Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print April 19, 2002, 10.1148/radiol.2233010792
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2233010792v1
223/3/672    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lehéricy, S.
Right arrow Articles by Marsault, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lehéricy, S.
Right arrow Articles by Marsault, C.
(Radiology 2002;223:672-682.)
© RSNA, 2002

Neuroradiology

Arteriovenous Brain Malformations: Is Functional MR Imaging Reliable for Studying Language Reorganization in Patients? Initial Observations1

Stéphane Lehéricy, MD, PhD, Alessandra Biondi, MD, Nader Sourour, MD, Michaela Vlaicu, MD, Sophie Tézenas du Montcel, MD, Laurent Cohen, MD, PhD, Elio Vivas, MD, Laurent Capelle, MD, Thierry Faillot, MD, Alfredo Casasco, MD, Denis Le Bihan, MD, PhD and Claude Marsault, MD

1 From the Departments of Neuroradiology (S.L., A.B., N.S., E.V., A.C., C.M.), Neurology (L. Cohen), Neurosurgery (M.V., L. Capelle, T.F.), and Biostatistics and Medical Informatics (S.T.d.M.), Hôpital de la Salpêtrière, 47 Boulevard de l’Hôpital, 75013 Paris, France; and Department of Medical Research, Service Hospitalier Frédéric Joliot, Orsay, and the IFR 49, France (S.L., D.L.B.). From the 2000 RSNA scientific assembly. Received April 16, 2001; revision requested May 25; final revision received November 8; accepted December 11. Supported in part by grants from the Délégation à la Recherche Clinique (DRC) and the Assistance Publique-Hôpitaux de Paris (CRC 96067). Address correspondence to S.L. (e-mail: stephane.lehericy@psl.ap-hop-paris.fr).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
PURPOSE: To determine whether the blood flow abnormalities frequently associated with arteriovenous malformations (AVMs) can alter functional magnetic resonance (MR) imaging evaluation of language lateralization and whether reorganization of language function occurs in patients with brain AVMs.

MATERIALS AND METHODS: Eleven patients with left-hemisphere brain AVMs and 10 age-matched control subjects were examined with 1.5-T blood oxygen level–dependent (BOLD) functional MR imaging. Verbal fluency, sentence repetition, and story listening tasks were performed. The functional MR imaging laterality index in the frontal and temporal lobes was defined as the (L - R)/(L + R) ratio, where L and R are the numbers of activated pixels in the left and right hemispheres, respectively. Statistical analyses were performed with Wilcoxon signed rank, Fisher exact, and Kruskal-Wallis tests.

RESULTS: Control subjects had left-sided language dominance, although symmetric pixel counts were observed in the frontal lobes in two subjects and in the temporal lobes in one subject. Six patients had left-sided language dominance similar to that observed in control subjects. Five of these patients had AVMs outside frontal or temporal language areas, without flow abnormalities. Five patients had abnormally right-sided asymmetric indexes (below mean control subject value - 2 SDs), which suggested language reorganization (P < .05). Results of Wada examination and/or postembolization functional MR imaging performed in two of these patients showed that the abnormal laterality indexes were at least partly due to severe flow abnormalities that impaired detection of BOLD MR imaging signal intensity.

CONCLUSION: These data suggest that flow abnormalities may interfere with language lateralization assessment with functional MR imaging.

© RSNA, 2002

Index terms: Arteriovenous malformations, cerebral, 13.75, 13.76 • Brain, function • Brain, MR, 13.121412, 13.121413, 13.121416, 13.12144 • Magnetic resonance (MR), functional imaging, 13.12144


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
Treatment options for arteriovenous malformations (AVMs) of the brain include surgical removal, endovascular treatment, and stereotactic radiosurgery. The clinical outcome following treatment depends on many factors, including AVM features (eg, nidus size and venous drainage) and the location of the lesion relative to the functional cortex (1). Mapping of brain function in areas adjacent to vascular malformations may be performed by using many pre- and intraoperative techniques, including functional magnetic resonance (MR) imaging. Functional MR imaging has proved to be a valuable alternative to the Wada examination for evaluation of hemispheric dominance for language (2–7), and the results of studies (8) with comparisons of functional MR imaging and cortical stimulation have suggested that functional MR imaging may help in localizing areas critical to language. By depicting the relationship between the vascular lesions and the functional cortex, functional MR imaging may facilitate therapeutic planning, whether it be resection or endovascular treatment, while minimizing the potential clinical deficits (9–12). Functional MR imaging may also improve target volume definition at radiosurgery (13).

Language mapping is even more necessary in patients with brain AVMs than it is in patients with other types of brain lesions, because AVMs are commonly thought to be congenital. Thus, the development of these lesions, as well as the associated brain damage due to hemorrhage or ischemia, could lead to a reorganization of language areas. The results of previously performed studies (11–15) have suggested that a reorganization of language circuits may occur in these patients. However, blood flow abnormalities associated with AVMs, such as the steal phenomenon with retrograde feeding of the distal territory, may interfere with the functional MR imaging signal intensity changes that result from blood oxygen level–dependent (BOLD) contrast (16), and to our knowledge, this possibility has not been specifically addressed in previous studies.

The purpose of our study was to determine whether the blood flow abnormalities frequently associated with AVMs can alter the functional MR imaging evaluation of language lateralization and whether a reorganization of language function occurs in patients with brain AVMs.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
Patients and Control Subjects
Eleven consecutive patients (mean age ± SD, 38.8 years ± 14.8) referred for surgical or endovascular treatment of AVMs of the left hemisphere were examined. Routine examination of these patients included functional MR imaging evaluation of hemispheric dominance for language. After receiving approval from the local ethics committee, we compared the findings in the 11 patients with those in 10 age-matched control subjects (mean age, 34.4 years ± 17.5) (P > .19, Wilcoxon exact test) who gave their informed consent for participation in the study. There were significantly more female subjects than male subjects in the control group (P < .03, Fisher exact test). All patients and control subjects were right handed.

Imaging Protocol
We performed BOLD functional MR imaging by using a 1.5-T unit (Signa; GE Medical Systems, Milwaukee, Wis). In each subject, after a sagittal scout sequence was performed, 12 transverse gradient-echo echo-planar images of the entire area of the frontal lobes (5,000/60 [repetition time msec/echo time msec], 90° flip angle, 6-mm section thickness with no intersection gap, 3.75 x 3.75-mm in-plane resolution) and 12 transverse inversion-recovery three-dimensional fast spoiled gradient-echo images for anatomic localization (10.0/2.1, 400-msec inversion time, 10° flip angle, 6-mm section thickness, no intersection gap, 0.94 x 0.94-mm in-plane resolution) were obtained. The images were acquired in less than 45 minutes. To prevent involuntary head movement inside the magnet, the subject’s forehead was taped and firmly restrained on either side with foam pads.

Functional MR Imaging Tasks
Three different tasks were performed by the subjects: semantic fluency, covert sentence repetition, and story listening. All tasks were administered by the same person (S.L.) in the same way. In the semantic fluency task, subjects had to generate mentally as many words as possible from a given semantic category (ie, fruits, vegetables, pieces of furniture, body parts, animals, sports). The name of a different category was presented orally at the beginning of each activation period. Fluency was contrasted with a control rest condition. At the beginning of each control period, subjects were asked to stop generating words and to rest. In the covert sentence repetition task, subjects were presented with short sentences (mean duration, 2.88 seconds; range, 1.82–4.22 seconds), each of which was followed by a silent period of the same duration. During the silent period, subjects had to repeat the sentence mentally. Repetition was contrasted with a control rest condition. In the story listening task, subjects listened to a story that was divided into three 30-second segments. Story listening was contrasted with a control condition in which the subjects listened to the same story segments played backward so that the auditory component of the task could be subtracted. Task order was varied pseudorandomly among the subjects.

During MR imaging, the subjects lay in the dark. The stimuli were recorded on a digital audiotape and presented through standard headphones that were customized for functional MR imaging experiments and inserted in a noise-protecting helmet that provided isolation from imaging noise. The paradigm consisted of seven 30-second epochs in which the control and language conditions were alternated. Forty-two volumes of 12 sections were acquired in 3 minutes 30 seconds. The first four volumes of each sequence were used to reach a signal intensity equilibrium and were excluded from analysis.

All patients were free of any language impairment at clinical examination at the time of functional MR imaging. They all responded adequately during the short instruction phase that preceded the MR imaging experiment. They all were capable of generating a number of words within the specified category and of repeating short sentences. They all were capable of recalling several items of the story after the MR examination. Tasks were chosen for the following reasons on the basis of results from a previous study (7): (a) The semantic fluency task has been proved to be strongly correlated with the Wada examination; (b) the story listening task enabled an estimation of language laterality in the posterior language areas (inconstantly activated during the fluency task), although the temporal laterality index (LI) was not correlated with the Wada LI; and (c) the repetition task provided robust, more symmetric frontal and temporal activation.

Functional MR Imaging Analysis
Data were motion corrected (17), temporally filtered, and analyzed with dedicated software (ACTIV; CEA, Orsay, France) written in interactive data language (RSI, Boulder, Colo) by means of pixel-by-pixel autocorrelation and cross correlation with a reference waveform (18) of the MR imaging signal intensity time course. Motion was evaluated by using the cine mode and calculating the displacement of the center of the mass on the functional MR images over time in the three planes (<0.5 mm in all subjects, corresponding to less than 15% of pixel size in the transverse plane). Activated clusters were defined as follows: more than three contiguous pixels, correlation coefficient greater than 0.40, and autocorrelation coefficient greater than 0.20 (P < .001).

Activated pixels were overlaid on transverse anatomic images with a color scale representing the correlation coefficient. One investigator (S.L.) then localized the pixels according to the individual anatomy of the subjects by performing multiplanar analysis and three-dimensional surface rendering of the cortex (Voxtool; GE Medical Systems, Milwaukee, Wis) and clustered the pixels in two regions of interest. These regions were the frontal lobe, including all pixels rostral to the central sulcus, and the temporal lobe, including the inferior parietal lobule and the superior, middle, and inferior temporal gyri.

For each region of interest, the investigator (S.L.) computed the total number of activated pixels in the left hemisphere (L), the total number of activated pixels in the right hemisphere (R), and an LI, defined as the (L - R)/(L + R) ratio (LI range, -1 to +1). A positive LI corresponded to a left-predominant activation: An LI of +0.50 to +1.00 indicated strong left lateralization, and an LI of +0.25 to +0.50 indicated weak left lateralization. A negative LI corresponded to a right-predominant activation: An LI of -1.00 to -0.50 indicated strong right lateralization, and an LI of -0.50 to -0.25 indicated weak right lateralization. LIs for bilateral activation ranged from -0.25 to +0.25.

Frontal language dominance was estimated on the basis of the frontal LI during the semantic fluency task, which correlated best with the Wada examination in a previous study (7), and temporal language dominance was estimated on the basis of the temporal LI during the story listening task, because temporal activation was inconstant during the fluency task and symmetric during the repetition task. Intrarate reproducibility of LI measurements was evaluated by performing the analysis twice in five randomly selected control subjects. Language lateralization was classified as strong or weak left lateralization, bilateral, or strong or weak right lateralization during the first and second MR imaging examinations in these five subjects.

Follow-up Functional MR Imaging
To assess the effect of blood flow changes on LIs, follow-up functional MR imaging was performed in all the patients with abnormal LIs (patients 7–11), except patient 7, who died 2 years after undergoing functional MR imaging. In patients 9–11, follow-up functional MR examinations were performed 4–24 months after the endovascular treatment by using the protocol described earlier. Patient 8 underwent follow-up angiography, which depicted features that were similar to those observed at the time of the first functional MR imaging study. Follow-up functional MR imaging yielded unchanged LIs in this patient, and, thus, these data are not shown. Patients with normal LIs were not reexamined.

Angiographic Assessment
All patients underwent digital subtraction angiography before and after treatment. The angiographic characteristics of the AVMs and other patient information are given in Table 1. Endovascular treatment was performed in 10 patients; four of these patients then underwent surgery, and two of these patients then underwent radiation therapy. One patient underwent surgery only. In each embolization session, occlusion of the AVM compartments was achieved with intranidal injections of a mixture of glue and lipiodol (Lipiodol Ultra-fluide; Guerbet, Roissy, France). The permanent liquid polymerizing agent used was n-butyl cyanoacrylate (Histoacryl; Braun, Melsungen, Germany).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Patients and Lesions

 
Statistical Comparisons
Data are expressed as means plus or minus SDs. Frontal language dominance was estimated on the basis of the frontal LI during the semantic fluency task, and temporal language dominance was estimated on the basis of the temporal LI during the story listening task. Because of the small samples, the Wilcoxon signed rank test was used to perform paired comparisons. The distributions of the variables tested between groups were compared by using the Fisher exact test for qualitative variables and the Kruskal-Wallis test for quantitative variables. A P value of less than .05 was considered to indicate a statistically significant difference. Statistical analyses were performed by using a statistical software package (SAS 8.1; SAS Institute, Cary, NC).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
Control Subjects: Location of Language Areas
The patterns of activation observed in the control subjects during the three tasks were as follows:

Semantic fluency.—Activation was observed in the frontal lobes, including the inferior frontal gyrus and sulcus, the anterior insular cortex, the middle frontal gyrus, and the supplementary motor and cingulate areas. In a few subjects, activation was observed in posterior temporoparietal areas.

Covert sentence repetition.—Activation was observed bilaterally in the superior temporal gyrus and sulcus, the inferior temporal gyrus, the inferior parietal areas, the inferior and middle frontal gyri, the supplementary motor and cingulate areas, and less frequently in other regions. Activation was also observed bilaterally in the planum temporale, the primary auditory areas, and the inferior part of the precentral gyrus.

Story listening.—Activation was observed in the superior temporal gyrus and the inferior parietal areas in all subjects and in the medial frontal gyri, the supplementary motor and cingulate areas, and the superior and inferior frontal gyri in a few subjects.

Control Subjects: Hemispheric Language Dominance
The LIs for hemispheric language dominance in the control subjects are listed in Table 2. Overall, with the exception of subjects 4, 5, and 10, the control subjects had strong or weak left lateralization in both the frontal and temporal lobes. Subject 4 had a symmetric pixel count in the frontal lobes and strong left lateralization in the temporal lobes. Subject 5 had a symmetric pixel count in the temporal lobes. Subject 10 had a symmetric pixel count in the frontal lobes and weak left lateralization in the temporal lobes. With the exception of the temporal LI during the covert repetition task, LIs were significantly higher than 0 (Wilcoxon signed rank test). During the word generation and story listening tasks, the degrees of leftward asymmetry in the frontal and temporal lobes were not significantly different (Wilcoxon signed rank test) (Table 2). During the covert repetition task, the LIs in the frontal lobes were significantly higher than those in the temporal lobes (P < .01, Wilcoxon signed rank test), and the LIs in the temporal lobes were not significantly different from 0 (Wilcoxon signed rank test).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Functional MR Imaging LIs in the Frontal and Temporal Lobes Observed during Three Tasks

 
Patients
Like the control subjects, all patients were right handed and were expected to have LIs within the control subject LI range with left-sided language dominance. Patients were divided into two groups: those with LIs within the range of the control subject LIs (group 1) and those with LIs outside the range of the control subject LIs (mean ± 2 SDs) (group 2) (Table 2). All lesions were left sided. Overall, the locations of activation during the three tasks were similar to those observed in the control subjects.

Group 1.—Group 1 comprised six patients with strong or weak left lateralization for language in the frontal and temporal lobes (Fig 1). Patients 1–5 had AVMs outside the classic language areas and no flow abnormalities (eg, no or only a slight distal steal phenomenon). The angiograms obtained in these patients showed small or medium-sized arteriolovenular nidi in the prefrontal (patients 1 and 4), precentral (patient 3), temporal pole (patient 2 [Fig 1]), and insular (patient 5) areas. These five patients presented with a history of recent seizures and had normal clinical examination results. Patient 5 had an LI at the lowest limit of control subject values in the frontal lobes and a strongly left-sided LI in the temporal lobes. In this patient, the nidus was close to the anterosuperior part of the insula and adjacent to the area that is normally activated in that area during the fluency task.



View larger version (130K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Patient 2. Top and bottom, first three images from left: Transverse functional MR images obtained while patient performed semantic fluency and story listening tasks show strong left-hemisphere dominance in both the frontal and temporal lobes. Activated pixels are overlaid on transverse anatomic images; color scale represents correlation coefficients from blue (-1) to red (+1). Far right: Lateral (top) and coronal (bottom) digital subtraction angiograms of left internal carotid artery show medium-sized AVM (arrow) of the anterior part of the left temporal lobe without flow abnormalities. Subject’s right is on reader’s left in coronal and transverse views.

 
Patient 6 had a pattern of activation in the temporal lobe that was different from that of the control subjects. The temporal LI was symmetric (0.22) during the story listening task. During this task, activation at the junction between the superior temporal and inferior parietal areas was left sided as in normal cases, but activation in the more rostral parts of the superior temporal gyrus and sulcus was strongly right sided, in contrast to activation in this area in the control subjects. This patient had a history of hemorrhage following the first surgical procedure performed 4 years previously, which resulted in aphasia and a contralateral sensorimotor deficit from which he recovered completely.

Group 2.—Group 2 comprised five patients with LIs below the control subject values in the temporal or frontal lobes. This group was further subdivided into patients with either no or only moderate flow abnormalities (ie, slight distal steal phenomenon) (patients 7–9) and patients with severe flow abnormalities (patients 10 and 11). Patients 7 and 8 had AVMs in posterior language areas (Fig 2). These patients had strong or weak left lateralization in the frontal lobes and symmetric or weakly right-sided lateralization in the temporal lobes. They had small or medium-sized arteriolovenular or arteriolovenous nidi in posterior temporoparietal areas, including the posterior temporobasal (patient 7) and the posterior temporal-inferior parietal (patient 8) areas. Patient 9 had a medium-sized arteriolovenular nidus in the central area. In this patient, a moderate steal phenomenon was observed in the territory of the middle cerebral artery, which was retrogradely supplied through leptomeningeal anterior cerebral artery anastomoses. Patient 9 had a strong left-sided dominance in the frontal lobe and a weak right-sided dominance in the inferior parietal and temporal lobes.



View larger version (133K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Patient 7. Top: Transverse functional MR images obtained while patient performed story listening task show strong left-sided dominance in the frontal lobe (left) and right-sided dominance in the temporal lobes (right); color scale represents correlation coefficients from blue (-1) to red (+1). Bottom: Coronal (left) and lateral (right) digital subtraction angiograms of left internal carotid artery show medium-sized AVM (arrow) of the left temporal lobe and, with the exception of a discrete steal phenomenon in the territory of the anterior cerebral artery, no flow abnormalities. Subject’s right is on reader’s left in coronal and transverse views.

 
Two patients in group 2 had severe blood flow abnormalities. Patient 10 had a medium-sized arteriolovenular nidus associated with direct arteriovenous shunts (Fig 3). The distal territory of the middle cerebral artery was supplied retrogradely through pial anastomoses of the anterior cerebral artery. Functional MR imaging depicted a strong left-sided dominance in the frontal lobes and a strong right-sided dominance in the inferior parietal and temporal lobes. Embolization and surgery were performed. Surgery resulted in dysphasic and dysgraphic disorders.



View larger version (136K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Patient 10. Top left: Preembolization digital subtraction angiogram (lateral view) of left internal carotid artery shows medium-sized arteriolovenular nidus (arrow) associated with direct arteriovenous shunts in the posterior insula (better analyzed on hyperselective angiograms) with severe hemodynamic disturbances. The distal territory of the middle cerebral artery was supplied retrogradely through pial anastomoses of the anterior cerebral artery (not shown). Bottom left: Follow-up digital subtraction angiogram (lateral view) shows that embolization reduced nidus size (arrow) and greatly normalized flow patterns. Top middle and right: Transverse preembolization functional MR images obtained at the level of the frontal (middle) and temporal (right) lobes while patient performed story listening task show left-hemisphere dominance in the frontal lobe and strong right-hemisphere dominance in the parietotemporal areas. Bottom middle and right: Transverse postembolization functional MR images obtained at the level of the frontal (middle) and temporal (right) lobes show reappearance of activated pixels in left hemisphere; color scale represents correlation coefficients from blue (-1) to red (+1). Subject’s right is on reader’s left in transverse views.

 
Patient 11 had a medium-sized AVM in the prefrontal lobe associated with severe hemodynamic disturbances in the middle cerebral artery territory due to a substantial steal effect and a stenosis of the proximal segment of the middle cerebral artery (Fig 4). The distal territory of the middle cerebral artery was supplied through pial anastomoses arising from the anterior cerebral artery. This patient’s language and sensorimotor deficits were considered to be due to these flow abnormalities. This patient had strong right-sided lateralization in the middle and inferior frontal gyri and in the inferior parietal and lateral temporal lobes (territory of the middle cerebral artery) and weak left-sided lateralization in the supplementary motor area (territory of the anterior cerebral artery). No activated pixel was found in the territory of the left middle cerebral artery, despite clinical evidence of language representation in the left hemisphere. Thus, this patient underwent a Wada examination, and the results confirmed that language was represented in the left hemisphere. In both patients with severe flow abnormalities, the LIs were right sided because of decreased activation volume in the left hemisphere rather than increased activation in the right hemisphere (Table 3).



View larger version (167K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Patient 11. Top left and middle: Transverse preembolization (Pre) functional MR images show strong right-hemisphere dominance with no pixels activated in the territory of the left middle cerebral artery; color scale represents correlation coefficients from blue (-1) to red (+1). Top right: Preembolization digital subtraction angiogram (coronal view) shows medium-sized left prefrontal AVM (twin arrows) associated with substantial steal effect (arrowheads) and stenosis of the middle cerebral artery (single arrow). Bottom left and middle: Transverse follow-up functional MR images show reappearance of activated pixels in left hemisphere. Bottom right: Postembolization (Post) digital subtraction angiogram (coronal view) shows that embolization and vasodilation of the middle cerebral artery greatly normalized flow patterns. Subject’s right is on reader’s left.

 

View this table:
[in this window]
[in a new window]

  		TABLE 3. Activation Volumes in the Frontal and Temporal Lobes during Three Tasks

 
To assess the effect of flow changes on BOLD contrast, follow-up functional MR imaging was performed in three patients (patients 9–11) 4–24 months after the endovascular treatment by using the protocol described herein earlier. Embolization greatly reduced nidus size and normalized flow patterns. In patient 9, posttreatment hemispheric dominance remained left sided in the frontal lobes and right sided in the temporal lobes. In patient 10, functional MR imaging was performed before the onset of language deficits that occurred following surgery. The temporal LI during the story listening and repetition tasks changed from strongly right sided before embolization to symmetric during story listening and weakly right sided during repetition after treatment. In patient 11, the LIs changed greatly: from strongly right sided in both the frontal and temporal lobes before embolization to symmetric in the temporal lobes (although still below control values) and slightly left sided in the frontal lobes after treatment.

Comparisons among subject groups.—Only the temporal LI during the story listening task was significantly higher in the control subjects than in the patients with AVMs (P < .02, Kruskal-Wallis test). Overall in the patients, only the frontal LIs during the word generation (P < .003) and story listening (P < .02) tasks were significantly different from 0 (Wilcoxon signed rank test). In group 1, only the frontal LI during word generation was significantly different from 0 (P < .04), whereas in group 2, the LIs were not significantly different from 0 (Wilcoxon signed rank test). The locations of the AVMs in groups 1 and 2, whether within or outside language areas, were similar (nonsignificant difference, Fisher exact test). No patient in group 1 had flow abnormalities, whereas all patients in group 2 had either discrete or severe flow abnormalities (P < .002, Fisher exact test). During the story listening task, the temporal LIs in group 1 were higher (mean LI ± SD, 0.53 ± 0.36) than those in group 2 (-0.68 ± 0.46) (P < .05, Kruskal-Wallis test). All other LIs were higher in group 1 than in group 2, but the difference was not significant.

The activation volumes in the control subjects, group 1 patients, and group 2 patients were compared (Table 3). The volume of activation in the left temporal lobe during the repetition task was the only variable that was significantly different among the three groups (P < .03, Kruskal-Wallis test). Results of two-by-two comparisons showed that this difference was due to only a lower volume of activation in the group 2 patients compared with the volume of activation in the control subjects (P < .02). To further understand the effect of treatment on LIs, we statistically compared the magnitude of functional MR imaging signal intensity changes before and after treatment in patients 9–11. There was no significant difference between the pre- and posttreatment signal intensity changes (mean magnitude of functional MR imaging signal intensity changes in these three patients ± SD: pretreatment left hemisphere, 1.79% ± 0.31; pretreatment right hemisphere, 1.66% ± 0.43; posttreatment left hemisphere, 1.56% ± 0.47; posttreatment right hemisphere, 1.69% ± 0.33; Wilcoxon signed rank test).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
Assessment of Language Dominance with Functional MR Imaging
Functional MR imaging has proved to be a reliable technique for assessing hemispheric dominance for language (2–7,19–22). The results of several studies (2–7) have shown a concordance between the degree of asymmetry of activation during language tasks and the hemispheric dominance as determined with the Wada examination. However, unlike the Wada examination (23), in which the successive injection of amobarbital into the two internal carotid arteries is required to induce temporary inactivation of each hemisphere, functional MR imaging is totally noninvasive. Thus, functional MR imaging appears to be well suited for assessment of hemispheric language dominance in patients with brain lesions and of reorganization of language circuits; however, this technique often cannot enable differentiation of the areas that are necessary for language function from those areas that are associated with task performance only. Production tasks, such as verbal fluency and word generation, which activate predominantly the frontal areas, have been better at indicating language lateralization than receptive tasks, such as story listening and repetition, which are associated with more prominent temporal activation (6,7).

Temporal activation during the covert repetition task was not asymmetric, because it included the planum temporale and primary auditory areas, which are poorly lateralized (7). In the present study, all control subjects had positive LIs in the frontal lobes and nine control subjects were left-hemisphere dominant in the temporal lobes; these findings are consistent with previous results. On the basis of these results, the patients were expected to have positive frontal and temporal LIs within the range of control subject values (mean ± 2 SDs).

Language Reorganization in Patients with AVMs
Vascular malformations are believed to form during gestation, and the development of these lesions and the associated brain damage due to hemorrhage or ischemia could lead to reorganization of not only the local anatomy but also the functional cortex. Such reorganization of brain function has been suggested recently for language with functional MR imaging in patients with epilepsy (24) or with superselective arterial injection of anesthetics during angiography in patients with AVMs (14,25). In these studies, patients with temporoparietal AVMs had imaging evidence that some of the language skills typically supported by posterior areas were displaced into anterior frontal areas (14), and patients with frontal AVMs had imaging evidence that features of expressive language were supported by homotopic right frontal areas (25). Left-to-right reorganization of language function has also been suggested with functional MR imaging (11,26). Two right-handed patients with temporal AVMs had larger activation in the right hemisphere than in the left hemisphere during a semantic judgment task (11). However, the effects of flow changes associated with AVM were not addressed in that study.

Functional MR Imaging Signal Intensity Detection in Patients with Severe Flow Abnormalities
In the present study, five patients had abnormal LIs—that is, indexes below the control subject value range. Data on patients 10 and 11 provide evidence that these abnormal LIs were secondary to altered BOLD signal intensity detection due to marked flow abnormalities. The LIs were abnormally right sided because of decreased activation volume in the left hemisphere in the patients with AVMs, not because of increased activation volumes in the right hemisphere. These patients had abnormally right-sided LIs in the temporal lobes despite clinical evidence of language representation in the left hemisphere. Furthermore, no pixel was activated in the left hemisphere in patient 11, although Wada examination results confirmed that this hemisphere was necessary for language. In these two patients, the postembolization LIs measured after the disappearance of the flow abnormalities were greatly increased compared with the preembolization values.

The impaired detection of activation in areas adjacent to AVMs may be explained in several ways. Hypotension or the presence of a so-called "steal phenomenon" may alter the normal functions in areas adjacent to an AVM (27–30). However, cerebral blood flow reductions do not necessarily cause cerebral dysfunction, as suggested in previous reports (14,31) and by the results of the Wada examination performed in patient 11. Alternatively, BOLD signal intensity variations may not be detected in areas adjacent to the nidus. The working model of BOLD contrast MR imaging postulates that an increase in neuronal activity causes blood flow to increase such that the amount of paramagnetic deoxyhemoglobin in the microvasculature is reduced, and this leads to an increase in T2* relaxation and thus in signal intensity (16). Changes in cerebral blood flow (27,32,33), perfusion pressure (28–30,33), oxygen metabolism (34), autoregulation process, and vasoreactivity (27–29,33,35–37), all of which have been reported in areas adjacent to AVMs, may alter BOLD signal intensity detection. The combined use of BOLD contrast and other perfusion techniques, such as arterial spin labeling, might yield different sensitivity in cases of flow abnormalities and thus help in determining language lateralization (38).

Language Lateralization in Patients with No or Mild Flow Abnormalities
Given the limitations of functional MR imaging in patients who have AVMs and severe blood flow disturbances, the possibility that the right-sided asymmetry indexes observed in patients 7–9, who had either no or moderate flow abnormalities, represented signs of language reorganization (11,26) remains speculative. Several arguments are in favor of this hypothesis, although none is sufficient to confirm it. AVMs were located in those temporal regions—specifically, the posterior parts of the superior and middle temporal gyri—where activation was commonly observed during the story listening and repetition tasks in the control subjects. These three patients had no clinical evidence of language dysfunction before or after treatment. Angiographic examination results showed mild flow abnormalities, evidence against the belief that flow changes may have interfered with BOLD signal intensity detection. At posttreatment functional MR imaging performed in patient 9 after the almost complete disappearance of the AVM and flow abnormalities, the LIs were similar to pretreatment values. In these three patients, the frontal LI was left sided as in normal cases, and this suggests that language reorganization due to AVM, if present, may occur specifically in homotopic brain areas contralateral to those affected by the AVM. A contribution of frontal areas in the reorganization of the language network also may be possible (14).

Sex-related differences among the groups cannot account for the differences in LIs that we observed. Language function typically has been described as being more strongly lateralized in the left hemisphere in male patients than in female patients (39,40). In the present study, the difference in language lateralization was the opposite of what would be expected in this case: Language was more lateralized in the control group—which included more female subjects, in whom language should be less lateralized—than in the patients—who included more male patients, in whom language should be more lateralized. However, the possibility that abnormal LIs in patients with either no or mild flow abnormalities may indicate language reorganization needs to be studied further by means of a reference standard, such as the Wada examination or cortical stimulation.

Last, six patients had normal LIs. Five of these patients (patients 1–5) had left-sided AVMs outside the classic frontal or temporal language areas and no flow abnormalities. This suggests that patients who have AVMs that spare the language areas and that are not associated with flow changes have a left-hemisphere dominance similar to that observed in control subjects. In patient 6, the right-sided temporal LI may have been secondary to the reorganization of language function following the recovery of a brain hemorrhage. Four years before undergoing functional MR imaging, patient 6 had a left temporal hemorrhage that resulted in language disturbances. This is consistent with the strongly left-sided frontal LI observed in this patient and suggests a left-hemisphere dominance for language in the frontal lobe. However, although the global temporal LI was symmetric during the story listening task, the pattern of temporal activation was different from that in the control subjects. Activation at the junction between the superior temporal and inferior parietal areas was left sided as in normal cases, but activation in the more rostral parts of the superior temporal gyrus and sulcus was strongly right sided; these findings differ from those observed in the control subjects. The results of previously performed positron emission tomographic studies have suggested that a redistribution of activity in homotopic contralateral temporal areas may be the central mechanism of functional recovery after brain lesion (41–44), although restoration of left-hemisphere activation has been more consistently associated with better recovery of language function (45–47).

In conclusion, in two patients with AVMs, follow-up functional MR imaging performed after embolization showed that the abnormal LIs were at least partly secondary to decreased BOLD signal intensity detection due to severe flow changes. These findings suggest that functional MR imaging does not enable reliable evaluation of language lateralization in patients with severe blood flow changes. On the other hand, the patients with AVMs outside the classic frontal or temporal language areas and without flow abnormalities in our study had the expected left-hemisphere dominance, similar to that observed in the control subjects. Last, three patients with either few or no flow abnormalities and AVMs within the language areas had altered language lateralization; these findings suggest language reorganization, but this theory needs to be confirmed by means of Wada examination or cortical stimulation.


   FOOTNOTES
 
Abbreviations: AVM = arteriovenous malformation, BOLD = blood oxygen level dependent, LI = laterality index

Author contributions: Guarantor of integrity of entire study, S.L.; study concepts, S.L., D.L.B., A.C., A.B., C.M.; study design, S.L., D.L.B., L. Cohen, A.C., C.M.; literature research, S.L., A.B.; clinical studies, A.B., N.S., M.V., E.V., T.F., L. Capelle, A.C.; experimental studies, S.L., D.L.B.; data acquisition, S.L.; data analysis/interpretation, S.L., D.L.B.; statistical analysis, S.L., S.T.d.M.; manuscript preparation, S.L.; manuscript definition of intellectual content, S.L., A.B., A.C., C.M.; manuscript editing, S.L., L. Cohen, A.B., N.S.; manuscript revision/review and final version approval, S.L.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 

  1. Spetzler RF, Martin NA. A proposed grading system for arteriovenous malformations. J Neurosurg 1986; 65:476-483.[Medline]
  2. Desmond JE, Sum JM, Wagner AD, et al. Functional MRI measurement of language lateralization in Wada-tested patients. Brain 1995; 118:1411-1419.[Abstract/Free Full Text]
  3. Binder JR, Swanson SJ, Hammeke TA, et al. Determination of language dominance using functional MRI: a comparison with the Wada test. Neurology 1996; 46:978-984.[Abstract/Free Full Text]
  4. Hertz-Pannier L, Gaillard WD, Mott SH, et al. Noninvasive assessment of language dominance in children and adolescents with functional MRI: a preliminary study. Neurology 1997; 48:1003-1012.[Abstract]
  5. Bahn MM, Lin W, Silbergeld DL, et al. Localization of language cortices by functional MR imaging compared with intracarotid amobarbital hemispheric sedation. AJR Am J Roentgenol 1997; 169:575-579.[Abstract/Free Full Text]
  6. Benson RR, FitzGerald DB, LeSueur LL, et al. Language dominance determined by whole-brain functional MRI in patients with brain lesions. Neurology 1999; 52:798-809.[Abstract/Free Full Text]
  7. Lehericy S, Cohen L, Bazin B, et al. Functional MR evaluation of temporal and frontal language dominance compared with the Wada test. Neurology 2000; 54:1625-1633.[Abstract/Free Full Text]
  8. FitzGerald DB, Cosgrove GR, Ronner S. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. AJNR Am J Neuroradiol 1997; 18:1529-1539.[Abstract]
  9. Baumann SB, Noll DC, Kondziolka DS, et al. Comparison of functional magnetic resonance imaging with positron emission tomography and magnetoencephalography to identify the motor cortex in a patient with an arteriovenous malformation. J Image Guid Surg 1995; 1:191-197.[CrossRef][Medline]
  10. Schlosser MJ, McCarthy G, Fulbright RK, Gore JC, Awad IA. Cerebral vascular malformations adjacent to sensorimotor cortex: functional magnetic resonance imaging studies before and after therapeutic intervention. Stroke 1997; 28:1130-1137.[Abstract/Free Full Text]
  11. Maldjian J, Atlas SW, Howard RS, et al. Functional magnetic resonance imaging of regional brain activity in patients with intracerebral arteriovenous malformations before surgical and endovascular therapy. J Neurosurg 1996; 84:477-483.[Medline]
  12. Stapleton SR, Kiriakopoulos E, Mikulis D, et al. Combined utility of functional MRI, cortical mapping, and frameless stereotaxy in the resection of lesions in eloquent areas of brain in children. Pediatr Neurosurg 1997; 26:68-82.[Medline]
  13. Schad LR, Bock M, Baudendistel K, et al. Improved target volume definition in radiosurgery of arteriovenous malformations by stereotactic correlation of MRA, MRI, blood bolus tagging, and functional MRI. Eur Radiol 1996; 6:38-45.[CrossRef][Medline]
  14. Lazar RM, Marshall RS, Pile-Spellman J, et al. Anterior translocation of language in patients with left cerebral arteriovenous malformation. Neurology 1997; 49:802-808.[Abstract/Free Full Text]
  15. Latchaw RE, Hu X, Ugurbil K, Hall WA, Madison MT, Heros RC. Functional magnetic resonance imaging as a management tool for cerebral arteriovenous malformations. Neurosurgery 1995; 37:619-625.[Medline]
  16. Ogawa S, Tank DW, Menon R, et al. Intrinsic signal changes accompanying sensory stimulation: functional brain mapping with magnetic resonance imaging. Proc Natl Acad Sci U S A 1992; 89:5951-5955.[Abstract/Free Full Text]
  17. Woods RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images. J Comput Assist Tomogr 1992; 16:620-633.[Medline]
  18. Bandettini PA, Jesmanowicz A, Wong EC, Hyde JS. Processing strategies for time-course data sets in functional MRI of the human brain. Magn Reson Med 1993; 30:161-173.[Medline]
  19. Binder JR, Rao SM, Hammeke TA, et al. Lateralized human brain language systems demonstrated by task subtraction functional magnetic resonance imaging. Arch Neurol 1995; 52:593-601.[Abstract]
  20. Dehaene S, Dupoux E, Mehler J, et al. Anatomical variability in the cortical representation of first and second language. Neuroreport 1997; 8:3809-3815.[Medline]
  21. Muller RA, Rothermel RD, Behen ME, Muzik O, Mangner TJ, Chugani HT. Receptive and expressive language activations for sentences: a PET study. Neuroreport 1997; 8:3767-3770.[Medline]
  22. Pujol J, Deus J, Losilla JM, Capdevila A. Cerebral lateralization of language in normal left-handed people studied by functional MRI. Neurology 1999; 52:1038- 1042.[Abstract/Free Full Text]
  23. Wada J, Rasmussen T. Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance: experimental and clinical observations. J Neurosurg 1960; 17:266-282.
  24. Springer JA, Binder JR, Hammeke TA, et al. Language dominance in neurologically normal and epilepsy subjects: a functional MRI study. Brain 1999; 122:2033-2046.[Abstract/Free Full Text]
  25. Lazar RM, Marshall RS, Pile-Spellman J, et al. Interhemispheric transfer of language in patients with left frontal cerebral arteriovenous malformation. Neuropsychologia 2000; 38:1325-1332.[CrossRef][Medline]
  26. Vikingstad EM, Cao Y, Thomas AJ, Johnson AF, Malik GM, Welch KM. Language hemispheric dominance in patients with congenital lesions of eloquent brain. Neurosurgery 2000; 47:562-570.[CrossRef][Medline]
  27. Barnett GH, Little JR, Ebrahim ZY, Jones SC, Friel HT. Cerebral circulation during arteriovenous malformation operation. Neurosurgery 1987; 20:836-842.[Medline]
  28. Fogarty-Mack P, Pile-Spellman J, Hacein-Bey L, et al. Superselective intraarterial papaverine administration: effect on regional cerebral blood flow in patients with arteriovenous malformations. J Neurosurg 1996; 85:395-402.[Medline]
  29. Young WL, Pile-Spellman J, Prohovnik I, Kader A, Stein BM. Evidence for adaptive autoregulatory displacement in hypotensive cortical territories adjacent to arteriovenous malformations. Neurosurgery 1994; 34:601-610.[Medline]
  30. Jungreis CA, Horton JA, Hecht ST. Blood pressure changes in feeders to cerebral arteriovenous malformations during therapeutic embolization. AJNR Am J Neuroradiol 1989; 10:575-577.[Abstract]
  31. Mast H, Mohr JP, Osipov A, et al. "Steal" is an unestablished mechanism for the clinical presentation of cerebral arteriovenous malformations. Stroke 1995; 26:1215-1220.[Abstract/Free Full Text]
  32. Young WL, Prohovnik I, Ornstein E, et al. The effect of arteriovenous malformation resection on cerebrovascular reactivity to carbon dioxide. Neurosurgery 1990; 27:257-266.[CrossRef][Medline]
  33. Hassler W, Steinmetz H. Cerebral hemodynamics in angioma patients: an intraoperative study. J Neurosurg 1987; 67:822-831.[Medline]
  34. Fink GR. Effects of cerebral angiomas on perifocal and remote tissue: a multivariate positron emission tomography study. Stroke 1992; 23:1099-1105.[Abstract/Free Full Text]
  35. Batjer HH, Devous MD. The use of acetazolamide-enhanced regional cerebral blood flow measurement to predict risk to arteriovenous malformation patients. Neurosurgery 1992; 31:213-217.[Medline]
  36. 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; 16:1865-1874.[Abstract]
  37. Takeshita G, Toyama H, Nakane K, et al. Evaluation of regional cerebral blood flow changes on perifocal brain tissue SPECT before and after removal of arteriovenous malformations. Nucl Med Commun 1994; 15:461-468.[Medline]
  38. Wong EC, Buxton RB, Frank LR. Implementation of quantitative perfusion imaging techniques for functional brain mapping using pulsed arterial spin labeling. NMR Biomed 1997; 10:237-249.[CrossRef][Medline]
  39. Kansaku K, Yamaura A, Kitazawa S. Sex differences in lateralization revealed in the posterior language areas. Cereb Cortex 2000; 10:866-872.[Abstract/Free Full Text]
  40. Shaywitz BA, Shaywitz SE, Pugh KR, et al. Sex differences in the functional organization of the brain for language. Nature 1995; 373:607-609.[CrossRef][Medline]
  41. Weiller C, Isensee C, Rijntjes M, et al. Recovery from Wernicke’s aphasia: a positron emission tomographic study. Ann Neurol 1995; 37:723-732.[CrossRef][Medline]
  42. Buckner RL, Corbetta M, Schatz J, Raichle ME, Petersen SE. Preserved speech abilities and compensation following prefrontal damage. Proc Natl Acad Sci U S A 1996; 93:1249-1253.[Abstract/Free Full Text]
  43. Mimura M, Kato M, Kato M, et al. Prospective and retrospective studies of recovery in aphasia: changes in cerebral blood flow and language functions. Brain 1998; 121:2083-2094.[Abstract/Free Full Text]
  44. Thulborn KR, Carpenter PA, Just MA. Plasticity of language-related brain function during recovery from stroke. Stroke 1999; 30:749-754.[Abstract/Free Full Text]
  45. Belin P, Van Eeckhout P, Remy P, et al. Recovery from nonfluent aphasia after melodic intonation therapy: a PET study. Neurology 1996; 47:1504-1511.[Abstract/Free Full Text]
  46. Cao Y, Vikingstad EM, George KP, Johnson AF, Welch KM. Cortical language activation in stroke patients recovering from aphasia with functional MRI. Stroke 1999; 30:2331-2340.[Abstract/Free Full Text]
  47. Heiss WD, Kessler J, Thiel A, Ghaemi M, Karbe H. Differential capacity of left and right hemispheric areas for compensation of poststroke aphasia. Ann Neurol 1999; 45:430-438.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 NeurologyHome page
W. D. Gaillard, M. M. Berl, E. N. Moore, E. K. Ritzl, L. R. Rosenberger, S. L. Weinstein, J. A. Conry, P. L. Pearl, F. F. Ritter, S. Sato, et al.
Atypical language in lesional and nonlesional complex partial epilepsy
Neurology, October 30, 2007; 69(18): 1761 - 1771.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
C. Stippich, N. Rapps, J. Dreyhaupt, A. Durst, B. Kress, E. Nennig, V. M. Tronnier, and K. Sartor
Localizing and Lateralizing Language in Patients with Brain Tumors: Feasibility of Routine Preoperative Functional MR Imaging in 81 Consecutive Patients
Radiology, June 1, 2007; 243(3): 828 - 836.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
D. Le Bihan, S.-i. Urayama, T. Aso, T. Hanakawa, and H. Fukuyama
Direct and fast detection of neuronal activation in the human brain with diffusion MRI
PNAS, May 23, 2006; 103(21): 8263 - 8268.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
W. D. Gaillard, L. Balsamo, B. Xu, C. McKinney, P. H. Papero, S. Weinstein, J. Conry, P. L. Pearl, B. Sachs, S. Sato, et al.
fMRI language task panel improves determination of language dominance
Neurology, October 26, 2004; 63(8): 1403 - 1408.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
J. Sailor, M. E. Meyerand, C. H. Moritz, J. Fine, L. Nelson, B. Badie, and V. M. Haughton
Supplementary Motor Area Activation in Patients with Frontal Lobe Tumors and Arteriovenous Malformations
AJNR Am. J. Neuroradiol., October 1, 2003; 24(9): 1837 - 1842.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2233010792v1
223/3/672    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lehéricy, S.
Right arrow Articles by Marsault, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lehéricy, S.
Right arrow Articles by Marsault, C.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY