Association between arterial tortuosity and early neurological deterioration in lenticulostriate artery infarction

Early neurological deterioration (END) in lenticulostriate artery (LSA) infarction is associated with perforating artery hypoperfusion. As middle cerebral artery (MCA) tortuosity may alter hemodynamics, we investigated the association between MCA tortuosity and END in LSA infarction. We reviewed patients with acute LSA infarction without significant MCA stenosis. END was defined as an increase of ≥ 2 or ≥ 1 in the National Institutes of Health Stroke Scale (NIHSS) total or motor score, respectively, within first 72 h. The MCA tortuosity index (actual /straight length) was measured. Stroke mechanisms were categorized as branch atheromatous disease (BAD; lesions > 10 mm and 4 axial slices) and lipohyalinotic degeneration (LD; lesion smaller than BAD). Factors associated with END in LD and BAD were investigated. END occurred in 104/390 (26.7%) patients. A high MCA tortuosity index (adjusted odds ratio, aOR 10.63, 95% confidence interval [2.57–44.08], p = 0.001) was independently associated with END. In patients with BAD, high initial NIHSS score (aOR 1.40 [1.03–1.89], p = 0.031) and presence of parental artery disease (stenosis < 50%; aOR 10.38 [1.85–58.08], p = 0.008) were associated with END. In patients with LD, high MCA tortuosity (aOR 41.78 [7.37–237.04], p < 0.001) was associated with END. The mechanism causing END in patients with LD and BAD may differ.


Results
During the study period, 3805 patients were admitted to our stroke center and 390 (10.2%) patients were diagnosed with acute ischemic stroke due to SVO in the LSA territory.The mean age of patients was 67 ± 12 years, 234 (54.5%) patients were men, and the median initial NIHSS score was 3 (1-5).Among these patients, 257 (65.9%) were diagnosed with LD and 133 (34.1%) with BAD.END occurred in 104 (26.7%) patients with 63 (24.5%) cases attributed to LD and 41 (30.8%) cases linked to BAD.
Presence of hypertension, parental artery disease, S-shaped MCA, severe initial NIHSS score, and a high MCA tortuosity index were associated with END.Multivariable analysis showed that a high MCA tortuosity index (adjusted odds ratio [aOR] 10.63; 95% confidence interval [CI] 2.57-44.08;p = 0.001) was independently associated with END in patients with LSA territory infarction (Table 2).3).
Although the percentage of patients with END (24.5% vs. 30.8%;p = 0.181) was similar, predictors of END differed between the LD and BAD subgroups.In patients with LD, a high MCA tortuosity index and S-shaped MCA were associated with END.In multivariable analysis, only a high MCA tortuosity index (aOR 41.78; 95% CI 7.37-237.04;p < 0.001) was independently associated with END in patients with LD (Table 4).In contrast, in patients with BAD, a high initial NIHSS score, and the presence of parental artery disease were associated with END.In multivariable analysis, a high initial NIHSS score (aOR 1.40; 95% CI 1.03-1.89;p = 0.031) and the presence of parental artery disease (aOR 10.38; 95% CI 1.85-58.08;p = 0.008) were independently associated with END in patients with BAD (Table 5).Nevertheless, there was no significant interaction between tortuosity index and stroke mechanism on END (p for interaction = 0.367).

Discussion
In the present study, the proportion of patients with END was 26.7% in patients with LSA territory infarction, which was consistent with the results of previous studies 2,3,7 .We found that high MCA tortuosity was independently associated with END.In particular, subgroup analysis of patients with LD showed that high MCA tortuosity was significantly associated with END, whereas the presence of parental artery disease was independently associated with END in patients with BAD (Table 5).
The mechanism associated with END in patients with LSA infarction is relatively homogenous and mainly involves growth of the infarction at the area of diffusion-perfusion mismatch 4,5,8 .As the perforator is an endartery, infarct growth is dependent on the flow through the perforator rather than through collaterals 9 .END was not rare in this population in the acute stage 2 .The proportion of patients showing END was still considerable among patients with LD.Although the LD pathology itself is stable, flow through the LSA in the acute stage of stroke may be influenced by arterial pressure, blood viscosity, and the tortuosity of the MCA and may lead to END 3,10 .
We have previously shown that the occurrence of LD in LSA territory was associated with a more tortuous shape of MCA 1 .Our current finding that the progression of LSA territory infarction with LD in the acute stage of stroke may be associated with MCA tortuosity can be explainable based on several hypotheses.First, tortuosity of a blood vessel is a systemic problem and the perforator itself may also show high tortuosity 11 .Perforators with a higher tortuosity may have a length longer than average, which is associated with END 12 .Second, matrix metalloproteinase (MMP) is associated with flow-induced vascular remodeling, resulting in high vascular tortuosity 13,14 .The acute inflammatory response regulated by MMP is also associated with the neurological worsening after acute lacunar infarction 15,16 .Third, the distortion and stretching of the perforators by the tortuous MCA may decrease the perfusion through the LSA, predisposing to END 6,10 .Finally, the tortuosity of the parent artery www.nature.com/scientificreports/may affect the local hemodynamics and induce local turbulence 17,18 .Local turbulence may reduce the perfusion through the perpendicularly branching perforators 6,10 .Contrarily, in patients with BAD, presence of parental artery disease was associated with END, which was consistent with the findings of a previous report 2 .This was explained by the atherosclerosis in the parent artery that blocked the orifice of the perforating artery 2 .High arterial tortuosity also enhances the progression of atheroma stemming from the low shear stress area distal to the atheroma 17,18 .However, in cases of BAD, the presence of atheroma in the parental artery, rather than the tortuosity, may be the rate-limiting factor influencing the flow through the perforator.Also, as high tortuosity index was significantly associated with END in overall patients, the relatively low tortuosity index in BAD patients may partially explain the lack of association between the tortuosity index and END.
Until recently, predictive factors associated with END were rarely studied in patients with specific mechanism of LD.Most reports have focused on the factors associated with END in patients with SVO, but not according to its detail mechanisms 4,5,7,8,10,19 .Based on the fact that END in LSA infarction leads to a high rate of functional disability, early identification of patients at risk of progression may improve their clinical and therapeutic management.
Our study has several limitations.First, the number of stroke cases was small, and the study was performed in a single center.Second, the diagnosis of BAD in the present study was defined using conventional MRI, which may not be sensitive enough to distinguish BAD from LD.A study with high-resolution vessel wall MRI may be helpful in distinguishing BAD from LD in future.Also, there is no standard etiological classification based on infarct morphology nor has this classification system been examined in large cohorts.Therefore, we additionally analyzed single subcortical infarction (SSIs) into 3 categories: SSI associated with parent artery disease (SSI-PAD), arteriosclerotic proximal perforator disease without PAD (pSSI-PAD), and distal perforator disease without PAD (dSSI-PAD) 20 .We found that high MCA tortuosity index was independently associated with distal perforator disease without PAD, which was the similar result from the present study.(Supplementary Tables 1 and  2) Third, vascular geometry data were based on a two-dimensional image of a three-dimensional reconstructed TOF-MRA.Recently, various methods involving automated high-performance techniques have been introduced www.nature.com/scientificreports/ to measure vascular tortuosity in three dimensions 17 .However, the inter-rater reliability of our method was fairly good (Cronbach's alpha: 0.799).Additionally, the simpler method of categorizing based on vessel shape can be more easily used in clinical practice.Finally, although we suggest an impaired hemodynamic status as one of the mechanisms of END, these issues could not be assessed in this retrospective study.Finally, although we suggest an impaired hemodynamic status as one of the mechanisms of END, these issues could not be assessed in this retrospective study due to the absence of perfusion imaging.Also, to achieve a more comprehensive relationship between atherosclerosis development and the specific tortuosity of perforating arteries and END, future prospective studies employing high-resolution MRI may be necessary.Despite the limitations of our study, our results suggest that high MCA tortuosity may be a factor associated with END, particularly in cases of SVO involving LD mechanism.Furthermore, the mechanism underlying END in cases with SVO related to LD and BAD may differ.As END in patients with LSA infarction leads to a high functional disability rate, early identification of patients who are at risk of progression may improve their clinical and therapeutic management.

Participants
We retrospectively reviewed patients with acute (< 7 days after stroke onset) LSA territory infarctions, confirmed by magnetic resonance imaging (MRI), who were admitted to the Asan Medical Center from January 2018 to June 2022.Patients were included in this study if they had isolated single small subcortical infarcts of no greater than 20 mm in diameter located within the territories of the LSA 21 and if they showed no or mild stenosis (< 50%) of MCA on magnetic resonance angiography (MRA).We excluded patients who had any of the following: (1) a significant (> 50%) stenosis of the corresponding extracranial or intracranial artery; (2) any potential causes of embolisms (i.e., embolic heart disease or coagulopathy), (3) other known causes of intracranial stenosis, regardless of the degree of stenosis (i.e., Moyamoya disease or intracranial arterial dissection).

Clinical data and END
Demographic data and risk factors were obtained by reviewing medical records and stroke registry database records.Hypertension was defined as receiving medication for hypertension or blood pressure > 140/90 mmHg on repeated measurements.Diabetes mellitus was defined as receiving medication for diabetes mellitus, fasting blood sugar ≥ 126 mg/dL, or 2-h postprandial blood sugar ≥ 200 mg/dL.Hyperlipidemia was defined as overnight fasting cholesterol level > 200 mg/dL or LDL ≥ 130 mg/dL.Previous stroke history was defined as history of stroke at least one diagnosed with a stroke by a physician.History of smoking was defined as people who had smoked before the current hospitalization episode, including current smokers and former smokers.The neurological deficit associated with stroke was evaluated using the National Institute of Health Stroke Scale (NIHSS) score at admission and at discharge.
END was defined as an increase, within the first 72 h of admission of ≥ 2 in the NIHSS total score or of ≥ 1 in the NIHSS motor score 10 , which was not considered to be caused by non-neurological conditions, such as worsening of the medical condition or body injury.The functional outcome was measured by using the modified Rankin scale score at 3 months after discharge and was categorized as good (score: 0-2) or poor (score: 3-6) by a stroke neurologist, either during the patients' outpatient visits or by telephone.

Stroke mechanisms
BAD was defined as infarcts caused by occlusion of the orifices or proximal portions of penetrating arteries.Based on diffusion-weighted imaging, BAD of the LSA was defined as infarcts > 10 mm in diameter on the axial slice and visible on four or more axial slices at a slice thickness of 7 mm 21 .Infarcts smaller than this were classified as LD 21 (Fig. 1B,C).

Imaging data and tortuosity
White matter lesions were defined as periventricular and deep white matter, and each region were rated using both the modified Fazekas scale.In periventricular white matter, severity was graded as 0 = absent; 1 = pencil-thin lining; 2 = halo of ≥ 5-mm thickness; 3 = irregular white matter hyperintensities extending into the deep white matter.In deep white matter, the severity was graded as 0 = absent, 1 = punctate foci, 2 = beginning confluence, 3 = large confluent areas, which was dichotomized into two groups: grades 0-1 and 2-3 22 .Cerebral microbleeds (CMBs) were identified as small perivascular hemosiderin deposits, which could be visualized as small, rounded, homogeneous, and hypointense lesions on T2*-weighed gradient-recalled echo or susceptibility-weighted images.For this study, all CMBs that had deep and lobar locations were included.Lacune was defined as a round or ovoid, subcortical, fluid-filled (similar signal as CSF) cavity, of between 3 mm and about 15 mm in diameter, consistent with a previous acute small deep brain infarct or hemorrhage in the territory of one perforating arteriole 23 .
Stenosis of the intracranial artery was evaluated based on the time-of-flight (TOF)-MRA.Parental artery disease was defined as a signal reduction of < 50% of the nearest normal sized vessel by referring to a prior segment

Figure 1 .
Figure 1.(A) Measuring of the MCA TI.The actual length of the MCA is (a + b + c), and the straight length is d.The TI was calculated as ([a + b + c]/d); BAD (B) and LD (C).MCA Middle cerebral artery, TI Tortuosity index, BAD Branch atheromatous disease, LD lipohyalinotic degeneration.

Table 1 .
Baseline characteristics of patients with and without early neurological deterioration.Results presented as number (%) or mean ± standard deviation or interquartile range.END Early neurological deterioration, BAD branch atheromatous disease, LD lipohyalinotic degeneration, NIHSS National Institutes of Health Stroke Scale, MCA middle cerebral artery.

Table 2 .
Factors associated with early neurological deterioration.Results are presented as odds ratio (OR) and 95% confidence intervals (CIs).BAD branch atheromatous disease, LD lipohyalinotic degeneration, NIHSS

Table 3 .
Baseline characteristics of patients according to the stroke mechanisms.Results are presented as number (%) or mean ± SD or IQR.END Early neurological deterioration, BAD branch atheromatous disease, LD lipohyalinotic degeneration, NIHSS National Institutes of Health Stroke Scale, MCA middle cerebral artery.

Table 4 .
Factors associated with early neurological deterioration in patients with lipohyalinotic degeneration.Results are presented as odds ratio (OR) and 95% confidence intervals (CIs).NIHSS National Institutes of Health Stroke Scale, MCA middle cerebral artery.*Multivariable logistic regression adjusted for age, sex, initial NIHSS score, and tortuosity index.

Table 5 .
Factors associated with early neurological deterioration in patients with branch atheromatous disease.Results are presented as odds ratio (OR) and 95% confidence intervals (CIs).NIHSS National Institutes of Health Stroke Scale, MCA middle cerebral artery.*Multivariablelogistic regression adjusted for age, sex, initial NIHSS score, hypertension, previous antiplatelet, and mild stenosis.The local ethics committee of Asan Medical Center, South Korea, approved this study (IRB number: 2021-1879).The need to obtain informed patient consent was waived due to the retrospective nature of the study.All methods of this study were performed following the relevant guidelines and regulations.