POSTOPERATIVE HYPERPERFUSION IN A PATIENT WITH A DURAL ARTERIOVENOUS FISTULA WITH RETROGRADE LEPTOMENINGEAL VENOUS DRAINAGE: CASE REPORT Yutaka Kai, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Jun-ichiro Hamada, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Motohiro Morioka, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Shigetoshi Yano, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Takamasa Mizuno, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Yukitaka Ushio, M.D. Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan Reprint requests: Yutaka Kai, M.D., Department of Neurosurgery, Kumamoto University Medical School, 1-1-1 Honjo, Kumamoto 860-8556, Japan. Email: ykai@kaiju.medic.kumamoto-u.ac.jp Received, October 30, 2002. Accepted, February 10, 2003. OBJECTIVE AND IMPORTANCE: Hyperperfusion has been reported after carotid endarterectomy or stenting for stenosis of the internal carotid artery. Because few reports have examined postoperative hyperperfusion after treatment for dural arteriovenous fistulae (DAVFs), we present a case describing a patient who manifested this clinical entity. CLINICAL PRESENTATION: The patient was a 63-year-old man with a DAVF in the left transverse sigmoid sinus with retrograde leptomeningeal venous drainage. He experienced slowly progressive disorientation lasting for several months. Preoperative single-photon emission computed tomography with 123I-labeled N-isopropyl-piodoamphetamine revealed an area of hyperintensity on T2-weighted magnetic resonance imaging (MRI) scans that coincided with the hypoperfusion area; it was not increased after acetazolamide challenge. Complete DAVF obliteration was achieved by embolization, then sinus isolation. After treatment, he experienced frequent generalized convulsions that were terminated by 2-day barbiturate therapy. INTERVENTION: On T2-weighted MRI scans obtained 3 days after surgery, the hyperintense area not only persisted but had expanded to the left parietal lobe. Moreover, a subcortical hyperintense lesion was recognized on T1-weighted MRI scans; this was considered to reflect cortical laminar necrosis. Single-photon emission computed tomography revealed hyperperfusion in the left parietal lobe; it changed to hypoperfusion a month after treatment. CONCLUSION: In patients with DAVFs with preoperative findings of marked low perfusion and a poor perfusion reserve, postoperative study may reveal hyperperfusion on single-photon emission computed tomography or cortical laminar necrosis on MRI. This may be evidence of severe perfusion disturbance as a result of venous infarction. In these patients, careful blood pressure control and early treatment of seizures are important after DAVF treatment. KEY WORDS: Dural arteriovenous fistulae, Magnetic resonance imaging, Retrograde leptomeningeal venous drainage, Single-photon emission computed tomography, Venous infarction Neurosurgery 53:228-233, 2003 S DOI: 10.1227/01.NEU.0000069536.64823.F9 urgical or endovascular treatment for dural arteriovenous fistulae (DAVFs) can resolve venous ischemia (11, 15). In such cases, the reduced regional cerebral blood flow (rCBF) also normalizes after treatment (11). To our knowledge, there is only one previous report of postoperative hyperperfusion (14). Hyperperfusion has been reported to occur after carotid endarterectomy or stenting for stenosis of the internal carotid artery (1, 3, 7, 12, 17–19, 21). Some studies have suggested that hyperperfusion after blood flow restoration is attributable to vasoparalysis in regions with impaired autoregulation resulting from long-standing ischemia (7, 12, 21). 228 | VOLUME 53 | NUMBER 1 | JULY 2003 www.neurosurgery-online.com We now describe a DAVF patient with postoperative hyperperfusion with retrograde leptomeningeal venous drainage (RLVD). We also studied rCBF disturbances in other DAVF patients by looking for changes on pre- and postoperative magnetic resonance imaging (MRI) and single-photon emission computed tomography (SPECT). CASE REPORT A 63-year-old man had a 7-month history of mental disturbance. MRI revealed areas of hyperintensity on T2-weighted MRI scans at the left temporo-occipital lobe (Fig. 1). On www.neurosurgery-online.com POSTOPERATIVE HYPERPERFUSION OF DURAL ARTERIOVENOUS FISTULA FIGURE 3. SPECT scan with 123I-IMP revealing a hypoperfusion area in the left temporo-occipital lesion (A) and a normal perfusion area in the left parietal lesion (B). FIGURE 1. T2-weighted MRI scans revealing a hyperintense lesion in the left temporo-occipital lobe (A) and no hyperintense lesion in the left parietal lobe (B). continuous-mode angiograms, DAVFs with RLVD were recognized in the left transverse sigmoid sinus. The feeders were the left occipital arteries. A venous-phase angiogram revealed RLVD in the left temporal area (Fig. 2). There was no accessory route in the drainage system. The accessory route was angiographic evidence of flow into another sinus besides the main draining sinus (9). During dynamic SPECT, 111 MBq of 123Ilabeled N-isopropyl-p-iodoamphetamine (123I-IMP) (Nihon Mediphysics, Nishinomiya, Japan) was intravenously injected at the start of imaging; 1 g of acetazolamide (Diamox; Lederle, Ltd., Tokyo, Japan) was slowly injected intravenously during a 1-minute period 9 minutes after the initial 123I-IMP injection. An additional 111 MBq of 123I-IMP was injected 27 minutes after the start of imaging. Two perfusion images, resting and vasodilated, were obtained with the subtraction technique. This facilitated semiquantitative measurements of the rCBF. The hypoperfusion area on SPECT coincided with the area of hyperintensity in the left temporo-occipital area on T2weighted MRI scans; it was not increased after acetazolamide challenge. The left parietal area manifested no hyperintensity on T2-weighted MRI scans or hypoperfusion on SPECT analysis (Figs. 3 and 4). FIGURE 2. A, left external carotid angiogram (arterial phase), lateral projection, revealing a DAVF adjacent to the left transverse sinus. B, the venous phase angiogram reveals RLVD in the left temporal area. No accessory route is recognized. NEUROSURGERY FIGURE 4. SPECT scan revealing that the hypoperfusion area (A) is not increased after acetazolamide challenge (B). The patient underwent preoperative transarterial embolization of the left occipital arteries with porous cellulose beads (200 and 400 ␮m) (5, 6), then sinus isolation. Two days later, the left transverse sigmoid sinus was isolated by dural incision. By use of a properly fashioned bone flap, two dural incisions, one superior, the other inferior, were made parallel to the long axis of the left transverse sigmoid sinus. The tentorium was cauterized and divided with extension toward the petrous ridge. After incision of the tentorium, the left transverse sigmoid sinus was isolated from the dura overlying the cerebellum, occipital lobe, and tentorium. The left transverse sigmoid sinus was not resected. Interruption of the draining veins was not performed to avoid disturbing the normal venous return. Because the patient experienced frequent generalized convulsions when he awoke from anesthesia, he received barbiturate therapy with induced hypotension for 2 days. Although postoperative angiography revealed no FIGURE 5. Left external carotid residual DAVF (Fig. 5), on angiogram, lateral projection, revealthe T2-weighted MRI scan ing no residual DAVF. VOLUME 53 | NUMBER 1 | JULY 2003 | 229 KAI ET AL. obtained 3 days after surgery, the area of hyperintensity not only persisted but had expanded to the left parietal lobe (Fig. 6). Moreover, a subcortical hyperintense lesion was recognized on the T1-weighted MRI scan; this lesion was visualized as a very hyperintense area on a diffusion-weighted MRI scan (Figs. 7 and 8). SPECT revealed an area of hyperperfusion in the left parietal lobe that was not increased after acetazolamide challenge (Figs. 9 and 10). The patient’s convulsions ceased 3 days after the operation, and disturbances of consciousness gradually improved. After 10 days, his consciousness was clear; however, the hyperperfusion area on SPECT implied disturbance of vasoreactivity. One month after treatment, SPECT analysis demonstrated that the hyperperfusion area had changed to hypoperfusion and that there was disturbance of vasoreactivity (Figs. 11 and 12). At 6 months after surgery, the patient continued to be slightly disoriented. DISCUSSION Patients with DAVFs and cerebral ischemia and/or cerebral hemorrhage manifest rCBF abnormalities attributable to disturbance of venous drainage (2, 8, 16). The angiographic characteristics of DAVF with disturbed rCBF include RLVD in the draining system (4). Willinsky et al. (23) reported that in DAVFs with cortical venous drainage, T2-weighted MRI scans revealed areas of hyperintensity in the venous territory; they proposed this as direct evidence of venous hypertension and congestive encephalopathy. We previously reported that there is a correlation between the pattern of venous drainage and MRI findings in DAVF patients with RLVD (9): patients with accessory routes manifested no hyperintense areas on preoperative T2-weighted MRI scans, and SPECT demonstrated no areas of hypoperfusion. We suggest that the areas of hyperintensity on T2-weighted MRI scans that coincide with areas of hypoperfusion on SPECT reflect the presence of rCBF disturbance attributable to venous congestion. FIGURE 6. T2-weighted MRI scans obtained after treatment, revealing that the hyperintense area persists (A) and expands to the left parietal lobe (B). 230 | VOLUME 53 | NUMBER 1 | JULY 2003 We retrospectively studied 22 DAVF patients with RLVD and examined their pre- and posttreatment MRI and SPECT findings. Of the 22 patients, 11 had no hyperintense areas on preoperative T2-weighted MRI scans and normal perfusion areas on preoperative SPECT. In the other 11 patients, preoperative SPECT analysis revealed hypoperfusion areas that FIGURE 7. T1-weighted MRI scan coincided with hyperperfu- showing appearance of a subcortical sion areas on T2-weighted hyperintense lesion after treatment. MRI scans. After treatment, the hyperintense area disappeared in 7 of these 11 patients. Their preoperative SPECT study had demonstrated preservation of vasoreactivity after acetazolamide challenge. In the other four patients, the hyperintense areas persisted after treatment; their preoperative SPECT study had revealed marked disturbance of vasoreactivity af- FIGURE 8. Diffusion-weighted MRI ter acetazolamide chal- scan showing a hyperintense lesion. lenge. Only one of these four patients manifested postoperative hyperperfusion after treatment of their DAVFs with RLVD. Hyperperfusion has been reported after carotid endarterectomy or stenting for stenosis of the internal carotid artery (1, 3, 7, 12, 17–19, 21). Some studies suggested that hyperperfusion after blood flow restoration may result from vasoparalysis in regions with impaired autoregulation as a result of longstanding ischemia (7, 12, 21). Patients with hyperperfusion after carotid endarterectomy or stenting tended to experience fatal complications consisting of cerebral edema and hemorrhage. These patients manifested areas of hyperperfusion on SPECT and no vasoreactivity after acetazolamide challenge (24). Kuroda et al. (13) reported that acetazolamide challenge is valuable for assessing the cerebral perfusion reserve and for predicting the long-term prognosis of patients with internal carotid artery occlusion. Postoperative hyperperfusion in patients with DAVFs is rare; to our knowledge, only one such case was previously reported by Kuroda et al. (14). In their patient, the hyperperfusion area returned to normal perfusion after treatment, and they suggested that the presence of the hyperperfusion area reflected venous congestion. However, their rCBF study did not include evaluation of vasoreactivity by acetazolamide challenge. In our patient, preoperative findings were indica- www.neurosurgery-online.com POSTOPERATIVE HYPERPERFUSION OF DURAL ARTERIOVENOUS FISTULA tive of marked hypoperfusion with poor vasoreactivity, and a transient hyperperfusion area on SPECT was recognized after treatment. The hyperperfusion did not return to normal perfusion, and the symptoms persisted after treatment. Therefore, hyperperfusion may result from vasoparaly- FIGURE 9. SPECT scan obtained sis in regions with impaired 3 days after treatment, revealing autoregulation as a result of hyperperfusion in the left parietal long-standing rCBF distur- lobe. bance by venous congestion. If venous congestion is severe, microvessels may dilate maximally to restore flow, and chronic dilation may induce organic disorders and dysautoregulation. If rCBF with normal perfusion pressure is suddenly restored to this area by vascular reconstruction, the inability of the long-term dilated microvessels to constrict results in an FIGURE 10. SPECT scan revealuncontrollable increase in ce- ing that the hyperperfusion area is rebral blood perfusion. We not increased after acetazolamide posit that in these patients, challenge. hyperperfusion and poor vasoreactivity after acetazolamide challenge is recognized on SPECT studies. In our patient, the T1weighted MRI scan revealed hyperintensity in the subcortical layer after treatment; the lesion was located in the left parietal lobe. Sawada et al. (20), who reported patients with anoxic encephalopathy who demonstrated a hyperintense lesion on T1weighted images, suggested FIGURE 11. SPECT scan obtained 1 month after treatment, demonthat the hyperintensity was strating hypoperfusion in the left attributable to cortical lami- parietal lobe. nar necrosis. Although the mechanism underlying cortical laminar necrosis remains unclear, it may be attributable to neuronal damage and reactive tissue change—that is, reactive glial changes and the deposition of fat-laden macrophages (10, 20). According to Takahashi et al. (22), hyperintensity on the MRI scan implies hypoxic encephalopathy attributable to cortical laminar necrosis, delayed white matter degeneration, and probably increased iron deposition in the white matter. NEUROSURGERY In our case, the area of laminar necrosis corresponded with the hyperperfusion lesion on posttreatment SPECT. Moreover, strong hyperintensity was recognized on the diffusionweighted MRI scan. Therefore, there was severe disturbance of the rCBF in the left parietal lobe after treatment. Hyperperfusion on SPECT FIGURE 12. SPECT scan revealand hyperintensity on the ing that the hypoperfusion area is diffusion-weighted MRI not increased after acetazolamide scan after treatment is evi- challenge. dence of disruption of the blood-brain barrier. The lesion was irreversible, and the patient’s symptoms persisted after treatment. This condition may be strongly indicative of venous infarction. Therefore, in DAVF patients with preoperative findings of marked low perfusion and poor perfusion reserve, posttreatment results may include hyperperfusion. Patients with symptoms suggestive of hyperperfusion syndrome should receive aggressive blood pressure control and should undergo imaging evaluation for cerebral edema, infarct, or hemorrhage. In our patient, barbiturate therapy effectively controlled blood pressure and led to the cessation of seizures. REFERENCES 1. Andrews BT, Levy ML, Dillon W, Weinstein PR: Unilateral normal perfusion pressure breakthrough after carotid endarterectomy: Case report. Neurosurgery 21:568–571, 1987. 2. Awad IA, Little JR, Akarawi WP, Ahl J: Intracranial dural arteriovenous malformations: Factors predisposing to an aggressive neurological course. J Neurosurg 72:839–850, 1990. 3. Bernstein M, Fleming JF, Deck JH: Cerebral hyperperfusion after carotid endarterectomy: A cause of cerebral hemorrhage. Neurosurgery 15:50–56, 1984. 4. Goto K, Sidipratomo P, Ogata N, Inoue T, Matsuno H: Combining endovascular and neurosurgical treatments of high-risk dural arteriovenous fistulas in the lateral sinus and the confluence of the sinuses. J Neurosurg 90:289–299, 1999. 5. Hamada J, Kai Y, Nagahiro S, Hashimoto N, Iwata H, Ushio Y: Embolization with cellulose porous beads: Part II—Clinical trial. AJNR Am J Neuroradiol 17:1901–1906, 1996. 6. Hamada J, Ushio Y, Kazekawa K, Tsukahara T, Hashimoto N, Iwata H: Embolization with cellulose porous beads: Part I—An experimental study. AJNR Am J Neuroradiol 17:1895–1899, 1996. 7. Higashi S, Matsuda H, Fujii H, Ito H, Yamashita J: Luxury perfusion syndrome confirmed by sequential studies of regional cerebral blood flow and volume after extracranial to intracranial bypass surgery: Case report. Neurosurgery 25:85–89, 1989. 8. Hurst RW, Bagley LJ, Galetta S, Glosser G, Lieberman AP, Trojanowski J, Sinson G, Stecker M, Zager E, Raps EC, Flamm ES: Dementia resulting from dural arteriovenous fistulas: The pathologic findings of venous hypertensive encephalopathy. AJNR Am J Neuroradiol 19:1267–1273, 1998. 9. Kai Y, Hamada J, Morioka M, Koga K, Hasegawa S, Ushio Y: Correlation between magnetic resonance images and draining patterns in dural arteriovenous fistulas with leptomeningeal venous drainage. Acta Neurochir (Wien) 142:413–418, 2000. VOLUME 53 | NUMBER 1 | JULY 2003 | 231 KAI ET AL. 10. Komiyama M, Nishikawa M, Yasui T: Cortical laminar necrosis in brain infarcts: Chronological changes on MRI. Neuroradiology 39:474–479, 1997. 11. Kurata A, Miyasaka Y, Yoshida T, Kunii M, Yada K, Kan S: Venous ischemia caused by dural arteriovenous malformation: Case report. J Neurosurg 80:552–555, 1994. 12. Kuroda S, Kamiyama H, Abe H, Asaoka K, Mitsumori K: Temporary neurological deterioration caused by hyperperfusion after extracranialintracranial bypass: Case report and study of cerebral hemodynamics. Neurol Med Chir (Tokyo) 34:15–19, 1994. 13. Kuroda S, Kamiyama H, Abe H, Houkin K, Isobe M, Mitsumori K: Acetazolamide test in detecting reduced cerebral perfusion reserve and predicting long-term prognosis in patients with internal carotid artery occlusion. Neurosurgery 32:912–918, 1993. 14. Kuroda S, Ushikoshi S, Houkin K, Saito H, Kikuchi Y, Abe H: Postoperative hyperperfusion in dural arteriovenous fistula associated with venous ischemia: Case report. Surg Neurol 49:406–411, 1998. 15. Lalwani AK, Dowd CF, Halbach VV: Grading venous restrictive disease in patients with dural arteriovenous fistulas of the transverse/sigmoid sinus. J Neurosurg 79:11–15, 1993. 16. Lasjaunias P, Chiu M, ter Brugge K, Tolia A, Hurth M, Bernstein M: Neurological manifestations of intracranial dural arteriovenous malformations. J Neurosurg 64:724–730, 1986. 17. Meyers PM, Higashida RT, Phatouros CC, Malek AM, Lempert TE, Dowd CF, Halbach VV: Cerebral hyperperfusion syndrome after percutaneous transluminal stenting of the craniocervical arteries. Neurosurgery 47:335– 343, 2000. 18. Phatouros CC, Meyers PM, Higashida RT, Malek AM, Lempert TE, Dowd CF, Halbach VV: Intracranial hemorrhage and cerebral hyperperfusion syndrome after extracranial carotid artery angioplasty and stent placement. AJNR Am J Neuroradiol 23:503–504, 2002. 19. Piepgras DG, Morgan MK, Sundt TM Jr, Yanagihara T, Mussman LM: Intracerebral hemorrhage after carotid endarterectomy. J Neurosurg 68:532– 536, 1988. 20. Sawada H, Udaka F, Seriu N, Shindou K, Kameyama M, Tsujimura M: MRI demonstration of cortical laminar necrosis and delayed white matter injury in anoxic encephalopathy. Neuroradiology 32:319–321, 1990. 21. Solomon RA, Loftus CM, Quest DO, Correll JW: Incidence and etiology of intracerebral hemorrhage following carotid endarterectomy. J Neurosurg 64:29–34, 1986. 22. Takahashi S, Higano S, Ishii K, Matsumoto K, Sakamoto K, Iwasaki Y, Suzuki M: Hypoxic brain damage: Cortical laminar necrosis and delayed changes in white matter at sequential MR imaging. Radiology 189:449–456, 1993. 23. Willinsky R, ter Brugge K, Montanera W, Mikulis D, Wallace MC: Venous congestion: An MR finding in dural arteriovenous malformations with cortical venous drainage. AJNR Am J Neuroradiol 15:1501–1507, 1994. 24. Yoshimoto T, Houkin K, Kuroda S, Abe H, Kashiwaba T: Low cerebral blood flow and perfusion reserve induce hyperperfusion after surgical revascularization: Case reports and analysis of cerebral hemodynamics. Surg Neurol 48:132–138, 1997. COMMENTS K ai et al. postulate that treatment of a dural arteriovenous fistula (DAVF) resulted in a hyperperfusion syndrome. The mechanism they suggest is chronic arteriolar dilation and loss of vasomotor reactivity attributable to venous congestion. I am not convinced. Hyperperfusion after carotid endarterectomy occurs (rarely) in the setting of chronic, high-grade internal carotid artery stenosis with loss of autoregulatory ability in the cerebral vasculature distal to the flow-restrictive obstruction. Hyperperfusion occurs when the proximal flowlimiting stenosis is removed by surgery or endovascular treatment. Thus, the occurrence of a hyperperfusion syndrome requires not only the loss of vasomotor tone but augmented 232 | VOLUME 53 | NUMBER 1 | JULY 2003 cerebral blood flow after treatment. The reason for augmented flow after carotid endarterectomy is immediately apparent. In this case, it is hypothesized that treatment of the DAVF resulted in decreased venous outflow resistance, which resulted in hyperperfusion. I think a more parsimonious explanation is that the patient experienced a postoperative stroke with misery perfusion and made a partial recovery. Robert E. Harbaugh Lebanon, New Hampshire K ai et al. report a case of transverse-sigmoid DAVF with retrograde cortical venous reflux; a postoperative regional cerebral blood flow study revealed parietal hyperperfusion and disturbed vasoreactivity to acetazolamide challenge. The authors speculate that the underlying mechanism of this phenomenon is vasoparalysis attributable to severe and longstanding venous congestion and that the restored normal perfusion pressure led to hyperperfusion and parietal cortical necrosis. These assumptions are doubtful because, although preoperative venous congestion was notable in the temporooccipital region, there was no evidence of venous congestion in the parietal region, as assessed by both magnetic resonance imaging and single-photon emission computed tomography. Preoperative angiograms did not imply the existence of cortical reflux in the parietal area. If the postoperative hyperperfusion occurred in the temporo-occipital area as well and the parietal hyperperfusion was its extension, the authors’ speculation might be reasonable. But the postoperative singlephoton emission computed tomographic scan of the temporooccipital area is not presented. This postoperative hyperperfusion in the area with preoperatively normal perfusion may have to be interpreted, instead, as a rare complication that occurred after surgical isolation of the affected sinuses. Akiyo Sadato Nobuo Hashimoto Kyoto, Japan K ai et al. report a case of postoperative hyperperfusion as a complication of surgical treatment for a DAVF of the transverse sigmoid sinuses. When treating a patient with a DAVF, we should remember the possibility of postoperative hyperperfusion. In this case, the hyperperfusion area was observed in the parietal lobe; analysis by single-photon emission computed tomography and magnetic resonance imaging indicated no hypoperfusion preoperatively. On the basis of these neuroimages, regional autoregulation would have been maintained in the parietal lobe. A question remains regarding the cause of the postoperative hyperperfusion in the parietal lobe, despite the authors’ discussion of this topic. Shigeaki Kobayashi Tetsuyoshi Horiuchi Matsumoto, Japan www.neurosurgery-online.com POSTOPERATIVE HYPERPERFUSION OF DURAL ARTERIOVENOUS FISTULA T he authors present an interesting case of a 63-year-old man who presented with a DAVF in the region of the left transverse sigmoid sinus. The multiple fistulae were fed by occipital artery branches and associated with retrograde leptomeningeal venous drainage. Before surgery, the authors noted a hyperdense signal in the left parietal area on T2weighted magnetic resonance imaging scans. This was associated with hypoperfusion seen on single-photon emission computed tomographic scans with and without Diamox (Lederle, Ltd., Tokyo, Japan) in the same parietal region. The remaining portion of the brain demonstrated increased perfusion after Diamox, the outcome one would expect in this situation. After embolization and surgery, the authors observed an area of T2-weighted hyperintensity. Three days after surgery, they found subcortical hyperintense signal in the region of the left parietal and occipital lobes. They also noted hyperintense signal in the subcortical region of the left parietal and occipital lobes on both T1-weighted and diffusion-weighted imaging. This is consistent with an infarct. In this setting, one must also consider hypertensive encephalopathy. The images in Figures 8 and 9 suggest that the left parietal lobe vasculature was maximally dilated, because there was actually a decrease in perfusion in the left parietal lobe after Diamox. Although there was temporary hyperperfusion before Diamox, one notes that the left parietal area had dimin- NEUROSURGERY ished perfusion after Diamox, suggesting a steal phenomenon. One month later, the computed tomographic and perfusion studies returned to the immediate postoperative state. This sequence of events suggests that the autoregulatory reserve in the left parietal lobe had always been altered secondary to the DAVF. Whether this was because of increased flow or venous congestion remains enigmatic. It is difficult to explain the pathogenesis of the transient increase in flow 3 days after surgery. The magnetic resonance imaging scan after surgery seems more consistent with an infarct. The cortical laminar necrosis is suggestive of cell injury secondary to ischemia. It seems that the magnetic resonance images are inconsistent with the single-photon emission computed tomographic findings. One must remember that single-photon emission computed tomography is semiquantitative. Thus, the suggested elevated cerebral blood flow may have been caused by venous congestion. This highlights the usefulness of quantitative techniques such as positron emission tomographic or computed tomographic perfusion studies. This is an interesting case report that once again highlights our need to continue refining our ability to adequately assess cerebral blood flow. Philip E. Stieg New York, New York VOLUME 53 | NUMBER 1 | JULY 2003 | 233