Application of 3.0T Intraoperative High-Field Magnetic Resonance Imaging Guidance for the Surgery of Arteriovenous Malformation Within Eloquent Areas.
Fengping Zhu,Yanlong Tian,Wei Zhu,Yuxiang Gu,Bin Xu,Jinsong Wu,Ying Mao,Liangfu Zhou
DOI: https://doi.org/10.3760/cma.j.issn.0366-6999.20131599
IF: 6.133
2014-01-01
Chinese Medical Journal
Abstract:Arteriovenous malformations (AVMs) are vascular abnormalities that consist of multiple fistulous connections between arteries and veins without the normal intervening capillary bed. Microsurgical resection is regarded as the gold standard treatment, as it can achieve complete elimination of AVM. However, it may be prone to postoperative complications, including fatality and neurological deficits, especially in cases involving mild or no neurological deficits.1 Therefore, an appropriate balance is necessary between the extent of removal of AVM and functional preservation in the resection of AVM within eloquent brain areas. Intraoperative MRI (iMRI), which provides real-time images for neuronavigation to compensate for errors caused by brain tissue deformation and brain shift, has been rapidly developing and widely applied in neurosurgery. However, the role of iMRI in the surgical treatment of AVMs remains controversial. In this study, we report on our preliminary clinical experiences with 3.0 T iMRI-guided surgeries in seven cases of AVM to determine whether the use of iMRI has a favorable impact on the surgical treatment of such vascular malformations. METHODS Patient population Patients with AVM in eloquent functional brain areas that necessitated iMRI-guided surgery from January 2012 to June 2012 were included. The diagnosis of AVM was made by T2-weighted images and confirmed by six-vessel digital subtraction angiography (DSA). Endovascular and surgical therapeutic options were evaluated based on the clinical and angiographic presentation of the AVM, with special attention to venous draining patterns and AVM localization. Individuals with any contraindications precluding iMRIguided surgery (e.g., those with implanted devices such as cardiac pacemakers, artificial heart valves, cochlear implants or other ferromagnetic implants) were excluded. All patient data, including AVM localization, initial symptoms, nidus size, Spetzler and Martin grading, and clinical neurological outcome, were recorded. The study protocol was approved by the Institutional Research and Ethics Committee of Huashan Hospital. Informed consent was obtained from all patients. iMRI series Each iMRI study was performed with a mobile MRI system (iMRIS, Innovative Magnetic Resonance Imaging Systems, Inc., Winnipeg, MB, Canada) interfaced to a 3.0 T MR System console (Verio, Siemens Medical Systems, Erlangen, Germany). Moreover, blood-oxygen-leveldependent functional MRI (BOLD-fMRI) and diffusion tensor imaging (DTI) series were performed in selected patients following the aforementioned protocol, while the following iMRI sequences were performed in all cases.2 Fluid attenuated inversion recovery (FLAIR) images were acquired using an axial turbo spin echo (TSE) pulse sequence (TR, 7600 ms; TE, 96 ms; TI, 900 ms; flip angle of 9 degrees; 60 slices; slice thickness of 2 mm; matrix size of 256×180; FOV of 240 mm × 240 mm; and voxel size of 1.3 mm × 0.9 mm × 2.0 mm). TOF MRA scans (4 slabs; 40 slices per slab; TR, 22 ms; TE, 4.2 ms; flip angle of 18 degrees; matrix size of 365×384; slice thickness of 0.5 mm; and FOV of 181 mm × 200 mm) were obtained over 8 minutes and 32 seconds. Surgical procedure and intraoperative guidance For the AVMs that were adjacent to the functional areas, we also used intraoperative neurophysiological monitoring (IONM) to preserve neurological function during the surgery. After anesthetization, the patient was positioned supine on the operation room (OR) table (iMRIS, Inc., Canada) and fixed with a three-pin, MR-safe carbon headholder (DORO Radiolucent Headrest System, Pro Med Instruments GmbH, Germany). After all MRI-unsafe surgical instruments and tables were moved outside the 5-G magnetic field and the safety checklist was completed, the radio frequency (RF) shielding door was opened, and the movable magnet was transferred into the OR. After iMRI series scanning was finished, the magnet was moved back to the diagnostic room (DR). The postprocessing (including 3D TOF-MRA, BOLD-fMRI and DTI tractography) procedure was performed at the workstation (Syngo MultiModality Workplace, Siemens Healthcare, Siemens AG, Germany) for further use in the neuronavigation system. Integrated data sets for both FLAIR and MRA maps by rigid registration were used for intraoperative surgical guidance. In all cases, the StealthStation Treon neuronavigator (Medtronic Sofamor Danek Co., Minneapolis, MN, USA) was used. The head position of each patient was registered by light-emitting diodes. The craniotomy was designed according to the location of AVM as determined on plain angiographic images and MRI sequences and guided by the neuronavigation system. The angioarchitecture of the AVM, including feeding arteries, nidus, and draining veins, was visualized via intraoperative MRA when the suspected pathological vessels were visible. The spatial relationship between AVM and surrounding brain tissues, especially the functional cortex and subcortical neural pathways, was delineated by BOLD-fMRI and DTI neuronavigation. Once AVM resection was performed as planned, the patients were prepared for the second iMRI. If continued resection was necessary, the latest iMRI sequences were loaded to update the navigational MRIs. When AVM resection was deemed to be complete and performed as safely as possible on the basis of intraoperative visualization, iMRI neuronavigation, and IONM, AVM resection was ceased. RESULTS The patients in this case series ranged in age from 23 to 58 years (mean age = 34.1 years), including three males and four females (Table 1). In two cases, preoperative endovascular treatment was attempted but was only partially successful. Microsurgical resection of the AVM was successful in all cases. No technical difficulties related to the navigation system occurred during the surgical procedures. iMRI provided high-quality images in each case. Intraoperative MRA series was successfully integrated with FLAIR images in each subject, which allowed the 3D reconstruction of the AVM and adjacent brain tissues. FLAIR images demonstrated brain anatomic features with visible shifts induced by surgery. MRA series clearly illustrated the supplying arteries and nidus of the AVM. BOLD-fMRI and DTI tractography neuronavigation demonstrated the proximity of the AVM to the language, motor, and visual cortex areas, as well as the subcortical pathways, thereby enhancing preoperative surgical planning and intraoperative surgical guidance. Intraoperative integrated MRI and MRA series revealed complete elimination of the AVM, meanwhile, DTI tractography delineated the subcortical motor pathway, which was well preserved (Figure 1). All patients recovered well without any neurological deficits. No further complications related to intraoperative imaging (i.e., infection, intracranial hematoma, anesthetic accident, and/or scalp burn) were recorded. Postoperative DSA was performed in all patients, which confirmed the intraoperative findings. No falsepositive or false-negative results were observed.Table 1: Clinical data for patients (n=7)Figure 1.: Representative case of a diffuse arteriovenous malformation in the left frontal lobe. A and B: Preoperative 3D TOF-MRA of the AVM, showing the angioarchitecture of the AVM (A: anteroposterior view; B: transverse view). C: Preoperative integrated constructed 3D image of language Bold-fMRI and DTI tractography, demonstrating the spatial relationship between the lesion and language cortex (orange spots), as well as the subcortical motor pathway. D and E: Intraoperative 3D TOF-MRA of the AVM, showing the complete elimination of the AVM (D: anteroposterior view; E: transverse view). F: Intraoperatively constructed 3D image of DTI tractography, demonstrating the obvious brain shift and well-preserved motor pathway.DISCUSSION Current treatment of AVMs includes microsurgical resection, stereotactic radiosurgery, endovascular embolization, combined use of the aforementioned treatment modalities, and observation only.1 With the development of microscopy and surgical techniques, some high-grade AVMs or those within eloquent functional areas are referred for microsurgical resection.1 Despite these advances, a large number of patients still suffer from postoperative intracranial bleeding caused by incomplete resection.1 The incidence of residual nidus has been reported to be as high as 19% when intraoperative assessment was limited to the judgment of a surgeon. As such, real-time intraoperative monitoring with AVM surgery is highly recommended to reduce postoperative complications. There are various intraoperative visualization techniques for AVM surgery, including intraoperative DSA, ultrasonography, and indocyanine green (ICG) angiography. Although DSA is considered as the “gold standard” for visualizing vascular anatomy, the use of intraoperative DSA is limited because it requires a specified angiography team, as well as exposes the patient and medical personnels to radiation. Ultrasonography is a useful and easy tool that does not require interruption of the surgery. However, ultrasonography is associated with low spatial resolution, thereby detecting only superficial architecture and potentially missing residual nidus. Intraoperative ICG angiography is vulnerable to surrounding brain tissues, nerves, vessels and operative instruments. It also requires injection of foreign materials, which may have severe side effects. iMRI in neurosurgery has been mainly used to assess tumor residue and functional neuronavigation. Few studies have reported the use of this approach for neurovascular surgery.3,4 Sakurada et al3 reported a case in which the utilization of iMRI helped to confirm complete removal of a left occipital AVM. 3.0 T iMRI has obvious advantages over 1.5 T iMRI in the structural and functional imaging of central nervous system, including faster imaging, thinner layers, and higher-resolution neurovascular images. In this case series, high-resolution TOF MRA clearly depicted morphologic features and vessels, including feeding arteries, nidus, and draining veins. Five patients underwent BOLD-fMRI and DTI tractography functional neuronavigation to locate functional cortices and subcortical neural pathways, which were requisite to critical functional preservation. Compared with other studies, the fact that all patients achieved total resection and good neurological outcome might demonstrate the benefits of combined use of 3T iMRI and IONM. Because of high-flow abnormal vessels and nidus found in AVM architecture, the neurovascular surgeon cannot quit and leave a piece, as one would with tumor or epilepsy surgery. However, with knowledge of the adjacent critical brain tissues, the surgeon can approach the AVM with more confidence. While showing promise, this study has certain limitations. First, the number of cases was limited, future prospective studies involving large sample sizes of cases with AVM of different Spetzler-Martin grades should be performed to more objectively assess benefits and limitations of these techniques. Whether 3.0 T iMRI is much more beneficial than 1.5 T iMRI for surgical resection of AVM within eloquent areas should be also well studied. The second limitation was associated with the nature of TOF-MRA. Because the flow velocity in the pathological draining veins of AVM is high, TOF-MRA with a 3.0 T unit can visualize most information regarding the angioarchitecture of an AVM with high quality. However, the nidus and tiny vessels with low flow velocity may not be depicted. High-resolution contrast-enhanced time-resolved TOFMRA or its integration with DSA may be helpful for the visualization of even the tiny architecture of AVM. Meanwhile, with the development of MRI techniques, four-dimensional time-resolved MRA has the potential to replace DSA, given its efficacy in the diagnosis across all grades of AVM.5