Comparison of three FLAIR vascular hyperintensities methodologies in patients with acute ischemic stroke

Abstract

Background

Multiple methods have been used to analyze fluid-attenuated inversion recovery (FLAIR) vascular hyperintensities (FVHs) which may represent collaterals in patients with acute ischemic stroke (AIS); however, there is no consensus between methods.

Purpose

To compare three frequently used FVH methods for predicting early infarct volume and clinical outcome in patients with AIS.

Material and Methods

Patients with AIS in middle cerebral artery territory were recruited. FVHs were evaluated using extensive FVHs, FVH-diffusion-weighted imaging (DWI) mismatch, and FVH-in/out-DWI. Infarct volume at baseline and day 7 were measured. Early neurological improvement (ENI) was assessed. Good outcomes were defined by modified Rankin Scale scores of 0–2 at 90 days.

Results

Fifty-one patients were included. ENI was 55.6% in patients with extensive FVHs and 23.3% in those without (P = 0.024). Patients with extensive FVHs had smaller infarct volume growth at seven days than those without (P = 0.041). ENI was 48.3% in patients with FVH-DWI mismatch and 15.8% in those without (P = 0.021). Patients with FVH-DWI mismatch had smaller infarct volumes at seven days than those without (P = 0.038). Patients with FVH-out-DWI had smaller baseline infarct volumes, smaller seven-day volumes, and smaller infarct growth than those with FVH-in-DWI (P<0.001, P<0.001, and P = 0.031, respectively). In multivariate logistic regression analysis, the infarct growth at seven days negatively independently predicted ENI (OR = 0.737, 95% CI 0.593–0.915, P = 0.006). However, none of the FVH classifications could predict a good 90-day outcome.

Conclusion

Patients with extensive FVHs or FVH-DWI mismatch tend to have early favorable clinical outcome. FVH-out-DWI being associated with smaller infarct growth may also indicate early favorable clinical outcome.

Keywords

Acute brain ischemia/infarction magnetic resonance imaging magnetic resonance angiography

Introduction

Infarctions caused by proximal arterial occlusion, such as the middle cerebral artery (MCA), ultimately demonstrate variable distribution and volume. This may be due to different leptomemingeal collaterals in individuals. Conventional digital subtraction angiography (DSA) remains the gold standard to assess the status of collaterals, but it is invasive. In non-invasive magnetic resonance (MR) examination, fluid-attenuated inversion recovery (FLAIR) vascular hyperintensities (FVHs) may represent slow retrograde flow in leptomeningeal collaterals. However, the effect of FVHs on the prognosis of acute ischemic stroke (AIS) remains controversial. There are studies suggested that FVHs indicate a good prognosis (1 –3); contrarily, studies also demonstrated that they indicate a poor prognosis (4 –6). There are also studies showing no relation between FVHs and prognosis (7). These discrepancies are likely caused by differences among populations, imaging time, and methodologies used to determine FVHs. Although various methods have been proposed to determine FVHs, there is no consensus on the optimal choice.

The aim of the present study was to compare the three most commonly used methods for assessing FVHs to determine their value in predicting clinical outcomes in symptomatic patients with AIS.

Material and Methods

Patients

Consecutive patients with AIS who were admitted to our hospital between January 2012 and December 2015 were retrospectively assessed in the present study. The inclusion criteria were as follows: time between stroke onset and having a MR scan study <8 h; male or female patients aged 18–80 years; the presence of MCA occlusion/severe stenosis; no history of previous stroke; and pre-stroke modified Rankin Scale (mRS) score ≤ 2. The exclusion criteria were as follows: intracranial hemorrhage or brain tumors; serum glucose level < 2.7 mmol/L or >22.2 mmol/L; and inadequate image quality including severe motion artifacts. The study protocol was approved by the ethics committee of our hospital. Written consent was obtained from patients or their next of kin.

Follow-up protocol and clinical assessments

Multimodal MR protocol at baseline and follow-up MR protocol at seven days were obtained. Treatment strategies included recanalization treatments (rt-PA intravenous thrombolysis and/or endovascular treatment) and routine treatment according to their MR results and clinical information. The routine treatment contained treatment of any complications and secondary prevention such as antiplatelet therapy and management of risk factors. The routine treatment was the only treatment in patients without recanalization treatments or followed recanalization treatments. Lifestyle modification was included in the secondary stroke prevention after discharge.

According to predefined criteria (8), vascular risk factors such as history of atrial fibrillation, transient ischemic attack (TIA), hypertension, diabetes mellitus, and hyperlipidemia were obtained on admission.

National Institutes of Health Stroke Scale (NIHSS) scores were assessed at baseline before treatment and at seven days as well as mRS scores at three months. Early neurological improvement (ENI) was defined as a decrease of the NIHSS of at least 8 points between baseline and seven days after onset or a NIHSS score of 0–1 at day 7. Good and poor functional outcomes at three months were defined by mRS scores of 0–2 and 3–6, respectively.

Imaging

MR imaging (MRI) has been routinely implemented in our hospital in patients with AIS presenting < 8 h from ictus. All the MR examinations were performed on 3.0-T machine (Trio-Tim, Siemens, Erlangen). The multimodal MR protocol included diffusion-weighted imaging (DWI), FLAIR, and time of flight (TOF) MR angiography (MRA). DWIs were acquired with a single-shot echo-planar imaging with b values of 0 s/mm² and 1000 s/mm², respectively, with the following parameters: repetition time (TR)/echo time (TE) =3000/75 ms; field of view (FOV) = 23 × 23 cm; and matrix = 128 × 128. Apparent diffusion coefficient maps were created from DWI images with b values of 0 s/mm² and 1000 s/mm². FLAIR parameters were: TR/TE/inversion time = 8000/94/2500 ms; FOV =20 × 17.6 cm; matrix = 256 × 179; and flip angle (FA) = 150°. TOF MRA parameters were: TR/TE =28/3.04 ms; FOV = 20 × 18 cm; matrix = 256 × 179; thickness = 0.7 mm; slices per slab = 40; and FA = 13°. The follow-up MR protocol included, but was not limited to, DWI, FLAIR, and TOF MRA.

Image analysis

All MR images were anonymized and reviewed at the same workstation, independently, by two experienced neuroradiologists who were blinded to the patients’ clinical information. FVHs were defined as focal, tubular, or serpentine hyperintensities in the subarachnoid space relative to cerebrospinal fluid signal which were distal to the M1 segment. Three predefined methods—FVH Alberta Stroke Program Early Computed Tomography Scores (ASPECTS), FVH-DWI mismatch, and FVH-in/out-DWI—were used to evaluate FVHs. Discordance between observers was resolved by consensus.

FVH ASPECTS

Seven ASPECTS cortical areas (insular, M1–M6) were used to assess FVH scores according to their spatial distribution (9). An ASPECTS cortical area was considered to be positive when a FVH was geographically associated with it. The FVH score ranged from 0 (no FVH) to 7 (FVHs abutting all ASPECTS cortical areas). For further analyses, FVHs were dichotomized into extensive FVHs (FVH scores 5–7) and non-extensive FVHs (FVH scores 0–4).

FVH-DWI mismatch

When FVHs were located in the region beyond the boundaries of the cortical DWI lesion, FVH-DWI mismatch was considered (10). If the ischemic lesion was in the basal ganglia but FVH ASPECTS were among M1–M6 area, FVH-DWI mismatch was also considered. FVH-DWI mismatch was not considered in the following situations: there was no FVH; all FVHs were involved in the hyperintense cortex on DWI; or the ischemic lesion was in the basal ganglia while FVH ASPECTS was in the “insular” area.

FVH-in/out-DWI

FVH-in/out-DWI indicates whether the FVH distribution was inside or outside the DWI abnormality (11). Liu et al. (11) assessed FVH using both FVH-in/out-DWI and FVH ASPECTS, and then divided FVH-in/out-ASPECTS into two subcatorgeries, respectively, which was complicated. In the present study, we simplified their method using FVH-in/out-DWI as an independent method in patients with FVHs. As long as FVHs extended into the sulci of the ischemic stroke region, FVH-in-DWI was considered; otherwise, FVH-out-DWI was considered. When the ischemic stroke occurred in the basal ganglia and/or deep white matter without cortical involvement, FVH-out-DWI was also considered.

Arterial occlusion or stenosis was evaluated by TOF MRA according to the modified Thrombolysis In Myocardial Infarction (TIMI) scale (12): TIMI 0 = complete occlusion; TIMI 1 = severe stenosis; TIMI 2 = mild or moderate stenosis; and TIMI 3 = normal. Vascular status was divided into two subgroups: TIMI 0–1 and TIMI 2–3. Both baseline infarct volumes (volume 1) and follow-up infarct volumes (volume 2) were measured by using DWI. Infarct growth was defined as the difference between volume 2 and volume 1. These operations were independently performed by a third neuroradiologist using 3D Slicer version 4.8 (http://www.slicer.org). For each patient, the semi-automatically measured volumes were assessed for subsequent quantitative comparison.

Statistical methods

All statistical analyses were performed using SPSS version 23.0 (IBM Corp., Armonk, NY, USA). The normally distributed continuous variables were reported as mean ± SD. The non-normally distributed continuous variables were reported as median (interquartile range). Categorical variables were presented as percentages. Continuous variables were compared by using Student’s t test or Mann–Whitney U test where appropriate. Categorical variables were tested by using the χ² test or Fisher’s exact test where appropriate. The independent factors associated with early good outcome (ENI) were analyzed by binary logistic regression analysis. The variables with P < 0.05 from the univariate analysis were entered into the logistic regression analysis. Odds ratio (OR) and 95% confidence interval (CI) were calculated. A value of P < 0.05 was considered statistically significant.

Results

Patient clinical characteristics

From 224 patients with a clinical diagnosis of AIS, infarction located in the MCA territory were found in 142 cases of which there are 80 cases with TIMI0 or TIMI1 of the MCA. Due to the absence of FLAIR sequence or artifacts on FLAIR images in 29 cases, 51 cases were finally included in the present study which are detailed in Table 1. Of 46 (90.2%) cases demonstrated FVHs’ positivity. Of 46 cases with FVHs, acute M1 segment occlusion (TIMI 0) of the MCA was found in 38 cases on TOF MRA, and severe M1 segment stenosis (TIMI 1) was found in the remaining eight cases. There were five cases with M1 segment occlusion without positive FVHs. Of 22 patients with disappearance of FVHs at seven days, 20/22 (90.9%) cases were found in patients with MCA TIMI 2–3 (P < 0.001).

Table 1.

Clinical and imaging characteristics of the study population (n = 51).

Characteristics	Values
Baseline
Age (years)	59.1 ± 11.9
Female	16 (31.4)
Risk factors
AF	11 (21.6)
Prior TIA	6 (11.8)
Hypertension	22 (43.1)
Diabetes	3 (5.9)
Hyperlipidemia	7 (13.7)
NIHSS	7 (4–11)
Onset to MRI time (min)	278.7 ± 97.4
FVHs
Positive	46 (90.2)
Negative	5 (9.8)
M1 segment of MCA
TIMI 0	43 (84.3)
TIMI 1	8 (15.9)
Volume 1 (mL)	7.6 (2.7–24.3)
Treatment
Intravenous thrombolysis	15 (29.4)
Endovascular treatment	13 (25.5)
Routine treatment only	23 (45.1)
Seven-day follow-up
NIHSS	4 (2–7)*
ENI	17/48 (35.4)
Volume 2 (mL)	26.3 (8.9–57.9)^†
Infarct growth (mL)	12.1 (0.4–27.1)^†
Disappearance of FVHs	22/44 (50.0)
MCA TIMI 2–3	31/50 (62.0)
MCA TIMI 2–3 after intravenous thrombolysis	11/15 (73.3)
MCA TIMI 2–3 after endovascular treatment	9/13 (69.2)
MCA TIMI 2–3 after routine treatment recanalization	11/22 (52.4)
90-day good outcome	38/50 (76.0)

Values are given as n (%), mean ± SD, or median (IQR).

*Values in 48 cases.

^†Values in 50 cases.

AF, atrial fibrillation; ENI, early neurological improvement; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity; IQR, interquartile range; MCA, middle cerebral artery; MRI, magnetic resonance imaging; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack; TIMI Thrombolysis In Myocardial Infarction.

Three FVHs methods in prediction of ischemic stroke outcome

According to FVH ASPECTS, there were 42/46 (91.3%) cases in which the insular region was involved, of which FVH was found merely in the insular region in five cases. There were 41/46 (89.1%) cases in which M1–M6 regions were involved variably. When compared with the 23.3% of patients without extensive FVHs, 55.6% of patients with extensive FVHs had an ENI at the seven-day follow-up (P = 0.024) (Table 2, Fig. 1). Patients with extensive FVHs had less infarct volume growth at seven days than patients without extensive FVHs (P = 0.041) (Fig. 2). Compared with 15.8% of patients without FVH-DWI mismatch, 48.3% of patients with FVH-DWI mismatch had ENI at seven days (P = 0.021) (Table 3, Fig. 1). Meanwhile, patients with FVH-DWI mismatch had smaller infarct volumes at seven days after presentation than those without FVH-DWI mismatch (P = 0.038) (Fig. 3). Patients with FVH-out-DWI had smaller infarct volumes at baseline, and smaller infarct volumes at seven days, as well as smaller infarct volume growth after stroke than those with FVH-in-DWI (P < 0.001, P < 0.001, P = 0.031, respectively) (Table 4, Figs. 4 and 5).

Table 2.

Comparison of patient demography and clinical presentation according to FVH ASPECTS.

Characteristics	Non-extensive FVHs (n = 31)	Extensive FVHs (n = 20)	P value
Baseline
Age (years)	60.3 ± 11.3	57.1 ± 12.7	0.439
Female sex	8 (25.8)	8 (40.0)	0.286
Risk factors
AF	4 (12.9)	7 (35.0)	0.085
Prior TIA	4 (12.9)	2 (10.0)	1.000
Hypertension	10 (32.3)	12 (60.0)	0.051
Diabetes	2 (6.5)	1 (5.0)	1.000
Hyperlipidemia	4 (12.9)	3 (15.0)	1.000
Onset to MRI time (min)	276.1 ± 103.9	282.6 ± 88.8	0.821
NIHSS	7 (4–9)	8 (2–12)	0.960
Volume 1 (mL)	6.2 (1.9–35.1)	8.3 (5.0–18.3)	0.877
Seven-day follow-up
NIHSS*	4 (2–8)	3 (1–5)	0.441
ENI*	7/30 (23.3)	10/18 (55.6)	0.024
Volume 2 (mL)^†	34.0 (10.5–78.6)	17.2 (7.4–34.5)	0.060
Infarct growth (mL)^†	15.1 (4.5–41.3)	4.6 (−0.3–19.7)	0.041
MCA TIMI 2–3^†	17/30 (56.7)	14/20 (70.0)	0.341
Disappearance of FVHs^‡	11/24 (45.8)	11/20 (55.0)	0.545
90-day good outcome^†	23 (74.2)	15/19 (79.0)	1.000

Values are given as n (%), mean ± SD, or median (IQR).

*Values in 48 cases.

^†Values in 50 cases.

^‡Values in 44 cases.

Fig. 1.

An illustrative case of FLAIR vascular hyperintensities (FVHs). DWI (a) and ADC (b) demonstrate acute infarction (arrow) with MCA TIMI 0 (c, MRA). FLAIR (d) demonstrates FVH ASPECTS = 5, which indicate extensive FVHs. FVHs (circle) extended beyond the infarction boundary which indicated FVH-DWI mismatch. Follow-up FLAIR (e) demonstrates FVHs disappearance with MCA recanalization (f, arrow). ADC, apparent diffusion coefficient; DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity; MCA, middle cerebral artery; MRA, magnetic resonance angiography.

Fig. 2.

Baseline infarct volume, follow-up infarct volume, and infarct growth in patients with and without extensive FLAIR vascular hypersensitivities. FLAIR, fluid-attenuated inversion recovery; FVHs, FLAIR vascular hyperintensities.

Table 3.

Comparison of patient demography and clinical presentation according to FVH-DWI mismatch.

Characteristics	FVH-DWI mismatch – (n = 20)	FVH-DWI mismatch + (n = 31)	P value
Baseline
Age (years)	62.6 ± 9.6	56.8 ± 12.8	0.087
Female	5 (25.0)	11 (35.5)	0.431
Risk factors
AF	3 (15.0)	8 (25.8)	0.493
Prior TIA	3 (15.0)	3 (9.7)	0.668
Hypertension	9 (45.0)	13 (41.9)	0.829
Diabetes	1 (5.0)	2 (6.5)	1.000
Hyperlipidemia	3 (15.0)	4 (12.9)	1.000
Onset to MRI time (min)	279.2 ± 107.1	278.3 ± 92.4	0.974
NIHSS	6 (4–12)	8 (3–11)	0.600
Volume 1 (mL)	17.8 (2.0–51.2)	6.9 (4.6–16.0)	0.255
Seven-day follow-up
NIHSS*	4 (2–9)	3 (1–7)	0.819
ENI*	3/19 (15.8)	14/29 (48.3)	0.021
Volume 2 (mL)^†	51.6 (11.0–89.2)	20.4 (8.9–33.6)	0.038
Infarct growth (mL)^†	15.8 (3.7–49.6)	6.1 (−0.2–26.9)	0.104
MCA TIMI 2–3^†	11 (55.0)	20/30 (66.7)	0.405
Disappearance of FVHs^‡	7/16 (43.8)	15/28 (53.6)	0.531
90-day good outcome^†	15 (75.0)	23/30 (76.7)	1.000

Values are given as n (%), mean ± SD, or median (IQR).

*Values in 48 cases.

^†Values in 50 cases.

^‡Values in 44 cases.

AF, atrial fibrillation; DWI, diffusion-weighted imaging; ENI, early neurological improvement; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity; IQR, interquartile range; MCA, middle cerebral artery; MRI, magnetic resonance imaging; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack; TIMI Thrombolysis In Myocardial Infarction.

Fig. 3.

Baseline infarct volume, follow-up infarct volume, and infarct growth in patients with and without FVH-DWI mismatch. DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity.

Table 4.

Comparison of patient demography and clinical presentation according to FVH-in/out-DWI.

Characteristics	FVH-in-DWI (n = 23)	FVH-out-DWI (n = 23)	P value
Baseline
Age (years)	60.4 ± 11.6	57.6 ± 12.8	0.437
Female	7 (30.4)	9 (39.1)	0.536
Risk factors
AF	6 (26.1)	5 (21.7)	0.730
Prior TIA	1 (4.4)	3 (13.0)	0.608
Hypertension	10 (43.5)	10 (43.5)	1.000
Diabetes	1 (4.4)	2 (8.7)	1.000
Hyperlipidemia	1 (4.4)	5 (21.7)	0.187
Onset to MRI time (min)	287.7 ± 114.5	263.4 ± 83.8	0.417
NIHSS	8 (5–13)	6 (2–8)	0.238
Volume 1 (mL)	24.3 (6.2–40.4)	5.5 (2.4–8.1)	<0.001
Seven-day follow-up
NIHSS*	5 (2–10)	3 (1–5)	0.284
ENI*	8/22 (36.4)	9/21 (42.9)	0.663
Volume 2 (mL)^†	42.7 (23.7–86.3)	11.0 (6.9–27.9)	<0.001
Infarct growth (mL)^†	16.4 (5.6–41.6)	5.4 (−1.1–16.5)	0.031
MCA TIMI 2–3^†	14 (60.9)	15/22 (68.2)	0.608
Disappearance of FVHs^‡	10/20 (50.0)	9/20 (45.0)	0.752
90-day good outcome^†	15 (65.2)	19/22 (86.4)	0.099

Values are given as n (%), mean ± SD, or median (IQR).

*Values in 43 cases.

^†Values in 45 cases.

^‡Values in 40 cases.

Fig. 4.

Illustrative cases of FVH-in/out-DWI. FVH-in-DWI is presented in case 1 (a–d). FVHs (arrows, c, d) is covered by infarction (DWI, a, b). FVH-out-DWI is presented in case 2 (e, f). FVHs (circle, f) is out of infarction area in the right basal ganglia. DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity.

Fig. 5.

Baseline infarct volume, follow-up infarct volume and infarct growth in patients with FVHs-in/out-DWI. DWI, diffusion-weighted imaging; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity.

In univariate analysis, patients with ENI had a higher percentage of extensive FVHs, FVH-DWI mismatch at baseline, MCA TIMI 2–3 at seven days, and smaller infarct volumes and infarct growth at seven days than those without ENI (P = 0.024, P = 0.021, P = 0.002, P = 0.014, P < 0.001, respectively (Table 5). In multivariate logistic regression analysis, the infarct growth at seven days negatively independently predicted ENI (OR = 0.737, 95% CI = 0.593–0.915, P = 0.006) (Table 6).

Table 5.

Univariate analysis according to the ENI.

Characteristics	ENI (n = 17)	No ENI (n = 31)	P value
Baseline
Age (years)	56.0 ± 12.1	60.7 ± 10.5	0.166
Female	6 (35.3)	8 (25.8)	0.522
Risk factors
AF	4 (23.5)	6 (19.4)	0.727
Prior TIA	3 (17.7)	3 (9.7)	0.651
Hypertension	9 (52.9)	12 (38.7)	0.342
Diabetes	1 (5.9)	2 (6.5)	1.000
Hyperlipidemia	4 (23.5)	2 (6.5)	0.167
Onset to MRI time (min)	282.9 ± 97.2	257.8 ± 89.2	0.385
NIHSS	8 (2–13)	7 (4–9)	0.754
Volume 1 (mL)	10.8 (5.7–24.9)	5.1 (1.4–32.5)	0.129
Extensive FVH	10 (58.8)	8 (25.8)	0.024
FVH/DWI mismatch	14 (82.4)	15 (48.4)	0.021
FVH-out-DWI	9 (52.9)	12/26 (46.2)	0.663
Recanalization therapies	11 (64.7)	15 (48.4)	0.278
Seven-day follow-up
Volume 2 (mL)*	13.6 (5.7–32.7)	32.6 (15.9–76.0)	0.014
Infarct growth (mL)*	−0.2 (−2.5–6.4)	21.8 (6.4–53.8)	<0.001
MCA TIMI 2–3	15/16 (93.5)	15 (48.4)	0.002

Values are given as n (%), mean ± SD, or median (IQR).

*Values in 47 cases.

Table 6.

Multivariate logistic regression analysis for the ENI.

Characteristics	P	OR	95% CI
Extensive FVHs	0.599	2.129	0.127–35.692
FVH/DWI mismatch	0.898	1.230	0.051–29.600
Volume 2 (mL)	0.120	1.045	0.988–1.106
Infarct growth (mL)	0.006	0.737	0.593–0.915
MCA TIMI 2–3	0.065	82.188	0.756–8931.362

CI, confidence interval; DWI, diffusion-weighted imaging; ENI, early neurological improvement; FLAIR, fluid-attenuated inversion recovery; FVH, FLAIR vascular hyperintensity; MCA, middle cerebral artery; OR, odds ratio; TIMI, Thrombolysis In Myocardial Infarction.

The mRS at 90 days were not, however, significantly different between patients with and without extensive FVHs (P = 0.967), with and without FVH-DWI mismatch (P = 1.000), with FVH-in-DWI and FVH-out-DWI (P = 0.099).

Discussion

Several approaches have been proposed to investigate the extent of FVHs. Kufner et al. (4) counted the number of axial FLAIR sections with FVHs to determine whether the FVHs were extensive or not. This approach is dependent on the number of slices and slice thickness. Moreover, it only provides the rostrocaudal extension of FVHs. Huang et al. (1) have dichotomously categorized patients into distal or proximal FVHs according to their sulcal location. In the present study, both the insular (proximal) and M1–M6 (distal) FVHs were simultaneously observed in the majority of patients according to their FVHs ASPECTS. Due to the limitations of the above methodologies, in the present study none of them were adopted to assess FVHs.

In the present study, three approaches were used to evaluate FVHs including FVHs ASPECTS (extensive FVHs) which may reflect the distribution of FVHs rostrocaudally and inferosuperiorly, and FVH-DWI mismatch and FVH-in/out-DWI, both of which assessed FVHs compared with infarct topography. Patients with extensive FVHs had a higher percentage of ENI and less infarct growth at day 7. Cases with FVH-DWI mismatch had a higher percentage of ENI and smaller infarct volumes at day 7, which was consistent with previous studies (10). Compared with extensive FVHs and FVH-DWI mismatch, FVH-out-DWI appears to be associated with infarct volumes including the infarct volumes at baseline and day 7 and infarct growth. It has previously been reported that the presence of good collateral flow on conventional angiography is associated with smaller infarcts and better clinical outcome in patients with AIS (13). Hence, extensive FVHs, FVH-DWI mismatch, and FVH-out-DWI may imply leptomeningeal collaterals. Liu et al. (11) reported that the number of distal FVH-ASPECTS had no prognostic value in patients with acute MCA occlusion, but FVH-ASPECTS combined with DWI showed good clinical outcome. In the present study, both extensive FVHs and FVH-DWI mismatch, which may reflect impaired yet viable tissue, were found to be valuable to be associated with early functional improvement. Although we did not find any significant difference in ENI between patients with FVH-in-DWI and FVH-out-DWI, the latter was strongly associated with either smaller infarct volumes at baseline and day 7 or infarct growth. As we knew, the infarct growth at day 7 could negatively predict the early outcome after the logistic regression, so FVH-out-DWI could reflect good clinical outcome. Meanwhile, extensive FVHs were found to be related with early favorable outcome in either direct or indirect ways. Therefore, compared with the two other methods mentioned above, extensive FVHs might be more effective and comprehensive in prediction of early clinical outcome after stroke.

Our results demonstrated that none of the FVH classifications at baseline could determine directly good clinical outcome 90 days after stroke onset. This result is similar with previous reports (14) despite good collaterals that FVHs have been proved to represent in their study. The controversial value of FVHs in assessing infarction volume and predicting post-stroke treatment outcome might be due to the following reasons. On one hand, FVHs are reported to be related to retrograde collateral flow, but blood reaches cortical areas later than with anterograde flow (14), which may indicate that collateral flow might be insufficient. Moreover, good collaterals may not be visible on FLAIR sequences (4), which have been reported rarely. It is hard to get the subsequent confirmation on DSA due to its invasiveness when there is no symptomatic infarction in cases with probable good collaterals. We demonstrated a 90.9% disappearance of FVHs occurring in cases with MCA TIMI 2–3 at follow-up, which indicated that the impaired hemodynamics could be recovered after recanalization. The absence of FVHs is not, however, equivalent to the absence of effective collaterals. On the other hand, the clinical outcome is related to many factors including differing therapeutic strategies, recanalization of the occluded MCA, hemorrhagic transformation, and so on. In the present study, the treatment plans for patients with MCA occlusion were quite variable, recanalization and routine treatments both being demonstrated. Due to the various factors mentioned above, it is very hard to predict the long-term clinical prognosis using a sole controversial imaging sign at baseline after ischemic stroke.

Despite these interesting findings, limitations exist. First, the sample size is small. Second, it is known that FVHs might change over a short period of time. In the present study, FVHs were evaluated only at baseline and on day 7. Third, 13/51 (25.5%) patients underwent DSA before endovascular therapy, meaning the association of FVHs with collateral status on angiography could not be analyzed.

In conclusion, we provide a comparative assessment of three methods for FVHs establishing infarct volume and clinical outcome in patients with AIS. Patients with extensive FVHs or FVH-DWI mismatch tend to have early favorable clinical outcome. Extensive FVHs and FVH-out-DWI were found to be associated with smaller infarct growth in patients with AIS, which may indicate early favorable outcome. However, none of FVHs classifications could predict good outcome at 90 days.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received the following financial support for the research, authorship, and/or publication of this article: This study was supported by grants from the National Natural Science Foundation of China (grant nos. 81301193 and 81361120402) and the Beijing Natural Science Foundation of China (grant nos. 7162056 and 7133238).

ORCID iD

Lina Jing

References

Huang

Liu

Zhu

, et al. Distal hyperintense vessels on FLAIR: a prognostic indicator of acute ischemic stroke. Eur Neurol 2012; 68:214–220.

Perez de la Ossa

Hernandez-Perez

Domenech

, et al. Hyperintensity of distal vessels on FLAIR is associated with slow progression of the infarction in acute ischemic stroke. Cerebrovasc Dis 2012; 34:376–384.

Nave

Kufner

Bucke

, et al. Hyperintense vessels, collateralization, and functional outcome in patients with stroke receiving endovascular treatment. Stroke 2018; 49:675–681.

Kufner

Galinovic

Ambrosi

, et al. Hyperintense vessels on FLAIR: hemodynamic correlates and response to thrombolysis. AJNR Am J Neuroradiol 2015; 36:1426–1430.

Ebinger

Kufner

Galinovic

, et al. Fluid-attenuated inversion recovery images and stroke outcome after thrombolysis. Stroke 2012; 43:539–542.

Nam

Kim

, et al. FLAIR vascular hyperintensities predict early ischemic recurrence in TIA. Neurology 2018; 90:e738–e744.

Schellinger

Chalela

Kang

, et al. Diagnostic and prognostic value of early MR Imaging vessel signs in hyperacute stroke patients imaged <3 hours and treated with recombinant tissue plasminogen activator. AJNR Am J Neuroradiol 2005; 26:618–624.

Kufner

Nolte

Galinovic

, et al. Smoking-thrombolysis paradox: recanalization and reperfusion rates after intravenous tissue plasminogen activator in smokers with ischemic stroke. Stroke 2013; 44:407–413.

Lee

Seo

Kim

, et al. Correlation between hyperintense vessels on FLAIR imaging and arterial circulation time on cerebral angiography. Magn Reson Med Sci 2016; 15:105–110.

10.

Legrand

Tisserand

Turc

, et al. Do FLAIR vascular hyperintensities beyond the DWI lesion represent the ischemic penumbra? AJNR Am J Neuroradiol 2015; 36:269–274.

11.

Liu

Scalzo

Rao

, et al. Fluid-attenuated inversion recovery vascular hyperintensity topography, novel imaging marker for revascularization in middle cerebral artery occlusion. Stroke 2016; 47:2763–2769.

12.

Fiebach

Al-Rawi

Wintermark

, et al. Vascular occlusion enables selecting acute ischemic stroke patients for treatment with desmoteplase. Stroke 2012; 43:1561–1566.

13.

Yeo

Paliwal

Teoh

, et al. Assessment of intracranial collaterals on CT angiography in anterior circulation acute ischemic stroke. AJNR Am J Neuroradiol 2015; 36:289–294.

14.

Mahdjoub

Turc

Legrand

, et al. Do fluid-attenuated inversion recovery vascular hyperintensities represent good collaterals before reperfusion therapy? AJNR Am J Neuroradiol 2018; 39:77–83.