More pronounced hemodynamic alterations in patients with brain arteriovenous malformation–associated epilepsy

View More View Less
  • 1 Department of Neurosurgery,
  • | 2 Clinical Neuroscience Center, and
  • | 3 Department of Neuroradiology, University Hospital Zurich, University of Zurich, Switzerland
Free access

OBJECTIVE

Epileptic seizures in patients with brain arteriovenous malformations (bAVMs) may be caused by hemodynamic alterations due to the complex angioarchitecture of bAVMs. In particular, an arterial steal phenomenon and venous outflow disruption may play an etiological role in seizure development but remain challenging to demonstrate quantitatively. Blood oxygenation level–dependent (BOLD) cerebrovascular reactivity (CVR) imaging is an emerging technique that can measure both arterial steal phenomenon (as a paradoxical BOLD signal decrease during a vasodilatory stimulus) and impaired perinidal BOLD-CVR (which has been found in the presence of venous congestion on conventional angiography in bAVM patients with epilepsy). By applying this innovative BOLD-CVR technique, the aim is to better study CVR patterns and their correlation with morphological features on conventional angiography in patients with bAVM with and without epilepsy.

METHODS

Twenty-two patients with unruptured and previously untreated bAVMs (8 with and 14 without epilepsy) were included in this case-control study. Quantitative CVR measurements were derived from BOLD functional MRI volumes using a novel standardized and precise hypercapnic stimulus (i.e., % BOLD/mm Hg CO2). In addition, 22 matched healthy controls underwent an identical BOLD-CVR study. Evaluation of venous congestion was performed on conventional angiography for all patients with bAVM.

RESULTS

Patients with bAVM-associated epilepsy showed impaired whole-brain BOLD-CVR compared to those in the nonepilepsy group, even after correction for AVM volume and AVM grade (epilepsy vs nonepilepsy group: 0.17 ± 0.07 vs 0.25 ± 0.07, p = 0.04). A BOLD-CVR–derived arterial steal phenomenon was observed in 2 patients with epilepsy (25%). Venous congestion was noted in 3 patients with epilepsy (38%) and in 1 patient without epilepsy (7%; p = 0.08).

CONCLUSIONS

These data suggest that whole-brain CVR impairment, and more pronounced hemodynamic alterations (i.e., arterial steal phenomenon and venous outflow restriction), may be more present in patients with bAVM-associated epilepsy. The association of impaired BOLD-CVR and bAVM-associated epilepsy will need further investigation in a larger patient cohort.

ABBREVIATIONS

AVM = arteriovenous malformation; bAVM = brain AVM; BOLD = blood oxygenation level–dependent; CBF = cerebral blood flow; CBV = cerebral blood volume; CVR = cerebrovascular reactivity; DSA = digital subtraction angiography; fMRI = functional MRI; MP-RAGE = magnetization-prepared rapid gradient-echo; MTT = mean transit time; PaCO2 = arterial partial pressure of CO2; PetCO2 = end-tidal partial pressure of CO2.

OBJECTIVE

Epileptic seizures in patients with brain arteriovenous malformations (bAVMs) may be caused by hemodynamic alterations due to the complex angioarchitecture of bAVMs. In particular, an arterial steal phenomenon and venous outflow disruption may play an etiological role in seizure development but remain challenging to demonstrate quantitatively. Blood oxygenation level–dependent (BOLD) cerebrovascular reactivity (CVR) imaging is an emerging technique that can measure both arterial steal phenomenon (as a paradoxical BOLD signal decrease during a vasodilatory stimulus) and impaired perinidal BOLD-CVR (which has been found in the presence of venous congestion on conventional angiography in bAVM patients with epilepsy). By applying this innovative BOLD-CVR technique, the aim is to better study CVR patterns and their correlation with morphological features on conventional angiography in patients with bAVM with and without epilepsy.

METHODS

Twenty-two patients with unruptured and previously untreated bAVMs (8 with and 14 without epilepsy) were included in this case-control study. Quantitative CVR measurements were derived from BOLD functional MRI volumes using a novel standardized and precise hypercapnic stimulus (i.e., % BOLD/mm Hg CO2). In addition, 22 matched healthy controls underwent an identical BOLD-CVR study. Evaluation of venous congestion was performed on conventional angiography for all patients with bAVM.

RESULTS

Patients with bAVM-associated epilepsy showed impaired whole-brain BOLD-CVR compared to those in the nonepilepsy group, even after correction for AVM volume and AVM grade (epilepsy vs nonepilepsy group: 0.17 ± 0.07 vs 0.25 ± 0.07, p = 0.04). A BOLD-CVR–derived arterial steal phenomenon was observed in 2 patients with epilepsy (25%). Venous congestion was noted in 3 patients with epilepsy (38%) and in 1 patient without epilepsy (7%; p = 0.08).

CONCLUSIONS

These data suggest that whole-brain CVR impairment, and more pronounced hemodynamic alterations (i.e., arterial steal phenomenon and venous outflow restriction), may be more present in patients with bAVM-associated epilepsy. The association of impaired BOLD-CVR and bAVM-associated epilepsy will need further investigation in a larger patient cohort.

Brain arteriovenous malformations (bAVMs) are the result of abnormal shunting between high-flow arteries and low-flow veins, forming a complex dysplastic vascular nidus within brain parenchyma. Approximately one-third to one-half of patients with newly diagnosed unruptured bAVMs present with epileptic seizures.1,2 Brain AVM–associated epilepsy typically occurs in a young patient population, is often disabling and involves lifelong medication use, and adversely impacts a patient’s quality of life.3–5 The etiology by which bAVMs cause seizures remains poorly understood.2,6–8 In a recent study with the aim to identify morphological MRI-based characteristics of unruptured bAVMs associated with seizures at presentation, peri-AVM edema, a venous pouch/varix, long draining vein, and larger bAVM size based on Spetzler-Martin grading were factors associated with epileptic seizures.9

In addition to standard MRI characteristics, hemodynamic alterations caused by the complex bAVM angioarchitecture have been associated with seizure presentation. Specifically, previous studies postulated that an arterial steal phenomenon may play an etiological role,7,10 whereas others have raised the possibility that seizures are more often related to a disrupted venous outflow pattern.11–13 However, these hemodynamic patterns remain challenging to demonstrate quantitatively.

Blood oxygenation level–dependent (BOLD) cerebrovascular reactivity (CVR) mapping is an emerging hemodynamic imaging method that can quantitatively measure an arterial steal phenomenon (as a paradoxical BOLD signal decrease during a vasodilatory stimulus, resulting in a negative CVR response).14–16 Furthermore, a recent BOLD-CVR study demonstrated impaired perinidal CVR and concomitant venous congestion on conventional angiography, which appeared to correlate with seizure susceptibility in patients with bAVM.11 By applying this innovative BOLD-CVR technique, our aim is to better study the extent of CVR impairment and its correlation with morphological features on conventional angiography in patients with bAVM with and without epilepsy.

Methods

This project is part of an ongoing BOLD-CVR study in patients with bAVMs, which was approved by the local research ethics board. Written informed consent was obtained from each participant before inclusion in the database. The study was conducted in accordance with the ethical standards stated in the 1964 Declaration of Helsinki and its later amendments.

Patient Inclusion

Patients with bAVMs who presented at the Clinical Neuroscience Center of the University Hospital Zurich, Switzerland, during the study period from January 2016 to June 2021 were consecutively screened for participation. The inclusion criteria for this study were 1) patients 18 years of age or older with unruptured and previously untreated bAVMs (including patients both with and without seizures), and 2) patients who underwent diagnostic digital subtraction angiography (DSA) for evaluation of venous congestion within 6 weeks of BOLD-CVR examination. Furthermore, patients with epilepsy underwent electroencephalography and clinical epilepsy management following clinical standards of care.

To determine the normal ranges of BOLD-CVR values, we extracted the data of 22 age- and sex-matched healthy subjects from our BOLD-CVR database to compare CVR patterns at the brain tissue level. Differences in the BOLD-CVR response in healthy subjects are known and are explained by age-related changes in vascular mechanical properties; therefore, we used age and sex matching from the healthy population. The criteria for inclusion in this database were healthy adult subjects with no known neurological symptoms or intracranial pathologies and not taking any medication. These healthy controls underwent an identical BOLD-CVR protocol and signed an informed consent form before the study. The data that support the findings of this study are available from the corresponding author (M.S.) upon reasonable request.

Image Acquisition and Analysis

BOLD functional MRI (fMRI) images were obtained from a 3-T Skyra VD13 machine (Siemens Healthcare) with CO2 as the vasoactive stimulus, modulated by a computer-controlled gas blender with prospective gas-targeting algorithms (RespirAct, Thornhill Research Institute). The sequence consisted of an initial 100-second patient baseline end-tidal partial pressure of CO2 (PetCO2) level, after which a PetCO2 step was increased for approximately 10 mm Hg above baseline for 80 seconds and then a return to baseline for 100 seconds, before free breathing was restored. During this sequence, iso-oxia is maintained at 100 mm Hg. A measurable increase in the PetCO2 can be used as a surrogate measure for the true independent stimulus, i.e., the arterial partial pressure of CO2 (PaCO2) in arterial blood. CVR is then defined as the percentage of change in the BOLD signal per mm Hg change in PaCO2.17

A high-resolution, T1-weighted, magnetization-prepared rapid gradient-echo (MP-RAGE) image was obtained for anatomical overlay of the BOLD-CVR images. All BOLD-CVR images were obtained using a previously published protocol,17 which has been applied in multiple studies.18–22

BOLD-CVR was calculated from the slope of a linear least-square fit of the BOLD signal time course to the CO2 time course during the BOLD scan.17,23,24 For BOLD-CVR, an arterial steal phenomenon is defined as a paradoxical BOLD signal decrease during hypercapnia, resulting in negative BOLD-CVR values.14

From the T1-weighted MP-RAGE scan, probability maps for the gray matter, white matter, and CSF were obtained. Each T1-weighted MP-RAGE image was then manually masked for the affected hemisphere, and in combination with a merged gray matter–white matter probability map (> 80% probability), CVR of the affected hemisphere was calculated.17,24

Brain AVM Masking

Three-dimensional bAVM masks were determined and manually drawn from the current state-of-the-art anatomical bAVM MRI protocol using iPlan software (Brainlab AG). The nidus of the bAVM with the most proximal part of the draining vein was drawn on every slice on which the bAVM was visible, combining the information of high-resolution T1-weighted, T1-weighted contrast-enhanced, T2-weighted, and time-of-flight sequences to obtain a 3D spherical volume of interest, representing a bAVM mask. These bAVM masks were overlaid on the BOLD-CVR maps to obtain mean intralesional values in those specific regions, as well as to subdivide the bAVM mask values from the whole-brain BOLD-CVR values to gain pure whole-brain BOLD-CVR values, without intralesional values of the masked bAVM. Furthermore, this allowed us to obtain bAVM volume, BOLD-CVR values of affected and unaffected hemispheres, as well as gray and white matter values of the whole brain and the affected and unaffected hemispheres. The affected hemisphere was defined as the hemisphere containing the bAVM. Data sets of patients with bAVMs affecting both hemispheres were excluded from calculations of BOLD-CVR values of affected and unaffected hemispheres.

DSA and Determination of Venous Congestion Pattern

All patients underwent diagnostic 6-vessel DSA within 6 weeks of the BOLD-CVR investigation. Venous congestion and nidus and feeding arteries were evaluated using 3D and 4D DSA reconstructions, according to the following criteria: 1) venous congestion: pseudophlebitic pattern, draining vein outflow restriction, remote outflow restriction; 2) nidus: glomus, fistulous, or both; and 3) feeding arteries: significant dilation, moderate dilation, flow-related aneurysm.11

Statistical Analysis

Twenty-two age-matched healthy controls were recruited as a reference population (or external control) to compare BOLD-CVR patterns at the brain tissue level. BOLD-CVR impairment ≥ −2 standard deviations away from the mean whole-brain BOLD-CVR of the healthy matched cohort (external cohort) was considered significant BOLD-CVR impairment for the AVM cohort. Variables from the AVM cohort and healthy cohort were compared using chi-square tests and paired Student t-tests.

The data from the AVM cohort were assessed for normality using the Shapiro-Wilk test. All normally distributed continuous variables were reported as means ± standard deviations. Categorical data were presented as proportions. Continuous variables from the epilepsy and nonepilepsy groups were compared using the Student t-test. To identify a relationship between the whole brain and BOLD-CVR affected and unaffected hemispheres and bAVM volume, regardless of epilepsy status, the Pearson correlation coefficient was used. Furthermore, a Spearman rank-order correlation analysis was used to identify the relationship between the whole brain, BOLD-CVR affected and unaffected hemispheres, and bAVM grade, as well as bAVM location.

The statistical analysis was performed using SPSS Statistics (version 26, IBM Corp.). A two-sided p value < 0.05 was considered significant. ANCOVA was used to statistically control the effect of covariates (two covariates: bAVM volume and bAVM grade) for BOLD-CVR findings between the epilepsy and nonepilepsy groups.

Results

Study Population Characteristics

A flowchart illustrating patient screening and inclusion is shown in Fig. 1. Twenty-two patients with bAVMs and 22 age- and sex-matched healthy subjects were included in the study. The healthy population included right-handed nonsmokers without a history of brain pathology, neurological disease, neurological symptoms, or medications. A detailed overview of the characteristics of bAVM patients and matched healthy controls is found in Table 1.

FIG. 1.
FIG. 1.

Study flowchart. From the prospective BOLD-CVR database with 135 subjects who underwent BOLD-CVR study, we extracted 55 patients with bAVMs who underwent BOLD-CVR imaging. From 55 patients with bAVMs, 28 patients had an unruptured and previously untreated bAVM. Of these 28 patients, 22 underwent BOLD-CVR imaging and DSA examination in a 6-week time frame and were eligible for further analysis. From 80 healthy subject who underwent the BOLD-CVR study, we extracted 22 age- and sex-matched controls eligible for inclusion. In the final analysis, 44 subjects were included (22 age- and sex-matched pairs).

TABLE 1.

Relevant characteristics of AVM patients and matched healthy controls

VariablebAVM Cohort (n = 22)Healthy Cohort (n = 22)p Value
Mean age ± SD, yrs40.2 ± 15.140.7 ± 14.40.92
Male sex, n (%)11 (50) 11 (50)NA
Mean whole-brain CVR ± SD0.22 ± 0.080.27 ± 0.070.07
Mean gray matter CVR ± SD0.25 ± 0.090.30 ± 0.090.08
Mean white matter CVR ± SD0.16 ± 0.060.19 ± 0.050.054
Mean affected hemisphere CVR ± SD*0.20 ± 0.070.26 ± 0.080.01
Mean unaffected hemisphere CVR ± SD*0.21 ± 0.060.27 ± 0.080.007

NA = not applicable. Boldface type indicates statistical significance.

Three data sets of patients with AVMs affecting both hemispheres were excluded from calculations.

For the healthy cohort, the right hemisphere was defined as affected and the left hemisphere as unaffected.

The bAVM cohort included 8 patients (36%) with and 14 (64%) without epilepsy. All patients with epilepsy were treated with antiepileptic medications and were seizure free by the time of BOLD-CVR imaging. Patient demographics and general bAVM characteristics are presented in Table 2.

TABLE 2.

Patient demographics and general AVM characteristics

Case No.Age (yrs), SexAVM LocationAVM Volume (cm3)Grade*Feeding ArteryInitial Presentation
ACAMCAPCA
125, FGyrus cingula, corpus callosum28.763++Headache, dizziness
243, MGenu corpus callosum2.172+Dizziness
336, FRt occipital3.371+Incidental finding by head injury
451, MRt temporal13.643++GTCS
520, FRt hippocampal16.863++Headache
638, MRt temporal1.851+Incidental finding
763, MLt central/postcentral3.882+Hypesthesia & paresis of rt hand
841, FRt paraventricular temporo-occipital1.353++Lt superior homonymous hemianopsia
939, FLt fronto-orbital9.702++GTCS
1021, MRt gyrus cinguli1.172+Incidental finding by head injury
1134, FRt postcentral19.542+GTCS
1260, FRt frontal6.571++Incidental finding by frontal sinusitis
1328, MLt gyrus rectus8.713+GTCS
1424, FRt frontal2.342+Incidental finding by Tourette syndrome
1522, MRt frontal68.594++GTCS
1671, MRt parietal1.361+Focal epileptic seizures
1740, MLt temporal24.234+Word-finding difficulty
1868, MRt temporal1.181+Episodes of transient global amnesia
1934, MLt temporal38.962++Focal epileptic seizures
2049, FRt frontal10.922+Incidental finding by hard of hearing
2144, FRt central27.952+Complex focal epileptic seizures
2234, FRt precuneus3.001++Hemihypesthesia on lt side

+ = yes; − = no; ACA = anterior cerebral artery; GTCS = generalized tonic-clonic seizure; MCA = middle cerebral artery; PCA = posterior cerebral artery.

According to Spetzler-Martin classification system.

BOLD-CVR Findings in Patients With bAVMs Versus Healthy Subjects

Patients with bAVMs showed a trend toward more impaired whole-brain CVR compared to the healthy cohort: mean whole-brain CVR = 0.22 ± 0.08 for bAVM patients versus 0.27 ± 0.07 in healthy subjects (p = 0.07). Similarly, nonsignificant differences in BOLD-CVR between the bAVM and healthy cohort values were noted for both gray and white matter (Table 1).

Correlation of BOLD-CVR Findings and bAVM Characteristics Regardless of Epilepsy Status

BOLD-CVR values of the whole brain, as well as for the affected and unaffected hemispheres, showed some evidence for correlations with the bAVM volume (Pearson coefficients −0.41, −0.42, and −0.46, with corresponding p values 0.06, 0.08, and 0.05, respectively). BOLD-CVR values of the whole brain, and the affected and unaffected hemispheres, also showed strong evidence for correlations with bAVM grade (Spearman’s rho 0.59, 0.62, and 0.58, with corresponding p values 0.01, 0.01, and 0.01, respectively). Weak evidence for correlations was observed between whole-brain BOLD-CVR and bAVM location (Spearman’s rho = 0.37, p = 0.09), whereas no correlation was found between BOLD-CVR affected and unaffected hemispheres and bAVM location (Spearman’s rho 0.10 and 0.15, p values 0.97 and 0.55, respectively).

BOLD-CVR Findings in bAVM Patients

Patients with bAVM-associated epilepsy showed significantly more impaired whole-brain CVR (epilepsy vs nonepilepsy = 0.17 ± 0.07 vs 0.25 ± 0.07, p = 0.04) as well as CVR for both gray matter (epilepsy vs nonepilepsy = 0.19 ± 0.07 vs 0.29 ± 0.08, p = 0.05) and white matter (epilepsy vs nonepilepsy = 0.12 ± 0.05 vs 0.18 ± 0.05, p = 0.05) compared with nonepilepsy bAVM patients, corrected for bAVM volume and bAVM grade (Spetzler-Martin grade) as possible biases (Table 3).

TABLE 3.

BOLD-CVR findings in patients with bAVM-associated epilepsy and nonepilepsy bAVM patients

Functional Measurement (mean ± SD)Total Cohort (n = 22)Epilepsy bAVM Group (n = 8)Nonepilepsy bAVM Group (n = 14)p Value
Whole-brain BOLD-CVR0.22 ± 0.080.17 ± 0.070.25 ± 0.070.04*
Gray matter CVR0.25 ± 0.090.19 ± 0.070.29 ± 0.080.05*
White matter CVR0.16 ± 0.060.12 ± 0.050.18 ± 0.050.05*
Affected hemisphere BOLD-CVR†0.20 ± 0.070.16 ± 0.070.22 ± 0.050.13*
Unaffected hemisphere BOLD-CVR†0.21 ± 0.060.17 ± 0.060.13 ± 0.050.07*
Volume of bAVM (mm3)13.36 ± 16.5123.56 ± 21.737.54 ± 9.260.03
CVR of bAVM0.14 ± 0.120.06 ± 0.110.18 ± 0.110.02

CVR is defined as percentage BOLD signal change per mm Hg CO2. Boldface type indicates statistical significance.

Using ANCOVA corrected for AVM volume and Spetzler-Martin grade as possible covariates.

Three data sets of patients with AVMs affecting both hemispheres were excluded from calculations.

Hemodynamic Alterations in Patients With bAVM-Associated Epilepsy

Patients with bAVM-associated epilepsy exhibited more pronounced hemodynamic alterations, defined as more impaired BOLD-CVR values, when compared with the nonepilepsy group. We used a BOLD-CVR cutoff of 0.13 for defining ≥ −2 standard deviations from the mean whole-brain BOLD-CVR of the healthy cohort (mean whole-brain CVR value ± SD = 0.27 ± 0.07).

After investigating ≥ −2 standard deviations away from the mean BOLD-CVR value of the healthy cohort, a significant BOLD-CVR impairment was noted overall. Some brain areas also exhibited a negative BOLD-CVR response (i.e., a paradoxical BOLD signal decrease during hypercapnia) indicating an arterial steal phenomenon, which was observed in the perinidal bAVM vicinity in 2 patients with bAVM-associated epilepsy (Fig. 2). None of the patients in the nonepilepsy group had a negative BOLD-CVR response.

FIG. 2.
FIG. 2.

Illustrative images of epilepsy patients with bAVM exhibiting more pronounced hemodynamic alterations. A: A 34-year-old male patient presented with focal epileptic seizures. On morphological MRI, a left-sided temporal bAVM (triangles) was detected. A few weeks after the diagnosis, the patient suffered an ischemic event with acute ischemic infarction around the bAVM lesion, as seen on diffusion-weighted imaging (DWI). CE = contrast enhanced. B: BOLD-CVR showed a significant hemispheric-impaired CVR clearly beyond the bAVM borders when compared to the age-matched healthy subjects. Furthermore, a clear arterial steal is seen in the immediate vicinity of the bAVM (arrows). The area of infarction (DWI) is in the region with arterial steal, as seen on BOLD-CVR images. C: On DSA images, no signs of venous congestion were identified. D: A 34-year-old female patient presented with complex focal epileptic seizures. Morphological MRI demonstrated a right-sided bAVM in the central region (triangles). E: BOLD-CVR showed a hemispheric-impaired CVR with areas of arterial steal in the vicinity of the bAVM (arrows). F: DSA images demonstrate venous outflow restriction of the major draining vein.

In Table 4, the results of morphological bAVM features evaluated on DSA are presented. Venous congestion was observed in 37.5% patients in the epilepsy group, compared with 7.1% patients in the nonepilepsy group. Although striking, this difference did not achieve statistical significance (p = 0.08).

TABLE 4.

Morphological bAVM features evaluated on 6-vessel DSA

FeatureEpilepsy AVM Group (n = 8)Nonepilepsy AVM Group (n = 14)
Venous congestion31
 Pseudophlebitic pattern01
 Draining vein outflow restriction30
 Remote outflow restriction30
Nidus
 Glomus21
 Fistulous38
 Both35
Feeding arteries
 Significant dilation00
 Moderate dilation814
 Flow-related aneurysm20

Discussion

By using conventional angiography and an emerging hemodynamic imaging technique (BOLD-CVR), our data suggest that whole-brain CVR impairment and more pronounced hemodynamic alterations (i.e., arterial steal phenomenon and venous outflow restriction) may correlate with seizure susceptibility in patients with bAVMs. This result, as well as findings by others, underscores the importance of advanced hemodynamic imaging to better study the risk for developing epilepsy in patients with bAVM. For example, using CT perfusion,25 different distinct patterns of extranidal brain parenchymal perfusion abnormalities of cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT) in patients with bAVM-associated epilepsy were reported.

The next step is to prospectively validate these results in a larger cohort and externally validate them. This study, supported by other literature,11,26,27 shows promising results, indicating that the noninvasive BOLD-CVR approach has the potential to be an imaging marker for predicting epilepsy in patients with bAVMs.

Morphological Characteristics and Hemodynamic Studies in Patients With bAVMs

Although bAVM is a rare neurovascular disease, it can present with devastating symptoms such as intracranial hemorrhage and epileptic seizures.3 Most bAVM studies are focused on determining risk of hemorrhage, whereas seizure origin in bAVMs remains poorly understood. Because similar-appearing bAVMs may or may not cause seizures, their etiology cannot be simply explained by the presence of the bAVM alone.2,5 Reports on associations of bAVM morphology, location, or prior hemorrhage with seizure risk showed conflicting results.2,6–8 Thus, additional pathophysiological mechanisms must be involved, underlining the importance of further studies identifying markers for seizure risk in those patients. Hemodynamic alterations caused by the complex angioarchitecture of bAVMs have been suggested to play an important role. Whereas an arterial steal phenomenon has been postulated by some authors,7,10 others have raised the possibility that seizures are more often related to a disrupted venous outflow11,13,28,29 rather than to an inadequate arterial blood supply. This theory supports the role of advanced hemodynamic investigations to assess the cerebrovascular reserve capacity. In this regard, BOLD-CVR allows for a noninvasive and quantitative assessment of cerebrovascular reserve capacity on a voxel-by-voxel basis. An impaired CVR can range from a blunted increase in CBF in mild impairment, to paradoxical reduction in CBF indicating an arterial steal phenomenon in severe CVR impairment.18,21

Pathophysiological Explanation of More Pronounced Hemodynamic Impairment in Patients With bAVM-Associated Epilepsy

Previous historical results of CVR testing in patients with bAVMs reported in the literature are limited and inconclusive due to the variety of methods applied and by arbitrary interpretation of the results.30–33 Arteriolar exhaustion resulting from a shunt-induced, chronically reduced cerebral perfusion pressure was thought to be the underlying cause of reduced vascular reserve capacity in xenon CT studies after acetazolamide challenges.34 One previous study reported on the utility of advanced novel BOLD-CVR imaging, showing perinidal impairment of the cerebrovascular autoregulatory reserve in patients with seizure-prone bAVM and highlighting a strong association with venous congestion in all seizure-prone patients.11 In our study, a clear association of venous congestion with impaired BOLD-CVR could not be shown in all seizure-prone patients, but a trend toward more evident venous congestion was discernible in the epilepsy group. Moreover, an arterial steal phenomenon as a pathophysiological mechanism was suspected in 2 seizure-prone bAVM patients with whole-brain BOLD-CVR values ≥ −2 standard deviations below the mean whole-brain BOLD-CVR of the healthy cohort, indicating significant BOLD-CVR impairment with partly negative BOLD-CVR response (i.e., arterial steal) in the perinidal bAVM vicinity (Fig. 2). In a recent study35 using perfusion CT, decreased CBF, CBV, and MTT were introduced as “functional” arterial steal and seizure was the most common presenting symptom in the patients with this pattern. Interestingly, 50% of those patients also presented with venous reflux,25 suggesting that arterial steal and venous congestion may even coexist. One of our 2 patients with arterial steal showed venous congestion on DSA, with outflow restriction in the draining vein and remote outflow restriction (Fig. 2). Although these are observations in individual cases, the findings are suggestive and should promote further research. The second perfusion pattern, introduced by Kim and Krings,25 is an ischemic arterial steal with decreased CBF and CBV and increased MTT. As presented in Fig. 2A and B, 1 of 2 patients with a BOLD-CVR–proven arterial steal phenomenon developed an ischemic stroke in the area of negative BOLD-CVR.

The arterial steal theory in bAVM patients is explained by a high flow through the arterial feeding vessels and the nidus that creates a low-resistance vascular bed, redirecting blood away from surrounding brain tissue.36–38 Moreover, it has been hypothesized that this arterial steal with subsequent local hypoxemia may lead to an increased perinidal angiogenesis, with blood being rerouted from the normal brain tissue due to the sump effect of bAVM, resulting in a larger area of cortex becoming hypoxemic.10,11,25 The association between the steal phenomenon and seizure activity is then potentially explained by chronic hypoxemia leading to gliosis, which may trigger epileptic activity.2,7,10

In contrast, venous congestion may be best explained by the potential high arterial inflow through the bAVM nidus, overwhelming the venous drainage capacity and leading to venous congestion and redirection of venous outflow, i.e., a pseudophlebitic pattern.39 This congestion may then limit the ability of arterioles in the perinidal tissue to vasodilate, which may subsequently result in impaired cerebrovascular reserve capacity. This pressure-related inability to respond to a vasoactive stimulus (elevated PaCO2) is detectable as a blunted BOLD-CVR response.17,40

Lastly, the association between impaired CVR and neurovascular uncoupling in the vicinity of the bAVM lesion may be of clinical relevance. Although we have not obtained resting-state fMRI in our current patient cohort, we have shown that with our BOLD-CVR technique and corroborating task-based fMRI, brain areas that exhibit neurovascular uncoupling can be identified.41 Our BOLD-CVR data from the current study may therefore suggest that neurovascular uncoupling may be more present in patients with bAVM-associated epilepsy because they exhibit more impaired BOLD-CVR values when compared to the nonepilepsy group. For seizure-prone bAVM patients specifically, the perinidal tissue that shows impaired BOLD-CVR may consecutively suffer from perinidal neurovascular uncoupling, potentially related to perinidal angiogenesis.42

Clinical Implications and Future Directions

The ability to better quantify hemodynamic alterations on advanced imaging in patients with bAVM-associated epilepsy, particularly impaired whole-brain CVR, opens new avenues to explore hemodynamic mechanisms regarding why some bAVMs cause epileptic seizures and others do not. Hence, the results of this study need to be interpreted as a basis for future studies. Moreover, future studies should also investigate whether bAVM treatment will result in normalization of BOLD-CVR values and whether this is associated with seizure-free outcome.

Limitations

Our data must be interpreted in the context of the study design. In this preliminary study, we studied only 22 patients with bAVM, a sample size that is similar to that in other studies investigating hemodynamic alterations in bAVMs.11,25 Therefore, the results should be interpreted with caution. Nevertheless, the finding of impaired BOLD-CVR and the trend toward the presence of arterial steal and venous congestion in patients with bAVM-associated epilepsy may open new avenues of hemodynamic imaging research. Because we included data of 22 patients with bAVM, we were able to correct for two variances that could influence our BOLD-CVR results; that is, we corrected for bAVM volume and bAVM grade as possible biases. We did not correct for bAVM location or number of antiepileptic medications, which could affect the whole-brain BOLD-CVR values. After correction for bAVM volume and bAVM grade, however, patients with bAVM-associated epilepsy exhibited significantly more impaired BOLD-CVR values of the whole brain as well as gray and white matter when compared to the nonepilepsy bAVM group. Future studies will need to be conducted in larger patient cohorts to confirm the generalizability of our findings before definitive conclusions can be drawn.

Conclusions

Our data indicate that whole-brain CVR impairment, and more pronounced hemodynamic alterations (arterial steal and venous congestion), may correlate with seizure susceptibility in patients with bAVM. BOLD-CVR may be considered an investigative imaging tool to better characterize hemodynamics in patients with bAVM and epilepsy. The association of impaired BOLD-CVR and bAVM-associated epilepsy, however, will need further investigation in a larger patient cohort.

Acknowledgments

This project was funded by the Theodor und Ida Herzog-Egli Stiftung and by the Stroke–Clinical Research Priority Program of the University of Zurich.

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: all authors. Acquisition of data: Sebök, van Niftrik. Analysis and interpretation of data: all authors. Drafting the article: Sebök, Germans, van Niftrik, Fierstra. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Sebök. Statistical analysis: Sebök, Germans. Administrative/technical/material support: Sebök. Study supervision: Regli, Fierstra.

References

  • 1

    Turjman F, Massoud TF, Sayre JW, Viñuela F, Guglielmi G, Duckwiler G. Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural characteristics. AJNR Am J Neuroradiol. 1995;16(2):345350.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Shankar JJ, Menezes RJ, Pohlmann-Eden B, Wallace C, terBrugge K, Krings T. Angioarchitecture of brain AVM determines the presentation with seizures: proposed scoring system. AJNR Am J Neuroradiol. 2013;34(5):10281034.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 3

    Soldozy S, Norat P, Yağmurlu K, et al. Arteriovenous malformation presenting with epilepsy: a multimodal approach to diagnosis and treatment. Neurosurg Focus. 2020;48(4):E17.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 4

    Jones JE, Berven NL, Ramirez L, Woodard A, Hermann BP. Long-term psychosocial outcomes of anterior temporal lobectomy. Epilepsia. 2002;43(8):896903.

  • 5

    Ding D, Starke RM, Quigg M, et al. Cerebral arteriovenous malformations and epilepsy, part 1: predictors of seizure presentation. World Neurosurg. 2015;84(3):645652.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 6

    Hoh BL, Chapman PH, Loeffler JS, Carter BS, Ogilvy CS. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcomes. Neurosurgery. 2002;51(2):303311.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 7

    Hacein-Bey L, Nour R, Pile-Spellman J, Van Heertum R, Esser PD, Young WL. Adaptive changes of autoregulation in chronic cerebral hypotension with arteriovenous malformations: an acetazolamide-enhanced single-photon emission CT study. AJNR Am J Neuroradiol. 1995;16(9):18651874.

    • PubMed
    • Search Google Scholar
    • Export Citation
  • 8

    Garcin B, Houdart E, Porcher R, et al. Epileptic seizures at initial presentation in patients with brain arteriovenous malformation. Neurology. 2012;78(9):626631.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 9

    Benson JC, Chiu S, Flemming K, Nasr DM, Lanzino G, Brinjikji W. MR characteristics of unruptured intracranial arteriovenous malformations associated with seizure as initial clinical presentation. J Neurointerv Surg. 2020;12(2):186191.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10

    Mast H, Mohr JP, Osipov A, et al. ‘Steal’ is an unestablished mechanism for the clinical presentation of cerebral arteriovenous malformations. Stroke. 1995;26(7):12151220.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11

    Fierstra J, Conklin J, Krings T, et al. Impaired peri-nidal cerebrovascular reserve in seizure patients with brain arteriovenous malformations. Brain. 2011;134(Pt1):100109.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12

    Krings T, Hans FJ, Geibprasert S, Terbrugge K. Partial “targeted” embolisation of brain arteriovenous malformations. Eur Radiol. 2010;20(11):27232731.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13

    Kosnik EJ, Hunt WE, Miller CA. Dural arteriovenous malformations. J Neurosurg. 1974;40(3):322329.

  • 14

    Poublanc J, Han JS, Mandell DM, et al. Vascular steal explains early paradoxical blood oxygen level-dependent cerebrovascular response in brain regions with delayed arterial transit times. Cerebrovasc Dis Extra. 2013;3(1):5564.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15

    Fierstra J, Poublanc J, Han JS, et al. Steal physiology is spatially associated with cortical thinning. J Neurol Neurosurg Psychiatry. 2010;81(3):290293.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    van Niftrik CHB, Piccirelli M, Bozinov O, et al. Impact of baseline CO2 on Blood-Oxygenation-Level-Dependent MRI measurements of cerebrovascular reactivity and task-evoked signal activation. Magn Reson Imaging. 2018;49:123130.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17

    van Niftrik CHB, Piccirelli M, Bozinov O, et al. Iterative analysis of cerebrovascular reactivity dynamic response by temporal decomposition. Brain Behav. 2017;7(9):e00705.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18

    Sebök M, Niftrik C, Lohaus N, et al. Leptomeningeal collateral activation indicates severely impaired cerebrovascular reserve capacity in patients with symptomatic unilateral carotid artery occlusion. J Cereb Blood Flow Metab. 2021;41(11):30393051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19

    Sebök M, van Niftrik CHB, Winklhofer S, et al. Mapping cerebrovascular reactivity impairment in patients with symptomatic unilateral carotid artery disease. J Am Heart Assoc. 2021;10(12):e020792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20

    van Niftrik CHB, Sebök M, Wegener S, et al. Increased ipsilateral posterior cerebral artery P2-segment flow velocity predicts hemodynamic impairment. Stroke. 2021;52(4):14691472.

    • Search Google Scholar
    • Export Citation
  • 21

    Fierstra J, van Niftrik C, Warnock G, et al. Staging hemodynamic failure with blood oxygen-level-dependent functional magnetic resonance imaging cerebrovascular reactivity: a comparison versus gold standard (15O-)H2O-positron emission tomography. Stroke. 2018;49(3):621629.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22

    van Niftrik CHB, Sebök M, Muscas G, et al. Characterizing ipsilateral thalamic diaschisis in symptomatic cerebrovascular steno-occlusive patients. J Cereb Blood Flow Metab. 2019;40(3):563573.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23

    Sebök M, van Niftrik CHB, Piccirelli M, et al. BOLD cerebrovascular reactivity as a novel marker for crossed cerebellar diaschisis. Neurology. 2018;91(14):e1328e1337.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24

    Sebök M, van Niftrik CHB, Wegener S, Luft A, Regli L, Fierstra J. Agreement of novel hemodynamic imaging parameters for the acute and chronic stages of ischemic stroke: a matched-pair cohort study. Neurosurg Focus. 2021;51(1):E12.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25

    Kim DJ, Krings T. Whole-brain perfusion CT patterns of brain arteriovenous malformations: a pilot study in 18 patients. AJNR Am J Neuroradiol. 2011;32(11):20612066.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Fierstra J, Spieth S, Tran L, et al. Severely impaired cerebrovascular reserve in patients with cerebral proliferative angiopathy. J Neurosurg Pediatr. 2011;8(3):310315.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27

    Busch KJ, Kiat H, Stephen M, Simons M, Avolio A, Morgan MK. Cerebral hemodynamics and the role of transcranial Doppler applications in the assessment and management of cerebral arteriovenous malformations. J Clin Neurosci. 2016;30:2430.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28

    Lasjaunias P, Chiu M, ter Brugge K, Tolia A, Hurth M, Bernstein M. Neurological manifestations of intracranial dural arteriovenous malformations. J Neurosurg. 1986;64(5):724730.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29

    Hurst RW, Hackney DB, Goldberg HI, Davis RA. Reversible arteriovenous malformation-induced venous hypertension as a cause of neurological deficits. Neurosurgery. 1992;30(3):422425.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Diehl RR, Henkes H, Nahser HC, Kühne D, Berlit P. Blood flow velocity and vasomotor reactivity in patients with arteriovenous malformations. A transcranial Doppler study. Stroke. 1994;25(8):15741580.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31

    De Salles AA, Manchola I. CO2 reactivity in arteriovenous malformations of the brain: a transcranial Doppler ultrasound study. J Neurosurg. 1994;80(4):624630.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Van Roost D, Schramm J. What factors are related to impairment of cerebrovascular reserve before and after arteriovenous malformation resection? A cerebral blood flow study using xenon-enhanced computed tomography. Neurosurgery. 2001;48(4):709717.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Schaller C, Schramm J, Haun D, Meyer B. Patterns of cortical oxygen saturation changes during CO2 reactivity testing in the vicinity of cerebral arteriovenous malformations. Stroke. 2003;34(4):938944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34

    Tarr RW, Johnson DW, Rutigliano M, et al. Use of acetazolamide-challenge xenon CT in the assessment of cerebral blood flow dynamics in patients with arteriovenous malformations. AJNR Am J Neuroradiol. 1990;11(3):441448.

    • Search Google Scholar
    • Export Citation
  • 35

    Kim SW, Kim YD, Chang HJ, et al. Different infarction patterns in patients with aortic atheroma compared to those with cardioembolism or large artery atherosclerosis. J Neurol. 2018;265(1):151158.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36

    Spetzler RF, Hargraves RW, McCormick PW, Zabramski JM, Flom RA, Zimmerman RS. Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg. 1992;76(6):918923.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37

    Norris JS, Valiante TA, Wallace MC, et al. A simple relationship between radiological arteriovenous malformation hemodynamics and clinical presentation: a prospective, blinded analysis of 31 cases. J Neurosurg. 1999;90(4):673679.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38

    Moftakhar P, Hauptman JS, Malkasian D, Martin NA. Cerebral arteriovenous malformations. Part 2: physiology. Neurosurg Focus. 2009;26(5):E11.

  • 39

    Geibprasert S, Pongpech S, Jiarakongmun P, Shroff MM, Armstrong DC, Krings T. Radiologic assessment of brain arteriovenous malformations: what clinicians need to know. Radiographics. 2010;30(2):483501.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40

    Fierstra J, Sobczyk O, Battisti-Charbonney A, et al. Measuring cerebrovascular reactivity: what stimulus to use?. J Physiol. 2013;591(23):58095821.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41

    van Niftrik CHB, Piccirelli M, Muscas G, et al. The voxel-wise analysis of false negative fMRI activation in regions of provoked impaired cerebrovascular reactivity. PLoS One. 2019;14(5):e0215294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42

    Li M, Liu Q, Guo R, et al. Perinidal angiogenesis is a predictor for neurovascular uncoupling in the periphery of brain arteriovenous malformations: a task-based and resting-state fMRI study. J Magn Reson Imaging. 2021;54(1):186196.

    • Crossref
    • Search Google Scholar
    • Export Citation

Illustration from Agosti et al. (E5). Used with permission of Mayo Foundation for Medical Education and Research. All rights reserved.

  • View in gallery

    Study flowchart. From the prospective BOLD-CVR database with 135 subjects who underwent BOLD-CVR study, we extracted 55 patients with bAVMs who underwent BOLD-CVR imaging. From 55 patients with bAVMs, 28 patients had an unruptured and previously untreated bAVM. Of these 28 patients, 22 underwent BOLD-CVR imaging and DSA examination in a 6-week time frame and were eligible for further analysis. From 80 healthy subject who underwent the BOLD-CVR study, we extracted 22 age- and sex-matched controls eligible for inclusion. In the final analysis, 44 subjects were included (22 age- and sex-matched pairs).

  • View in gallery

    Illustrative images of epilepsy patients with bAVM exhibiting more pronounced hemodynamic alterations. A: A 34-year-old male patient presented with focal epileptic seizures. On morphological MRI, a left-sided temporal bAVM (triangles) was detected. A few weeks after the diagnosis, the patient suffered an ischemic event with acute ischemic infarction around the bAVM lesion, as seen on diffusion-weighted imaging (DWI). CE = contrast enhanced. B: BOLD-CVR showed a significant hemispheric-impaired CVR clearly beyond the bAVM borders when compared to the age-matched healthy subjects. Furthermore, a clear arterial steal is seen in the immediate vicinity of the bAVM (arrows). The area of infarction (DWI) is in the region with arterial steal, as seen on BOLD-CVR images. C: On DSA images, no signs of venous congestion were identified. D: A 34-year-old female patient presented with complex focal epileptic seizures. Morphological MRI demonstrated a right-sided bAVM in the central region (triangles). E: BOLD-CVR showed a hemispheric-impaired CVR with areas of arterial steal in the vicinity of the bAVM (arrows). F: DSA images demonstrate venous outflow restriction of the major draining vein.

  • 1

    Turjman F, Massoud TF, Sayre JW, Viñuela F, Guglielmi G, Duckwiler G. Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural characteristics. AJNR Am J Neuroradiol. 1995;16(2):345350.

    • Search Google Scholar
    • Export Citation
  • 2

    Shankar JJ, Menezes RJ, Pohlmann-Eden B, Wallace C, terBrugge K, Krings T. Angioarchitecture of brain AVM determines the presentation with seizures: proposed scoring system. AJNR Am J Neuroradiol. 2013;34(5):10281034.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 3

    Soldozy S, Norat P, Yağmurlu K, et al. Arteriovenous malformation presenting with epilepsy: a multimodal approach to diagnosis and treatment. Neurosurg Focus. 2020;48(4):E17.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4

    Jones JE, Berven NL, Ramirez L, Woodard A, Hermann BP. Long-term psychosocial outcomes of anterior temporal lobectomy. Epilepsia. 2002;43(8):896903.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5

    Ding D, Starke RM, Quigg M, et al. Cerebral arteriovenous malformations and epilepsy, part 1: predictors of seizure presentation. World Neurosurg. 2015;84(3):645652.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6

    Hoh BL, Chapman PH, Loeffler JS, Carter BS, Ogilvy CS. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcomes. Neurosurgery. 2002;51(2):303311.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 7

    Hacein-Bey L, Nour R, Pile-Spellman J, Van Heertum R, Esser PD, Young WL. Adaptive changes of autoregulation in chronic cerebral hypotension with arteriovenous malformations: an acetazolamide-enhanced single-photon emission CT study. AJNR Am J Neuroradiol. 1995;16(9):18651874.

    • Search Google Scholar
    • Export Citation
  • 8

    Garcin B, Houdart E, Porcher R, et al. Epileptic seizures at initial presentation in patients with brain arteriovenous malformation. Neurology. 2012;78(9):626631.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9

    Benson JC, Chiu S, Flemming K, Nasr DM, Lanzino G, Brinjikji W. MR characteristics of unruptured intracranial arteriovenous malformations associated with seizure as initial clinical presentation. J Neurointerv Surg. 2020;12(2):186191.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10

    Mast H, Mohr JP, Osipov A, et al. ‘Steal’ is an unestablished mechanism for the clinical presentation of cerebral arteriovenous malformations. Stroke. 1995;26(7):12151220.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11

    Fierstra J, Conklin J, Krings T, et al. Impaired peri-nidal cerebrovascular reserve in seizure patients with brain arteriovenous malformations. Brain. 2011;134(Pt1):100109.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12

    Krings T, Hans FJ, Geibprasert S, Terbrugge K. Partial “targeted” embolisation of brain arteriovenous malformations. Eur Radiol. 2010;20(11):27232731.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13

    Kosnik EJ, Hunt WE, Miller CA. Dural arteriovenous malformations. J Neurosurg. 1974;40(3):322329.

  • 14

    Poublanc J, Han JS, Mandell DM, et al. Vascular steal explains early paradoxical blood oxygen level-dependent cerebrovascular response in brain regions with delayed arterial transit times. Cerebrovasc Dis Extra. 2013;3(1):5564.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 15

    Fierstra J, Poublanc J, Han JS, et al. Steal physiology is spatially associated with cortical thinning. J Neurol Neurosurg Psychiatry. 2010;81(3):290293.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    van Niftrik CHB, Piccirelli M, Bozinov O, et al. Impact of baseline CO2 on Blood-Oxygenation-Level-Dependent MRI measurements of cerebrovascular reactivity and task-evoked signal activation. Magn Reson Imaging. 2018;49:123130.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 17

    van Niftrik CHB, Piccirelli M, Bozinov O, et al. Iterative analysis of cerebrovascular reactivity dynamic response by temporal decomposition. Brain Behav. 2017;7(9):e00705.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18

    Sebök M, Niftrik C, Lohaus N, et al. Leptomeningeal collateral activation indicates severely impaired cerebrovascular reserve capacity in patients with symptomatic unilateral carotid artery occlusion. J Cereb Blood Flow Metab. 2021;41(11):30393051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19

    Sebök M, van Niftrik CHB, Winklhofer S, et al. Mapping cerebrovascular reactivity impairment in patients with symptomatic unilateral carotid artery disease. J Am Heart Assoc. 2021;10(12):e020792.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20

    van Niftrik CHB, Sebök M, Wegener S, et al. Increased ipsilateral posterior cerebral artery P2-segment flow velocity predicts hemodynamic impairment. Stroke. 2021;52(4):14691472.

    • Search Google Scholar
    • Export Citation
  • 21

    Fierstra J, van Niftrik C, Warnock G, et al. Staging hemodynamic failure with blood oxygen-level-dependent functional magnetic resonance imaging cerebrovascular reactivity: a comparison versus gold standard (15O-)H2O-positron emission tomography. Stroke. 2018;49(3):621629.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22

    van Niftrik CHB, Sebök M, Muscas G, et al. Characterizing ipsilateral thalamic diaschisis in symptomatic cerebrovascular steno-occlusive patients. J Cereb Blood Flow Metab. 2019;40(3):563573.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23

    Sebök M, van Niftrik CHB, Piccirelli M, et al. BOLD cerebrovascular reactivity as a novel marker for crossed cerebellar diaschisis. Neurology. 2018;91(14):e1328e1337.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24

    Sebök M, van Niftrik CHB, Wegener S, Luft A, Regli L, Fierstra J. Agreement of novel hemodynamic imaging parameters for the acute and chronic stages of ischemic stroke: a matched-pair cohort study. Neurosurg Focus. 2021;51(1):E12.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 25

    Kim DJ, Krings T. Whole-brain perfusion CT patterns of brain arteriovenous malformations: a pilot study in 18 patients. AJNR Am J Neuroradiol. 2011;32(11):20612066.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Fierstra J, Spieth S, Tran L, et al. Severely impaired cerebrovascular reserve in patients with cerebral proliferative angiopathy. J Neurosurg Pediatr. 2011;8(3):310315.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 27

    Busch KJ, Kiat H, Stephen M, Simons M, Avolio A, Morgan MK. Cerebral hemodynamics and the role of transcranial Doppler applications in the assessment and management of cerebral arteriovenous malformations. J Clin Neurosci. 2016;30:2430.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 28

    Lasjaunias P, Chiu M, ter Brugge K, Tolia A, Hurth M, Bernstein M. Neurological manifestations of intracranial dural arteriovenous malformations. J Neurosurg. 1986;64(5):724730.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29

    Hurst RW, Hackney DB, Goldberg HI, Davis RA. Reversible arteriovenous malformation-induced venous hypertension as a cause of neurological deficits. Neurosurgery. 1992;30(3):422425.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Diehl RR, Henkes H, Nahser HC, Kühne D, Berlit P. Blood flow velocity and vasomotor reactivity in patients with arteriovenous malformations. A transcranial Doppler study. Stroke. 1994;25(8):15741580.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 31

    De Salles AA, Manchola I. CO2 reactivity in arteriovenous malformations of the brain: a transcranial Doppler ultrasound study. J Neurosurg. 1994;80(4):624630.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Van Roost D, Schramm J. What factors are related to impairment of cerebrovascular reserve before and after arteriovenous malformation resection? A cerebral blood flow study using xenon-enhanced computed tomography. Neurosurgery. 2001;48(4):709717.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Schaller C, Schramm J, Haun D, Meyer B. Patterns of cortical oxygen saturation changes during CO2 reactivity testing in the vicinity of cerebral arteriovenous malformations. Stroke. 2003;34(4):938944.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34

    Tarr RW, Johnson DW, Rutigliano M, et al. Use of acetazolamide-challenge xenon CT in the assessment of cerebral blood flow dynamics in patients with arteriovenous malformations. AJNR Am J Neuroradiol. 1990;11(3):441448.

    • Search Google Scholar
    • Export Citation
  • 35

    Kim SW, Kim YD, Chang HJ, et al. Different infarction patterns in patients with aortic atheroma compared to those with cardioembolism or large artery atherosclerosis. J Neurol. 2018;265(1):151158.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36

    Spetzler RF, Hargraves RW, McCormick PW, Zabramski JM, Flom RA, Zimmerman RS. Relationship of perfusion pressure and size to risk of hemorrhage from arteriovenous malformations. J Neurosurg. 1992;76(6):918923.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37

    Norris JS, Valiante TA, Wallace MC, et al. A simple relationship between radiological arteriovenous malformation hemodynamics and clinical presentation: a prospective, blinded analysis of 31 cases. J Neurosurg. 1999;90(4):673679.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 38

    Moftakhar P, Hauptman JS, Malkasian D, Martin NA. Cerebral arteriovenous malformations. Part 2: physiology. Neurosurg Focus. 2009;26(5):E11.

  • 39

    Geibprasert S, Pongpech S, Jiarakongmun P, Shroff MM, Armstrong DC, Krings T. Radiologic assessment of brain arteriovenous malformations: what clinicians need to know. Radiographics. 2010;30(2):483501.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 40

    Fierstra J, Sobczyk O, Battisti-Charbonney A, et al. Measuring cerebrovascular reactivity: what stimulus to use?. J Physiol. 2013;591(23):58095821.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 41

    van Niftrik CHB, Piccirelli M, Muscas G, et al. The voxel-wise analysis of false negative fMRI activation in regions of provoked impaired cerebrovascular reactivity. PLoS One. 2019;14(5):e0215294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 42

    Li M, Liu Q, Guo R, et al. Perinidal angiogenesis is a predictor for neurovascular uncoupling in the periphery of brain arteriovenous malformations: a task-based and resting-state fMRI study. J Magn Reson Imaging. 2021;54(1):186196.

    • Crossref
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 446 447 298
PDF Downloads 359 359 243
EPUB Downloads 0 0 0