Intraoperative Ultrasonography Combined with Indocyanine Green Video-Angiography in Patients with Cerebral Arteriovenous Malformations Hui Wang, Zhuo-peng Ye, Zhen-chao Huang, Lun Luo, Chuan Chen, Ying Guo From the Department of Neurosurgery, the Third Affiliated Hospital, Sun Yat-Sen University, Guangzhou 510630, China.

ABSTRACT BACKGROUND AND PURPOSE: During the operation, accurately identifying the boundary of cerebral arteriovenous malformation (AVM) and discriminating between feeding arteries and draining veins is the key to successful surgical treatment of cerebral AVM. We evaluated the application of intraoperative ultrasonography (IOU) combined with intraoperative indocyanine green video-angiography (IOICGA) in the patients with cerebral AVM. METHODS: The effects of IOU combined with IOICGA on AVM surgery were observed in 12 patients with cerebral AVM. RESULTS: The lesions of cerebral AVM were completely removed in the 12 patients. IOU could clearly visualize the boundary of AVM, so no patients had massive hemorrhage caused by rupture of malformed vessels. IOU also could detect the location of deep vessels and a total of 11 deep vessels were identified in the 12 patients. IOICGA was performed 41 times altogether in the 12 patients, and 31 feeding arteries and 10 draining veins were identified, so there was no massive hemorrhage caused by misjudgment of feeding arteries or draining veins. CONCLUSIONS: IOU combined with IOICGA can identify the boundary of AVM, detect deep vessels, and discriminate between feeding arteries and draining veins, reducing operation difficulty, decreasing mortality and disability rate, and increasing the rate of complete excision.

Keywords: Intraoperative ultrasonography, intraoperative indocyanine green video-angiography, cerebral arteriovenous malformation, microsurgery. Acceptance: Received August 23, 2014, and in revised form January 23, 2015. Accepted for publication January 29, 2015. Correspondence: Address correspondence to Ying Guo, Department of Neurosurgery, the Third Affiliated Hospital, Sun Yat-Sen University, No. 600, Tianhe Road, Guangzhou 510630, China. E-mail: [email protected]. Hui Wang, Zhuo-peng Ye, and Zhen-chao Huang equally contributed to this study. J Neuroimaging 2015;25:916-921. DOI: 10.1111/jon.12232

Introduction Although microsurgery, endovascular embolization, stereotactic radiosurgery and combined therapy have been used in the treatment of cerebral arteriovenous malformation (AVM), surgical excision is still the most important therapeutic method. During the operation, accurately identifying the boundary of AVM and discriminating between feeding arteries and draining veins is the key to successful surgical treatment of AVM. From January 2011 to January 2014, 12 patients with cerebral AVM underwent surgical treatment assisted by intraoperative ultrasonography (IOU) and intraoperative indocyanine green video-angiography (IOIGCA), obtaining satisfied outcomes.

Patients and Methods All study methods were approved by Institutional Review Board and Ethics Committee of the Third Affiliated Hospital, Sun Yat-Sen University. All the subjects or their guardians enrolled into the study gave written formal consent to participate.

Surgical Procedures

Patients From January 2011 to January 2014, 12 patients with cerebral AVMs receiving surgical treament in our hospital were consecutively enrolled in this study. Of the 12 patients, 7 were men 916

and 5 women, with a mean age of 33.5 years (range 11-65). The first symptoms of cerebral AVMs included intracerebral hemorrhage in 8 patients (the intracerebral hemorrhage occurred in the frontal lobe in 2 patiens, the occipital lobe in 3 patients, the temporal lobe in 1 patient and the cerebellum in 2 patients; the range of hemorrhage was between 2 × 3 × 2 cm and 4.5 × 3.5 × 5 cm), seizures in 3 patients and intermittent headache in 1 patient. Preoperative Glasgow Coma Scale (GCS) score was 15 in 6 patients, 14 in 2 patients, 12 in 2 patients, 10 in 1 patient and 9 in 1 patient. Lesions were located in frontal lobe in 5 patients, occipital lobe in 4 patients, temporal lobe in 1 patient and cerebellum in 2 patients. The mean diameter of lesions was 41.2 mm (range 21-66). Three patients had Spetzler-Martin grade II AVM, 6 grade III AVM and 3 grade IV AVM. All patients underwent computed tomography angiography (CTA), magnetic resonance angiography (MRA) plus magnetic resonance venography (MRV), and digital subtraction angiography (DSA) after admission.

Copyright

The surgical procedures were as follows: (1) Patients underwent tracheal intubation under general anesthesia, and then an incision was made depending upon the location of

◦ 2015 by the American Society of Neuroimaging C

Fig 1. Large circuitous malformed vessel mass with a uniform color after opening the dura mater.

Fig 2. Visualized arteries in early phase of intraoperative indocyanine green video-angiography. Arrows indicate arteries. AVM lesion. After removing an appropriate bone flap, the dura mater was opened to expose AVM lesions (Fig 1). (2) After 25 mg of indocyanine green (ICG) was intravenously injected, IOICGA was performed followed by observation under a surgical microscope (Leica OH-4, FL800 fluorescence camera, Leica Microsystems Heidelberg GmbH, Germany). ICG video angiograms were recorded and used to discriminate between feeding arteries and draining veins. Generally, feeding arteries are first visualized followed by malformed vessels or draining veins (Figs 2–4). (3) After IOICGA examination, Color Doppler ultrasound diagnostic apparatus (Prosound SSD 3500, Aloka, Japan) combined an avariable angle-linear array probe (UST-5546, Aloke, Japan) was used in the operation. The probe had a frequency of 5-10 MHz, a maximum detection depth of 24 cm, an area of .8 cm × 4 cm, 176 db dynamic range, a scanning frame rate of 95

and a color display frame rate of 34 f/s. After an aseptic plastic film smeared with coupling medium was put on the probe, color Doppler ultrasonic continuous scanning was performed on the surface of lesions to detect the boundary, range and depth of lesions (Fig 5). During IOU, the surface of the probe was washed with sterile physiological saline, and then parameters such as the size of sampling frame, focal depth, rate of repetitive pulse and chroma gain were adjusted to obtain good-quality images. (4) Starting from the feeding artery away from the functional area, the lesion was separated at 1-2 mm away from the boundary of AVM under ultrasound guidance. The identified feeding arteries should be occluded by electrocoagulation. During the separation, IOICGA might be repeatedly performed to discriminate between feeding arteries and draining veins, which was conducive to complete separation for the shallow focus.

Wang et al: Cerebral Arteriovenous Malformation

917

Fig 3. Visualized malformed vessel mass in late phase of intraoperative indocyanine green video-angiography. Arrows indicate arteries.

Fig 4. Feeding arteries under general condition. Arrows indicate arteries same as that in Figure 2. (5) The deep portion of lesion was separated following the separation of the superficial portion. During the separation, deep vessels were identified based on preoperative imaging and intraoperative ultrasonography. Then, these vessels were exposed followed by IOICGA (Fig 6). The feeding arteries were occluded, while the draining veins were temporarily retained. After other deep vessels were treated with the same method, the lesion was completely separated. (6) Afterwards, all draining veins were respectively occluded by electrocoagulation, and then the entire AVM was removed. (7) After the surgery, thorough hemostasis was achieved without application of any packing material. Saline was injected into the lesion cavity, and then ultrasonography showed no residual lesions. The cranial window was closed using routine procedures.

918

All the procedures were performed by the same one surgeon.

Results Surgical Outcomes and Prognosis of Patients The 12 patients were diagnosed with AVM by postoperative pathology. Postoperative images also confirmed that AVMs were completely removed in the 12 patients. No patients died from surgical treatment. However, 2 patients had contralateral hemiparesis and 1 hydrocephalus after operation. The patient with hydrocephalus recovered after ventriculoperitoneal shunt therapy. The follow-up period was between 6 and 27 months.

Journal of Neuroimaging Vol 25 No 6 November/December 2015

Fig 5. (A)-(F) The boundary, range and depth of the cerebral arteriovenous malformation showed by multilevel and multidimensional ultrasonography.

Fig 6. During separation, IOU can localize deep vessels; and IOICGA can discriminate between feeding arteries and draining veins. (A) The arrow indicates a deep vessel visualized by IOU; (B) in the preoperative imaging data, the arrow indicates the origin of the deep vessel shown in (A); (C) in a surgical field, the arrow indicates the exposed deep vessel shown in (A); (D) during IOICGA, imaging is visualized earlier in the exposed deep vessel (vertical arrow) than in malformed vessel mass (horizontal arrow), suggesting that it is a feeding artery (as shown by arrows). Notes: IOU = intraoperative ultrasonography; IOICGA = intraoperative indocyanine green video-angiography.

Of the 12 patients, 1 patient lost follow-up, and other 11 patients were able to look after themselves. Of the 11 patients, 10 returned to work. Postoperative Glasgow Outcome Scale (GCS) score was 5 in 9 patients and 4 in 2 patients, and 1 patient lost follow-up.

IOU Under gray scale, AVM showed heterogeneous strong echo without clear boundary from normal brain tissue.

Intraoperative color Doppler ultrasonography was able to detect the boundary and range of AVM. Under color Doppler, AVM showed irregular and multicolored image with clear boundary. The judgment of direction of blood flow depended on vascular color. The blood in the red vessels flowed toward the probe, and the blood in the blue vessels flowed toward the opposite direction. The tortuous and thick vessels in which blood flowed toward the vascular mass were regarded as feeding arteries. The huge and dilated vessels in which there was red and blue turbulent eddy, and blood flowed away from Wang et al: Cerebral Arteriovenous Malformation

919

the vascular mass were regarded as draining veins. In this study, lesions were separated at 1-2 mm away from the boundary of AVM under ultrasound guidance. During the operation, no patients had massive hemorrhage caused by rupture of malformed vessels. IOU was also able to identify the position of deep vessels in AVMs, facilitating exposure of these deep vessels. In the 12 patients, a total of 11 deep vessels were identified with IOU.

IOICGA Based on the speed of imaging, feeding artery and draining vein could be clearly discriminated by IOICGA either in superficial vessels or in exposed deep vessels. IOICGA was performed 41 times altogether in the 12 patients, and 31 feeding arteries and 10 draining veins were found, so there was no massive hemorrhage caused by misjudgment of feeding arteries or draining veins.

Discussion IOU Combined with IOICGA Improving the Efficacy of Surgery for AVM Although the therapeutic methods for cerebral AVM include surgical excision, endovascular embolization, stereotactic radiosurgery and combined therapy, microsurgery is the most effective for AVM.1,2 Preoperative DSA, CTA, and MR can provide information about AVM lesions including the size, location, hemodynamics, feeding artery and draining vein, which helps surgeons to make a detailed microsurgical protocol.3 However, large cerebral AVM is still a great challenge for neurosurgeons because it is quite difficult to determine the boundary of AVM lesion and discriminate between feeding arteries and draining veins.4 Currently, the solution of these two difficult problems mainly depends on the experience of surgeons, and experience easily leads to misjudgment. Nevertheless, IOU combined with IOICGA can resolve the two problems, reducing operation difficulty, decreasing the incidence of massive hemorrhage, and increasing the rate of complete excision of AVM lesion.

Application of IOU in AVM Microsurgery Complete excision is an essential requirement for AVM, and the best course to separate AVM is located at 1-2 mm away from the boundary of AVM. In case of rupture of malformed vessels, uncontrollable hemorrhage may occur, bringing extremely dangerous outcomes to patients.5 Therefore, it is necessary to identify and determine the boundary of AVM lesions. Endovascular navigation is conducive to the localization of AVM lesions to some extent, but it fails to achieve precise real-time localization of AVM lesions due to inevitable brain shift during operation. On color Doppler ultrasonography, however, AVM lesion can be visualized as a colorful vessel mass which can be easily identified.6,7 Therefore, the boundary of lesion can be clearly determined by multilevel and multidimensional ultrasonography. In additionally, ultrasonography can be performed repeatedly during operation to get the real-time information regarding AVM lesion. Compared with complex endovascular navigation, ultrasonography has plenty of advantages and is not affected by brain shift. In this study, we first performed repeated multidimensional IOU to determine the boundary of AVM lesion, and then started to separate from the feeding artery away from the functional area. In the 12 patients, no massive 920

hemorrhage caused by misjudgment of the boundary of AVM lesions occurred. In addition, IOU also can localize deep vessels.8 After separation of the superficial portion of AVM, deep portion should be further separated. At this time, besides judgment of lesion boundary, identifying the location of deeper vessels is also very important. Severe outcomes may occur if deep draining veins are prematurely occluded. IOU can help us easily to find and expose deep vessels. In the 12 patients, a total of 11 deep vessels were identified with IOU, no deep vessels were occluded in error. In summary, the advantages of IOU include: (1) clear visualization of the boundary and range of lesions; (2) good visualization of deep vessels; (3) repeated use when necessary. But IOU can not discriminate between feeding arteries and draining veins. Discriminating between feeding arteries and draining veins is another key to successful surgical treatment of AVM. IOICGA, an adjuvant technique, can discriminate between feeding arteries and draining veins.

Application of IOICGA in AVM Microsurgery AVM excision should be performed in the following order: (1) occlusion of feeding artery; (2) separation of malformed vessels; and (3) occlusion of draining veins. Therefore, it is important to identify feeding arteries and draining veins during operation.9 In January 2003, Raabe et al10 used ICG in intraoperative cerebral angiography for the first time and found that this technique had important significance for the removal of AVM lesions. In intraoperative cerebral angiography, ICG, an imaging agent, can bind tightly to plasma proteins; the conjugation product emits fluorescence under near-infrared light. The collected images can reflect the dynamic changes of blood circulation. Currently, ICG-angiography is mainly used to observe the direction of blood flow and the degree of vascular patency.11 Therefore, ICG-angiography can accurately discriminate between feeding arteries and draining veins for superficial vessels or exposed deep vessels.12 In the 12 patients of this study, IOICGA was performed 41 times, and 31 feeding arteries, and 10 draining veins were identified. There was no massive hemorrhage caused by misjudgment of feeding arteries or draining veins. Therefore, we concluded the advantages of IOICGA as follows: (1) discriminating between feeding arteries and draining veins for superficial vessels or exposed deep vessels; (2) directly observing blood flow velocity and turbulence in vessels; (3) repeated use when required. However, IOICGA can not discriminate deep unexposed or hidden vessles.13 In conclusion, IOU combined with IOICGA is of great assistance to surgeons in identifying vessels and boundary of lesions, reducing operation difficulty. The procedures for the removal of AVM are summarized as follows: (1) after opening the dura mater, IOICGA is first performed to discriminate between feeding arteries and draining veins for the superficial vessels, and then IOU is used to identify the boundary of AVM lesions; (2) separation should be started from the feeding artery located in non-functional area, and the course of deep vessels is detected with IOU in order to carry out the separation of deep portion of AVM lesions; (3) for exposed deep vessels, ICGA may be used to discriminate between feeding arteries and draining veins; (4) the above mentioned procedures can be alternately performed until lesion is completely removed; (5) finally, IOU

Journal of Neuroimaging Vol 25 No 6 November/December 2015

combined with IOICGA are used to check whether there is residual lesions still. However, there are some limitations in this study. For example, the sample in this study was small, and most AVM in this study were superficial and large. How to remove deep cerebral AVMs and achieve good therapeutic effects remains to be further investigated. This study was supported by grants from the National Natural Science Foundation of China (30901542), Chinese Ministry of Education for young teacher training of Sun Yat-Sen University(12ykpy44), by Science and Technology Project from Guangzhou City (12c002061756) and by Science and Technology Project from Guangdong Province (20120314).

References 1. Chang SD, Marcellus ML, Marks MP, et al. Multimodality treatment of giant intracranial arteriovenous malformations. Neurosurgery 2003;53:1-13. 2. Englot DJ, Young WL, Han SJ, et al. Seizure predictors and control after microsurgical resection of supratentorial arteriovenous malformations in 440 patients. Neurosurgery 2012;71:572-80. 3. Graham GD. Arteriovenous malformations in the brain. Curr Treat Options Neurol 2002;4:435-44. 4. Novakovic RL, Lazzaro MA, Castonguay AC, et al. The diagnosis and management of brain arteriovenous malformations. Neurol Clin 2013;31:749-63.

5. Toulgoat F, Lasjaunias P. Vascular malformations of the brain. Handb Clin Neurol 2013;112:1043-51. 6. Zhang Y, Wang S, He W. Intraoperative Doppler ultrasonography and ultrasound angiography in operation for brain arteriovenous malformations. Zhonghua Yi Xue Za Zhi 2008;88:2461-4. 7. Dempsey RJ, Moftakhar R, Pozniak M. Intraoperative Doppler to measure cerebrovascular resistance as a guide to complete resection of arteriovenous malformations. Neurosurgery 2004;55:15561. 8. Akdemir H, Oktem S, Menku¨ A, et al. Image-guided microneurosurgical management of small arteriovenous malformation: role of neuronavigation and intraoperative Doppler sonography. Minim Invasive Neurosurg 2007;50:163-9. 9. van Beijnum J, van der Worp HB, Buis DR, et al. Treatment of brain arteriovenous malformations: a systematic review and metaanalysis. J Am Med Acad 2011;306:2011-9. 10. Raabe A, Beck J, Gerlach R, et al. Near-infrared indocyanine green video angiography: a new method for intraoperative assessment of vascular flow. Neurosurgery 2003;52:132-9. 11. Balamurugan S, Agrawal A, Kato Y, et al. Intra operative indocyanine green video-angiography in cerebrovascular surgery: An overview with review of literature. Asian J Neurosurg 2011;6:8893. 12. Chen SF, Kato Y, Oda J, et al. The application of intraoperative near-infrared indocyanine green videoangiography and analysis of fluorescence intensity in cerebrovascular surgery. Surg Neurol Int 2011;2:42. 13. Ng YP, King NK, Wan KR, et al. Uses and limitations of indocyanine green videoangiography for flow analysis in arteriovenous malformation surgery. J Clin Neurosci 2013;20:224-32.

Wang et al: Cerebral Arteriovenous Malformation

921

Intraoperative Ultrasonography Combined with Indocyanine Green Video-Angiography in Patients with Cerebral Arteriovenous Malformations.

During the operation, accurately identifying the boundary of cerebral arteriovenous malformation (AVM) and discriminating between feeding arteries and...
779KB Sizes 0 Downloads 6 Views