FLAIR vascular hyperintensities and dynamic 4D angiograms for the estimation of collateral blood flow in posterior circulation occlusion

Fluid-attenuated inversion recovery vascular hyperintensities in anterior circulation acute ischemic stroke: associations with cortical brain infarct volume and 90-day prognosis

BACKGROUND AND PURPOSE

Fluid-attenuated inversion recovery vascular hyperintensity (FVH) is often observed in conjunction with acute ischemic stroke (AIS) of the carotid system. However, the significance of FVH in patients with AIS has not been fully elucidated. The purpose of this study is to investigate the effects of FVH on the final infarct volume (including cortical and deep brain infarct volume) and on 90-day prognosis in AIS patients.

MATERIAL AND METHODS

We analyzed data of 160 patients who had AIS of anterior circulation. FVH was identified and the cortical brain infarct volume (CBIV) and deep brain infarct volume (DBIV) were calculated. We assessed 90-day clinical outcome using the modified Rankin Scale (mRS).

RESULTS

FVH was identified in 83 of the 160 patients (51.88%). Patients with FVH showed larger CBIV (13.94 ± 25.55 vs. 6.56 ± 13.49 ml; p = 0.025), more frequent intracranial-large artery disease (74.70 vs. 27.27%; p < 0.001), and more severe clinical impairment on admission (NIHSS 7.22 ± 4.01 vs. 5.42 ± 4.52; p = 0.009). Considering the factors influencing prognosis, FVH positivity (OR = 2.12, 95% CI 1.13-3.99; p = 0.02) and NIHSS (at discharge) (OR = 2.14, 95% CI 1.64-2.78; p < 0.001) were independently associated with 90-day clinical outcome of AIS patients.

CONCLUSION

FVH is a more common finding associated with larger CBIV, intracranial-large artery disease, and more severe strokes on admission. In the presence of good collateral circulation, FVH may be a predictor of better outcome in anterior circulation AIS patients at 90 days.

Hippocampal infarction: Identification of three new types

BACKGROUND AND PURPOSE

Hippocampal infarction (HI) is common but yet still not comprehensively studied. In the present study, we aimed to identify novel HI patterns and to describe additional ischemic lesions outside the hippocampus to draw conclusions regarding the underlying vessel occlusion.

METHODS

In 222 patients (mean age 69.9 (±13.6) years; 129 (58.1%) male, 93 (41.9%) female) with HI, diffusion-weighted images were analyzed with emphasis on HI patterns and associated ischemic lesions outside the hippocampus. HI were classified as type 1 (complete), 2 (lateral), 3 (dorsal), and 4 (circumscribed). Further possible HI patterns were defined and classified as type 5 (ventral), 6 (ventrolateral), and 7 (dorsolateral).

RESULTS

Unilateral HI was found in 218 (98.2%) patients. In these, type 5 and 6 were identified in 5 (2.3%) patients, and type 7 in 8 (3.7%) patients respectively. Type 1 was found in 62 (28.4%), 2 in 53 (24.3%), 3 in 57 (26.1%), and 4 in 28 (12.8%) patients. Further ischemic lesions were found in the territory of the anterior cerebral artery (3.6%), middle cerebral artery (14.9%), anterior choroidal artery (AChA) (7.2%), posterior cerebral artery (89.6%), and in the brainstem (6.3%) and cerebellum (20.3%). Type 5 and 6 were significantly associated with acute ischemic lesions in the AChA territory (6/10 (60%) vs. 11/200 (5.5%), P<0.001).

CONCLUSIONS

We identified three novel HI types. Probably, type 5 and 6 can be attributed to occlusion of the AChA. Overall, these HI types are rare, possibly due to a better collateralization in the case of AChA occlusion.

FLAIR vascular hyperintensities and 4D MR angiograms for the estimation of collateral blood flow in anterior cerebral artery ischemia

PURPOSE

To assess FLAIR vascular hyperintensities (FVH) and dynamic (4D) angiograms derived from perfusion raw data as proposed magnetic resonance (MR) imaging markers of leptomeningeal collateral circulation in patients with ischemia in the territory of the anterior cerebral artery (ACA).

METHODS

Forty patients from two tertiary care university hospitals were included. Infarct volumes and perfusion deficits were manually measured on DWI images and TTP maps, respectively. FVH and collateral flow on 4D MR angiograms were assessed and graded as previously specified.

RESULTS

Forty-one hemispheres were affected. Mean DWI lesion volume was 8.2 (± 13.9; range 0-76.9) ml, mean TTP lesion volume was 24.5 (± 17.2, range 0-76.7) ml. FVH were observed in 26/41 (63.4%) hemispheres. Significant correlations were detected between FVH and TTP lesion volume (ρ = 0.4; P<0.01) absolute (ρ = 0.37; P<0.05) and relative mismatch volume (ρ = 0.35; P<0.05). The modified ASITN/SIR score correlated inversely with DWI lesion volume (ρ = -0.58; P<0.01) and positively with relative mismatch (ρ = 0.29; P< 0.05). ANOVA of the ASITN/SIR score revealed significant inter-group differences for DWI (P<0.001) and TTP lesion volumes (P<0.05). No correlation was observed between FVH scores and modified ASITH/SIR scores (ρ = -0.16; P = 0.32).

CONCLUSIONS

FVH and flow patterns on 4D MR angiograms are markers of perfusion deficits and tissue at risk. As both methods did not show a correlation between each other, they seem to provide complimentary instead of redundant information. Previously shown evidence for the meaning of these specific MR signs in internal carotid and middle cerebral artery stroke seems to be transferrable to ischemic stroke in the ACA territory.

Perfusion-weighted imaging and dynamic 4D angiograms for the estimation of collateral blood flow in lacunar infarction

Although lacunar infarction accounts for approximately 25% of ischemic strokes, collateral blood flow through anastomoses is not well evaluated in lacunar infarction. In 111 lacunar infarction patients, we analyzed diffusion-weighted images, perfusion-weighted images, and blood flow on dynamic four-dimensional angiograms generated by use of Signal Processing In NMR-Software. Blood flow was classified as absent (type 1), from periphery to center (type 2), from center to periphery (type 3), and combination of type 2 and 3 (type 4). On diffusion-weighted images, lacunar infarction was found in the basal ganglia (11.7%), internal capsule (24.3%), corona radiata (30.6%), thalamus (24.3%), and brainstem (9.0%). In 58 (52.2%) patients, perfusion-weighted image showed a circumscribed hypoperfusion, in one (0.9%) a circumscribed hyperperfusion, whereas the remainder was normal. In 36 (62.1%) patients, a larger perfusion deficit (>7 mm) was observed. In these, blood flow was classified type 1 in four (11.1%), 2 in 17 (47.2%), 3 in 9 (25.0%), and 4 in six (16.7%) patients. Patients with lacunar infarction in the posterior circulation more often demonstrated blood flow type 2 and less often type 3 (p = 0.01). Detailed examination and graduation of blood flow in lacunar infarction by use of dynamic four-dimensional angiograms is feasible and may serve for a better characterization of this stroke subtype.

Sparing of the hippocampus indicates better collateral blood flow in acute posterior cerebral artery occlusion

BACKGROUND

In acute posterior cerebral artery, occlusion involvement of the hippocampus is a common finding. Nevertheless, until today, infarction and ischemic lesion evolution in the hippocampus has not been studied systematically.

AIM

Evaluation of hippocampal infarction patterns in posterior cerebral artery occlusion in the very early phase (≤six-hours) and ischemic lesion evolution on follow-up magnetic resonance imaging in relation to collateral blood flow assessed by a magnetic resonance imaging-based approach was conducted.

METHODS

In 28 patients [mean age 69·4 ± 13·8 years, 19 (67·9%) males, 10 (32·1%) females] with proximal posterior cerebral artery occlusion, magnetic resonance imaging findings were analyzed, with emphasis on hippocampal infarction patterns on diffusion-weighted images and collateralization on dynamic 4D angiograms derived from perfusion-weighted raw images.

RESULTS

On initial diffusion-weighted images, we identified all known hippocampal infarction patterns: type 1 (complete) in 6/18 (33·3%) patients, type 2 (lateral) in 10/18 (55·6%) patients, and type 3 (dorsal) and type 4 (circumscribed) in 1/18 (5·6%) patient respectively. On dynamic 4D angiograms, the grade of collateralization was classified as 1 in 9 (32·1%), 2 in 1 (3·6%), 3 in 10 (35·7%), and 4 in 8 (28·6%) patients. On follow-up diffusion-weighted images, we found new ischemic lesions in three and infarction growth in the hippocampus in five patients. Patients with better collateralization (grades 3 and 4) less often had hippocampal infarctions on initial (P = 0·003)/follow-up diffusion-weighted images (P = 0·046) as well as type 1 on initial (P = 0·007)/follow-up diffusion-weighted images (P = 0·005).

CONCLUSIONS

Involvement of the hippocampus in proximal posterior cerebral artery occlusion is frequently but not obligatorily observed and highly dependent on the extent of collateralization. The same holds true for hippocampal infarction patterns.