardiography, complete blood counts, blood lipid profiles, renal and liver function, and coagulation factors were conducted. Although it was not mandatory, BV was examined as part of the systemic investigation. Except for those who did not give informed consent, patients underwent transthoracic echocardiography and 24-hour Holter monitoring. The techniques employed to measure BV in this study have been documented in previous report.¹⁰ A scanning capillary-tube viscometer (SCTV; Hemovister; Pharmode Inc., Seoul, Korea) was used to assess the whole BV. The SCTV is capable of measuring both systolic BV (SBV) and diastolic BV (DBV), which correspond to viscosities at high and low shear rates, respectively. In this study, SBV was evaluated at a shear rate of 300 s^–1, while DBV was gauged at 1 s^–1. All BV samples were collected before the initiation of hydration therapy, and the measurements were conducted within 24 hours of sample collection. The Research Ethics Committee (IRB No. SGPAIK 2022-03-011) approved this study. The requirement for informed consent was waived as the database was only accessed for analytical purposes.

For the study, patients were categorized into two groups for analysis: TIA-AT and TIA-E. In TIA-AT group, patients should have one or more stenosis of a major brain artery such as carotid, vertebral, basilar arteries, or proximal segments of the anterior, middle, and posterior cerebral arteries (Fig. 1A). In TIA-E group, patients should have normal major brain artery (Fig. 1B). The presence of AF was not considered in two grouping. Separately, patients were classified according to the ABCD² score and analyzed. The ABCD² score is a seven-point, risk-stratification tool used to predict the early risk of stroke following TIA. It is based on five parameters: age, blood pressure (BP), clinical features, duration of TIA, and presence of diabetes. The score classifies TIA patients at low, moderate or high risk using cut-off points of <4, 4–5 and >5.

A total of 153 patients (13% of the ischemic stroke or TIA patients during the study period) who had TIA were screened for enrolment. Of these, 67 patients (44%) were excluded from the study (37 patients with prior antithrombotics use, 22 patients due to no BV measurement, 8 patients due to Hct level <30 or >50%). Consequently, 86 patients were included for the final analysis.

The total study population had a mean age of 62.6 years, with 47% being female. The prevalence of hypertension was 59%, diabetes 21%, and dyslipidemia 45%. Stroke history was present in 6% of the population, CAD in 5%, and AF in 13%. Current smoking was reported by 24% and 34% were using statins. Twelve patients (14%) had cerebral infarction on MR DWI. Table 1 shows the baseline characteristics and laboratory findings of 86 patients according to the presence of cerebral artery stenosis: TIA-E (n=30) and TIA-AT (n=56) group. No differences were observed in the baseline characteristics between the groups. Though there was no statistical difference, the mean age in the TIA-E group tended to be slightly higher than the TIA-AT group (65.5 years vs. 61.1 years, p=0.101). Statin use was more common in the TIA-AT group (39.3% vs 23.3%, p=0.136). The presence of AF was were similar between the groups (p=0.431). No differences were observed in the baseline characteristics between the groups except for BV. In relation to BV, TIA-AT group had a higher SBV (p=0.033) and DBV (p=0.031) levels, demonstrating that the TIA-AT group had higher BV compared to the TIA-E group within 24 hours of symptoms onset. In multivariable logistic regression analysis using age, sex, statins use, DWI positivity, diastolic BP, SBV, and DBV showed that SBV (odds ratio [OR] 3.61; 95% confidence interval [CI], 1.06–12.27; p=0.04), DBV (OR 1.13; 95% CI, 1.03–1.23; p=0.01) were related to TIA-AT (Table 2). A partial correlation analysis, adjusted for Hct, showed an increased SBV (r=0.279, p=0.010) and DBV (r=0.258, p=0.017) levels in the TIA-AT group.

For the comparison, patients were also categorized into three groups according to the presence of AF: TIA-AF (n=11), TIA-E without AF (n=25), and TIA-AT without AF (n=50). There were no differences in the baseline characteristics and laboratory findings among the groups. The BV levels were comparable among the groups but there was a trend of having higher BV levels in TIA-AT without AF group (SBV: 4.49 vs. 4.34 vs. 4.60, respectively, p=0.103, DBV: 30.74 vs. 29.58 vs. 32.97, respectively, p=0.118).

Separately, patients were divided into three groups based on the ABCD² score: low-risk (n=25), medium-risk (n=45), and high-risk (n=16) (Table 3). The age increased with risk level, being statistically significant (p=0.004). The prevalence of diabetes was significantly higher in the high-risk group (p=0.004). Hypertension and dyslipidemia were more prevalent in the medium- and high-risk groups, but the differences were not significant. The use of statins was slightly higher in the medium- and high-risk groups. The laboratory findings of the study population according to the ABCD² score are presented in Table 3. No significant differences were observed in the baseline characteristics among the groups, except for high sensitive C-reactive protein (hs-CRP). The hs-CRP level was significantly higher in low-risk group (p=0.042).

The evaluation of TIA necessitates comprehensive brain and vascular imaging. About half of patients with TIA and DWI lesions have stenosis or occlusion of either extra or intracranial arteries.¹¹ The primary objective of vascular imaging is to identify patients with high grade cervical carotid stenosis who may be candidates for revascularization treatments. Symptomatic intracranial atherosclerosis is also related to an elevated risk of recurrent stroke.¹² Consequently, both extra and intracranial vascular imaging should be incorporated into the diagnostic workup of TIA.

Our study explored the potential differences in BV levels in relation to the presence cerebral artery stenosis in TIA patients, yielding several significant insights. First, we found that patients in the TIA-AT group had a higher BV levels within 24 hours of symptoms onset than those in the TIA-E group. This suggested that the underlying pathophysiological mechanisms of TIA may influence BV levels. The pathophysiology of TIA varies by subtype, all involving transient cerebral arterial blood flow interruption such as large artery atherosclerosis due to reducing blood flow due to major brain arterial stenosis or artery-to-artery embolism, small vessel disease resulting from lipohyalinosis or arteriolosclerosis, cardiac embolism involving a clot usually due to AF, cryptogenic often referring to as the embolic stroke of undetermined source, and other causes including arterial dissection or hypercoagulable states. In this study, the subtype classification of TIA-AT is supported by the presence of CAD. This was based on an idea that atherosclerosis is a systemic disease impacting multiple vascular beds.^7,13 AT is common etiologic mechanism of TIA. The cerebral artery stenosis is one of the common causes of TIA and accounts for 30–50% of strokes in Asian people. The proposed TIA mechanisms associated with cerebral artery stenosis may include hypoperfusion distal to the stenotic artery, artery to artery embolism, and branch atheromatous disease. The stability of plaques is critical as well as the degree of stenosis. One study shows that recurrent strokes are associated with stenosis <50% in the relevant artery or ≥50% stenosis in a nonrelevant artery. This highlights the importance of plaque stability and atherothrombotic mechanisms. Endothelial damage, impaired blood flow, and hypercoagulability collectively contribute to the initiation of thrombus formation. High BV may promote endothelial shear stress, endothelial inflammation and vascular remodeling.^14,15 High BV provokes plaque rupture, causes shear-induced platelet activation and provokes blood cell aggregation, reducing blood flow and ultimately leading to thrombosis.¹⁶ Within this context, BV plays a pivotal role as a primary mechanism for thrombus development. Considering the Virchow triad, which encompasses endothelial damage, altered blood flow, and hypercoagulability, it can be one plausible biological mechanism that thromboembolic susceptibility in cerebral artery stenosis may be influenced by these factors. Additionally, hemorheological alterations leading to hyperviscosity could be related to high BV within the TIA-AT group.³

Secondly, our findings highlighted the role of BV as a potential risk factor for TIA. Elevated BV can increase resistance against vessel walls, place additional strain on the heart, and decrease oxygen delivery to tissues, potentially triggering a TIA.¹⁷ This is in line with prior studies that have identified elevated BV in patients with stroke risk factors such as hypertension, diabetes, CAD, and tobacco use.³

Finally, TIA is recognized as a medical emergency, with a 7-day stroke risk as high as 30% in some patients.¹⁸ The ABCD² score attempt to quantify stroke risk after TIA and patients with the highest scores are significantly more likely to experience early stroke. In this study, the age increased with risk level and the prevalence of diabetes was significantly higher in the high-risk group. Controversy surrounds the application of the ABCD² score. While attempts to validate ABCD²-based scoring in independent populations have been generally successful, some studies have raised questions. One study showed that the ABCD² score results in only minor adjustments to baseline stroke risk, particularly in situations of very low initial risk and when used by nonspecialists.¹⁹ Patients defined as low risk by the ABCD² score had a negligible event rate, suggesting that the tool effectively distinguishes true TIA from other noncerebrovascular disease at lower scores. While the ABCD² score provides valuable risk stratification, additional studies may be needed for clinical judgment and complementary diagnostic tools in TIA assessment.

Our study had several limitations. Firstly, due to the retrospective collection of data from a single center, we were unable to establish a causal relationship between the onset of TIAs and BV levels. To overcome this limitation, conducting a large-scale prospective study across multiple centers could clarify the temporal relationship between BV levels before TIA onset and vascular stenosis. Secondly, the small sample size prevented us from further classifying the TIA-AT group according to extra or intracranial arterial stenosis, thus leaving the relationship between BV levels and the locations of cerebral artery stenosis unresolved. Increasing the sample size and further categorizing TIA patients based on the location of arterial stenosis could help compare BV levels more effectively. This approach would provide better insights into how the location of cerebral artery stenosis influences BV levels. Third, due to inadequate patient enrollment, our study did not perform further analysis based on the severity of cerebral artery stenosis but relied solely on visual inspection by researchers to confirm the presence of cerebral artery stenosis. Future studies with a sufficient number of participants are necessary to analyze hemorheological variables according to the degree of cerebral artery stenosis. Fourth, the presence of AF was not considered when classifying into two groups in this study. There was a trend of having higher BV levels in TIA-AT without AF group than TIA-AF and TIA-E with AF group (p=0.118). Though the presence of AF was were similar between the groups, this potential selection bias could impact the validity of our results. These limitations should be taken into account when interpreting our study results.

In conclusion, this study showed that TIA patients with cerebral artery stenosis had higher BV levels than those without stenosis within 24 hours of symptoms. Thromboembolic susceptibility in cerebral artery stenosis may be influenced by endothelial damage, altered blood flow, and hypercoagulability. Hemorheological alterations leading to hyperviscosity could be related to high BV within the TIA-AT group. The role of BV in the pathophysiological mechanism of TIA remains an area that requires further investigation. To gain deeper insights, additional research involving larger patient populations is essential, potentially leading to more conclusive findings.