Neuropsychological improvements following endovascular embolization of arteriovenous malformation

J Stroke Cerebrovasc Dis 1994;4:194-199 © 1994 National Stroke Association

Neuropsychological Improvements Following Endovascular Embolization of Arteriovenous Malformation Matthew Bowen, Ph.D., Gary K. Steinberg, M.D., Ph.D., Michael Marks, M.D., and Yvonne Acosta, OTR

Two right-handed patients with large, left-hemisphere arteriovenous malformations (AVM) underwent serial endovascular embolization and neuropsychological testing procedures. One patient presented with hemorrhage, hemiplegia, and aphasia; the other patient presented with a seizure only with multiple cognitive impairments. Case 1 was assessed for motor-sensory skills in the left hand and for visual recall (nondominant hemisphere). Case 2 was administered a more global assessment of cognitive functions. Both patients demonstrated significant improvements in neuropsychological functions 1 month following endovascular embolization with >50% reduction in AVM volume. The nature and extent of these improvements are discussed with respect to intervention-related, reversed impairments. Key Words: Arteriovenous malformation-CerebrovascularstealTranshemispheric diaschisis.

A small literature indicates that surgical resection can enhance higher cortical and motor-sensory skills impaired by arteriovenous malformation (AVM) (16). No studies have investigated the effect of endovascular embolization on recovery of function. Furthermore, only surgical study has shown postoperative improvements in abilities controlled by the hemisphere contralateral to the AVM (4). Identifying the presence and potential reversability of "vascular steal" due to AVM is an important topic requiring more data. In this article, we present two patients assessed for higher cortical and motor-sensory functions before and 1 month after endovascular embolization. The patients differed in their initial presentation (hernor-

From the Department of Neurosurgery and Division of Neuroradiology, Stanford University School of Medicine, Stanford, CA, U.s.A. Address correspondence and reprint requests to Dr. M. Bowen at 235 22nd Avenue, San Francisco, CA 94121, U.s.A. 194

J STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994

rhage versus seizure) and functional symptoms. Both patients showed significant improvement in neuropsychological testing following endovascular embolization of their AVM. These findings are discussed with regard to the phenomenon of cerebrovascular steal (7), lateralized hemisphere dysfunction, and recovery from cerebrovascular insults (8-11).

Case Reports Case 1 A 38-year-old, right-handed man suffered acute rupture of an AVM. He presented with right hemiplegia and moderately severe expressive aphasia. He did not lose consciousness. Computed tomography (CT) demonstrated a 3-cm left temporoparietal hematoma with minimal midline shift. Cerebral angiography revealed a 3.5-cm X 3.0-cm X 2.5-cm AVM in the left deep frontal region, including the basal ganglia, which was fed by the lenticulostriate and insular branches of the left middle cerebral artery and left

EUROPSYCHOLOGICALIMPROVEMENTSAFTEREMBOLIZATION

Figure 1. Leftcarotid anteroposterior angiograms. Pre-embolization angiogram (Jeft) showslarge AVM indeep leftfrontal region, fedby lenticulostriate insular branches of middle urethral and anterior cerebral arteries . Angiogram following endouascular embolization with N-butylcyanoacrylate (right) shows 50% reduction in AVMvolume.

anterior cerebral artery (Fig.1,left). Several intranidus aneurysms were also present. The venous drainage was both deep and superficial. Magnetic resonance imaging (MRI) confirmed the AVM to be situated in the external capsule and basal ganglia (Fig. 1, left). The patient was placed on -phenobarbltel and phenytopin prophylaxis. After 10 days, recovery included slight return of movement in the right leg and shoulder. Speech improved to a functional level for single word to command and spontaneous phrase production with response latency and verbal dyspraxia . A neurosurgeon evaluated the patient, who at that time was determined to be a candidate for endovascular embolization followed by helium ion, Bragg peak stereotactic radiosurgery because of the high risk of potential damage to language and motor functions with conventional microsurgery. Sixty days posthemorrhage and 1 day prior to embolization (February 18, 1992), the patient was administered standardized neuropsychological tests. Measures assessed visual recall and complex integrative motor-sensory skills in the unaffected limb and abilities harbored in the intact right hemisphere,

contralateral to the AVM. The dominant hemisphere, harboring the AVM, was unable to be assessed by these standardized measures due to the persistence of hemiparesis. Endovascular embolization of three lenticulostriate arteries feeding the AVM was performed on February 19, using N-butyl cyanocrylate (NBCA, Avacryl). Four venous aneurysms within the nidus were also obliterated. Approximately 50% of the AVM volume was reduced by the embolization procedure (Fig. 1, right). Repeat neuropsychological assessment of the left upper extremity was conducted 30 days after embolization . Table 1 provides the scores on the standardized neuropsychological measures administered before and after embolization. 'These data show significant improvement in a range of abilities, either strictly motor-sensory or also demanding higher cortical functions. Importantly, the skills assessed are welldocumented as relying primarily on the right hemisphere, in this case opposite the AVM. These data suggest that the patient demonstrated more dramatic gains relative to the complexity of the measure. Furthermore, visual memory, the sole measure not re/ STROKECEREBROVASCDIS,VOL. 4, NO.3, 1994

195

M BOWENETAL

Table 1.

Scores on neuropsqchologica! measures localized to tile rig1lt hemisphere (contralateral to tile AVM) Preembolization'

Pegboard Grip Finger Tapping Trail Making A Trail Making B TPT (sensorimotor) Visual Reproduction

94 s 26 kgs. 48

53 s 160 s 5:37 c

Normal range

Postembolization"

79s (+15%) 31 kgs. (+20%) 59 (+25%) 35 s (+33%) 104 s (+33%) 2:23 (+143%) Not reassessed

All tests utilized the hand ipsilateral to the AVM. 'Sixty days posthemorrhage, 1 day prior to embolization. bNinety days posthemorrhage, 30 days after embolization. cRecaIl of shapes and location was within normal limits, although time to completion was severely impaired.

quiring motor and kinesthetic function, was the only score in the normal range on initial assessment, suggesting the relatively focal nature of the anatomical deficit. The patient's residual AVM was subsequently treated with stereotactic, helium-ion radiosurgery (Table 1).

Case 2 A 41-year-old, right-handed college-educated man presented with a grand mal seizure. CT, MRI, and angiography demonstrated a large left anterior temporal AVMwith the nidus measuring approximately 3 X 4 X 4 em. The venous drainage was predominantly cortical, and drainage demonstrated a massively enlarged varix with extensive pulsation artifact and mass effect. The AVM was supplied via the left internal carotid and left posterior cerebral artery. There were anterior temporal branches from the middle cerebral artery and the anterior branch of the posterior cerebral artery. There was no evidence of acute or subacute hemorrhage with CT or MRI examinations . The patient presented clinically with impaired recent memory, speech production, word finding, speed of information processing, and thought organization. Initial embolization was conducted on December 9, 1992, and reduced the AVM by approximately 33%. Nonetheless, the patient showed no apparent functional gains in cognitive-linguistic abilities and reported no subjective sense of improvement. Neuropsychological testing on January 9, 1993, yielded a range of highly significant global impairments in mental speed, sustained attention, verbal 196

J STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994

and visual memory, fine motor speed, and higher visuoperceptual analysis . A second embolization was done on January 25,1993. This achieved an overall reduction of approximately 65% of the AVM volume (Fig. 2). Within days immediately thereafter, the patient reported a significant subjective sense of overall, functional improvements. He was administered a formal repeat neuropsychological examination of February 24, 1993. Highly significant improvements were found in a range of higher cortical functions. Table 2 gives scores on the standardized cognitive measures comparing results before and after the second embolization. The patient subsequently underwent microsurgical resection of his AVM (Table 2).

Discussion Overbeeke et al. (6) performed presurgical and serial postoperative neuropsychological assessments on a female adolescent AVM patient. The patient's large AVM was located in the posterior left hemisphere, resulting in severe visual-processing deficits but without other neurological deficits. Neurocognitive measures similar to this study were administered, but only results with the right had were reported. At 3 weeks postsurgery, Trail Making and Tactual Performance were rated as improved from below average to average, with Finger Tapping decreased from below average to "much" below average, and Pegboard time maintained at the preoperative level of average. The patient's improvement in Trail Making was likely directly related to partial resolution of her visuoperceptual impairment. However, the improvement in Tactual Performance, a blindfolded measure, would indicate return of more parietal-based neural function. The actual degree of improvement, as represented by below-average to average-type ratings, may actually be slight, given that these standardized measures employ strict criteria for such ratings in the form of cutoff scores. The patient's regression in Finger Tapping, particularly in light of maintained levels on the more complex Pegboard task, are a curiosity, although subsequent improvements occurred at later testing intervals. Since Case 1 of the present study had an anterior locus of his AVM, it is difficult to make direct anatomical and statistical comparisons with Overbeeke's subject. Other case and group studies (1,2,8,9) have reported various improvements in neuropsychological abilities originating in the hemisphere harboring the AVM. Mahalick et al. (3) examined the neuropsychological sequelae of AVM in a preoperative population. Their study confirmed impairments indicating de-

EUROPSYCHOLOGICALIMPROVEMENTSAFTEREMBOLIZATION

..,....

Figure 2. Leftvertebral angiograms (Towne projection). Pre-embolization angiogram (left) shows a lefttemporal AVM suppliedby the temporal artery. Post-embolization angiogram (right) shows significantlydecreased flow and volume.

creased abilities both ipsilateral and contralateral to the AVMsite. The investigators were most impressed with impairments reflecting decreased efficiency in abilities based in the hemisphere contralateral to the AVM and suggested this indicated the presence of an

extensive cerebrovascular steal syndrome. Their subsequent study (S) suggests postoperative reversal of the syndrome. Case 1 of the current studies is believed to be the first report documenting reduction or frank reversal

Table 2. Scores on neuropsychological measures (Case 2) Pre-embolization Grip Tapping Pegboard Trail-Making A Trail-Making B Kaufman Hand Test Stroop Color-Word PASAT Speech-Sounds Verbal Learning Visual Learning Card Sorting Token Test Naming COWA Calculations WAIS Performance WAIS Verbal Verbal Stories Stories Delayed

Both normal Both normal Both normal 41 s = lower 5% 91 s = lower 8% 9 = lower 2% 30 = lower 8% 119 = lower 10% -7 = lower 15% 51 = low normal 9 = lower 7% Normal Normal Normal 30 = lower 1% Normal 102 = normal 130 = superior 13 = lower 8% 11 = lower 9%

Post-embolization

29 s = 70 s = 10 = 42 = 138 =

normal low normal lower 7% low normal normal o = high normal 56 = normal 17 = normal

32 = lower 3% 121 = upper 85% 31 = normal 21 = normal

J STROKECEREBROVASCDIS, VOL.4, NO.3, 1994

197

M. BOWENET AL.

of impairments in higher-level neurobehavioral functions represented in the hemisphere contralateral to the AVM, following partial embolization of the AVM. Two specific neurobehavioral phenomena seemed most apparent in this subject. First, preoperative assessment, although not extensive, indicated neurobehavioral impairments were pronounced only in the anterior cerebrum, bilaterally. Measures of verbal, visual, tactile, and spatial memory were within normal limits, as were constructions. However, the patient was severely impaired on the Wisconsin Card Sorting Test, a valid measure of frontal lobe function. In addition to the right hemiparesis, the motor functions of the left hand (ipsilateral to the AVM) ranged from normal to severely impaired. These findings regarding the ipsilateral motor abilities lend to the second principal neurobehavioral finding, i.e., in skills based in the hemisphere contralateral to the AVM, the patient showed a trend of making greater postembolization improvements relative to the complexity of the task. Overall, levels in right hemisphere motor functions improved to average or above average, even where pre embolization levels were severely impaired. There are several possible explanations for bilateral hemisphere improvement in neurobehavioral measures following partial embolization of a large AVM. Reduction in cerebrovascular steal with bilateral improvement in blood flow might underlie the observed benefit. This mechanism was suggested by Okabe et aI. (6) who found the decrease in interhemispheric steal and improvement in blood flow bilaterally was related to extent of AVM excision. Another possibility for the improved neurobehavioral function contralateral to a partial AVM obliteration is a reduction in transhemispheric diaschisis. This phenomenon has been demonstrated under other conditions of clinical cerebral ischemia, as well as in experimental ischemic models (7). A diaschisis effect may be mediated via corpus callosum transhemispheric connections (7). Case 2 demonstrated a range of significant cognitive improvements that were bilaterally represented in the cerebrum. The lesion-bearing left hemisphere demonstrated highly significant gains in auditory attention, processing speed, and immediate and delayed memory for narratives, whereas deficits persisted in speed of categorical word production and moderately complex ideational praxis. Both of these functions are harbored in language centers in the left hemisphere, with the latter known to have extensive connections with the caudate nucleus. The results of this study raise at least two other primary questions. The first concerns the natural course 198

J STROKE CEREBROVASCDIS, VOL. 4, NO.3, 1994

of recovery of motor function after cerebral injury such as an infarct or hemorrhage. This was addressed most recently and comprehensively by Duncan et aI. (9). Utilizing a variety of measures emphasizing the function of the hemiparetic side contralateral to lesion, their results concluded that regardless of initial severity of stroke, the most dramatic recovery occurs within the first 30 days. Our hemorrhagic patient (Case 1) showed major improvements in ipsilateral motor skills for the period beginning at day 60 and ending at day 90. Improvement of this degree, in the third month and on the nonhemiplegic/hemiparetic side, is unlikely to be due to spontaneous recovery of motor function after stroke. The second concern regards a more specific understanding of the presence and nature of ipsilateral motor dysfunction after cerebrovascular accident. Hom and Reitan (11) showed that with chronic, cerebrovascular lesions, the right-lesion group had significantly greater ipsilateral sensorimotor impairment than the left-lesion group. This finding was also present, but to a lesser degree, with both neoplastic and traumatic lesions, respectively. Since our patient has a left-sided lesion, the presence of ipsilateral dysfunction would be expected to be minimal or absent. Instead, it was both present and significant. This would suggest a marked abnormality of vascular perfusion of his right hemisphere prior to embolization. A more specific understanding is needed concerning the role cerebrovascular steal and diaschisis play in recovery of function after AVM obliteration. In the future, quantitative detailed cerebral blood flow measurements, in conjunction with neuropsychological testing, should be conducted before and after embolization. A final note is due with respect to the influence of serial neurobehavioral assessment on performance. Among such studies, two of the more recent examples (12,13) are instructive. Those studies utilized tests that duplicated or were similar to those of the present case studies, although their subjects were normals and cranial-irradiated small cell lung cancer patients, respectively. Thus, utilizing subjects without primary brain lesions and with a much shorter interval between assessments than the current cases, those studies showed more specific and less significant improvements than the current case reports.

References 1. Carter PL, Morgan M, Urrea D. Psychological improvement following arteriovenous malformation excision.

J Neurosurg 1975;42:452-6.

2. Conley F, Moses lA, Helle TL Deficits of higher cor-

EUROPSYCHOLOGICAL IMPROVEMENTS AfTER EMBOLIZATION

3. 4. 5.

6.

7.

tical functioning in two patients with posterior parietal arteriovenous malformations: use of the standardized Luria Nebraska Neuropsychological Battery for preand postoperative assessment. Neurosurgery 1980;7: 230-7. Mahalick DM, Ruff RM, U Hoi Sang. Neuropsychological sequelae of arteriovenous malformations. Neurosurgery 1991;29:351-7. Mahalick DM, Ruff RM, Heary RF, Sang H. Preoperative versus postoperative sequelae of arteriovenous malformations, Neurosurgery 1993;33:563-71. Okabe T, Meyer JS, Okayasu H, et at. Xenon-enhanced CT CBF measurements in cerebral AVMs before and after excision: contribution to pathogenesis and treatment. J Neurosurg 1983;59:21-37. Overbeeke J], Bosma NJ, Verdonck AF, Huffelen AC. Higher cortical disorders: an unusual presentation of arteriovenous malformation. Neurosurgery 1987;21: 839-42. Andrews RJ. Transhemispheric diaschisis; a review and comment. Stroke 1991;22:943-8.

8. Dunn M, Rosen WG, Malmut B, Zapulla A. Memory deficits in left hemisphere AVMs. Submitted for presentation at the International Neuropsychological Society 13th Annual Meeting, San Diego, CA, February 1985. 9. Waltimo 0, Putkoken AR Intellectual performance of patients with intercranial arteriovenous malformations. Brain 1974;97:511-20. 10. Duncan PW, Goldstein LB, Matchar D, Divine GW, Feussner J. Measurement of motor recovery after stroke. Stroke 1992;23:1084-9. 11. Hom J, Reitan R Effect of lateralized cerebral damage upon contralateral and ipsilateral sensorimotor performances. J Clin Neuropsychol1982;4:249-68. 12. McCaffrey RJ,Ortega A, Haase RF. Effects of repeated neuropsychological assessments. Arch Clin Neuropsychol1993;8:519-24. 13. Youngjohn JR, Larrabee GJ, Crook TH. Test-retest reliability of everyday and traditional memory tests. Clin Neuropsychol1992;6:276-86.

J STROKE CEREBROVASC DIS, VOL. 4, NO.3,

1994

199