- Email: [email protected]
Usefulness of Dual and Fully Automated Measurements of Cerebral Blood Flow during Balloon Occlusion Test of the Internal Carotid Artery
Usefulness of Dual and Fully Automated Measurements of Cerebral Blood Flow during Balloon Occlusion Test of the Internal Carotid Artery Takeshi Torigai, MD,* Mitsuhito Mase, MD, PhD,* Takayuki Ohno, MD,* Hiroyuki Katano, MD, PhD,* Yusuke Nisikawa, MD,* Keita Sakurai, MD,† Shigeru Sasaki, MD,† Junko Toyama, MD,† and Kazuo Yamada, MD, PhD*
Background: This study was conducted to show the reliability of fully automated quantification of regional cerebral blood flow (rCBF) in balloon occlusion test (BOT) of the internal carotid artery (ICA). We also shows the usefulness of ratio of rCBF during BOT to rCBF at rest (BOT/rest ratio 5 rCBF during BOT/rCBF at rest) rather than asymmetry index (AI) during BOT (AI 5 occluded-side rCBF/contralateral rCBF). Methods: In the last 2 years, we performed the BOT on 10 consecutive patients (4 with intracranial aneurysms and 6 with head and neck tumors). During the BOT, mean stump pressure (MSTP) of the ICA was monitored. We measured cerebral blood flow (CBF) with technetium-99m hexamethylpropylene amine oxime single-photon emission computed tomography at rest and during BOT. rCBF was determined using 3-dimensional stereotaxic region of interest template (3DSRT) which automatically divided CBF into 12 segments. We defined hypoperfusion segment as BOT/rest ratio ,0.9 or AI ,0.9. Results: When the BOT/rest ratio was used as a hypoperfusion parameter, the number of hypoperfusion segments was significantly greater in patients with an MSTP #50 mm Hg than in patients with an MSTP .50 mm Hg. However, only AI during BOT did not reflect MSTP significantly. Conclusions: The evaluation of CBF changes in BOT using 3DSRT and the BOT/rest ratio were useful because of objective comparison. Key Words: Balloon occlusion test—cerebral blood flow—dual single-photon emission computed tomography— mean stump pressure—technetium-99m hexamethylpropylene amine oxime— 3-dimensional stereotactic region of interest template. Ó 2013 by National Stroke Association
Patients with aneurysms of the internal carotid artery (ICA) or tumors invading the ICA at the neck or base of the skull require permanent sacrifice or transient occlusion
From the *Departments of Neurosurgery and Restorative Neuroscience; and †Quantum Radiology, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan. Received December 1, 2010; revision received July 11, 2011; accepted July 21, 2011. Address correspondence to: Mitsuhito Mase, MD, PhD, Department of Neurosurgery and Restorative Neuroscience, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2011.07.015
of the ICA. The balloon occlusion test (BOT) of the ICA has been used to predict whether the patient can tolerate either temporary or permanent occlusion of the ICA. The complication rate of ICA occlusion is lowered when the BOT is used.1 Previous reports have proposed various techniques to predict the likelihood of ischemia after ICA occlusion, such as angiographically synchronous venous filling,2,3 the measurement of stump pressure,4 the measurement of cerebral blood flow (CBF) with single-photon emission computed tomography (SPECT),5-7 dynamic computed tomography (CT),8 perfusion magnetic resonance imaging,9 regional oxygen saturation,10 and transcranial Doppler ultrasonography,11,12 xenon-enhanced CT,13-16 and monitoring of short latency somatosensory evoked potentials and electroencephalography.17,18
Journal of Stroke and Cerebrovascular Diseases, Vol. 22, No. 3 (April), 2013: pp 197-204
197
T. TORIGAI ET AL.
198
The BOT however, has several problems. The falsenegative rate ranges from 3.3% to 10%, and it is difficult to predict postoperative thromboembolic stroke.13 Several sensitive monitoring methods have been reported, but the clinical threshold for symptoms is still controversial. Of these monitoring methods, CBF study is the most reliable parameter. To evaluate CBF more objectively, we used a fully automated regional CBF (rCBF) quantification software 3-dimensional stereotaxic region of interest (ROI) template (3DSRT).19 To evaluate CBF reduction in a more sensitive manner, we applied dual SPECT study with baseline state and BOT state and found a reasonable correlation with mean stump pressure (MSTP).
Methods Patients In last 2 years, BOT of the ICA was performed in 10 consecutive patients (5 males and 5 females; mean age 57 years [range 37-65 years]; Table 1). The group was comprised of 4 patients with intracranial aneurysms and 6 patients with cervical tumors. The BOT has been approved in our institutional ethics committee, and written informed consent was obtained from all patients after a detailed explanation of the BOT procedures. All patients in this study were able to tolerate the BOT clinically.
Procedure and Measurement Before the BOT, all patients underwent baseline CBF studies using technetium-99m hexamethylpropylene amine oxime SPECT (99mTc-HMPAO SPECT) as described previously.20 The noninvasive method used by Matsuda et al21 (Patlak plot method) was used for the quantitative measurement of rCBF with 99mTc-HMPAO SPECT. Intravenous radionuclide angiography was first performed
99m
by bolus injection of 740 MBq of Tc-HMPAO from the right brachial vein, and afterward the usual brain perfusion SPECT image was obtained. The passage of the tracer from the aortic arch to the brain was monitored in a 64 3 64 format for 100 seconds at 1-second intervals using a rectangular large-field gamma camera (Shimadzu SNC-500R; Shimadzu Inc, Tokyo, Japan). ROIs were drawn over the aortic arch and bilateral brain hemispheres, and time-activity curves for these ROIs were processed. The Patlak plot was established to determine a hemispheric brain perfusion index (BPI). The BPI was converted to the hemispheric mean CBF (mCBF) values measured by 133Xe inhalation SPECT using the Matsuda regression equation. Regional CBF was calculated from mCBF using the Lassen correction algorithm. The interval between baseline CBF studies and BOT ranged from 1 to 16 days (mean interval 3 days). Because of the short half-life of 99mTc, a repeat scan could be performed 24 to 48 hours after the initial study.6 BOT was carried out in the angiographic suite. At first, diagnostic cerebral angiography was performed using a 4F catheter by the Seldinger method through a 6F sheath in the right femoral artery. Then a 5.5F triple-lumen balloon catheter (Endeavor or Sentry; Boston Scientific, Freemont, CA) was placed in the cervical portion of the ICA. To prevent embolic events during inflation of the balloon, we injected 5000 U of heparin intravenously. The ICA was occluded by inflating the balloon. Five minutes after the occlusion, an intravenous injection of 740 MBq of 99mTc-HMPAO was performed. 99mTc-HMPAO can be injected in the angiographic suite, which is also registered and approved in the operating room of our University Hospital. The administration of radioactive materials is permitted. The balloon remained inflated for 15 minutes, and patients were neurologically examined. The stump pressure was monitored continuously for these 15 minutes. Systemic blood pressure at the brachial artery was also monitored
Table 1. Characteristics of 10 patients indicated for balloon occlusion test Case no.
Age (y)/Sex
Diagnosis
MSTP (mm Hg)
1 2 3 4 5
37/F 65/M 58/M 61/M 62/M
49 42 64 40 100
6 7 8 9 10
59/M 50/F 59/F 63/F 51/F
L giant ICA ophthalmic artery aneurysm L PcomA aneurysm R neurinoma R carotid body tumor L lymph node metastasis of the thyroid adenocarcinoma L recurrent pharyngeal carcinoma L glomus tumor R ICA-PcomA aneurysm L squamous cell carcinoma of the parotid gland R ICA-PcomA and AcomA aneurysms
68 50 39 58 55
Treatment Observation Coiling Resection Resection Resection Resection Observation Coiling Chemotherapy Trapping with STA-MCA bypass
Abbreviations: AcomA, anterior communicating artery; F, female; ICA, internal carotid artery; L, left; M, male; MCA, middle cerebral artery; MSTP, mean stump pressure; PcomA, posterior communicating artery; STA, superficial temporal artery.
MEASUREMENTS OF CBF DURING BOT OF THE ICA
using an automatic sphygmomanometer at 1-minute intervals. After the balloon catheter was deflated and removed, the patient was transported to the nuclear medicine department. Then SPECT data were acquired as described above.
Statistical Analysis SPECT data were analyzed with 3DSRT. 3DSRT is fully automated whole-brain ROI analysis software. SPECT images were anatomically standardized using SPM99 followed by the quantification of 318 constant ROIs, grouped into 12 segments in each hemisphere (callosomarginal, precentral, central, parietal, angular, temporal, posterior cerebral, pericallosal, lenticular nucleus, thalamus, hippocampus, and cerebellum). The former 8 segments reflected perfusion areas irrigated by primary branches of cerebral arteries. Using these procedures, segmental CBF can be calculated as the area-weighted mean value for each of the respective 12 segments based on the rCBF in each ROI. The SPM99 converted the patient’s brain data into those in the Talairach stereotactic brain atlas. The ROI template is automatically placed on the brain. The 3DSRT, which could be identically set on the anatomically standardized images, allowed objective assessment of rCBF at rest and during BOT, which is otherwise not easy with conventional ROI settings.19,22 The parameters of CBF decrease were expressed as a ratio of CBF during BOT to CBF at rest (BOT/rest ratio) and an asymmetry index (AI) during BOT (AI 5 occludedside CBF/contralateral CBF). There are no strict criteria to define hypoperfusion, but generally it is estimated as an abnormal segment where CBF decrease is ,90% of baseline.5 Heys et al23 reported that a stump pressure of .50 mm Hg was a reliable index of safety because no permanent neurologic deficit caused by ischemia developed in patients who fulfilled this criterion and underwent endarterectomy without internal shunting. In the present study, we defined hypoperfusion as a segment with BOT/ rest ratio ,0.9 or with an AI during BOT of ,0.9. We set a cut-off value of the MSTP at 50 mm Hg. The relationship between the 2 parameters and MSTP was evaluated with the Mann–Whitney U test or Spearman rank correlation. P , .05 was considered statistically significant.
Results Two patients (cases 4 and 10) underwent extracranial to intracranial vascular bypass grafting before sacrifice of the ICA: 1 resulting in infarct because of occlusion of the graft, and the other had no complications. One patient (case 5) underwent permanent therapeutic carotid occlusion without revascularization and had no complications. Four patients underwent aneurysmal coil embolization or tumor resection without sacrifice of the ICA. The remainder of patients were under observation.
199
CBF Study By 3DSRT analysis, 5 patients (50%) developed hypoperfusion in the angular segment during BOT, because the angular segment includes the watershed zone of the middle cerebral artery (MCA). The ICA balloon occlusion caused CBF decrease in the ipsilateral watershed zone of cortex of the MCA territory. The lenticular nucleus segment also revealed hypoperfusion during BOT (Fig 1A). Regarding CBF on the nonoccluded side, 4 patients (40%) developed hypoperfusion in the lenticular nucleus segment, although there was no cortical hypoperfusion on the nonoccluded side of the MCA territory (Fig 1A). CBF changes in the angular segment of the MCA territory on the nonoccluded side correlated well with those on the occluded side (Fig 1B), but the BOT/rest ratio on the nonoccluded side was within normal limits (.0.9; Fig 1B).
Relationship between CBF and MSTP When the CBF decrease was expressed as a BOT/rest ratio, higher MSTP values correlated with fewer numbers of hypoperfused segments (Fig 2A). The number of hypoperfusion segments parallels the extent of ischemia.24 When the cut-off value of MSTP was set at 50 mm Hg, there was significant difference in the hypoperfused areas between the MSTP above 50 mm Hg group and below 50 mm Hg group (P , .05 using the Mann–Whitney U test; Fig 2B). Especially in the temporal segment, which included the watershed zone, the BOT/rest ratio was significantly correlated with MSTP (Fig 2C) (r 5 0.66; P ,.05 using Spearman rank correlation). However, when the CBF decrease was expressed as an AI during BOT, there was no significant correlation between the MSTP and the extent of hypoperfused areas (Fig 3A) nor was there any significant difference between the MSTP above 50 mm Hg group and the below 50 mm Hg group (Fig 3B; P . .05).
Crossed Cerebellar Diaschisis Associated with BOT We defined crossed cerebellar diaschisis (CCD) during BOT as diminished cerebellar perfusion during BOT (ie, a cerebellar BOT/rest ratio on the nonoccluded side of ,1). The number of cerebral hypoperfused segments was significantly greater in patients with CCD than in patients without CCD (P , .05 using the Mann–Whitney U test; Fig 4). CCD was associated with the extent of hypoperfusion.
Representative Cases Case 4 A 61-year-old man underwent a BOT of the right ICA because his carotid body tumor invaded the right ICA. A dual and fully automated SPECT study revealed hypoperfusion in the precentral, central, parietal, angular, and temporal segments, which meant diffuse ischemia in the
200
T. TORIGAI ET AL.
Figure 1. Hypoperfusion induced by the balloon occlusion test (BOT). In the cortical segment of the middle cerebral artery territory, hypoperfusion was induced only on the occluded side, but in the perforator territory, especially the lenticular nucleus segment, hypoperfusion was induced on both sides (A). In angular and lenticular nucleus segments, there was significant correlation of the BOT/rest ratio between on the occluded side and nonoccluded side (r 5 0.76 [P , .05] and r 5 0.78 [P , .05], respectively, using the Spearman rank correlation) (B).
MCA territory. The BOT/rest ratios were 0.82, 0.87, 0.82, 0.82, and 0.85, respectively. Moreover, short latency somatosensory evoked potentials abnormality (reduction
of N20 amplitude) and a low MSTP value (,45 mm Hg) during BOT were detected. These results were an indication of high-flow bypass. Before resection of the neck
Figure 2. Relationship between regional cerebral blood flow and mean stump pressure of the internal carotid artery (MSTP). The cerebral blood flow decrease was expressed as a balloon occlusion test (BOT)/rest ratio. There was a significant correlation between the MSTP and the number of hypoperfused segments (r 5 20.67; P ,.05 using the Spearman rank correlation) (A). The number of hypoperfused segments was significantly greater in patients with the MSTP #50 mm Hg than in patients with MSTP .50 mm Hg (P , .05 using the Mann–Whitney U test) (B). In the temporal segment, the BOT/rest ratio correlated well with MSTP (r 5 0.66; P , .05 using the Spearman rank correlation) (C).
MEASUREMENTS OF CBF DURING BOT OF THE ICA
201
Figure 3. Relationship between asymmetry index (AI) and the mean stump pressure of the internal carotid artery (MSTP). The cerebral blood flow decrease expressed as AI during the balloon occlusion test was compared to MSTP. There was no significant correlation between the MSTP and the number of hypoperfused segments (A). There was no statistical significant between AI and MSTP (B).
tumor, the ICA was occluded, and we placed bypass graft with saphenous vein between common carotid artery and distal ICA. However, the bypass graft was occluded because of compression from the postoperative subcutaneous hematoma. He developed infarction in the watershed zone of the MCA because of low CBF related to
bypass occlusion. The infarct area was included in the hypoperfused segments induced by the BOT. Case 5 A 62-year-old man underwent BOT of the left ICA because of thyroid adenocarcinoma invading the left ICA, which needed to be sacrificed. His MSTP on the BOT was 100 mm Hg. The number of hypoperfusion segments was none when the CBF decrease was counted according to the BOT/rest ratio (Table 2). If we counted his hypoperfusion segment during BOT according to the AI, 4 areas were defined as hypoperfused segments (AI ,0.9). During resection of the neck tumor, the ICA was sacrificed without any bypass. He did not develop any ischemic events postoperatively.
Discussion Accuracy of BOT Figure 4. Crossed cerebellar diaschisis associated with the balloon occlusion test (BOT). The number of hypoperfused segments was significantly greater in patients with BOT/rest ratios on the nonoccluded side of the cerebellum ,1 than in patients with those $1 (P , .05 using the Mann– Whitney U test).
Therapeutic occlusion of the ICA remains an important procedure, but the risk of ischemic events remains a matter of discussion. To overcome this drawback, temporary BOT is imperative to evaluate ischemic risks before definitive
T. TORIGAI ET AL.
202
Table 2. Cerebral blood flow of case 5 CBF at rest
CBF during BOT
Segment
R side
L side
R side
L side
AI during BOT
BOT/rest ratio
Anterior Precentral Central Parietal Angular Temporal Occipital Pericallosal Lenticular nucleus Thalamus Hippocampus Cerebellum
40.95 41.82 37.06 42.26 44.55 40.39 44.56 43.21 47.83 55.93 39.99 52.14
40.51 42.04 36.57 42.24 39.84 40.32 42.76 42.35 55.15 52.78 35.04 52.05
44.38 45.99 42.02 46.75 45.11 44.55 49.24 48.12 53.03 64.54 46.02 55.78
42.8 40.15 39.45 41.43 40.73 39.99 44.94 44.42 51.99 62.3 40.31 58.89
0.96 0.87 0.93 0.88 0.9 0.89 0.91 0.92 0.98 0.96 0.87 1.05
1.05 0.95 1.07 0.98 1.02 0.99 1.05 1.04 0.94 1.18 1.15 1.13
Abbreviations: AI, asymmetry index; BOT, balloon occlusion test; CBF, cerebral blood flow (mL/100 g/min); L, left; R, right.
occlusion is performed. To increase the diagnostic accuracy of BOT, many reports have proposed procedures and modalities. SPECT is a useful tool for detecting hypoperfusion during BOT. SPECTcan depict clinically silent areas of decreased cerebral perfusion in patients who had no symptoms during the BOT.5 In these areas, quantification of the absolute value in CBF was important,25 and semiquantitative analysis provided a more objective tool to support the visual impression, especially in mild perfusion abnormalities.8 However, the specificity of changes between baseline and occlusion SPECTwas reportedly poor.12 3DSRT is a piece of fully automated rCBF quantification software that enables fully automated accurate ROI delineation to minimize interoperator variability and objective quantification of rCBF.22 For example, it was used for ROI analysis in the territory of the MCA to evaluate preoperative cerebrovascular reactivity measured by quantitative SPECT.26 Because we used 3DSRT registration, anatomic location was standardized and location-related artifacts were omitted. The present study revealed that the redistribution of CBF occurred during BOT more segmentally. BOT caused hypoperfusion in the occluded-side cortical area of the MCA territory. At the same time, hemodynamic compensation via collateral flow through the anterior communicating artery (AcomA) might cause subtle CBF decrease on the nonoccluded side of MCA territory. On the both sides of perforator territory, hypoperfusion developed, because lenticulostriate arteries had relatively little collateral blood supply27 and interhemispheric compensation via AcomA might have caused ischemia on the nonoccluded side of the lenticular nucleus segment (Fig 1A). The segmental analysis of case 4 showed that because the SPECT study revealed a hypoperfused area, the watershed infarction was predictable when the graft was occluded, although most of the causes of graft occlusion are
technical and can be improved by experience and careful intraoperative monitoring.8
Usefulness of Dual SPECT Study To maintain homogeneity of the present study, we conducted baseline SPECT in all patients as the reference. However, questions arise about whether baseline SPECT should always be performed.28 Additional preocclusion SPECT study, when compared to the postocclusion study alone, might provide more information and sensitivity as compared to post-BOT AI analysis.25 SPECT, however, is expensive, cumbersome, and frequently unavailable.2 If a BOT revealed no clinical deficits, a baseline study was not always obtained.29 Witt et al14 reported that when the CBF decrease was expressed as an AI during BOT, qualitative flow analysis failed to identify the patients at high risk who had reduced CBF values between 20 and 30 mL/ 100 g/minute, because raising the asymmetry threshold increased the specificity as the sensitivity decreased. The current study indicated that the CBF decrease was underestimated with AI measurement than with the BOT/rest ratio measurement because CBF on the nonoccluded side decreased in parallel to the CBF on the occluded side. When there was asymmetry at rest (Table 2), dual SPECT was more accurate.
Relationship between MSTP and rCBF Erba et al30 reported that rCBF in the territory of the MCA did not correlate with MSTP. On the other hand, Steed et al15 reported that xenon computed tomographic CBF mapping was correlated with ICA stump pressure and clinical neurologic assessment during temporary ICA occlusion. The present study shows that AI during BOT has a disadvantage because it was not correlating with MSTP (Fig 3). However, Fig 2C shows in the temporal segment that the
MEASUREMENTS OF CBF DURING BOT OF THE ICA
BOT/rest ratio was significantly correlated with MSTP. We revealed that MSTP reflected the magnitude of CBF decrease in the watershed zone measured with the BOT/ rest ratio.
Threshold of Hypoperfusion during BOT No definite quantitative criteria are available for defining hypoperfusion. Monsein et al25 considered that asymmetry ratios different by .10% were important because the left/right hemisphere ratio in normal patients has been founded to be approximately 1.00 6 0.10 (mean 6 SD). Lorberboym et al8 defined mild, moderate, and severe hypoperfusion as diminished perfusion of 10% to 19%, 20% to 29%, and $30% compared to the contralateral ROI, respectively. They reported 2 infarcts caused by graft occlusion among 8 patients with moderate or severe hypoperfusion and 3 infarcts after carotid occlusion without revasularization despite normal perfusion. In the present study, 9 of 10 patients had mildly hypoperfused segments whose BOT/rest ratios were ,0.9 in some areas, but they were .0.8. One patient (case 2) had moderately hypoperfused segments in MCA territory where the BOT/rest ratio was ,0.8. The patient underwent coil embolization without ICA sacrifice and had no complications. From these data, we defined the threshold of the BOT/rest ratio as 0.9, because 1 case whose BOT/ rest ratio was ,0.9 became an infarct caused by graft occlusion (case 4). When the BOT/rest ratio was .0.9—as seen in case 5—permanent therapeutic carotid occlusion without revascularization was successful.
CCD during BOT In the present study, CCD occurred during BOT when it was defined as cerebellar BOT/rest ratio ,1. The extent of cerebral hypoperfused segments was significantly greater in patients with CCD than in patients without CCD. The reduction of cerebellar CBF was subtle because the cerebellar BOT/rest ratio was .0.9. Brunberg et al31 reported a case of CCD encountered during BOT with positron emission tomography. They suggested that the rapid development of CCD was an almost instantaneous response to diminished cortico-ponto-cerebellar and/or spinocerebellar pathway excitatory input and might be a useful indicator of remote neuronal dysfunction. In the present study, CCD was associated with the extent of cerebral hypoperfusion. CCD might be attributed to diffuse cortical hypoperfusion and be an indicator of high risk.
Conclusion Our results suggest that the redistribution of CBF might occur during the BOT, not only in the occluded side but also in the nonoccluded side. Additional study with a greater numbers of patients is required to review the clinical meanings.
203
In conclusion, quantitative analysis of SPECT images using 3DSRT was more objective and accurate than visual analysis alone because it minimized interoperator variability. The evaluation of rCBF with dual SPECT revealed the ability of cerebral collateral-flow redistribution more accurately than with asymmetry index alone during the BOT. Defining the threshold of hypoperfusion as BOT/ rest ratio ,0.9 was a sensitive predictor for developing ischemia after therapeutic occlusion of the ICA.
References 1. Brunberg JA, Frey KA, Horton JA, et al. [15O]H2O positron emission tomography determination of cerebral blood flow during balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 1994;15:725-732. 2. van Rooij WJJ, Sluzewski M, Metz NH, et al. Carotid balloon occlusion for large and giant aneurysms: Evaluation of a new test occlusion protocol. Neurosurgery 2000; 47:116-122. 3. Abud DG, Spelle L, Piotin M, et al. Venous phase timing during balloon test occlusion as a criterion for permanent internal carotid artery sacrifice. AJNR Am J Neuroradiol 2005;26:2602-2609. 4. Kurata A, Miyasaka Y, Tanaka C, et al. Stump pressure as a guide to the safety of permanent occlusion of the internal carotid artery. Acta Neurochir (Wien) 1996;138: 549-554. 5. Sugawara Y, Kikuchi T, Ueda T, et al. Usefulness of brain SPECT to evaluate brain tolerance and hemodynamic changes during temporary balloon occlusion test and after permanent carotid occlusion. J Nucl Med 2002; 43:1616-1623. 6. Peterman SB, Taylor A Jr, Hoffman JC Jr. Improved detection of cerebral hypoperfusion with internal carotid balloon test occlusion and 99mTc-HMPAO cerebral perfusion SPECT imaging. AJNR Am J Neuroradiol 1991; 12:1035-1041. 7. Simonson TM, Ryals TJ, Yuh WT, et al. MR imaging and HMPAO scintigraphy in conjunction with balloon test occlusion: Value in predicting sequelae after permanent carotid occlusion. AJR Am J Roentgenol 1992;159: 1063-1068. 8. Lorberboym M, Pandit N, Machac J, et al. Brain perfusion imaging during preoperative temporary balloon occlusion of the internal carotid artery. J Nucl Med 1996; 37:415-419. 9. Michel E, Liu H, Remley KB, et al. Perfusion MR neuroimaging in patients undergoing balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 2001;22:1590-1596. 10. Kaminogo M, Ochi M, Onizuka M, et al. An additional monitoring of regional cerebral oxygen saturation to HMPAO SPECT study during balloon test occlusion. Stroke 1999;30:407-413. 11. Sorteberg W, Sorteberg A, Lindegaard KF, et al. Transcranial Doppler ultrasonography-guided management of internal carotid artery closure. Neurosurgery 1999;45:76-87. 12. Schneweis S, Urbach H, Solymosi L, et al. Preoperative risk assessment for carotid occlusion by transcranial Doppler ultrasound. J Neurol Neurosurg Psychiatry 1997;62:485-489. 13. Linskey ME, Jungreis CA, Yonas H, et al. Stroke risk after abrupt internal carotid artery sacrifice: Accuracy of preoperative assessment with balloon test occlusion and
T. TORIGAI ET AL.
204
14.
15.
16.
17.
18.
19.
20.
21.
22.
stable xenon-enhanced CT. AJNR Am J Neuroradiol 1994; 15:829-843. Witt JP, Yonas H, Jungreis C. Cerebral blood flow response pattern during balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 1994; 15:847-856. Steed DL, Webster MW, DeVries EJ, et al. Clinical observations on the effect of carotid artery occlusion on cerebral blood flow mapped by xenon computed tomography and its correlation with carotid artery back pressure. J Vasc Surg 1990;11:38-43. Okudaira Y, Arai H, Sato K. Cerebral blood flow alteration by acetazolamide during carotid balloon occlusion: Parameters reflecting cerebral perfusion pressure in the acetazolamide test. Stroke 1996;27:617-621. Murphy KJ, Payne T, Jamadar DA, et al. Correlation of continuous EEG monitoring with [O-l5]H2O positron emission tomography determination of cerebral blood flow during balloon test occlusion of the internal carotid artery. Experience in 34 cases. Interv Neuroradiol 1998; 4:51-55. Liu AY, Lopez JR, Do HM, et al. Neurophysiological monitoring in the endovascular therapy of aneurysms. AJNR Am J Neuroradiol 2003;24:1520-1527. Takeuchi R, Yonekura Y, Matsuda H, et al. Usefulness of a three-dimensional stereotaxic ROI template on anatomically standardised 99mTc-ECD SPET. Eur J Nucl Med 2002;29:331-341. Umemura A, Suzuka T, Yamada K. Quantitative measurement of cerebral blood flow by 99mTc-HMPAO SPECT in acute ischaemic stroke: Usefulness in determining therapeutic options. J Neurol Neurosurg Psychiatry 2000;69:472-478. Matsuda H, Tsuji S, Shuke N, et al. Noninvasive measurements of regional cerebral blood flow using technetium99m hexamethylpropylene amine oxime. Eur J Nucl Med 1993;20:391-401. Takeuchi R, Yonekura Y, Takeda SK, et al. Fully automated quantification of regional cerebral blood flow with three-dimensional stereotaxic region of interest
23.
24.
25.
26.
27.
28.
29.
30.
31.
template: validation using magnetic resonance imaging— Technical note. Neurol Med Chir (Tokyo) 2003;43:153-162. Hays RJ, Levinson SA, Wylie EJ. Intraoperative measurement of carotid back pressure as a guide to operative management for carotid endarterectomy. Surgery 1972; 72:953-960. Tomura N, Omachi K, Takahashi S, et al. Comparison of technetium Tc 99m hexamethylpropyleneamine oxime single-photon emission tomograph with stump pressure during the balloon occlusion test of the internal carotid artery. AJNR Am J Neuroradiol 2005;26:1937-1942. Monsein LH, Jeffery PJ, van Heerden BB, et al. Assessing adequacy of collateral circulation during balloon test occlusion of the internal carotid artery with 99mTcHMPAO SPECT. AJNR Am J Neuroradiol 1991;12: 1045-1051. Aso K, Ogasawara K, Sasaki M, et al. Preoperative cerebrovascular reactivity to acetazolamide measured by brain perfusion SPECT predicts development of cerebral ischemic lesions caused by microemboli during carotid endarterectomy. Eur J Nucl Med Mol Imaging 2009; 36:294-301. Yong SW, Bang OY, Lee PH, et al. Internal and cortical border-zone infarction: Clinical and diffusion-weighted imaging features. Stroke 2006;37:841-846. Askienazy S, Lebtahi R, Meder JF. SPECT HMPAO and balloon test occlusion: interest in predicting tolerance prior to permanent cerebral artery occlusion. J Nucl Med 1993;34:1243-1245. Mathews D, Walker BS, Purdy PD, et al. Brain blood flow SPECT in temporary balloon occlusion of carotid and intracerebral arteries. J Nucl Med 1993;34:1239-1243. Erba SM, Horton JA, Latchaw RE, et al. Balloon test occlusion of the internal carotid artery with stable xenon/CT cerebral blood flow imaging. AJNR Am J Neuroradiol 1988;9:533-538. Brunberg JA, Frey KA, Horton JA, et al. Crossed cerebellar diaschisis: occurrence and resolution demonstrated with PET during carotid temporary balloon occlusion. AJNR Am J Neuroradiol 1992;13:58-61.
Background: This study was conducted to show the reliability of fully automated quantification of regional cerebral blood flow (rCBF) in balloon occlusion test (BOT) of the internal carotid artery (ICA). We also shows the usefulness of ratio of rCBF during BOT to rCBF at rest (BOT/rest ratio 5 rCBF during BOT/rCBF at rest) rather than asymmetry index (AI) during BOT (AI 5 occluded-side rCBF/contralateral rCBF). Methods: In the last 2 years, we performed the BOT on 10 consecutive patients (4 with intracranial aneurysms and 6 with head and neck tumors). During the BOT, mean stump pressure (MSTP) of the ICA was monitored. We measured cerebral blood flow (CBF) with technetium-99m hexamethylpropylene amine oxime single-photon emission computed tomography at rest and during BOT. rCBF was determined using 3-dimensional stereotaxic region of interest template (3DSRT) which automatically divided CBF into 12 segments. We defined hypoperfusion segment as BOT/rest ratio ,0.9 or AI ,0.9. Results: When the BOT/rest ratio was used as a hypoperfusion parameter, the number of hypoperfusion segments was significantly greater in patients with an MSTP #50 mm Hg than in patients with an MSTP .50 mm Hg. However, only AI during BOT did not reflect MSTP significantly. Conclusions: The evaluation of CBF changes in BOT using 3DSRT and the BOT/rest ratio were useful because of objective comparison. Key Words: Balloon occlusion test—cerebral blood flow—dual single-photon emission computed tomography— mean stump pressure—technetium-99m hexamethylpropylene amine oxime— 3-dimensional stereotactic region of interest template. Ó 2013 by National Stroke Association
Patients with aneurysms of the internal carotid artery (ICA) or tumors invading the ICA at the neck or base of the skull require permanent sacrifice or transient occlusion
From the *Departments of Neurosurgery and Restorative Neuroscience; and †Quantum Radiology, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan. Received December 1, 2010; revision received July 11, 2011; accepted July 21, 2011. Address correspondence to: Mitsuhito Mase, MD, PhD, Department of Neurosurgery and Restorative Neuroscience, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2013 by National Stroke Association doi:10.1016/j.jstrokecerebrovasdis.2011.07.015
of the ICA. The balloon occlusion test (BOT) of the ICA has been used to predict whether the patient can tolerate either temporary or permanent occlusion of the ICA. The complication rate of ICA occlusion is lowered when the BOT is used.1 Previous reports have proposed various techniques to predict the likelihood of ischemia after ICA occlusion, such as angiographically synchronous venous filling,2,3 the measurement of stump pressure,4 the measurement of cerebral blood flow (CBF) with single-photon emission computed tomography (SPECT),5-7 dynamic computed tomography (CT),8 perfusion magnetic resonance imaging,9 regional oxygen saturation,10 and transcranial Doppler ultrasonography,11,12 xenon-enhanced CT,13-16 and monitoring of short latency somatosensory evoked potentials and electroencephalography.17,18
Journal of Stroke and Cerebrovascular Diseases, Vol. 22, No. 3 (April), 2013: pp 197-204
197
T. TORIGAI ET AL.
198
The BOT however, has several problems. The falsenegative rate ranges from 3.3% to 10%, and it is difficult to predict postoperative thromboembolic stroke.13 Several sensitive monitoring methods have been reported, but the clinical threshold for symptoms is still controversial. Of these monitoring methods, CBF study is the most reliable parameter. To evaluate CBF more objectively, we used a fully automated regional CBF (rCBF) quantification software 3-dimensional stereotaxic region of interest (ROI) template (3DSRT).19 To evaluate CBF reduction in a more sensitive manner, we applied dual SPECT study with baseline state and BOT state and found a reasonable correlation with mean stump pressure (MSTP).
Methods Patients In last 2 years, BOT of the ICA was performed in 10 consecutive patients (5 males and 5 females; mean age 57 years [range 37-65 years]; Table 1). The group was comprised of 4 patients with intracranial aneurysms and 6 patients with cervical tumors. The BOT has been approved in our institutional ethics committee, and written informed consent was obtained from all patients after a detailed explanation of the BOT procedures. All patients in this study were able to tolerate the BOT clinically.
Procedure and Measurement Before the BOT, all patients underwent baseline CBF studies using technetium-99m hexamethylpropylene amine oxime SPECT (99mTc-HMPAO SPECT) as described previously.20 The noninvasive method used by Matsuda et al21 (Patlak plot method) was used for the quantitative measurement of rCBF with 99mTc-HMPAO SPECT. Intravenous radionuclide angiography was first performed
99m
by bolus injection of 740 MBq of Tc-HMPAO from the right brachial vein, and afterward the usual brain perfusion SPECT image was obtained. The passage of the tracer from the aortic arch to the brain was monitored in a 64 3 64 format for 100 seconds at 1-second intervals using a rectangular large-field gamma camera (Shimadzu SNC-500R; Shimadzu Inc, Tokyo, Japan). ROIs were drawn over the aortic arch and bilateral brain hemispheres, and time-activity curves for these ROIs were processed. The Patlak plot was established to determine a hemispheric brain perfusion index (BPI). The BPI was converted to the hemispheric mean CBF (mCBF) values measured by 133Xe inhalation SPECT using the Matsuda regression equation. Regional CBF was calculated from mCBF using the Lassen correction algorithm. The interval between baseline CBF studies and BOT ranged from 1 to 16 days (mean interval 3 days). Because of the short half-life of 99mTc, a repeat scan could be performed 24 to 48 hours after the initial study.6 BOT was carried out in the angiographic suite. At first, diagnostic cerebral angiography was performed using a 4F catheter by the Seldinger method through a 6F sheath in the right femoral artery. Then a 5.5F triple-lumen balloon catheter (Endeavor or Sentry; Boston Scientific, Freemont, CA) was placed in the cervical portion of the ICA. To prevent embolic events during inflation of the balloon, we injected 5000 U of heparin intravenously. The ICA was occluded by inflating the balloon. Five minutes after the occlusion, an intravenous injection of 740 MBq of 99mTc-HMPAO was performed. 99mTc-HMPAO can be injected in the angiographic suite, which is also registered and approved in the operating room of our University Hospital. The administration of radioactive materials is permitted. The balloon remained inflated for 15 minutes, and patients were neurologically examined. The stump pressure was monitored continuously for these 15 minutes. Systemic blood pressure at the brachial artery was also monitored
Table 1. Characteristics of 10 patients indicated for balloon occlusion test Case no.
Age (y)/Sex
Diagnosis
MSTP (mm Hg)
1 2 3 4 5
37/F 65/M 58/M 61/M 62/M
49 42 64 40 100
6 7 8 9 10
59/M 50/F 59/F 63/F 51/F
L giant ICA ophthalmic artery aneurysm L PcomA aneurysm R neurinoma R carotid body tumor L lymph node metastasis of the thyroid adenocarcinoma L recurrent pharyngeal carcinoma L glomus tumor R ICA-PcomA aneurysm L squamous cell carcinoma of the parotid gland R ICA-PcomA and AcomA aneurysms
68 50 39 58 55
Treatment Observation Coiling Resection Resection Resection Resection Observation Coiling Chemotherapy Trapping with STA-MCA bypass
Abbreviations: AcomA, anterior communicating artery; F, female; ICA, internal carotid artery; L, left; M, male; MCA, middle cerebral artery; MSTP, mean stump pressure; PcomA, posterior communicating artery; STA, superficial temporal artery.
MEASUREMENTS OF CBF DURING BOT OF THE ICA
using an automatic sphygmomanometer at 1-minute intervals. After the balloon catheter was deflated and removed, the patient was transported to the nuclear medicine department. Then SPECT data were acquired as described above.
Statistical Analysis SPECT data were analyzed with 3DSRT. 3DSRT is fully automated whole-brain ROI analysis software. SPECT images were anatomically standardized using SPM99 followed by the quantification of 318 constant ROIs, grouped into 12 segments in each hemisphere (callosomarginal, precentral, central, parietal, angular, temporal, posterior cerebral, pericallosal, lenticular nucleus, thalamus, hippocampus, and cerebellum). The former 8 segments reflected perfusion areas irrigated by primary branches of cerebral arteries. Using these procedures, segmental CBF can be calculated as the area-weighted mean value for each of the respective 12 segments based on the rCBF in each ROI. The SPM99 converted the patient’s brain data into those in the Talairach stereotactic brain atlas. The ROI template is automatically placed on the brain. The 3DSRT, which could be identically set on the anatomically standardized images, allowed objective assessment of rCBF at rest and during BOT, which is otherwise not easy with conventional ROI settings.19,22 The parameters of CBF decrease were expressed as a ratio of CBF during BOT to CBF at rest (BOT/rest ratio) and an asymmetry index (AI) during BOT (AI 5 occludedside CBF/contralateral CBF). There are no strict criteria to define hypoperfusion, but generally it is estimated as an abnormal segment where CBF decrease is ,90% of baseline.5 Heys et al23 reported that a stump pressure of .50 mm Hg was a reliable index of safety because no permanent neurologic deficit caused by ischemia developed in patients who fulfilled this criterion and underwent endarterectomy without internal shunting. In the present study, we defined hypoperfusion as a segment with BOT/ rest ratio ,0.9 or with an AI during BOT of ,0.9. We set a cut-off value of the MSTP at 50 mm Hg. The relationship between the 2 parameters and MSTP was evaluated with the Mann–Whitney U test or Spearman rank correlation. P , .05 was considered statistically significant.
Results Two patients (cases 4 and 10) underwent extracranial to intracranial vascular bypass grafting before sacrifice of the ICA: 1 resulting in infarct because of occlusion of the graft, and the other had no complications. One patient (case 5) underwent permanent therapeutic carotid occlusion without revascularization and had no complications. Four patients underwent aneurysmal coil embolization or tumor resection without sacrifice of the ICA. The remainder of patients were under observation.
199
CBF Study By 3DSRT analysis, 5 patients (50%) developed hypoperfusion in the angular segment during BOT, because the angular segment includes the watershed zone of the middle cerebral artery (MCA). The ICA balloon occlusion caused CBF decrease in the ipsilateral watershed zone of cortex of the MCA territory. The lenticular nucleus segment also revealed hypoperfusion during BOT (Fig 1A). Regarding CBF on the nonoccluded side, 4 patients (40%) developed hypoperfusion in the lenticular nucleus segment, although there was no cortical hypoperfusion on the nonoccluded side of the MCA territory (Fig 1A). CBF changes in the angular segment of the MCA territory on the nonoccluded side correlated well with those on the occluded side (Fig 1B), but the BOT/rest ratio on the nonoccluded side was within normal limits (.0.9; Fig 1B).
Relationship between CBF and MSTP When the CBF decrease was expressed as a BOT/rest ratio, higher MSTP values correlated with fewer numbers of hypoperfused segments (Fig 2A). The number of hypoperfusion segments parallels the extent of ischemia.24 When the cut-off value of MSTP was set at 50 mm Hg, there was significant difference in the hypoperfused areas between the MSTP above 50 mm Hg group and below 50 mm Hg group (P , .05 using the Mann–Whitney U test; Fig 2B). Especially in the temporal segment, which included the watershed zone, the BOT/rest ratio was significantly correlated with MSTP (Fig 2C) (r 5 0.66; P ,.05 using Spearman rank correlation). However, when the CBF decrease was expressed as an AI during BOT, there was no significant correlation between the MSTP and the extent of hypoperfused areas (Fig 3A) nor was there any significant difference between the MSTP above 50 mm Hg group and the below 50 mm Hg group (Fig 3B; P . .05).
Crossed Cerebellar Diaschisis Associated with BOT We defined crossed cerebellar diaschisis (CCD) during BOT as diminished cerebellar perfusion during BOT (ie, a cerebellar BOT/rest ratio on the nonoccluded side of ,1). The number of cerebral hypoperfused segments was significantly greater in patients with CCD than in patients without CCD (P , .05 using the Mann–Whitney U test; Fig 4). CCD was associated with the extent of hypoperfusion.
Representative Cases Case 4 A 61-year-old man underwent a BOT of the right ICA because his carotid body tumor invaded the right ICA. A dual and fully automated SPECT study revealed hypoperfusion in the precentral, central, parietal, angular, and temporal segments, which meant diffuse ischemia in the
200
T. TORIGAI ET AL.
Figure 1. Hypoperfusion induced by the balloon occlusion test (BOT). In the cortical segment of the middle cerebral artery territory, hypoperfusion was induced only on the occluded side, but in the perforator territory, especially the lenticular nucleus segment, hypoperfusion was induced on both sides (A). In angular and lenticular nucleus segments, there was significant correlation of the BOT/rest ratio between on the occluded side and nonoccluded side (r 5 0.76 [P , .05] and r 5 0.78 [P , .05], respectively, using the Spearman rank correlation) (B).
MCA territory. The BOT/rest ratios were 0.82, 0.87, 0.82, 0.82, and 0.85, respectively. Moreover, short latency somatosensory evoked potentials abnormality (reduction
of N20 amplitude) and a low MSTP value (,45 mm Hg) during BOT were detected. These results were an indication of high-flow bypass. Before resection of the neck
Figure 2. Relationship between regional cerebral blood flow and mean stump pressure of the internal carotid artery (MSTP). The cerebral blood flow decrease was expressed as a balloon occlusion test (BOT)/rest ratio. There was a significant correlation between the MSTP and the number of hypoperfused segments (r 5 20.67; P ,.05 using the Spearman rank correlation) (A). The number of hypoperfused segments was significantly greater in patients with the MSTP #50 mm Hg than in patients with MSTP .50 mm Hg (P , .05 using the Mann–Whitney U test) (B). In the temporal segment, the BOT/rest ratio correlated well with MSTP (r 5 0.66; P , .05 using the Spearman rank correlation) (C).
MEASUREMENTS OF CBF DURING BOT OF THE ICA
201
Figure 3. Relationship between asymmetry index (AI) and the mean stump pressure of the internal carotid artery (MSTP). The cerebral blood flow decrease expressed as AI during the balloon occlusion test was compared to MSTP. There was no significant correlation between the MSTP and the number of hypoperfused segments (A). There was no statistical significant between AI and MSTP (B).
tumor, the ICA was occluded, and we placed bypass graft with saphenous vein between common carotid artery and distal ICA. However, the bypass graft was occluded because of compression from the postoperative subcutaneous hematoma. He developed infarction in the watershed zone of the MCA because of low CBF related to
bypass occlusion. The infarct area was included in the hypoperfused segments induced by the BOT. Case 5 A 62-year-old man underwent BOT of the left ICA because of thyroid adenocarcinoma invading the left ICA, which needed to be sacrificed. His MSTP on the BOT was 100 mm Hg. The number of hypoperfusion segments was none when the CBF decrease was counted according to the BOT/rest ratio (Table 2). If we counted his hypoperfusion segment during BOT according to the AI, 4 areas were defined as hypoperfused segments (AI ,0.9). During resection of the neck tumor, the ICA was sacrificed without any bypass. He did not develop any ischemic events postoperatively.
Discussion Accuracy of BOT Figure 4. Crossed cerebellar diaschisis associated with the balloon occlusion test (BOT). The number of hypoperfused segments was significantly greater in patients with BOT/rest ratios on the nonoccluded side of the cerebellum ,1 than in patients with those $1 (P , .05 using the Mann– Whitney U test).
Therapeutic occlusion of the ICA remains an important procedure, but the risk of ischemic events remains a matter of discussion. To overcome this drawback, temporary BOT is imperative to evaluate ischemic risks before definitive
T. TORIGAI ET AL.
202
Table 2. Cerebral blood flow of case 5 CBF at rest
CBF during BOT
Segment
R side
L side
R side
L side
AI during BOT
BOT/rest ratio
Anterior Precentral Central Parietal Angular Temporal Occipital Pericallosal Lenticular nucleus Thalamus Hippocampus Cerebellum
40.95 41.82 37.06 42.26 44.55 40.39 44.56 43.21 47.83 55.93 39.99 52.14
40.51 42.04 36.57 42.24 39.84 40.32 42.76 42.35 55.15 52.78 35.04 52.05
44.38 45.99 42.02 46.75 45.11 44.55 49.24 48.12 53.03 64.54 46.02 55.78
42.8 40.15 39.45 41.43 40.73 39.99 44.94 44.42 51.99 62.3 40.31 58.89
0.96 0.87 0.93 0.88 0.9 0.89 0.91 0.92 0.98 0.96 0.87 1.05
1.05 0.95 1.07 0.98 1.02 0.99 1.05 1.04 0.94 1.18 1.15 1.13
Abbreviations: AI, asymmetry index; BOT, balloon occlusion test; CBF, cerebral blood flow (mL/100 g/min); L, left; R, right.
occlusion is performed. To increase the diagnostic accuracy of BOT, many reports have proposed procedures and modalities. SPECT is a useful tool for detecting hypoperfusion during BOT. SPECTcan depict clinically silent areas of decreased cerebral perfusion in patients who had no symptoms during the BOT.5 In these areas, quantification of the absolute value in CBF was important,25 and semiquantitative analysis provided a more objective tool to support the visual impression, especially in mild perfusion abnormalities.8 However, the specificity of changes between baseline and occlusion SPECTwas reportedly poor.12 3DSRT is a piece of fully automated rCBF quantification software that enables fully automated accurate ROI delineation to minimize interoperator variability and objective quantification of rCBF.22 For example, it was used for ROI analysis in the territory of the MCA to evaluate preoperative cerebrovascular reactivity measured by quantitative SPECT.26 Because we used 3DSRT registration, anatomic location was standardized and location-related artifacts were omitted. The present study revealed that the redistribution of CBF occurred during BOT more segmentally. BOT caused hypoperfusion in the occluded-side cortical area of the MCA territory. At the same time, hemodynamic compensation via collateral flow through the anterior communicating artery (AcomA) might cause subtle CBF decrease on the nonoccluded side of MCA territory. On the both sides of perforator territory, hypoperfusion developed, because lenticulostriate arteries had relatively little collateral blood supply27 and interhemispheric compensation via AcomA might have caused ischemia on the nonoccluded side of the lenticular nucleus segment (Fig 1A). The segmental analysis of case 4 showed that because the SPECT study revealed a hypoperfused area, the watershed infarction was predictable when the graft was occluded, although most of the causes of graft occlusion are
technical and can be improved by experience and careful intraoperative monitoring.8
Usefulness of Dual SPECT Study To maintain homogeneity of the present study, we conducted baseline SPECT in all patients as the reference. However, questions arise about whether baseline SPECT should always be performed.28 Additional preocclusion SPECT study, when compared to the postocclusion study alone, might provide more information and sensitivity as compared to post-BOT AI analysis.25 SPECT, however, is expensive, cumbersome, and frequently unavailable.2 If a BOT revealed no clinical deficits, a baseline study was not always obtained.29 Witt et al14 reported that when the CBF decrease was expressed as an AI during BOT, qualitative flow analysis failed to identify the patients at high risk who had reduced CBF values between 20 and 30 mL/ 100 g/minute, because raising the asymmetry threshold increased the specificity as the sensitivity decreased. The current study indicated that the CBF decrease was underestimated with AI measurement than with the BOT/rest ratio measurement because CBF on the nonoccluded side decreased in parallel to the CBF on the occluded side. When there was asymmetry at rest (Table 2), dual SPECT was more accurate.
Relationship between MSTP and rCBF Erba et al30 reported that rCBF in the territory of the MCA did not correlate with MSTP. On the other hand, Steed et al15 reported that xenon computed tomographic CBF mapping was correlated with ICA stump pressure and clinical neurologic assessment during temporary ICA occlusion. The present study shows that AI during BOT has a disadvantage because it was not correlating with MSTP (Fig 3). However, Fig 2C shows in the temporal segment that the
MEASUREMENTS OF CBF DURING BOT OF THE ICA
BOT/rest ratio was significantly correlated with MSTP. We revealed that MSTP reflected the magnitude of CBF decrease in the watershed zone measured with the BOT/ rest ratio.
Threshold of Hypoperfusion during BOT No definite quantitative criteria are available for defining hypoperfusion. Monsein et al25 considered that asymmetry ratios different by .10% were important because the left/right hemisphere ratio in normal patients has been founded to be approximately 1.00 6 0.10 (mean 6 SD). Lorberboym et al8 defined mild, moderate, and severe hypoperfusion as diminished perfusion of 10% to 19%, 20% to 29%, and $30% compared to the contralateral ROI, respectively. They reported 2 infarcts caused by graft occlusion among 8 patients with moderate or severe hypoperfusion and 3 infarcts after carotid occlusion without revasularization despite normal perfusion. In the present study, 9 of 10 patients had mildly hypoperfused segments whose BOT/rest ratios were ,0.9 in some areas, but they were .0.8. One patient (case 2) had moderately hypoperfused segments in MCA territory where the BOT/rest ratio was ,0.8. The patient underwent coil embolization without ICA sacrifice and had no complications. From these data, we defined the threshold of the BOT/rest ratio as 0.9, because 1 case whose BOT/ rest ratio was ,0.9 became an infarct caused by graft occlusion (case 4). When the BOT/rest ratio was .0.9—as seen in case 5—permanent therapeutic carotid occlusion without revascularization was successful.
CCD during BOT In the present study, CCD occurred during BOT when it was defined as cerebellar BOT/rest ratio ,1. The extent of cerebral hypoperfused segments was significantly greater in patients with CCD than in patients without CCD. The reduction of cerebellar CBF was subtle because the cerebellar BOT/rest ratio was .0.9. Brunberg et al31 reported a case of CCD encountered during BOT with positron emission tomography. They suggested that the rapid development of CCD was an almost instantaneous response to diminished cortico-ponto-cerebellar and/or spinocerebellar pathway excitatory input and might be a useful indicator of remote neuronal dysfunction. In the present study, CCD was associated with the extent of cerebral hypoperfusion. CCD might be attributed to diffuse cortical hypoperfusion and be an indicator of high risk.
Conclusion Our results suggest that the redistribution of CBF might occur during the BOT, not only in the occluded side but also in the nonoccluded side. Additional study with a greater numbers of patients is required to review the clinical meanings.
203
In conclusion, quantitative analysis of SPECT images using 3DSRT was more objective and accurate than visual analysis alone because it minimized interoperator variability. The evaluation of rCBF with dual SPECT revealed the ability of cerebral collateral-flow redistribution more accurately than with asymmetry index alone during the BOT. Defining the threshold of hypoperfusion as BOT/ rest ratio ,0.9 was a sensitive predictor for developing ischemia after therapeutic occlusion of the ICA.
References 1. Brunberg JA, Frey KA, Horton JA, et al. [15O]H2O positron emission tomography determination of cerebral blood flow during balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 1994;15:725-732. 2. van Rooij WJJ, Sluzewski M, Metz NH, et al. Carotid balloon occlusion for large and giant aneurysms: Evaluation of a new test occlusion protocol. Neurosurgery 2000; 47:116-122. 3. Abud DG, Spelle L, Piotin M, et al. Venous phase timing during balloon test occlusion as a criterion for permanent internal carotid artery sacrifice. AJNR Am J Neuroradiol 2005;26:2602-2609. 4. Kurata A, Miyasaka Y, Tanaka C, et al. Stump pressure as a guide to the safety of permanent occlusion of the internal carotid artery. Acta Neurochir (Wien) 1996;138: 549-554. 5. Sugawara Y, Kikuchi T, Ueda T, et al. Usefulness of brain SPECT to evaluate brain tolerance and hemodynamic changes during temporary balloon occlusion test and after permanent carotid occlusion. J Nucl Med 2002; 43:1616-1623. 6. Peterman SB, Taylor A Jr, Hoffman JC Jr. Improved detection of cerebral hypoperfusion with internal carotid balloon test occlusion and 99mTc-HMPAO cerebral perfusion SPECT imaging. AJNR Am J Neuroradiol 1991; 12:1035-1041. 7. Simonson TM, Ryals TJ, Yuh WT, et al. MR imaging and HMPAO scintigraphy in conjunction with balloon test occlusion: Value in predicting sequelae after permanent carotid occlusion. AJR Am J Roentgenol 1992;159: 1063-1068. 8. Lorberboym M, Pandit N, Machac J, et al. Brain perfusion imaging during preoperative temporary balloon occlusion of the internal carotid artery. J Nucl Med 1996; 37:415-419. 9. Michel E, Liu H, Remley KB, et al. Perfusion MR neuroimaging in patients undergoing balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 2001;22:1590-1596. 10. Kaminogo M, Ochi M, Onizuka M, et al. An additional monitoring of regional cerebral oxygen saturation to HMPAO SPECT study during balloon test occlusion. Stroke 1999;30:407-413. 11. Sorteberg W, Sorteberg A, Lindegaard KF, et al. Transcranial Doppler ultrasonography-guided management of internal carotid artery closure. Neurosurgery 1999;45:76-87. 12. Schneweis S, Urbach H, Solymosi L, et al. Preoperative risk assessment for carotid occlusion by transcranial Doppler ultrasound. J Neurol Neurosurg Psychiatry 1997;62:485-489. 13. Linskey ME, Jungreis CA, Yonas H, et al. Stroke risk after abrupt internal carotid artery sacrifice: Accuracy of preoperative assessment with balloon test occlusion and
T. TORIGAI ET AL.
204
14.
15.
16.
17.
18.
19.
20.
21.
22.
stable xenon-enhanced CT. AJNR Am J Neuroradiol 1994; 15:829-843. Witt JP, Yonas H, Jungreis C. Cerebral blood flow response pattern during balloon test occlusion of the internal carotid artery. AJNR Am J Neuroradiol 1994; 15:847-856. Steed DL, Webster MW, DeVries EJ, et al. Clinical observations on the effect of carotid artery occlusion on cerebral blood flow mapped by xenon computed tomography and its correlation with carotid artery back pressure. J Vasc Surg 1990;11:38-43. Okudaira Y, Arai H, Sato K. Cerebral blood flow alteration by acetazolamide during carotid balloon occlusion: Parameters reflecting cerebral perfusion pressure in the acetazolamide test. Stroke 1996;27:617-621. Murphy KJ, Payne T, Jamadar DA, et al. Correlation of continuous EEG monitoring with [O-l5]H2O positron emission tomography determination of cerebral blood flow during balloon test occlusion of the internal carotid artery. Experience in 34 cases. Interv Neuroradiol 1998; 4:51-55. Liu AY, Lopez JR, Do HM, et al. Neurophysiological monitoring in the endovascular therapy of aneurysms. AJNR Am J Neuroradiol 2003;24:1520-1527. Takeuchi R, Yonekura Y, Matsuda H, et al. Usefulness of a three-dimensional stereotaxic ROI template on anatomically standardised 99mTc-ECD SPET. Eur J Nucl Med 2002;29:331-341. Umemura A, Suzuka T, Yamada K. Quantitative measurement of cerebral blood flow by 99mTc-HMPAO SPECT in acute ischaemic stroke: Usefulness in determining therapeutic options. J Neurol Neurosurg Psychiatry 2000;69:472-478. Matsuda H, Tsuji S, Shuke N, et al. Noninvasive measurements of regional cerebral blood flow using technetium99m hexamethylpropylene amine oxime. Eur J Nucl Med 1993;20:391-401. Takeuchi R, Yonekura Y, Takeda SK, et al. Fully automated quantification of regional cerebral blood flow with three-dimensional stereotaxic region of interest
23.
24.
25.
26.
27.
28.
29.
30.
31.
template: validation using magnetic resonance imaging— Technical note. Neurol Med Chir (Tokyo) 2003;43:153-162. Hays RJ, Levinson SA, Wylie EJ. Intraoperative measurement of carotid back pressure as a guide to operative management for carotid endarterectomy. Surgery 1972; 72:953-960. Tomura N, Omachi K, Takahashi S, et al. Comparison of technetium Tc 99m hexamethylpropyleneamine oxime single-photon emission tomograph with stump pressure during the balloon occlusion test of the internal carotid artery. AJNR Am J Neuroradiol 2005;26:1937-1942. Monsein LH, Jeffery PJ, van Heerden BB, et al. Assessing adequacy of collateral circulation during balloon test occlusion of the internal carotid artery with 99mTcHMPAO SPECT. AJNR Am J Neuroradiol 1991;12: 1045-1051. Aso K, Ogasawara K, Sasaki M, et al. Preoperative cerebrovascular reactivity to acetazolamide measured by brain perfusion SPECT predicts development of cerebral ischemic lesions caused by microemboli during carotid endarterectomy. Eur J Nucl Med Mol Imaging 2009; 36:294-301. Yong SW, Bang OY, Lee PH, et al. Internal and cortical border-zone infarction: Clinical and diffusion-weighted imaging features. Stroke 2006;37:841-846. Askienazy S, Lebtahi R, Meder JF. SPECT HMPAO and balloon test occlusion: interest in predicting tolerance prior to permanent cerebral artery occlusion. J Nucl Med 1993;34:1243-1245. Mathews D, Walker BS, Purdy PD, et al. Brain blood flow SPECT in temporary balloon occlusion of carotid and intracerebral arteries. J Nucl Med 1993;34:1239-1243. Erba SM, Horton JA, Latchaw RE, et al. Balloon test occlusion of the internal carotid artery with stable xenon/CT cerebral blood flow imaging. AJNR Am J Neuroradiol 1988;9:533-538. Brunberg JA, Frey KA, Horton JA, et al. Crossed cerebellar diaschisis: occurrence and resolution demonstrated with PET during carotid temporary balloon occlusion. AJNR Am J Neuroradiol 1992;13:58-61.