C L IN IC A L STU D IES Mass Effect Caused by Clinically Unruptured Cerebral Arteriovenous Malformations Departm ent of Neurosurgery, Kitasato U niversity School of M ed icin e, Sagam ihara, Kanagaw a, Japan Downloaded from https://academic.oup.com/neurosurgery/article-abstract/41/5/1060/2846173 by INSEAD user on 09 December 2018 Yoshio M iyasaka, M .D ., Akira Kurata, M .D ., Ryusui Tanaka, M .D ., Shigeki Nagai, M .D ., Masaru Yamada, M .D ., Katsumi Irikura, M .D ., Kiyotaka Fujii, M .D . OBJECTIVE: It is generally considered that mass effect caused by arteriovenous malformations (AVMs) is evidence of ruptures. In the present study, the incidence of mass effect in clinically unruptured AVMs was evaluated, and the underlying causative factors and pathophysiological mechanisms were studied. METHODS: Twenty-seven patients with clinically unruptured supratentorial pial AVMs were examined. The majority were suffering from epilepsy, and frontal lobe involvement was revealed in approximately half of the patients Angiographic studies, computed tomographic scans, and magnetic resonance images were obtained for all patients. Twenty-one patients underwent removal of AVMs. In 10 of the surgically treated patients, intraoperative vascular pressure measurements were obtained before removal of the AVMs. RESULTS: Mass effect was detected in 12 (44% ) of the 27 patients. Cortical sulci obliteration (eight patients) and lateral ventricle displacement (seven patients) were frequently noted. The volume of AVMs was significantly larger in patients with mass effect than in those without mass effect (P < 0.001). Large dilated venous sacs or ectatic veins were observed to be associated with mass effect (P < 0.001). In only one patient was gross displacement related to a surrounding massive brain edema. Draining vein pressure in patients with mass effect was significantly elevated as compared to the average value in patients without mass effect (22 ± 5 versus 12 ± 3 mm Hg) (P < 0.01). CO NCLUSIO N: The present study suggests that mass effect is not infrequent in clinically unruptured AVMs Furthermore, multiple causative factors were detected, including the large size of AVMs, marked d r a i n i n g vein dilatation, and brain edema around the AVMs. Findings also indicated that a pathophysiologically high pressure in the venous drainage system may contribute to mass effect. ( N e u r o s u r g e r y 4 1 : 1 0 6 0 - 1 0 6 4 , 1 9 9 7 ) Key words: Arteriovenous malformations, Draining vein pressure, Mass effect, Unruptured ass effect with unruptured arteriovenous malforma­ tions (AVMs) has been reported in the literature only since the beginning of the computed tomo­ graphic (CT) scan era (1-4, 6, 7, 11). Kumar et al. (4) first discussed various causes of mass effect in patients dem on­ strated to have unruptured AVMs by CT scanning. Using magnetic resonance (MR) images, Smith et al. (11) assessed the effect of AVMs on adjacent brain tissue. Most of the articles that have been published, however, focused attention on the morphological characteristics of AVMs, and discussion of the underlying pathophysiology has been limited. In the present study, we reevaluated the incidence of mass effect and the causative factors producing this finding in clinically M 1060 unruptured AVMs, concentrating attention on pathophysiological mechanisms. PATIENTS A N D METHODS Since the advent of MR imaging, 112 consecutive patient' supratentorial pial AVMs were treated at K itasato Uni­ versity Hospital between January 1986 and D ecem ber 1 ^ Of these patients, the 27 patients with clinically unruptum AVMs were selected for the study. The patients' ages ranged from 10 to 61 years (a v e ra g e ^ 1 35 yr). Angiograms, CT scans, and MR images were avails^for all patients, and resection of AVM s was performed in w it h Neurosurgery, Vol. 41, No. 5, Novem ber 1997 A V M s and M ass Effect TABLE 2. Factors Contributing to Mass Effect in Arteriovenous Malformations'1 Factors Size of AVMs (maximum diameter) Small (< 3 cm) Medium (3 -6 cm) Large (> 6 cm) Volume of AVMs (mean ± SD; cm3) Venous sac/ectatic veins White-matter edema With Mass Without Mass Effect Effect (n = 12) (n = 15) 0 fa 3 9C 80 ± 4 7d 10 d 1 6 fe 8 1c 22 ± 2 2 d 1d 0 a Volume of AVMs according to Pasqual in et al. (10). SD , standard deviation; AVM s, arteriovenous malformations. b ( P < 0.05). C (P < 0.01). d (P < 0.001). RESULTS Mass effect caused by AVMs were detected in 12 (44%) of the 27 patients. The affected brain structures are summarized in Table 1 . Cortical sulci obliteration and lateral ventricle dis­ placement were frequently observed. Table 2 presents the factors contributing to mass effect. Findings confirmed that the volume of AVMs was one of the im portant determinants for the production of compression of the surrounding brain tissue. Furthermore, an association with large dilated venous sacs/ectatic veins (Fig. 1) was also observed. Gross displace­ ment was related to surrounding massive white-matter edema in only one patient, who had been included in one of our previous reports (9). No patients experienced objective neurological symptoms suggestive of the presence of increased intracranial pressure TABLE 1. Structures Affected by AVM s Producing Mass Effect Mass Effect No. of Patients Cortical sulci obliterated Lateral ventricles displaced Lateral ventricles only compressed Third ventricle displaced Pineal gland displaced Thoroid plexus displaced 8 7 3 3 2 1 FIG U R E 1. T1 -weighted MR images (coronal views) depict­ ing large AVMs with a dilated venous sac and marked dilata­ tion of the draining vein compressing the ipsilateral lateral ventricle and the cingulate gyrus. such as papilledema. The interval from the onset of seizure to admission without anticonvulsants (range, 1 d -20 yr) and the absolute numbers of seizures during these intervals (range, 1 to approximately 50 times) varied from one patient to another. We calculated the numbers of seizures in terms of 1 month, as follows: 17 ± 47 (mean ± standard deviation, n = 9) times per month in patients with mass effect and 6 ± 1 8 ( n = l l ) i n those without. There was a tendency for the incidence of seizures to be higher in patients with mass effect, but this finding was not statistically significant. The results of intraoperative vascular pressure measure­ ments are summarized in Table 3. Pressure measurements were obtained in four patients with mass effect and in six without mass effect. No significant differences in central ve­ nous pressure were determined between patients with mass effect and those without. Draining vein pressure was, how­ ever, significantly higher in patients with mass effect (22 ± 5 versus 12 ± 3 mm Hg) (P < 0.01), indicating that this may contribute pathophysiologically to the development of com­ pressive effect caused by AVMs. N e u r o s u rg e r y , Vol. 41, No. 5, Novem ber 1997 1061 Downloaded from https://academic.oup.com/neurosurgery/article-abstract/41/5/1060/2846173 by INSEAD user on 09 December 2018 patients. Twenty of the patients were epileptic, one exhibited the cerebral steal phenomenon, and five experienced no symptoms. Clinically silent hem orrhage (old small hemor­ rhage caused by AVMs) was detected at the time of surgery in seven patients. These patients were nevertheless still included in the present study population, because they experienced no clinical evidence of recent intra-axial hematoma formation. Fifteen AVMs were located in the frontal lobe, four in the temporal lobe, two in the frontoparietal lobe, two in the occipital lobe, two in the thalamus, one in the parieto-occipital lobe, and one in the frontotemporal lobe. AVMs were defined as "large" when their maximum diam eter was greater than 6 cm, "medium-sized" when it was between 3 and 6 cm, and "small" when it was less than 3 cm. We noted 10 patients with large AVMs, 11 patients with medium-sized lesions, and 6 patients with small lesions. The size of the AVMs was also calculated in terms of their volume, obtained as the product of their three diameters and 0.52 (10). Intraoperative vascular pressure measurements before re­ section of the AVMs were obtained in 10 patients (4 patients with mass effect and 6 w ithout mass effect) to clarify the pathophysiological mechanisms underlying the mass effect, using methods previously described in detail (8). The statistical methods used were y2 analysis, with Yates' correction for expected values less than 10, and the unpaired t test. The results were considered significant at P < 0.05. M iyasaka et al. 1062 TABLE 3. Intravascular Pressure in the Venous Draining System Mass Effect3 Draining Vein Pressure Central venous pressure 1 2 3 4 + + + + 29 20 19 18 3 4 3 2 22 ± 5 mm Hgfa 3 ± 1 mm Hg 15 14 13 12 10 6 9 3 6 2 4 2 12 ± 3 mm Hgfa 4 ± 3 mm Hg 5 6 7 8 9 10 — - a +, present; —, not present. b Statistically significant differences: a) (P < 0.01); mean ± standard deviation. ILLUSTRATIVE CASES Patient 1 A 23-year-old w o m a n presented w ith to n ico clo nic se izu re 1 m onth before a d m issio n . N e u ro lo g ic a l e xa m in a tio n re ve a le d no ab n o rm al fin d in g s. L e ft caro tid a n g io g ra p h y dem o n strated m e d iu m -size d A V M s (v o lu m e , 29 cm 3) in the rig h t fro n ta l lobe (Fig. 2). T h e lesion w a s su p p lie d b y the a n te rio r cereb ral a rte ry an d d ra in e d into the d ilated co rtica l v e in s w ith a v a r ix (Fig. 2). T h e re ve rse d flo w th ro u g h the co rtica l v e in w a s also noted (Fig. 2). M R im ag es re ve a le d com ­ p ressio n of the ip s ila te ra l c o rp u s c a llo su m and the cin g u la te g y ru s b y the fro n tal A V M s (Fig. 3). F o u r m o n th s afte r the onset, total re m o v a l o f the A V M s w a s p e rfo rm e d an d w a s u n e v e n tfu l. In tra o p e ra tiv e v a s c u la r p ressu re m easu re m en t d e m o n strated a m a rk e d ly elevated d ra in in g v e in p re ssu re (29 m m H g ) and m o d e ra te ly elevated sag ittal FIG U RE 2. Left carotid angiograms (oblique view), arterial phase (left pa nel) and venous phase (right panel), showing medium-sized AVMs in the frontal lobe fed by the anterior cerebral artery and draining into dilated cortical veins with a varix. Note the reversed flow through the cortical vein (arrow). Downloaded from https://academic.oup.com/neurosurgery/article-abstract/41/5/1060/2846173 by INSEAD user on 09 December 2018 Patient No. FIG U R E 3. T1-weighted MR images (sagittal view) showing compression of the corpus callosum and the cingulate gyrus by AVMs. s in u s p re s su re (18 m m H g ) co m p are d to c e n tra l ven o us pressure (3 m m H g ). Patient 2 A 44-year-old m an presented w it h fre q u e n t fo cal m otor seizures in the rig h t u p p e r lim b an d face d u rin g a 10-year p e rio d . Neurological e x a m in a tio n at the tim e of a d m is sio n re ve a le d m ild weakness in the rig h t u p p e r lim b . A n g io g ra m s o f b oth c a ro tid a rte rie s demonstrated larg e fro n to p arie tal A V M s (v o lu m e , 112 c m 3) fed b y the left anterior and the left m id d le cereb ral a rte rie s. T h e A V M s d rain e d into mark­ e d ly elong ated an d d ila te d c o rtica l v e in s . M R im a g es demonstrated c o m p ressio n of the ip s ila te ra l late ra l v e n tric le an d the cingulate gyrus caused b y larg e A V M s w ith a d ila te d v e n o u s sac (Fig. 1). Unfortu­ n a te ly , no in tra v a s c u la r p re ssu re m e asu re m e n t w a s obtained. Patient 3 A 44-year-old m an w a s a d m itte d at o u r in s titu tio n because of a te m p o ral lobe se iz u re exp e rie n ce d 2 m o n th s p re v io u s ly . Neurological e xa m in a tio n at the tim e of a d m is sio n re ve a le d no abnorm al findings R ig h t caro tid an g io g ram s re ve a le d larg e A V M s (vo lu m e, 62 cm3) situated in the te m p o ral lobe, s u p p lie d b y b ran ch e s of the middle cereb ral a rte ry (Fig. 4). T h e A V M s d ra in e d in to m a rk e d ly elongated and d ilate d su p e rio r cereb ral v e in s , the s u p e rfic ia l m id dle cerebral v e in , and the tw o an a sto m o tic v e in s o f La b b e (Fig. 4). Preoperative CT scan s d e m o n strated d isp la ce m e n t of m id lin e stru ctu re s (Fig. 5). De­ sp ite the te m p o ral lobe lo catio n o f the A V M s , o b literatio n of cortical FIG U R E 4. Right carotid angiograms (lateral view), arterial phase (left panel) and venous phase (right panel), showing large AVMs in the temporal lobe supplied by branches from the middle cerebral artery and draining into the markedly elongated and dilated superior cerebral veins, the superficial middle cerebral vein, and the two anastomotic veins of Labbe. Neurosurgery, Vol. 41, No. 5, Novem ber 1997 A V M s and M ass Effect sulci was depicted not o n ly th ere, b u t also in the fro n to p a rie ta l lobe (Fig. 5). Although a p ressu re m e a su re m e n t stu d y w a s not u n d e rta k e n , it was speculated that a m a rk e d ly h ig h e r ve n o u s p re ssu re in a d ilated functional derangement of the adjacent brain parenchyma (11). The findings in our study demonstrated that draining vein pressure was significantly elevated in patients with mass effect as compared to the average value in those without mass effect. It is possible that AVMs with venous hyperten­ sion may produce venous ischemia with neurological signs or symptoms (5). Thus, AVMs that produce mass effect probably represent a subset with a greater malignant potential, and, therefore, affected patients should be urgently considered for treatment with a combination of preoperative embolization, microsurgery, and radiosurgery before neurological deficits appear. Furthermore, for patients with mass effect, we need to avoid measures at the time of surgery that would further increase cerebral venous pressure. R e c e iv e d , A u g u s t 27, 1996. A c c e p te d , Ju n e 16, 1997. draining vein m ay h av e co n trib u te d to the d e ve lo p m e n t o f m ass effect. R e p r in t re q u e s ts : Y o s h io M iy a s a k a , M .D ., D e p a rtm e n t o f N e u r o ­ s u rg e r y , K ita s a to U n iv e r s it y S ch o o l o f M e d ic in e , 1-15-1 K ita s a to , S a g a m ih a ra , K a n a g a w a 228, Ja p a n . DISCUSSION REFERENCES In the angiographic literature from the pre-CT scan era, AVMs were considered to display mass effect only in hemor­ rhagic cases. However, since the advent of computed tomog­ raphy and MR imaging, a number of authors have reported mass effect of unruptured AVM s (1-4, 6, 7, 11). Kumar et al. (4) observed such findings obtained by CT scans for 55% of 60 patients with AVMs. Smith et al. (11) reported compressive effect based on MR images in 40% of 15 lesions. The present study, using both MR images and CT scans showing displace­ ment of surrounding brain structures in 44% of 27 patients, is thus in agreement with earlier investigations in indicating that hemorrhage is not a requirement for causing mass effect 1 of AVMs. Our analysis of the factors producing and/or contributing to mass effect of unruptured AVMs determined that the vol­ ume of AVMs, dilated draining veins, and varices are consid­ ered to play an important causative role. Massive edema surrounding the AVMs, although not frequent, can also cause mass effect in unruptured AVMs (9). Kumar et al. (4) similarly concluded that the magnitude of the mass effect largely par­ allels the size of the AVMs, and enlarged tortuous veins contribute greatly. 1 he reason why some AVMs cause mass effect and others produce atrophic changes in the surrounding brain tissue rather than compression requires explanation. Fortunately, in * the present study, we were able to obtain intraoperative vas­ cular pressure measurements in 10 patients. As a result, we observed the draining vein pressure in patients with mass effect (22 ± 5 mm Hg) to be significantly higher than in those r wilhout (12 ± 3) (P < 0.01). We, therefore, speculate that this p. lncrease in draining vein pressure, along with the volume of AVMs and the marked draining vein dilatation, may play the I, determining role. The clinical significance of mass effect caused by clinically Ur|ruptured AVMs is not known (11). However, it seems rcasonable that they could play a contributory role in causing 1. B ra d le y W G Jr: P a th o p h y sio lo g ic co rrelates o f sig n a l alte ra tio n s, in B ra n tz -Z a w a d z k i M , N o rm a n D (ed s): Magnetic Resonance Im­ aging of the Central Nervous System. N e w Y o r k , R a v e n , 1987, p p 23-42. 2. H a y w a rd R D : In tra c ra n ia l a rte rio ve n o u s m a lfo rm a tio n s: O b se r­ v a tio n s after exp erien ce w ith co m p u te rize d to m o g rap h y. J N e u ­ r o l N e u r o s u rg P s y c h ia t r y 39:1027-1033, 1976. 3. K e n d a ll B E , C la v e ria L E : T h e u se of co m p uted a x ia l to m o g rap h y ( C A T ) for the d ia g n o sis and m anag em ent of in tra c ra n ia l ang io ­ m as. N e u r o r a d io lo g y 12:141-160, 1976. 4. K u m a r A J , V in u e la F , F o x A J , R o sen b au m A E : U n ru p tu re d in tra ­ c ra n ia l arte rio ve n o u s m a lfo rm a tio n s do cause m ass effect. A J N R A m J N e u r o r a d io l 6:29-32, 1985. 5. K u ra ta A , M iy a sa k a Y , Y o s h id a T , K u n ii M , Y a d a K , K a n S: V e n o u s isch em ia caused b y d u ra l a rte rio ve n o u s m a lfo rm a tio n : C ase rep o rt. J N e u r o s u rg 80:552-555, 1994. 6. Le b la n c R , Le ve sq u e M , C o m a ir Y , E th ie r R : M ag n etic resonance im a g in g o f cereb ral arte rio ve n o u s m a lfo rm a tio n s. N e u r o s u rg e ry 21:15-20, 1987. 7. L e e B C P , H e rzb e rg L , Z im m e rm a n R D , D e c k M D F : M R im a g in g of cerebral v a s c u la r m a lfo rm a tio n s. A J N R A m J N e u r o r a d io l 6:863-870, 1985. 8. M iy a sa k a Y , K u ra ta A , T o k iw a K , T a n a k a R , Y a d a K , O h w a d a T : D ra in in g v e in p re ssu re incre ase s and h em o rrh ag e in p atients w ith a rte rio ve n o u s m alfo rm atio n . S tro k e 25:504-507, 1994. 9. M iy a sa k a Y , Y a d a K , K u ra ta A , T o k iw a K , T a n a k a R , O h w a d a T : A n u n ru p tu re d a rte rio ve n o u s m a lfo rm a tio n w ith ed em a. A J N R A m J N e u r o r a d io l 15:385-388, 1994. 10. P a sq u a lin A , B aro n e G , C io ffi F , R osta L , Scie n za R , D a P ia n R: T h e rele van ce of anato m ic an d h e m o d yn a m ic factors to a c la ssi­ fica tio n of cereb ral a rte rio ve n o u s m a lfo rm a tio n s. N e u r o s u rg e ry 28:370-379, 1991. 11. S m ith H J, Stro ther C M , K ik u c h i Y , D u ff T , R a m ire z L , M erless A , T o u ta n t S: M R im a g in g in the m anag em ent of su p ra te n to ria l in tra c ra n ia l A V M s . A J N R A m J N e u r o r a d io l 9:225-235, 1988. COMMENTS The authors describe 27 patients with clinically unruptured brain arteriovenous malformations (AVMs); of these, 12 pa­ N e u r o s u rg e r y , Vol. 41, No. 5, Novem ber 1997 Downloaded from https://academic.oup.com/neurosurgery/article-abstract/41/5/1060/2846173 by INSEAD user on 09 December 2018 1 FIGURE 5. Contrast-enhanced C T scans (axial view) depict­ ing AVMs in the temporal lobe and the displacement of i midline structures (left and center panels). Note the obliteration of the cortical sulci in the frontoparietal lobe (right panel, arrows). 1063 1064 M iyasaka et al. Christopher S. O gilvy Boston, M assachusetts The main focus of this report is to determine what ana­ tomic, physiological, or hemodynamic variables may contrib­ ute to the presence of radiological mass effect in unruptured AVMs. The authors evaluated 27 patients harboring AVMs, most of whom presented with seizures. None of the patients presented with clinical hemorrhage. Twenty-one of the pa­ tients underwent surgical resection of the AVMs, and in 10 of these, intraoperative vascular pressure measurements were assessed before the resection of the AVMs. The authors de­ tected mass effect in 12 (44%) of their 27 patients, selected by cortical sulci obliteration and lateral ventricular displacement. Radiological mass effect associated with the AVMs was sig­ nificantly correlated with the larger volume of AVMs and the presence of large dilated venous sacs or ectatic veins. Drain­ ing vein pressures in patients with radiographically apparent mass effect were significantly elevated compared to those without mass effect. The association between radiological mass effect of AVMs and large-sized or enlarged tortuous veins has been previ­ ously reported, but its clinical relevance remains uncertain. Although there was a trend toward increasing frequency of seizures in patients with AVMs having mass effect compared with those without mass effect, this difference was not statis­ tically significant, and it is not clear that presentation with seizures is related to radiological mass effect. It is interesting that none of the 12 patients noted as having mass effect from the AVMs were symptomatic with signs of raised intracranial pressure, such as papilledema, headache, or focal neurologi­ cal signs. Of particular note is the authors' novel finding that of 10 patients with intraoperative vascular pressure measure­ ments, draining vein pressures were significantly elevated in those 4 patients with mass effect compared to those 6 patients without mass effect. However, the num ber of patients in­ volved is small, and other variables, including venous steno­ ses, the number of draining veins, and the anatomy of arterial feeders, may contribute to determ ining these pressures. Fur­ thermore, although there is an association between elevated venous pressures, the volume of AVMs, and the presence of dilated or ectatic veins and mass effect, it is not necessarily the case that any of these factors cause or contribute to the radio­ logical mass effect phenomena observed. Downloaded from https://academic.oup.com/neurosurgery/article-abstract/41/5/1060/2846173 by INSEAD user on 09 December 2018 tients were noted to sustain mass effect. We have observed this phenomena of mass effect caused by AVMs in several patients during the last few years. In general, I agree with the authors that this is caused mostly by increased venous pres­ sure from venous outflow obstruction and relative venous hypertension. The authors were able to measure intravascular pressure in the venous drainage systems of four patients with mass effect from their AVMs and in six patients in whom there was no mass effect. There is a dramatic difference in the venous pressure measured. The authors present the finding, which is becoming increasingly common, that the role of the venous outflow in determining the pathophysiology of AVMs is paramount. We have observed the treatment of patients who present with AVMs with edema or mass effect to be potentially dan­ gerous. Usually, we have elected to perform intranidal transarterial embolization to reduce arterial inflow through the AVMs, which typically is in a state of severe imbalance, with inflow overwhelming venous outflow. In a few of the patients we have treated, the venous outflow occlusion has been acute. In addition to transarterial embolization, we have used heparin anticoagulation to prevent further venous out­ flow occlusions. Gary K. Steinberg Stanford, California Reid C. Thompson L o s A n g eles, California This article documents in a small, but well-studied, group of patients what intuitively seems obvious to neurosurgeons, i.e., that cerebral AVM s can be associated with mass effect and that the factors that correlate positively with mass effect are the size and volume of the AVMs, the presence of large venous varices, and the presence of surrounding brain edema Interestingly, but not surprisingly, in a small group of patients in whom the authors measured the pressure in the draining veins intraoperatively, there was a positive correlation be­ tween the presence of mass effect in preoperative studies and the finding of relatively high pressures in the venous drainage system. W hether this was an independent factor, or whether it was related to the finding that these were the larger AVMs remains speculative. Roberto C. Heros Miami, Florida This clinical study of a previously recognized but poorlv understood phenomena, i.e., mass effect associated with un­ ruptured AVMs, draws further attention to the previously neglected venous aspect of these malformations whose ana­ tomic and pathophysiological features remain incomplete^ understood. The authors document a statistical correlation among mass effect, large lesion size, and the presence ot venous varices and/or ectasias and have also reported, propensity for elevated venous pressures in such m a l f o r m a tions. The clinical relevance of such findings is unknown although the authors postulate a possible causative correla tion between mass effect and seizures; the patient p o p u la tio n as defined precludes any current statem ents regarding hem­ orrhagic tendencies, etc. Neurosurgery, Vol. 41, No. 5, Novem ber 1997 Duke S. Samson Dallas, Texas