Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Chinese Medical Journal 2012;125(): 1 Original article Clinical experience of 3T intraoperative magnetic resonance imaging integrated neurosurgical suite in Shanghai Huashan Hospital QIU Tian-ming, YAO Cheng-jun, WU Jin-song, PAN Zhi-guang, ZHUANG Dong-xiao, XU Gen, ZHU Feng-ping, LU Jun-feng, GONG Xiu, ZHANG Jie, YANG Zhong, SHI Jian-bin, HUANG Feng-ping, MAO Ying and ZHOU Liang-fu Keywords: intraoperative magnetic resonance imaging; glioma; pituitary adenoma Background Intraoperative magnetic resonance imaging (iMRI) dates back to the 1990s and has been successfully applied in neurosurgery but they were low-field iMRI (<1.0T). This paper reports the clinical experience with a 3T iMRI-integrated neurosurgical suite in Huashan Hospital, Shanghai, China. Methods From September 2010 through March 2012, 373 consecutive patients underwent neurological surgery under guidance with 3T iMRI. A retrospective analysis was conducted regarding clinical efficiency. Results All surgery in the 373 patients was safe. The ratio of gross total resection for cerebral gliomas (n=161) was increased from 55.90% to 87.58%. The ratio of benefit in extent of resection was 39.13%. One hundred and fifty eight of the 161 glioma patients accomplished follow-up at 3 months postoperatively. Twenty of 161 patients (12.42%) suffered from early motor deficit after surgery. Late motor deficit was however observed in five of 158 patients (3.16%). Twenty-one of 161 patients (13.04%) had early speech deficit and late speech deficit was only observed in six of 158 patients (3.8%). The ratio of gross total resection for pituitary adenomas (n=49) was increased from 77.55% to 85.71%. The ratio of benefit in extent of resection was 10.2%. There were no iMRI-related adverse events even for patients who underwent awake craniotomy. Conclusion The 3T iMRI integrated neurosurgical suite provides high-quality intraoperative structural functional and functional imaging for real-time tumor resection control and accurate functional preservation, resulting in an improvement in maximal safe brain surgery. Chin Med J 2012;125(): T he emergence of image-guided neurosurgery has drastically revolutionized neurosurgical procedures, yet brain shift is one of its limitations, decreasing both accuracy and safety of surgery.1,2 The introduction of intraoperative magnetic resonance Imaging (iMRI) and various pioneering efforts in the 1990s has led to development of minimal invasive neurosurgery.3 Between 2006 and 2012, 0.15 T PoleStar™ N-20 iMRI-guided neurosurgery was performed >400 times at Huashan Hospital in Shanghai, with a satisfactory outcome. In 2009, People’s Liberation Army General Hospital in Beijing began working with 1.5 T iMRI, which was the first neurosurgical institute in China to adapt high-field iMRI.4 The utilization of a 3.0 T iMRI at Huashan Hospital in 2010 prompted us to establish the iMRI Integrated Neurosurgical Suite at Huashan Hospital, ushering in a new digital era of minimally invasive neurosurgery, which also became the foundation of precision neurosurgery at our institute. Between September 2010 and March 2012, a total of 373 procedures of consecutive iMRI-guided neurosurgery were performed at our institute, and we take this opportunity to share our experiences with this state-of-the-art technology. METHODS Patient population Patients were selected based on the following inclusion criteria: (1) age ≥4 years; (2) potential benefit from the use of iMRI as indicated by preoperative MRI such as lesions nearby or locating at eloquent areas; and (3) no contraindications according to the security screening table for iMRI. The exclusion criteria included: (1) age <4 years; (2) contraindications for MRI scanning according to the security screening table for iMRI; and (3) patient refusal. Written informed consent was obtained from each patient. This study was undertaken at the Huashan DOI: 10.3760/cma.j.issn.0366-6999.2012.01.001 Glioma Surgery Division, Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai 200040, China (Qiu TM, Yao CJ, Wu JS, Pan ZG, Zhuang DX, Xu G, Zhu FP, Lu JF, Gong X, Zhang J, Yang Z, Shi JB, Huang FP, Mao Y and Zhou LF) Correspondence to: Dr. ZHOU Liang-fu, Department of Neurosurgery, Huashan Hospital, Fudan University, Shanghai 200040, China (Tel: 86-21-52887206. Fax: 86-21-62492884. Email: [email protected]) This study was supported by grants of the Ministry of Health of China (2010-2012), National Natural Science Foundation of China (No. 81071117, No. 81171295), and Shanghai Municipal Health Bureau (No. XBR2011022). Conflicts of interest: none. Chin Med J 2012;125(): 2 Hospital (Shanghai, China) with approval from the Huashan Institutional Review Board. 3T iMRI integrated neurosurgical suite The 3T iMRI integrated neurosurgical suite by IMRIS (Winnipeg, MB, Canada) includes an intraoperative room (OR) and a diagnostic room (DR); the core equipment in which is a 1.73-m moveable 3.0 T magnet with a 70-cm working aperture (MAGNETOM Verio; Siemens Healthcare, Erlangen, Germany). The magnet incorporated into a ceiling mount enables the scanner to move in and out of the OR, instead of moving the operating table. The IMRIS Matrix surgical information management system integrates all the information in the suite including: the picture archiving and communication system (PACS), images for neuronavigation, videos for the operating microscope, videos for neuroendoscopy, videos for the operating field monitor, OR full-viewing monitor, signals for electrocardiography (ECG) monitor, signals for the neurophysiological monitor, and other digital information. This suite covers an area >300 m2 consisting of OR (72 m2), DR (48 m2), iMRI control room (23 m2), post-anesthesia care unit (PACU) (10 m2), data processing room (9 m2), and patient locker room (25 m2) (Figure 1). The OR and DR are both shielded from radiofrequency (RF) radiation (Figure 1). OR equipment comprises: (1) OR table (IMRIS) constructed from high-grade, MR-compatible stainless steel and composites, and can be rotated 90° from the magnetic field axis; (2) MR-safe carbon fiber head holder (DORO Radiolucent Headrest System; PRO-MED Instrumente GmbH, Tuttlingen, Germany); (3) head RF coil (IMRIS) matched to the head holder, separated into two components with four channels each, the bottom of which is left beneath the sterile field during surgery; (4) MRI-safe anesthetic apparatus (Aestiva/5 MRI, GE Healthcare, Wauwatosa WI, USA) and ECG monitor (MAGLIFE C Plus, Schiller AG, Baar, Switzerland); (5) ceiling-mounted neuronavigation system (TRIA i7; Medtronic Navigation, Inc., Minneapolis, MN, USA); (6) operating microscope with high-digital video and navigation system (OPMI Pentero; Carl Zeiss AG, Jena, Germany); (7) Multifunctional Neurological Workstation (Epoch XP; AXON, New York, NY, USA); and (8) Multimedia Interactive System for awake surgery manufactured by ourselves. Preoperative and postoperative evaluation For all patients, history, preoperative and postoperative neurological signs were recorded. Muscle strength and language function were evaluated before surgery and discharge. All patients were followed up for 3 months after surgery. Glioma patients with lesions adjacent to the language cortex underwent language function assessment using the Aphasia Battery of Chinese (ABC),5 which is the Chinese standardized adaptation of the Western Aphasia Battery. Speech deficit was defined by language function deterioration compared to preoperative status according to ABC. Motor deficit was used to describe those patients whose muscle strength decreased postoperatively. Early deficits referred to those existing before discharge and late deficits to those deficits found 3 months after surgery. MRI acquisition and volumetric analysis Preoperative MR images were obtained in the DR 1 day prior to surgery and postoperative images were obtained within 3 days after surgery. The MRI sequences included a contrast-enhanced, 3D, magnetization-prepared, rapid-gradient echo (MPRAGE) sequence, fluid-attenuated inversion recovery (FLAIR) sequence; diffusion tensor imaging (DTI); blood oxygen-level-dependent functional MRI (BOLD fMRI); time-of-flight MR angiography (MRA); multi-voxel MR spectroscopy (MRS); and MR perfusion images. The motor or language pathways and activation of the motor or language areas were then reconstructed into 3D objects and integrated into the 3D structural data using a postprocessing workstation (Syngo MultiModality Workplace; Siemens AG). Figure 1. Plane chart of iMRI integrated neurosurgical suite in Huashan Hospital. For high-grade glioma (HGG) or pituitary adenoma, 3D MPRAGE images (plus gadolinium) were used for volume determination before and after resection. FLAIR was considered for low-grade Chinese Medical Journal 2012;125(): glioma (LGG). All pre-, intra- and postoperative tumor segmentations were manually performed using the OSIRIS software tool (Pixmeo, Switzerland) by a neurosurgeon and were verified by an additional neurosurgeon. The volumes of original or residual tumor were measured. In this case series, gross total resection (GTR) was defined as 100% resection of the tumor volume. Intraoperative procedure The full-time team working in the 3T iMRI integrated surgical suite includes six neurosurgeons, three nurses, one neuronavigation technician, two anesthetists, one radiologist, one radiology technician, and one OR worker. All staff and patients were required to sign the security screening table for iMRI. After endotracheal intubation with intravenous anesthesia, the OR table was rotated 90° to the magnetic field axis. The patient’s head was fixed in position using a carbon fiber head holder, integrated with the bottom part of the eight-channel coil and navigation reference frame. The navigation technician registered the patient’s head space to the MR images in the navigation system. After craniotomy, we used functional neuronavigation and electrophysiological techniques to localize the eloquent brain cortex. After tumor resection, all instruments and tables, with the exception of MRI-safe equipment, were withdrawn by the nurses beyond the 5-G boundary line. After draping the patient, the top part of the coil was matched to the bottom part. Then the neurosurgeons, nurses and radiology technician completed the MRI safe checklist together. The RF shielding door was opened and the scanner was moved into the OR from the DR. When the scanner was moved, all staff in the OR were required to ensure the safety of the patient until his/her head was situated in the center of the working aperture. All medical personnel then left the OR and closed the RF shielding door. During intraoperative MRI, the signals from the anesthetic apparatus and ECG monitor were transmitted wirelessly to the control room. When intraoperative MRI acquisition was finished, the RF shielding door was reopened. The wound was re-draped. If further resection were indicated, the navigation technician would update the navigation using the real-time intraoperative MR images. RESULTS There were 373 cases, including 125 male and 248 female, aged from 5 to 78 years (mean: 45.5 years). When classified by surgical approach there were: 288 craniotomies, 57 trans-sphenoidal approaches, 25 biopsies, and three neuroendoscopic surgery. When classified by pathology there were: 203 gliomas, 55 pituitary adenomas, 20 meningiomas, 19 moyamoya disease, and 76 others. When classified by main lesion location there were: 146 cases in the left hemisphere, 76 in the right hemisphere; 79 in the saddle area, 87 in the frontal lobe, 46 in the temporal lobe, 32 in the parietal lobe, 22 in the insular lobe, 17 in the occipital lobe, and 3 14 in the basal ganglia area. When classified by anesthesia pattern there were: 339 with general anesthesia, and 34 with awake craniotomy (Table 1). All 373 patients received 1–4 intraoperative scans (mean: 1.88) at OR, and during the same time span a total of 6500 diagnostic MRI scans were performed for more patients at DR. The image quality was sufficient in all cases. Histopathological diagnosis confirmed 203 cases of glioma, consisting of 112 HGGs (72 glioblastomas, 29 anaplastic astrocytomas, six anaplastic oligodendrogliomas, three gliosarcomas, one anaplastic oligoastrocytoma, and one anaplastic ependymoma), and 91 LGGs (58 astrocytomas, 20 oligodendrocytomas, six oligodendroastrocytomas, four gangliogliomas, two pilocytic astrocytomas and one ependymoma). Both preoperative and postoperative tumor volumes of 161 patients who underwent craniotomy for glioma resection were acquired. The mean preoperative lesion volume was 59.24 cm3. iMRI revealed that 90 of 161 cases (55.9%) achieved GTR and 71 had glioma residue, of which 63 cases (88.73%) required further resection, which gave a benefit ratio was of 39.13% (63/161). Although eight cases revealed tumor residue in eloquent areas, the amount of tumor resected reached preoperative surgical planning expectations and required no further resection. Among the 63 cases that required further resection, iMRI confirmed that 45 cases achieved GTR, with the remaining 18 cases still suggesting tumor residue. Six cases achieved GTR and 12 had tumor residue adjacent to critical areas that warranted no additional surgery. Table 1. Distribution of 373 cases with regard to histopathological diagnosis, surgical approach, main lesion location, and anesthesia pattern Items Histopathological Diagnoses Glioma Pituitary adenoma Meningioma Moyamoya disease Craniopharyngioma Lymphoma Vascular malformation Metastatic tumor Chordoma Others Surgical approach Craniotomy Transsphenoidal Biopsy Endoscopic Lesion mainly location Frontal lobe Saddle area Temporal lobe Parietal lobe Insular lobe Occipital lobe Basal ganglia area Others Anesthesia pattern General anesthesia Awake craniotomy n=373 203 55 20 19 13 11 8 6 3 35 288 57 25 3 87 79 46 32 22 17 14 76 339 34 Chin Med J 2012;125(): 4 Eventually, 141 gliomas (87.58%, 141/161) achieved GTR in 161 craniotomies. Therefore, the GTR rate for patients improved from 55.9% (90/161) to 87.58% (141/161). One hundred and fifty-one operations required functional neuronavigation guidance due to the proximity of the tumor and surrounding eloquent cortex and subcortical tracts. Twenty of 161 patients (12.42%) suffered from early motor deficit after surgery. Follow-up at 3 months for 158 patients suggested five patients (3.16%) had late motor deficit. Twenty-one of 161 patients (13.04%) presented with early speech deficit. Follow-up at 3 months revealed that late language deficits were observed in six of 158 patients (3.80%). Seven of 161 patients (4.35%) presented with severe postoperative intracranial hematoma which required reoperation; one died from severe infection and multiple organ failure. Of the 49 trans-sphenoidal pituitary tumor resection cases, histopathology confirmed that 31 were nonfunctioning pituitary adenomas. Thirty-eight of 49 cases (77.55%) achieved GTR, while 11 cases showed tumor residue on iMRI; five of which required further resection and four eventually achieved GTR. Finally, 42 of 49 pituitary adenomas achieved GTR (85.71%). The benefit ratio was 10.20% (5/49). The total resection rate improved from 77.55% (38/49) to 85.71% (42/49). Among 25 iMRI-guided biopsies, all histopathological diagnoses were confirmed, comprising 14 gliomas, seven lymphomas, three inflammatory granulomas, and one germinoma. Preoperative and intraoperative MRI lengthened surgery by 1–3 hours (mean: 1.9 hours). Preoperative MRI sequences included: T1 with contrast, T2 FLAIR, DTI, multi-voxel MRS, BOLD fMRI, resting state fMRI, and MR perfusion. Intraoperative MR sequences included: T1 with contrast, T2 FLAIR, and DTI. Superior images were acquired resulting from the improved signal-to-noise ratio due to increased field strength and redesigned RF coils. There were no MR-related adverse events. Average hospital stay for all cases was 18.81 days and the average total expense for hospitalization was RMB 72043.20 Yuan. No iMRI-related infectious occurred in the consecutive cases. Illustrative case A 39-year-old woman presented with a 2-month history of repeated seizure attacks without motor or language deficit. Preoperative MRI revealed a lesion located in the right frontal lobe, with proximity to the motor cortex and pyramidal tracts. The lesion was not enhanced and LGG was considered to be the preliminary diagnosis based on the increased ratio of choline/N-aceytl aspartate (Cho/NAA) on the multi-voxel MRS (Figure 2C). Presurgical planning was made following tractography of the pyramidal tracts (Figure 3C) and mapping of the motor cortex (Figure 2D) by fMRI. We performed an awake craniotomy and revealed the hand and mouth areas by brain mapping (Figure 4). The first iMRI scan showed a small amount of residual deep tumor (Figure 3D, 3E). Further resection was followed by updated neuronavigation. Final iMRI suggested GTR (Figure 3G, Figure 2. Preoperative series of MRI sequences for patient-specific surgical planning in the illustrative case. 2A and 2B: Lesion in the right frontal lobe. LGG was diagnosed according to the increased ratio of Cho/NAA (2C). Motor cortex was indicated on BOLD fMRI activation (orange spots in 2D). Figure 3. Pre- (3A–3C), intra- (3D–3F), and postoperative (3G–3I) series of the illustrative case. First intraoperative MRI indicated a small amount of residual tumor beneath the surgical cavity (3D, 3E). GTR was achieved after further resection (3G, 3H) and pyramidal tracts were protected (3I). Chinese Medical Journal 2012;125(): 5 clinical practice suggests that these advantages are most prominent in glioma craniotomy, pituitary tumor resection, and biopsies. More than one-third glioma cases and more than one-fifth pituitary adenoma cases acquired further resection under updated neuronavigation. The use of 3 T iMRI benefits the patients by improving overall GTR rates, while preserving the eloquent areas of the brain. Figure 4. Intraoperative braining mapping (A) was performed to locate the hand area (label H) and mouth area (label M) of the motor cortex, which were protected after tumor removal (B). 3H) without any injury to the motor cortex or pathway (Figure 3I). Histopathology confirmed an astrocytoma. No motor deficit was observed in this patient at 3 months postoperatively. DISCUSSION High-field MRI is the priority for image-guided neurosurgery due to its high resolution in brain imaging, precise temporal and spatial resolution, random flat 3D imaging, flow and temperature sensitivity, functional brain imaging, nonionizing radiation, and other advantages. The first iMRI system, based on a 0.5 T magnet, was introduced 15 years ago at Brigham and Women’s Hospital, Boston, USA.3 The following decade has witnessed enormous developments in MRI equipment and technology.6,7 Gone are the days when surgical beds had to be brought into an MRI suite, and MRI equipment and operating room have evolved into a single unit. The ability to image the brain during surgery has become a reality. Our institute started utilizing the 0.15 T PoleStar ™ N20 in 2006, which has good mobility and easy installation. However, suboptimal image quality, longer scanning times, and lack of functional applications are the limitations of this device.8 High-field iMRI brings real-time diagnostic quality imaging, and such functional applications as MR spectroscopy, fMRI and DTI may be carried out during surgery.9,10 This has brought substantial improvement to microsurgical and minimally invasive neurosurgery. In this study, all cases benefited from real-time diagnostic quality imaging and most of them benefited from functional applications. iMRI has the following advantages: (1) real-time diagnostic quality imaging, correcting the dynamic brain shift and improving the overall accuracy of surgery; (2) improving GTR rates while preserving vital neurovascular structures and eloquent areas; and (3) giving accurate real-time guidance for biopsies and implantations, minimizing surgical damage.6,10 Our initial Pioneering efforts in high-field iMRI have brought revolutionary changes to minimally invasive neurosurgery. An increasing number of studies11,12 have demonstrated that the extent of resection is a predictor of survival for glioma patients despite a lack of class I evidence. Some studies have even shown that supratotal resection might improve the outcome of LGGs.13,14 However, extensive resection may carry the risk of neurological morbidity, which affects the quality of life and subsequent overall survival.15 The evidence-based grade for recommendation of the clinical efficiency of iMRI for glioma resection still stands at level II.7,11 We are now performing a randomized controlled trial for investigating the application of iMRI in patients with glioma, which were registered at clinicaltrials.gov. This randomized controlled study, instead of the present study, will help us answer the following questions. What are the indications for iMRI operation? What is the efficacy of iMRI operation? Does iMRI operation have a higher incidence rate of unplanned second operations? Our preliminary results suggest that iMRI benefits both the patient and the surgeon. The utilization of iMRI dramatically decreases surgical morbidity, especially in those who have a lesion adjacent to eloquent areas. Dynamic tissue shifting occurs during the course of the operation and neuronavigation alone decreases the safety and accuracy of surgery. Radical resection may even cause late severe deficits, forcing many neurosurgeons to take a more conservative approach in order to preserve function, which may result in more residual tumor than anticipated. We also observed an increase in early speech and/or motor deficit rates, but both decreased during 3 months follow-up after surgery. This may be a result of a more radical surgical approach than previous conventional craniotomy, causing more edema in the eloquent areas adjacent to the resected lesions. There is no damage to the eloquent cortex or subcortical tracts themselves, thus, the function could be better preserved than with conventional craniotomy, and most patients motor and verbal abilities improve to preoperative baseline levels. We believe that 3 T iMRI provides neurosurgeons with dynamic functional and anatomical maps of brain structures, which improves the overall accuracy and safety of surgery, and helps decrease permanent surgical morbidity. At the same time, the fact that iMRI increases both the cost and length of surgical treatment should also be taken into account. We advocate further research and investigation in understanding the implications of iMRI on surgery and patient outcome, and Chin Med J 2012;125(): 6 potential trials in integrating iMRI with additional technologies such as endoscopic and robotic surgical devices in the near future. 8. Acknowledgments: The authors would like to thank all the nurses in the operating rooms and wards for their co-operation during data collection; SHI Jian-bing and YANG Zhong for management and collection of the image database; XU Gen for volumetric measurement, WANG Zhen-xiao; and WANG Yan for the input of clinical information. 9. 10. REFERENCES 11. 1. 2. 3. 4. 5. 6. 7. Samii M, Brinker T, Samii A. Image-guided neurosurgery--state of the art and outlook. Wien Klin Wochenschr 1999; 111: 618-628. Galloway RL Jr, Maciunas RJ, Edwards CA 2nd. Interactive image-guided neurosurgery. IEEE Trans Biomed Eng 1992; 39: 1226-1231. Black PM, Moriarty T, Alexander E 3rd, Stieg P, Woodard EJ, Gleason PL, et al. Development and implementation of intraoperative magnetic resonance imaging and its neurosurgical applications. Neurosurgery 1997; 41: 831-842; discussion 842-835. Chen X, Xu BN, Meng X, Zhang J, Yu X, Zhou D. Dual-room 1.5-T intraoperative magnetic resonance imaging suite with a movable magnet: implementation and preliminary experience. Neurosurg Rev 2012; 35: 95-109; discussion 109-110. Wang Y. Relations between the sides of linguistic cerebral dominance and manuality in Chinese aphasics. Chin Med J 1996; 109: 572-575. Lang MJ, Greer AD, Sutherland GR. Intra-operative MRI at 3.0 Tesla: a moveable magnet. Acta Neurochir Suppl 2011; 109: 151-156. Kubben PL, ter Meulen KJ, Schijns OE, ter Laak-Poort MP, 12. 13. 14. 15. van Overbeeke JJ, van Santbrink H. Intraoperative MRI-guided resection of glioblastoma multiforme: a systematic review. Lancet Oncol 2011; 12: 1062-1070. Acevedo CV, Vinas FC, Dujovny M. Intraoperative use of magnetic resonance imaging: neurosurgical applications and technical limitations. Crit Rev Neurosurg 1999; 9: 279-286. Zhao Y, Chen X, Wang F, Sun G, Wang Y, Song Z, et al. Integration of diffusion tensor-based arcuate fasciculus fibre navigation and intraoperative MRI into glioma surgery. J Clin Neurosci 2012; 19: 255-261. Pamir MN. 3 T ioMRI: the Istanbul experience. Acta Neurochir Suppl 2011; 109: 131-137. Kuhnt D, Becker A, Ganslandt O, Bauer M, Buchfelder M, Nimsky C. Correlation of the extent of tumor volume resection and patient survival in surgery of glioblastoma multiforme with high-field intraoperative MRI guidance. Neuro Oncol 2011; 13: 1339-1348. Sanai N, Berger MS. Operative techniques for gliomas and the value of extent of resection. Neurotherapeutics 2009; 6: 478-486. Duffau H. Awake surgery for incidental WHO grade II gliomas involving eloquent areas. Acta Neurochir (Wien) 2012; 154: 575-584; discussion 584. Yordanova YN, Moritz-Gasser S, Duffau H. Awake surgery for WHO Grade II gliomas within "noneloquent" areas in the left dominant hemisphere: toward a "supratotal" resection. Clinical article. J Neurosurg 2011; 115: 232-239. McGirt MJ, Mukherjee D, Chaichana KL, Than KD, Weingart JD, Quinones-Hinojosa A. Association of surgically acquired motor and language deficits on overall survival after resection of glioblastoma multiforme. Neurosurgery 2009; 65: 463-469; discussion 469-470. (Received August 14, 2012) Edited by JI Yuan-yuan