Utility and Pitfalls of High field 3 Tesla Intraoperative MRI in Neurosurgery: A Single Centre Experience of 100 Cases
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.284359
Source of Support: None, Conflict of Interest: None
Keywords: 3Tesla intraoperative MRI, adjuncts in neurosurgery, image-guided neurosurgery, technological advancements in neurosurgeryKey Message: Intraoperative MRI is a valuable adjunct to improve the extent of resection of ICSOLs. Maximum benefit was seen in resection of intra-axial SOL and Pituitary macroadenomas. This study provides a preliminary information on setting up and utilizing iMRI to guide resection in ICSOLs. Safe, successful and effective utilization of iMRI requires multiple i
“Primum non nocere” is a Latin phrase that means, “above all, do no harm”. This idea forms the basis of treatment for any medical illness. Similar way, for decades neurosurgery for intra axial brain tumors revolved around subtotal resection (STR) and tumor decompression followed by adjuvant therapy in order to prevent neurological deficits due to lack of technological armamentarium to differentiate pathological tissue and surrounding eloquent structures with certainty safely.
The extent of resection has been remarkably improved in past few years with integration of technology in the neurosurgical operating room (OR), like advanced microscopes and endoscopes for better Intraoperative illumination and visualization, advanced microsurgical instruments like cavitron ultrasonic aspirators (CUSA) that gives better precision and finesse to neurosurgeons, fluorescein dyes like sodium fluorosceine (NaF) and 5 ALA that helps in better identification of brain tumor interface, neuronavigation systems which helps in precisely defining anatomical landmarks, thus localizing extent of intracranial lesions and routine use of Intraoperative neural monitoring which help in identifying its relation to critical brain structures. All the modalities used have their own limitations. Neuronavigation which is used anatomically to navigate through the brain structures has the problem of brain shift during the surgery. Intraoperative brain shift caused by medications, Cerebrospinal fluid (CSF) release, placement of retractors, tissue manipulation and tumor resection is a very well-known phenomenon which confounds the information gathered from preoperative imaging and is a major factor for suboptimal resection of intracranial space occupying lesions (ICSOLs). Multiple studies have proved the direct correlation of extent of resection (EOR) and post-operative remnants with overall survival (OS), progression-free survival (PFS), lesser chances of recurrence and lesser need for adjuvant treatment in ICSOLs.,,,, To maximize the extent of safe resections, there is a gradual transition from post-operative imaging to Intraoperative imaging.
Multiple Intraoperative imaging modalities have been tried and tested namely Intraoperative computed tomography (iCT), Intraoperative ultrasound (iUS) and Intraoperative magnetic resonance imaging (iMRI) with or without using fluorescein dyes (NaF, 5-ALA) as an adjunct. iMRI has proved its efficacy in various studies as the best tool amongst all for Intraoperative orientation, exact determination of brain shift and detecting immediate Intraoperative complications. The information gathered helps a surgeon to analyze real time resection in order to make a decision of ending or continuing the surgery. Like every other interesting and useful medical breakthroughs till date, iMRI also has its own shades of grey, mainly an extremely costly setup which restricts it to only a few neurosurgical centers and an increase in overall surgical time.
Although first and most widely described use of iMRI was for intraaxial tumors, today neurosurgeons worldwide have trained themselves to efficiently apply this technique for other neurosurgical procedures like resection of extra-axial lesions, epilepsy surgeries, placement of deep brain stimulation electrodes and even in stereotactic aspirations of intracerebral hematoma.,,,,,,,,
In India, several centers are using iMRI system. Ours is the first center to acquire high field 3 Tesla iMRI system. The aim of this paper is to share our one year experience with the implementation of 3T iMRI in neurosurgical procedures and to evaluate its use to improve our surgical outcomes in neuro oncological conditions.
Prospective observational study conducted from August 2017 to July 2018 at Yashoda Hospital, Secunderabad. All patient undergoing 3T iMRI guided resection of ICSOLs were included in the study.
All iMRI were conducted using a 3 Tesla machine (MAGNETOM SKYRA, Siemens medical system, Erlangen, Germany). At our center, the magnet is located in the room adjacent to neurosurgery operating room (OR) i.e. “NEAR BY OT TYPE”. Before the start of procedure, the patient is positioned on MRI compatible head pins (DORO LUECENT®) attached to sliding patient board on an operating table (MAQUET MAGNUS OR TABLE SYSTEM™). All skin to skin contact surfaces is covered with cotton rolls. Navigation system is placed and virtual position of the patient inside gantry is anticipated using Bore gauze which mimics MRI gantry size. MRI staff is informed 30 minutes prior to expected arrival for intraoperative imaging. Routine imaging is withheld, and floor of MRI suite and corridor along with MRI surface is sterilized using Disinfectant (BACILLOCID®, Ramen and Weil). When the need for Intraoperative imaging is felt, the surgery is suspended and all the ferromagnetic instruments are removed from patient's body and surgical site is packed with antibiotic incise drape (IOBAN™) after filling the surgical cavity with saline. The patient is wrapped in a sterile plastic drape and checked with a metal detector for any accidentally left metal instruments on the patient's body. The MRI machine has a detachable trolley which is wheeled in and docked to operating table.
An iMRI checklist is filled by a team of neuroanesthetists before transfer of patient to ensure patient, personnel and equipment safety. After checklist patient is transferred to iMRI trolley and shifted to MRI room through a dedicated sterile corridor. Once in the imaging suite, patient is connected to MRI compatible ventilator and monitors using compatible ECG electrodes and oximeters. A team of dedicated neuroanesthetists and neurosurgeons observe the patient throughout and strict asepsis is maintained. The acquired images are analyzed by experienced neuroradiologists and neurosurgeons for any residual tumor and complications like bleeding, infarcts, etc. [Figure 1] and [Figure 2].
The following variables were recorded: preoperative imaging diagnosis, presence or absence of residue in Intraoperative imaging, whether or not iMRI modified our surgical decision, complications and mishaps attributed to iMRI, the time required to shift and time required for image acquisition.
The resection status based on iMRI was divided into three categories. Patient with no residue in iMRI were classified as “primary gross total resection (GTR)”, patients with residues seen in iMRI and surgery was continued with GTR confirmed in post-operative scan were classified as “Secondary GTR”, patients with residue seen but the decision of aborting the surgery or partial resection of residue was taken were classified as “Subtotal resection (STR)”, and the total number of primary and secondary GTR was termed as “Total GTR”.
The data was recorded using a spreadsheet software (EXCEL, Microsoft, Redmond, USA) and was analyzed statistically.
A total of 100 patients with various intracranial SOLs were included in the study. Primary GTR was achieved in 44% (44/100) and residue was detected in 56% (56/100), secondary GTR was achieved in 37% (37/100) and decision of discontinuing surgery was taken in 19% (19/100), due to presence of tumor remnant in eloquent cortex or adjacent to major vascular structures [Graph 1].
Out of 100 cases, the most common surgical indications were intraaxial SOLs (42%) and pituitary macroadenomas (30%) followed by other extraaxial lesions. iMRI was able to detect residues in 59.52% (25/42) intraaxial SOL, 60.00% (18/30) pituitary adenomas, 45.45% (5/11) meningioma, 71.42% (5/7) CPA mass lesions, 50% (1/2) craniopharyngioma and 33.33% (2/6) intraventricular SOLs [Table 1] and [Table 2].
iMRI also helped us to improve our extent of resection (to achieve secondary GTR) in 76% (19/25) detected residues in patients with intraaxial SOLs, 55.55% (10/18) pituitary adenoma residues, 60% (3/5) meningioma residues, 60% (3/5) CPA residues, 100% (2/2) residues of intraventricular SOL [Table 3].
We also noted and analyzed the mishaps that occured during imaging and complications in the early postoperative period. Coil induced and contact Radiofrequency burns were seen in 3 cases (3%), circuit disconnection and transient rise in ETCO2 occurred in 1 patient (1%) and minor easily resolvable technical issues like problems in docking the MRI trolley and sliding the patient on MRI trolley was recorded in total 18 patients (18%). Post-operative infections were seen in 2% which was comparable to our infection rate in non-iMRI guided surgeries.
The mean time required for shifting and image acquisition in the first 20 cases was 85.6 minutes which was reduced to 37.4 minutes in the next 80 cases due to multiple repetitions and adaption to the shifting process and reduction in number of MRI sequences to identify residual tumor.
A successful neurosurgical procedure for brain tumor resection hinges around defining accurate tumor margin and viable brain tumor interface which is considered challenging to even well trained and experienced surgeon. This problem along with continuous ongoing Intraoperative brain shift calls for the need of a better Intraoperative assessment of real time tumor resection, which is best done by an iMRI system. For our ease and better understanding of the data, we classified the EOR into primary GTR, secondary GTR and STR. In our work, we found that iMRI was an excellent tool to guide and improve our extent of safe resections by 37% in brain tumor surgeries. The maximum residues were detected in cases of pituitary adenomas and maximum iMRI aided secondary GTR was achieved in intraaxial lesions.
Pamir et al. in their study assessed extent of 3T iMRI guided resection in 56 patients with low grade glioma and concluded a 32.3% increase in total GTR from a 55.4% primary GTR in first look iMRI. Similarly, Senft et al. in their first of its kind randomized controlled trial compared iMRI guided resection versus conventional microneurosurgical resections in patients with gliomas of all grades and concluded a 33% secondary GTR and 96% total GTR rate in intervention group, when compared to the 68% total GTR rate in control group. In our series, out of 42 cases of intraaxial sols of all grade, we were able to achieve 40.47% (17/42) primary GTR and were able to detect 59.52% (25/42) residues. Intraoperative MRI helped us to achieve secondary GTR in 45.23% (19/42) and our total GTR was 85.71% (36/42), thus high field iMRI helped us to increase our extent of resections in glioma by 45.24%. The findings from our early experience show that iMRI can accurately assess tumor remanant and can be a guiding tool to continue surgery and achieve maximal resection safely in glioma surgeries.
Zaidi et al. in their study of 27 patients compared GTR rates of pituitary adenoma using endoscope alone vs endoscope with iMRI as an adjunct and concluded an improvement of GTR rates by 20% when iMRI was used. Serra et al. in their study concluded that the use of high definition iMRI helped in precise visualization and quantification of adenoma volume and thus helped in increasing EOR and total GTR rates in transnasal endoscopic surgeries for pituitary macroadenoma. The findings were confirmed in our study out of 30 functioning and non-functioning adenomas, iMRI helped us to achieve secondary GTR in 33.33% and thus increased over total GTR rates to 76.66%.
In our work, we found iMRI helpful even in maximizing the extent of tailored resections in extraaxial lesions like meningioma, CPA mass lesions, craniopharyngiomas, orbital tumors and in intraventricular SOLs. To the best of author knowledge, there are no studies till date to establish role of iMRI in extraxial lesions, and thus further studies are needed to analyze the impact of the same.
Installing and implementing an iMRI is a technically challenging and expensive task to accomplish. After installation, we found many pitfalls associated with iMRI mainly an increase in surgical time. Jankovski A et al. In their study, found an average increase in surgical time by 78 ± 20 minutes with an average of 34.1 minutes for image acquisition. Similar findings were given by chicoine et al. We conducted various mock drills and devised a pre shift checklist to train our team in shifting and bringing the patient back to OR table without any inadvertent mishap. A sequence of entry of shifting personnel, instruments and MRI trolley was changed 3 times to get the most streamlined and fast results. In our study, we found an average increase in surgical time which included time to shift patient to gantry, time for image acquisition and time to shift patient back to OR was 85.4 minutes in first 20 cases, which was reduced to 37.6 minutes in next 80 cases due to reduction in number of sequences acquired. We have set a desired imaging protocol with a minimum number of sequences to effectively detect residues with only minor variations on case to case basis [Table 4]. Thus, iMRI proved its utility by improving our total GTR rates but with a pitfall of increase in total surgical time, which we believe can be further reduced in near future as the team becomes accustomed to the procedure.
Wood et al. in their systematic review compared 1.5 tesla and 3 Tesla MRI and described pros and cons of a high field strength magnet. Advantages of 3 Tesla setup were high signal to noise ratio (SNR) and ability of parallel imaging (PI) thus improving image quality and reduced scan time, the difference in T1 relaxation time of solid tissue and blood gives a better blood tissue contrast, and thus more sensitive in identifying residues, excellent and robust BOLD imaging, DTI and MRS, high contrast to noise ratio (CNR) results in better visualization of vessels. The main cons of 3T magnet described were higher gradient noise and increased sound pressure in imaging suite, significantly higher heating potential and chances of thermal burns to the patient, data on safety of 3T MRI with implanted device is scarce, thus cannot be used confidently in patients with implants and a perceived higher upfront expenditure to setup a high field Intraoperative MRI unit.
Setting up and effective utilization of any new surgical adjunct has its own challenges. iMRI success depends on multi-departmental efforts, good communication between group persons involved and sincere team work. The team of neuroanesthetists and technical staff plays a pivotal role for successful and safe image acquisition. Our experience shows that it takes multiple iteration of the shifting process, along with initial training session and mock drill, proper education of neurosurgical technicians and nursing staff, meticulous data collection and auditing to analyze and smoothen the work flow. Institution protocols and checklists should be prepared to reduce any untoward events.
Limitations of our work include not describing the efficacy of iMRI in various grades of gliomas separately and not comparing it with the control conventional resection group along with no analysis of long term surgical outcome, overall survival and progression free survival rates of the patient which needs further attention.
So, to conclude, 3TiMRI is a valid and state of the art technology which can help us achieve better extent of safe resections, and in turn can improve prognosis in patients with intracranial SOLs.
Our sincere thanks to Dr. Venu Gopal, Consultant Neurosurgeon, Yashoda Hospital, Malakpet, Hyderabad, for sharing his experience and valuable data of patients who underwent iMRI guided resections and to Dr. Astha Palan, Fellow in neuroanesthesia, Yashoda Hospital, Secunderabad, for her hard and dedicated work towards safe and streamlined patient motion in and out of MRI suite.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4]