Use of High-Field Intraoperative Magnetic Resonance Imaging to Enhance the Extent of Resection of Enhancing and Nonenhancing Gliomas
Mohammadi, Alireza Mohammad MD; Sullivan, T. Barrett BSE; Barnett, Gene H. MD; Recinos, Violette MD; Angelov, Lilyana MD; Kamian, Kambiz MD; Vogelbaum, Michael A. MD, PhD
Rose Ella Burkhardt Brain Tumor and Neuro-Oncology Center, Department of Neurosurgery, Neurological Institute, Cleveland Clinic, Cleveland, Ohio
Correspondence: Michael A. Vogelbaum, MD, PhD, FAANS, FACS, Burkhardt Brain Tumor and NeuroOncology Center and Department of Neurosurgery, Cleveland Clinic, Mail Code ND40, 9500 Euclid Avenue, Cleveland, OH 44195. E-mail: firstname.lastname@example.org
Received July 11, 2013
Accepted December 13, 2013
BACKGROUND: Intraoperative magnetic resonance imaging (IoMRI) is used to improve the extent of resection of brain tumors. Most previous studies evaluating the utility of IoMRI have focused on enhancing tumors.
OBJECTIVE: To report our experience with the use of high-field IoMRI (1.5 T) for both enhancing and nonenhancing gliomas.
METHODS: An institutional review board–approved retrospective review was performed of 102 consecutive glioma patients (104 surgeries, 2010-2012). Pre-, intra-, and postoperative tumor volumes were assessed. Analysis was performed with the use of volumetric T2 images in 43 nonenhancing and 13 minimally enhancing tumors and with postcontrast volumetric magnetization-prepared rapid gradient-echo images in 48 enhancing tumors.
RESULTS: In 58 cases, preoperative imaging showed tumors likely to be amenable to complete resection. Intraoperative electrocorticography was performed in 32 surgeries, and 14 cases resulted in intended subtotal resection of tumors due to involvement of deep functional structures. No further resection (complete resection before IoMRI) was required in 25 surgeries, and IoMRI showed residual tumor in 79 patients. Of these, 25 surgeries did not proceed to further resection (9 due to electrocorticography findings, 14 due to tumor in deep functional areas, and 2 due to surgeon choice). Additional resection that was performed in 54 patients resulted in a final median residual tumor volume of 0.21 mL (0.6%). In 79 patients amenable to complete resection, the intraoperative median residual tumor volume for the T2 group was higher than for the magnetization-prepared rapid gradient-echo group (1.088 mL vs 0.437 mL; P = .049), whereas the postoperative median residual tumor volume was not statistically significantly different between groups.
CONCLUSION: IoMRI enhances the extent of resection, particularly for nonenhancing gliomas.
ABBREVIATIONS: 5-ALA, 5-aminolevulinic acid
CRDT, complete resection of the detectable tumor
CRET, complete resection of the enhancing tumor
EBRT, external beam radiation therapy
FLAIR, fluid-attenuated inversion recovery
GBM, glioblastoma multiforme
IoMRI, intraoperative magnetic resonance imaging
KPS, Karnofsky Performance Status
LGG, low-grade glioma
MP-RAGE, magnetization-prepared rapid gradient-echo
Microsurgical resection is considered to be the mainstay of treatment for most patients with gliomas.1 There is some debate in the literature regarding the magnitude of impact that the extent of resection has on survival, and there are no adequately powered prospective, randomized clinical trials available to address this question. It is well recognized that removal of the tumor mass, in cases of WHO grade II to IV gliomas, is not curative, and traditional oncologic, margin-free resections are not performed due to the fact that tumor cells invariably infiltrate the margin beyond the bulk tumor, and removal of all affected brain would lead to potentially significant loss of neurological functioning.2 Nonetheless, a growing body of evidence indicates a link between more radical surgical resection and longer life expectancy3-15 in addition to improved quality of life16 for patients with gliomas. Various methods have been used to enhance intraoperative visualization of the margins of bulk tumor, thereby allowing the surgeon to maximize the extent of resection. Intraoperative visual aids that have been used include ultrasound,17 computed tomography,18 image-guidance navigation systems,19 tumor-specific fluorophors (eg, 5-aminolevulinic acid [5-ALA]20 or sodium fluorescein21), computer-navigated microscopy,22 and intraoperative magnetic resonance imaging (IoMRI).23
The surgical goal and terminology referring to the completeness of resection are different in high-grade vs low-grade gliomas (LGGs). In high-grade gliomas, removal of all enhancing tumor is considered to be the primary goal of an intended complete resection, and it is not typical to remove the full extent of T2/fluid-attenuated inversion recovery (FLAIR) hyperintense tissue surrounding the enhancing tumor. In contrast, the surgical goal for LGGs is to remove all of the T2/FLAIR hyperintense tissue. Accordingly, for high-grade gliomas, assessment of completeness of resection refers to the enhancing tumor, and a complete resection is termed complete resection of the enhancing tumor (CRET). For LGGs, assessment of completeness of resection refers to the detectable tumor (that seen on T2/FLAIR sequences), and a complete resection is termed complete resection of the detectable tumor or (CRDT).24
There has been increasing interest in the use of IoMRI in neurosurgery over the course of the past 2 decades.25 Retrospective and prospective series that have evaluated the contribution of IoMRI to increasing the extent of resection of gliomas have primarily examined its use for surgery for enhancing tumors, particularly glioblastoma multiforme (GBM).10,25-33 In contrast to high-grade gliomas, which are often grossly visible, LGGs are frequently similar in appearance to adjacent normal parenchyma.34 Consequently, IoMRI may have an even more important role in maximizing the extent of resection of nonenhancing, LGGs.4,34,35 To date, there are no studies that directly compare the use of IoMRI for surgical resection of nonenhancing and enhancing gliomas.
We used a high-field IoMRI (1.5-T) system to perform maximal safe resection of both nonenhancing and enhancing gliomas. In this study, we evaluated the impact of this approach on the extent of resection by performing volumetric measurements of enhancing and nonenhancing disease on preoperative, intraoperative, and postoperative MRI.
PATIENTS AND METHODS
Study Size and Setting
A total of 102 consecutive glioma patients underwent high-field IoMRI-guided surgeries (104 surgical cases) at a single institution from 2010 to 2012 and were retrospectively reviewed after approval by the Cleveland Clinic Institutional Review Board.
Preoperative MRI showed a completely nonenhancing tumor in 43 (41%), minimal or patchy enhancement in 13 (13%), and diffuse or ring enhancement in the remaining 48 cases (46%). Final pathology was determined to be GBM (WHO grade IV) in 46 (44%), anaplastic glioma (WHO grade III) in 17 (17%), and LGG (WHO grade II) in 41 cases (39%). Further details regarding tumor enhancement and pathology are shown in Table 1. IoMRI surgery was performed using a high-field (1.5-T) IoMRI system (IMRIS, Winnipeg, Manitoba, Canada). Based on the evaluation of preoperative anatomic studies, 58 cases were considered to be amenable to complete resection of enhancing or detectable tumor.24 Evaluation of functional MRI studies led to the decision to use intraoperative electrocorticography (ECoG) in 32 surgeries (awake craniotomy in 12 patients). In 28 patients, monitoring of motor function was performed first with median nerve somatosensory evoked potentials, which confirmed the location of the motor strip, followed by motor cortex and white matter stimulation. Speech/language monitoring was performed in 4 patients with cortical stimulation performed to find area(s) of speech arrest. The remaining 14 surgeries were planned as intended subtotal resections.
Preoperative, intraoperative, and postoperative volumetric magnetic resonance images were imported to iPlan (BrainLAB, Feldkirchen, Germany) for evaluation and volume measurement. Postoperative imaging was completed either in the operating room using the IoMRI (for patients who had no additional resection) or outside of the operating room in the standard imaging suite within 24 hours of closure (for patients who had additional resection after first IoMRI evaluation). We refer to this aggregation of postoperative image sets as the final MRI. For nonenhancing and minimally enhancing tumors, a volumetric T2 sequence was used for the measurement of tumor volume (T2 group). We did not have a routine practice of performing the final MRI before the patient was awakened from surgery; in 33 patients in the T2 group, the final MRI was done before reversal of anesthesia (26 with no additional resection after IoMRI), and in 23, it was done during the postoperative hospitalization (all had additional resection after IoMRI), within 72 hours of surgery. Because postoperative edema can be mistaken for residual tumor in patients with a delayed final MRI, we performed a coregistered comparison of preoperatively and postoperatively to attempt to distinguish between preexisting T2 change (presumably mostly tumor) and new postoperative T2 change (presumably mostly surgical manipulation related) and thereby minimize the confounding effects of postoperative edema. For enhancing tumors, a volumetric magnetization-prepared rapid gradient-echo (MP-RAGE) sequence with and without contrast enhancement was used (MP-RAGE group). After image fusion, tumor contours were determined manually for each imaging series, and volumes were measured with use of the iPlan software. Contouring was performed by the first author, and it was performed with the consideration of the final neuroradiology report for each study. The senior author reviewed the contours. In brief, any type of enhancement was considered residual volume (nodular or linear), except when it was reported as a vessel by the neuroradiologist. Of note, we routinely performed pre- and post-gadolinium imaging so as to be able to subtract out nonspecific hyperintensity produced by blood products. Investigators were blinded to clinical outcomes during the volumetric evaluation process and to the volumetric results during the clinical evaluation process. Residual tumor volumes less than 0.175 mL were considered a complete resection in accordance with the methods used in 2 recent trials regarding extent of resection.20,25
Primary endpoints were defined as the residual tumor volume at IoMRI and final MRI, the extent of resection at IoMRI and final MRI, and the percentage of surgeries achieving complete resection of the enhancing tumor (CRET in MP-RAGE group) or complete resection of detectable tumor (CRDT in T2 group) at final MRI relative to IoMRI (among patients in whom complete resection was achievable). Secondary endpoints were defined as the percentage of surgeries with additional resection performed after ECoG + IoMRI (among patients for whom complete resection was uncertain based on tumor proximity to functional structures) and the percentage of surgeries achieving complete resection after ECoG + IoMRI in this group.
Patients' clinical and tumor characteristics (age, sex, tumor pathology, tumor location, neurological function status at surgery, Karnofsky Performance Status [KPS] score at surgery, upfront vs salvage surgery, preoperative, intraoperative, and postoperative tumor volumes) were evaluated using the median and range for continuous variables and the frequency count for categorical factors. The Student t test was used for comparison of the mean between the T2 and MP-RAGE groups in terms of residual volume and extent of resection at IoMRI and residual volume and extent of resection at final MRI. The χ2 test was used in the evaluation of the impact of intraoperative ECoG and additional resection on postoperative neurological complications. All tests of statistical significance were 2 sided, and P values <.05 were considered statistically significant.
The median age was 48 years, and the female:male ratio was 0.65. The most common presenting symptom was a seizure, observed in 38 patients (37%), followed by headaches in 26 patients (25%) and weakness in 17 patients (16%). Four patients were asymptomatic at time of treatment. The median KPS score at the time of surgery was 90 (range, 60-100). Twenty-five patients (24%) had no neurological deficit before surgery, 53 patients (51%) had mild neurological impairment, 18 patients (17%) had moderate neurological impairment, and 6 patients (6%) had severe neurological impairment. Thirty-four patients (33%) had previous surgery. The WHO grade at the previous surgery was the same as at IoMRI surgery in 24 patients, and WHO grade at previous surgery was lower than at IoMRI surgery in 10 patients. Previous external beam radiation therapy (EBRT) had been used in 19 patients (18%) and previous chemotherapy was used in 23 patients (22%). Comparison of the 2 groups for previous treatments revealed that previous surgery was performed in 22 cases (39%) in the T2 group vs 12 cases (25%) in the MP-RAGE group (P = .12), previous EBRT in 7 cases (13%) in the T2 group vs 12 cases (25%) in the MP-RAGE group (P = .10), and previous chemotherapy in 13 cases (23%) in the T2 group vs 10 cases (21%) in the MP-RAGE group (P = .77). The most common tumor location was the frontal lobe (42 patients, 40%), followed by temporal lobe in 35 patients (34%) and parietal lobe in 22 patients (21%). Based on the presurgical navigational MRI, 56 surgeries (54%) were determined to be in the T2 group and 48 surgeries (46%) were determined to be in the MP-RAGE group. The median duration of surgery, including anesthesia induction and initial recovery, was 6.6 hours (range, 3.1-12.5 hours), and the median postoperative hospital stay was 2 days (range, 1-26 days). After IoMRI-guided surgery, EBRT was performed in 44 patients (42%) and postoperative chemotherapy was performed in 52 patients (50%). The median follow-up after surgery was 7.2 months (range, 1-29 months), and 14 patients died during extended follow-up.
Preoperative evaluation of anatomic and functional imaging datasets indicated that 58 surgeries would be for tumors amenable to complete resection without any further intraoperative evaluation. Of the other 46 surgeries, 32 were considered to be candidates for intraoperative ECoG based on a review of functional MRI studies. For these 32 surgeries that included intraoperative ECoG and functional monitoring, complete resection was thought to be safe in 23. The other 14 surgeries (of 46) were for tumors involving deep functional structures (eg, internal capsule, basal ganglia), and the preoperative goal of these surgeries was safe maximal subtotal resection.
Surgery was terminated after IoMRI despite evidence of residual tumor in a total of 25 cases (24%): 14 surgeries due to proximity to or involvement of deep functional structures as determined by preoperative imaging alone, 9 surgeries due to proximity to eloquent cortex confirmed by intraoperative ECoG, and 2 surgeries due to surgeon discretion. In 1 of these 2 cases, there was a technical issue (poor IoMRI quality), which precluded further resection, and in the other one, the onset of poorly controlled brain edema before IoMRI led to early termination of the operation.
Among the 79 cases (76%) considered amenable to complete resection (42 from the T2 group and 37 from the MP-RAGE group), 25 initial resections had achieved the goal of a complete resection based on evaluation of the IoMRI, and in 54 cases, additional resection after the IoMRI was performed.
Outcome Data: Volumetrics
The median preoperative tumor volume was 20.123 mL (range, 0.9-216 mL) for the complete series of 104 surgeries. For each of the T2 and MP-RAGE groups, the median preoperative volume was 22.27 mL and 19.57 mL, respectively (P = .077). The median residual tumor volume improved from 1.22 mL at the time of IoMRI to 0.2 mL at the final MRI in the whole series of 104 surgeries (P = .14), and the median extent of resection increased from 92.9% at the time of IoMRI to a final extent of resection of 99.4% (P = .01).
Tumors Not Amenable to Complete Resection
A review of preoperative MRI studies (including functional MRI- and diffusion tensor imaging-based fiber tracking) along with intraoperative ECoG indicated that 25 cases were not amenable to complete resection (Figure 1). The median preoperative volume for these cases was 34.29 mL. The median residual tumor volume at the time of IoMRI was 5.22 mL. Fourteen cases were in the T2 group and 11 were in the MP-RAGE group (Table 2).
Tumors Amenable to Complete Resection
In the 79 cases that were amenable to complete resection, the median preoperative volume was 17.463 mL. After initial resection, IoMRI revealed a median residual tumor volume of 0.779 mL and a median extent of resection of 94.9%. After additional resection, the final MRI showed a median final residual tumor volume of 0.061 mL and a median final extent of resection of 99.6%. The reduction in residual tumor volume after additional resection was statistically significant (P = .002). The improvement in extent of resection after additional resection was also statistically significant (P < .001).
At the time of IoMRI, 52 cases (66%) had more than 90% resection, 31 (39%) had more than 98% resection, and 27 (34%) had already undergone a complete resection (Figure 2). At final MRI, 76 cases (96%) had more than 90% resection, 61 (77%) had more than 98% resection, and 51 (65%) had a complete resection. Further details, subdivided by T2 and MP-RAGE groups, are shown in Table 3.
Tumors Undergoing Additional Resection
In the 54 cases in which additional resection was performed (Figure 3), the median intraoperative residual tumor volume was 1.902 mL and the median intraoperative extent of resection was 89.8%. After additional resection, the final MRI showed the median residual tumor volume to be 0.207 mL, resulting in a median extent of resection of 99.4%. The reduction in residual tumor volume after additional resection was significant (P = .002). The improvement in the extent of resection after additional resection also was statistically significant (P < .001). Complete resection was attained in 26 cases (48%). Further details, subdivided by T2 and MP-RAGE groups, are shown in Table 4 and Figure 4.
Comparison of T2 and MP-RAGE Groups
Tumors Amenable to Complete Resection
Among the 79 cases amenable to complete resection, the median residual tumor volume at the time of IoMRI was found to be 1.088 mL for the T2 group and 0.437 mL for the MP-RAGE group. This difference was statistically significant (P = .049). At the time of the final MRI, median residual tumor volume for the T2 group was found to be 0.166 mL, whereas the median residual tumor volume for the MP-RAGE group was found to be 0.016 mL. This difference in residual tumor volume was not statistically significant (P = .14).
The median extent of resection at the time of IoMRI was found to be 91.5% for the T2 group and 98% for the MP-RAGE group. This difference in the extent of resection was statistically significant (P = .008). At the time of the final MRI, the median extent of resection was 99.3% in the T2 group and 99.9% in the MP-RAGE group. This difference in the extent of resection also was statistically significant (P = .045).
Tumors That Underwent Additional Resection After IoMRI
Among the 54 cases in which additional resection was performed after IoMRI, the median residual tumor volume at the time of IoMRI was found to be 3.933 mL for the T2 group and 0.905 mL for the MP-RAGE group. This difference was marginally significant (P = .05). At the time of the final MRI, the median residual tumor volume for the T2 group was found to be 0.420 mL, whereas the median residual tumor volume for the MP-RAGE group was found to be 0.016 mL. This difference in residual tumor volume was not a statistically significant difference (P = .14).
The median extent of resection at the time of IoMRI was 83.7% for the T2 group and 91.8% for the MP-RAGE group. This difference in extent of resection was statistically significant (P = .006). At the time of the final MRI, the median extent of resection was 98.2% in the T2 group and 99.95% in the MP-RAGE group. This difference in extent of resection also was statistically significant (P = .047).
Cases in the T2 group in which additional resection was performed had a median reduction in residual tumor volume from the time of IoMRI to the final MRI of 2.874 mL, whereas cases in the MP-RAGE group had a median reduction in residual tumor volume of 0.882 mL. This difference in reduction in residual tumor volume was not found to be significant (P = .07).
The median change in the extent of resection from the time of IoMRI to the final MRI in the T2 group was 11.8%. The median change in the extent of resection in the MP-RAGE group was 7.0%. This difference in improvement in the extent of resection was statistically significant (P = .03).
Patients Who Underwent Intraoperative ECoG
Among the 32 patients in whom intraoperative ECoG was performed, 7 (22%) had complete resection on IoMRI, 9 (28%) had no additional resection despite having residual tumor on IoMRI due to the tumor's involvement of or proximity to functional structures confirmed on ECoG, and 16 (50%) had additional resection after IoMRI, which resulted in complete resections in 4 more patients. Hence, a total of 23 surgeries (72%) that included intraoperative ECoG were able to achieve complete resection.
Eighteen surgeries that included ECoG (56%) achieved more than 90% extent of resection, 7 surgeries (22%) achieved more than 98% extent of resection, and 7 surgeries (22%) achieved complete resection at the time of IoMRI. The final extent of resection after additional resection was performed was more than 90% in 24 surgeries (75%) and more than 98% in 13 surgeries (41%), and complete resections were achieved in 11 surgeries (34%).
Twenty-five patients had an associated complication. The most common complication was neurological deterioration, which occurred in 12 patients (11%). New or worsened preexisting deficits were temporary and resolved spontaneously over the course of days in 8 patients (7%) but were permanent in 4 (4%). After surgery, a postoperative infection developed in 7 patients (7%), 4 patients (4%) had perioperative deep venous thrombosis, and 2 (2%) experienced other medical complications postoperatively. The median operating times for patients with and without postoperative complications were 6.7 vs 6.5 hours, respectively (P = .8). No perioperative mortality occurred in this series.
The 4 patients who sustained permanent neurological deficits were in the MP-RAGE subgroup. One of these patients had a left temporal GBM with associated baseline language impairment that did not respond to steroids and experienced symptoms after surgery. This patient underwent no additional resection despite having residual tumor seen on the IoMRI due to ECoG results. The remaining 3 patients with permanent neurological deficits all underwent intraoperative ECoG and additional resection after IoMRI. The median final residual tumor volume in these 4 patients was 0.291 mL (range, 0-1.21 mL), and the median final extent of resection was 99.4% (range, 95.4%-100%).
Among the 8 patients who had a spontaneously resolving postoperative neurological deficit, 4 underwent additional resection after IoMRI. ECoG was performed in 2 of these patients who underwent additional resection and 2 who did not.
Seven surgeries (13%) that were associated with any type of postoperative neurological deficit included additional resection after IoMRI compared with 5 surgeries (10%) that did not include additional resection (P = .86). For those cases associated with a permanent deficit, 3 (6%) included additional resection and 1 (2%) did not (P = .66).
Of all surgeries with any type of postoperative neurological deterioration, ECoG was used in 7 cases (22%) compared with 5 cases (7%) in which ECoG was not used (P = .062). Among the cases associated with permanent neurological deficit, 3 included ECoG (9%) and 1 (1%) did not (P = .16).
In this series of 102 patients (104 surgeries; 56 non-minimally enhancing and 48 enhancing gliomas), we found that the use of IoMRI led to additional resection in more than half of the patients (52%). Additional resection was associated with an improvement in the median residual tumor volume from 0.779 mL at the time of IoMRI to a final median residual tumor volume of 0.061 mL in the 79 cases amenable to complete resection (P = .002) and from a median residual tumor volume of 1.22 mL at the time of IoMRI to 0.2 mL of final residual tumor volume in the entire series of 104 operations. Use of IoMRI also was associated with an improvement in the extent of resection from 94.9% at the time of IoMRI to a final extent of resection of 99.6% (P < .001) in 79 cases in which complete resection was considered feasible and from an extent of resection of 92.9% at the time of IoMRI to a final extent of resection of 99.4% in all 104 surgeries.
Multiple retrospective series have supported the hypothesis that extensive resection improves the survival of patients with GBM,11-15,36-41 anaplastic glioma,11,14 and LGGs.7,8,42-47 Many groups have evaluated the use of IoMRI to improve the extent of resection in high-grade, enhancing gliomas. Senft et al25 performed a randomized, controlled trial of patients with enhancing gliomas that compared IoMRI and conventional surgery. They reported a statistically significant improvement in the percentage of patients with CRET as well as improved progression-free survival at 6 months in the IoMRI group relative to the conventional group without any difference in postoperative neurological deficits.25 Kuhnt et al48 retrospectively reviewed GBM patients who had undergone high-field IoMRI-guided surgery and showed an increase in the CRET rate from 35% to 45% of their data group after additional resection. Kubben et al26 performed a systematic review of IoMRI-guided GBM surgery and concluded that IoMRI-guided surgery is more effective than conventional surgery in increasing the extent of resection, prolonging survival, and enhancing quality of life after resection of a GBM. Our series showed a statistically significant improvement in the median residual tumor volume after additional resection.
Lacroix et al15 analyzed their experience with GBM patients undergoing craniotomy and reported that resection of 89% or more was necessary to significantly improve survival. They further noted that the most significant survival advantage was observed in cases in which extent of resection of 98% or more could be achieved. More recently, a retrospective review of IoMRI GBM surgeries showed that an extent of resection of 98% or greater significantly improved patient survival.48 In our study, we found that among patients in the MP-RAGE group amenable to CRET, the number of patients with an extent of resection greater than 98% increased from 18 patients (49%) at the time of IoMRI to 35 patients (95%) on the final MRI. Also, we found that the number of patients with CRET increased from 14 (38%) at the time of IoMRI to 30 (81%) on final MRI. Given the relationship between increments of extent of resection and survival that has been demonstrated by other groups, our results suggest that use of IoMRI to improve the extent of resection produced an increased number of patients in better prognostic groups.14,15,33
Several studies evaluated the use of IoMRI for resection of LGGs. Claus et al4 reported that IoMRI-guided surgical resection of LGGs resulted in CRDT in 36% of patients at a mean extent of resection of more than 80%. They also suggested a possible association between the extent of resection and survival after a median follow-up of 3 years. Pamir et al35 reported results of the 3-T IoMRI-guided surgery in LGG patients and showed that IoMRI increased the number of patients with CRDT from 55% at IoMRI to 73% at final MRI. In our series of non- or minimally enhancing tumors, additional resection after IoMRI produced significantly improved median residual tumor volume from 3.933 mL at IoMRI to a final residual tumor volume of 0.419 mL (P = .01).
The relationship between the extent of resection and survival in LGG is more controversial than that proposed for GBM. In a review of recently published literature on glioma surgery, Sanai and Berger49 identified a threshold of 75% to 100% extent of resection to have a significant impact on survival in 5 studies that used volumetric image analysis techniques. Among the 42 patients with LGG amenable to CRDT in our dataset, we observed an increase of 48% of cases with an extent of resection greater than 90% after additional resection as well as a 19% increase in the number of patients with CRDT. Our result stands in contrast to those of the study by Hatiboglu et al,32 which failed to show that the use of IoMRI resulted in a statistically significant increase in the extent of resection of nonenhancing tumors. However, the Hatiboglu et al study included only 7 patients amenable to complete resection, and this small sample size may have been an unreliable indicator of the more general utility of IoMRI.
No study to date directly compared the relative benefit of IoMRI-guided surgery in the setting of nonenhancing and enhancing gliomas. Hatiboglu et al32 reported results of IoMRI-guided surgery in 44 patients with 31 enhancing and 13 nonenhancing gliomas, but they did not directly compare the 2 groups with each other. Kuhnt et al33 reported the results of 293 IoMRI-guided surgeries for grade I to IV gliomas, but their analysis was based on WHO grade rather than enhancement. Overall, 76 patients (26%) needed additional resection after IoMRI, which resulted in improvement in the extent of resection from 63% at IoMRI to 94% at final MRI in WHO grade I gliomas (P < .01), from 76% to 89% in WHO grade II gliomas (P < .01), from 65% to 75% in WHO grade III gliomas (P < .01), and from 66% to 99% in GBMs (P < .01). In our series, we performed a direct comparison of the T2 and MP-RAGE groups that revealed that the intraoperative median residual tumor volume was higher in the T2 group than in the MP-RAGE group (1.088 mL in T2 and 0.437 mL in MP-RAGE; P = .049) and that the final residual tumor volumes showed no statistically significant difference (0.165 mL in T2 and 0.016 mL in MP-RAGE; P = .14). Our findings could indicate that IoMRI facilitates improved resection of non- and minimally enhancing tumors more so than of enhancing tumors.
It has been well recognized that tumors in or adjacent to eloquent area of the brain are less amenable to complete resection. Orringer et al38 reported a lesser extent of resection for tumors located in eloquent areas compared with that of tumors in noneloquent locations in a series of conventional craniotomies for GBM. However, aggressive resection is associated with a risk of at least temporary neurological deterioration. In a supplemental analysis to their study of the use of 5-ALA for resection of gliomas, Stummer et al50 used the National Institute of Health Stroke Scale score and the KPS score to assess postoperative neurological function status. A deterioration of 1 point or more of National Institute of Health Stroke Scale score at 48 hours was observed in 26% of the 5-ALA group, whereas only 14% of the control group had a similar deterioration (P = .002). By postoperative day 7, however, the 5-ALA group had recovered to a point where there was no longer a statistically significant difference in neurological functioning.
Intraoperative neurophysiological monitoring, diffusion tensor imaging MRI, and fiber tracking, and neuronavigation are tools that are used to improve the maximally safe extent of resection of tumors located in eloquent areas.19,51,52 The combined use of IoMRI and intraoperative ECoG may be considered as an optimal strategy for improving the extent of resection without necessarily increasing the risk of new or worsened neurological deficit. Senft et al53 reported a series of patients who had tumors located in eloquent cortex that were resected with use of IoMRI and ECoG. This study found no correlation between occurrence of new or severe neurological deficits and the decision to perform additional resection after IoMRI. In the current study, we performed intraoperative ECoG in 32 patients, of whom 23 were classified as amenable to complete resection after ECoG. Eleven patients (34%) ultimately had a complete resection, 4 of them after additional resection. Overall, additional resection after IoMRI was performed in 16 of the patients (50%) who underwent ECoG evaluation. There was no statistically significant correlation between the onset of permanent neurological deficits and the decision to perform additional resection after IoMRI or intraoperative ECoG. When temporary and permanent neurological deficits were considered together (12 patients), there remained no statistically significant relationship between the use of IoMRI and the decision to perform additional resection, but there was a trend toward an increased incidence of neurological deficits in patients who also required intraoperative ECoG (22% in ECoG group vs 7% in others; P = .062). We believe that the combination of IoMRI and intraoperative ECoG, which was used in one-third of the patients in our series, optimizes the surgeon's ability to achieve a maximal safe resection of gliomas.
There are obvious limitations related to the retrospective nature of our study. This is a mixed group of glioma patients with a range of WHO grades (grades II, III, and IV) and consequently different expectations from surgery. We have not performed an analysis of the impact of resection on survival. The heterogeneity of WHO grades in our series and the short follow-up time would make it difficult to meaningfully relate the extent of resection or residual tumor volume to outcomes in this series of patients. Additionally, this study includes patients operated on by 5 different senior surgeons (G.H.B., L.A., K.K., V.R., M.A.V.), and each is likely to have his or her own threshold for terminating the resection of a tumor that involves or is adjacent to functional cortex. In the surgeries of 25 patients, no further resection was performed despite the appearance of residual volume on IoMRI. There were no prospectively and uniformly adopted criteria that determined whether further resection could have or should have been done. Rather than viewing this lack of uniformity as a limitation of the study, instead we believe that our collective ability to achieve near-complete or complete resections in the majority of patients indicates that IoMRI is a useful tool in the hands of surgeons who have a wide range of experience performing glioma surgery and not just a specialized tool best used by only the most experienced glioma surgeons.
IoMRI appears to be an important tool that allows neurosurgeons to maximize the extent of surgical resection of gliomas, particularly for lower grade (nonenhancing) tumors. Despite our success in maximizing the extent of resection and minimizing residual tumor volume, we did not observe an increased rate of new or worsened permanent neurological deficits. We attribute this lack of additional morbidity to our use of preoperative functional mapping and intraoperative ECoG in cases of tumors located in eloquent areas.
The authors have no personal financial or institutional interest in any of the drugs, materials, or devices described in this article.
Many thanks to Christine Moore for her editorial help in preparing this manuscript.
1. Black P. Management of malignant glioma: role of surgery in relation to multimodality therapy. J Neurovirol. 1998;4(2):227–236.
2. Woodward DE, Cook J, Tracqui P, Cruywagen GC, Murray JD, Alvord EC Jr. A mathematical model of glioma growth. Cell Prolif. 1996;29(6):269–288.
3. Sanai N, Berger MS. Glioma extent of resection and its impact on patient outcome. Neurosurgery. 2008;62(4):753–766.
4. Claus EB, Horlacher A, Hsu L, et al.. Survival rates in patients with low-grade glioma after intraoperative magnetic resonance image guidance. Cancer. 2005;103(6):1227–1233.
5. Hardesty DA, Sanai N. The value of glioma extent of resection in the modern neurosurgical era. Front Neurol. 2012;3:140.
6. van Veelen ML, Avezaat CJ, Kros JM, van Putten W, Vecht C. Supratentorial low grade astrocytoma: prognostic factors, dedifferentiation, and the issue of early versus late surgery. J Neurol Neurosurg Psychiatry. 1998;64(5):581–587.
7. Smith JS, Chang EF, Lamborn KR, et al.. Role of extent of resection in the long-term outcome of low-grade hemispheric gliomas. J Clin Oncol. 2008;26(8):1338–1345.
8. McGirt MJ, Chaichana KL, Attenello FJ, et al.. Extent of surgical resection is independently associated with survival in patients with hemispheric infiltrating low-grade gliomas. Neurosurgery. 2008;63(4):700–707; author reply 707-708.
9. Karim AB, Maat B, Hatlevoll R, et al.. A randomized trial on dose-response in radiation therapy of low-grade cerebral glioma: European Organization for Research and Treatment of Cancer (EORTC) Study 22844. Int J Radiat Oncol Biol Phys. 1996;36(3):549–556.
10. Knauth M, Wirtz CR, Tronnier VM, Aras N, Kunze S, Sartor K. Intraoperative MR imaging increases the extent of tumor resection in patients with high-grade gliomas. AJNR Am J Neuroradiol. 1999;20(9):1642–1646.
11. Keles GE, Chang EF, Lamborn KR, et al.. Volumetric extent of resection and residual contrast enhancement on initial surgery as predictors of outcome in adult patients with hemispheric anaplastic astrocytoma. J Neurosurg. 2006;105(1):34–40.
12. Sanai N, Polley MY, McDermott MW, Parsa AT, Berger MS. An extent of resection threshold for newly diagnosed glioblastomas. J Neurosurg. 2011;115(1):3–8.
13. Stummer W. Extent of resection and survival in glioblastoma multiforme: identification of and adjustment for bias. Neurosurgery. 2008;62(3):564–576.
14. McGirt MJ, Chaichana KL, Gathinji M, et al.. Independent association of extent of resection with survival in patients with malignant brain astrocytoma. J Neurosurg. 2009;110(1):156–162.
15. Lacroix M, Abi-Said D, Fourney DR, et al.. A multivariate analysis of 416 patients with glioblastoma multiforme: prognosis, extent of resection, and survival. J Neurosurg. 2001;95(2):190–198.
16. Brown PD, Maurer MJ, Rummans TA, et al.. A prospective study of quality of life in adults with newly diagnosed high-grade gliomas: the impact of the extent of resection on quality of life and survival. Neurosurgery. 2005;57(3):495–504.
17. Hammoud MA, Ligon BL, elSouki R, Shi WM, Schomer DF, Sawaya R. Use of intraoperative ultrasound for localizing tumors and determining the extent of resection: a comparative study with magnetic resonance imaging. J Neurosurg. 1996;84(5):737–741.
18. Engle DJ, Lunsford LD. Brain tumor resection guided by intraoperative computed tomography. J Neurooncol. 1987;4(4):361–370.
19. Barnett GH. The role of image-guided technology in the surgical planning and resection of gliomas. J Neurooncol. 1999;42(3):247–258.
20. Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F, Reulen HJ. Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: a randomised controlled multicentre phase III trial. Lancet Oncol. 2006;7(5):392–401.
21. Schebesch KM, Proescholdt M, Höhne J, et al.. Sodium fluorescein-guided resection under the YELLOW 560 nm surgical microscope filter in malignant brain tumor surgery—a feasibility study. Acta Neurochir(Wien). 2013;155(4):693–699.
22. Roessler K, Ungersboeck K, Aichholzer M, et al.. Frameless stereotactic lesion contour-guided surgery using a computer-navigated microscope. Surg Neurol. 1998;49(3):282–288; discussion 288-289.
23. Lewin JS. Interventional MR imaging: concepts, systems, and applications in neuroradiology. AJNR Am J Neuroradiol. 1999;20(5):735–748.
24. Vogelbaum MA, Jost S, Aghi MK, et al.. Application of novel response/progression measures for surgically delivered therapies for gliomas: Response Assessment in Neuro-Oncology (RANO) working group. Neurosurgery. 2012;70(1):234–243; discussion 243-244.
25. Senft C, Bink A, Franz K, Vatter H, Gasser T, Seifert V. Intraoperative MRI guidance and extent of resection in glioma surgery: a randomised, controlled trial. Lancet Oncol. 2011;12(11):997–1003.
26. Kubben PL, ter Meulen KJ, Schijns OE, ter Laak-Poort MP, van Overbeeke JJ, van Santbrink H. Intraoperative MRI-guided resection of glioblastoma multiforme: a systematic review. The Lancet Oncol. 2011;12(11):1062–1070.
27. Lenaburg HJ, Inkabi KE, Vitaz TW. The use of intraoperative MRI for the treatment of glioblastoma multiforme. Technol Cancer Res Treat. 2009;8(2):159–162.
28. Schneider JP, Trantakis C, Rubach M, et al.. Intraoperative MRI to guide the resection of primary supratentorial glioblastoma multiforme—a quantitative radiological analysis. Neuroradiology. 2005;47(7):489–500.
29. Nimsky C, Fujita A, Ganslandt O, Von Keller B, Fahlbusch R. Volumetric assessment of glioma removal by intraoperative high-field magnetic resonance imaging. Neurosurgery. 2004;55(2):358–370.
30. Bohinski RJ, Kokkino AK, Warnick RE, et al.. Glioma resection in a shared-resource magnetic resonance operating room after optimal image-guided frameless stereotactic resection. Neurosurgery. 2001;48(4):731–742; discussion 742-744.
31. Wirtz CR, Knauth M, Staubert A, et al.. Clinical evaluation and follow-up results for intraoperative magnetic resonance imaging in neurosurgery. Neurosurgery. 2000;46(5):1112–1120; discussion 1120-1122.
32. Hatiboglu MA, Weinberg JS, Suki D, et al.. Impact of intraoperative high-field magnetic resonance imaging guidance on glioma surgery: a prospective volumetric analysis. Neurosurgery. 2009;64(6):1073–1081; discussion 1081.
33. Kuhnt D, Ganslandt O, Schlaffer SM, Buchfelder M, Nimsky C. Quantification of glioma removal by intraoperative high-field magnetic resonance imaging: an update. Neurosurgery. 2011;69(4):852–862; discussion 862-863.
34. Pamir MN, Ozduman K, Dinçer A, Yildiz E, Peker S, Ozek MM. First intraoperative, shared-resource, ultrahigh-field 3-Tesla magnetic resonance imaging system and its application in low-grade glioma resection. J Neurosurg. 2010;112(1):57–69.
35. Pamir MN, Özduman K, Yıldız E, Sav A, Dinçer A. Intraoperative magnetic resonance spectroscopy for identification of residual tumor during low-grade glioma surgery. J Neurosurg. 2013;118(6):1191–1198.
36. Keles GE, Anderson B, Berger MS. The effect of extent of resection on time to tumor progression and survival in patients with glioblastoma multiforme of the cerebral hemisphere. Surg Neurol. 1999;52(4):371–379.
37. Pope WB, Sayre J, Perlina A, Villablanca JP, Mischel PS, Cloughesy TF. MR imaging correlates of survival in patients with high-grade gliomas. AJNR Am J Neuroradiol. 2005;26(10):2466–2474.
38. Orringer D, Lau D, Khatri S, et al.. Extent of resection in patients with glioblastoma: limiting factors, perception of resectability, and effect on survival. J Neurosurg. 2012;117(5):851–859.
39. Bloch O, Han SJ, Cha S, et al.. Impact of extent of resection for recurrent glioblastoma on overall survival: clinical article. J Neurosurg. 2012;117(6):1032–1038.
40. Senft C, Franz K, Blasel S, et al.. Influence of iMRI-guidance on the extent of resection and survival of patients with glioblastoma multiforme. Technol Cancer Res Treat. 2010;9(4):339–346.
41. Dea N, Fournier-Gosselin MP, Mathieu D, Goffaux P, Fortin D. Does extent of resection impact survival in patients bearing glioblastoma? Can J Neurol Sci. 2012;39(5):632–637.
42. Philippon JH, Clemenceau SH, Fauchon FH, Foncin JF. Supratentorial low-grade astrocytomas in adults. Neurosurgery. 1993;32(4):554–559.
43. Scerrati M, Roselli R, Iacoangeli M, Pompucci A, Rossi GF. Prognostic factors in low grade (WHO grade II) gliomas of the cerebral hemispheres: the role of surgery. J Neurol Neurosurg Psychiatry. 1996;61(3):291–296.
44. Nicolato A, Gerosa MA, Fina P, Iuzzolino P, Giorgiutti F, Bricolo A. Prognostic factors in low-grade supratentorial astrocytomas: a uni-multivariate statistical analysis in 76 surgically treated adult patients. Surg Neurol. 1995;44(3):208–221; discussion 221-223.
45. Ito S, Chandler KL, Prados MD, et al.. Proliferative potential and prognostic evaluation of low-grade astrocytomas. J Neurooncol. 1994;19(1):1–9.
46. Keles GE, Lamborn KR, Berger MS. Low-grade hemispheric gliomas in adults: a critical review of extent of resection as a factor influencing outcome. J Neurosurg. 2001;95(5):735–745.
47. Ius T, Isola M, Budai R, et al.. Low-grade glioma surgery in eloquent areas: volumetric analysis of extent of resection and its impact on overall survival. A single-institution experience in 190 patients: clinical article. J Neurosurg. 2012;117(6):1039–1052.
48. 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(12):1339–1348.
49. Sanai N, Berger MS. Recent surgical management of gliomas. Adv Exp Med Biol. 2012;746:12–25.
50. Stummer W, Tonn JC, Mehdorn HM, et al.. Counterbalancing risks and gains from extended resections in malignant glioma surgery: a supplemental analysis from the randomized 5-aminolevulinic acid glioma resection study. Clinical article. J Neurosurg. 2011;114(3):613–623.
51. Berger MS, Rostomily RC. Low grade gliomas: functional mapping resection strategies, extent of resection, and outcome. J Neurooncol. 1997;34(1):85–101.
52. Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A. In vivo fiber tractography using DT-MRI data. Magn Reson Med. 2000;44(4):625–632.
53. Senft C, Forster MT, Bink A, et al.. Optimizing the extent of resection in eloquently located gliomas by combining intraoperative MRI guidance with intraoperative neurophysiological monitoring. J Neurooncol. 2012;109(1):81–90.
The authors have rightly identified an onerous challenge in glioma surgery, specifically, the intraoperative discrimination between low-grade glioma tissue and adjacent normal parenchyma. Any intraoperative adjunct, like intraoperative MRI, that might aid the identification of tumor tissue is certainly appealing to the neurosurgical community. However, careful reflection of its potential merits and drawbacks must be considered. The authors provide an in-depth description of their experience with IoMRI for all grades of infiltrative gliomas. Specifically, 104 consecutive craniotomies for resection of an intracranial glioma for which IoMRI was used to assess the extent of surgical resection were retrospectively reviewed. A volumetric analysis of residual tumor was used for both low-grade (nonenhancing) and high-grade (enhancing) tumors. Intraoperative ECoG was also used in select cases when adjacent eloquent tissue was of particular concern based on preoperative evaluation.
The authors achieved excellent resections for their series with a final EOR of 99.4%, up from 92.9% at IoMRI. The rates of resection were then compared for the enhancing and nonenhancing subgroups. They found a median residual volume of 1.088 mL (91.5% EOR) for the nonenhancing group, whereas the enhancing tumors had only 0.437 mL (98% EOR). This difference was statistically significant. The final EOR for both groups was greater than 99%.
The authors conclude that IoMRI safely maximizes the extent of glioma resection and may be particularly beneficial for low-grade, nonenhancing tumors as these tumors are more difficult to distinguish from normal brain by direct visual inspection. Interesting, although IoMRI certainly extends operating time, no relationship between length of surgery and postoperative complications was seen in this series. Although mainly descriptive, these data contribute to the growing literature regarding the impact of IoMRI on glioma management. Prospective data addressing the impact of IoMRI on patient survival remain wanting. Such studies will perhaps appear in the near future.
Devon H. Haydon
Ralph G. Dacey, Jr
St. Louis, Missouri
Intra-operative (iMRI) is increasingly employed for maximizing surgical resection of intrinsic brain tumors. In this paper the authors present a retrospective, single institution series of 104 surgical cases performed with the aid of iMRI to define the tumor margins, as well as intraoperative cortical mapping to define functional brain. They show that iMRI improves the extent of resection and reduces the amount of residual tumor for both enhancing gliomas (based on evaluation of postcontrast T1-weighted images) and nonenhancing gliomas (based on evaluation of T2/FLAIR images). Importantly, the impact of iMRI was most pronounced for the nonenhancing gliomas (WHO grade II and III tumors). This paper adds to the growing literature espousing the value of iMRI. The authors are to be congratulated for their careful analysis which emphasizes that iMRI is especially useful for nonenhancing tumors where the physical characteristics of the tumor compared with normal brain are often visually indistinct and are best defined by MR imaging. Indeed, I think most would agree that if there is an ideal application for iMRI, it is with non-enhancing grade II or III gliomas.
Given the expanding literature supporting the value of iMRI, should iMRI be considered the standard of care for resecting gliomas, particularly nonenhanincing ones? I would submit that the answer to this question is “probably not.” Ultimately, iMRI is a tool, much like computer-assisted image guidance or intraoperative ultrasound. In this context, based on current evolving literature, the surgical standard of care for enhancing intraparenchymal tumors is maximal, safe resection of the enhancing portion of the tumor, and for nonenhancing tumors the standard is maximal, safe resection of the hyperintensity as defined by T2/FLAIR-images. From this perspective, iMRI is a means to these ends and, therefore, its use is ultimately at the discretion of the surgeon.
From a personal perspective, I use iMRI very selectively. I find that for the majority of cases the anatomical borders of the tumor can be defined very precisely by the surrounding sulci, obviating the need for iMRI. Specifically, careful study of MRI images almost invariably shows that the edges of nonenhancing gliomas (as defined by the T2/FLAIR hyperintensity) correlates with the position of the surrounding sulci, probably because most nonenhancing tumors do not grow across or violate the sulci. These sulci, ie, the tumor borders, can be determined accurately in the operating room before any tumor is resected using intraoperative computer-assisted stereotactic guidance based on the preoperative MRIs. Intraoperative ultrasound also can be used as a real-time confirmation of the location of the tumor and its relationship to the surrounding sulci. Moreover, the bottoms of the sulci often define the depth of the tumor. Critical to this technique, is that the tumor is then resected in a circumferential “en bloc” fashion using the surrounding sulci as the tumor margins. When en bloc resection is employed brain shift is minimal and intraoperative computer guidance remains useful during the entire tumor resection, in most cases. iMRI is reserved for the cases where en bloc resection may be difficult, usually for large or deep tumors, features that can be ascertained preoperatively. In contradistinction to en bloc resection, resecting tumors using an “inside-out” technique may lead to a greater need for iMRI because brain shift is unavoidable as the tumor is removed.
Regardless, we must keep in mind that we really do not know whether removing the last cubic centimeter of a nonenhancing glioma actually impacts patient survival. This outcome can only be evaluated in a prospective randomized trial. Also critical to this discussion is the impact of the molecular biology on the extent of resection and the subsequent treatment. Indeed, the IDH-1 status, CIMP status, 1p/19q co-deletion status, and the p53 mutation status are known to influence survival and responsiveness to therapy. How these molecular features influence outcome after varying degrees of resection needs to be ascertained. Clearly there remains much to be learned, but fortunately due to careful studies such as the one presented here, the future of glioma treatment looks increasingly bright.
Frederick F. Lang
What is the possible outcome of the use of intraoperative MRI during surgical resection of a primary brain tumor?
A. Increased rate of complete resection of enhancing tumor (CRET)
B. Decrease in six month progression free survival (PFS)
C. Increase in postoperative deficits
D. No change in quality of life
What extent of surgical resection of low grade gliomas contributes to a survival benefit?
What is an adjunctive intraoperative tool that has been shown to improve maximum safe resection for primary brain tumors?
A. Somatosensory evoked potentials (SSEPs)
B. Electrocorticography ECoG
C. Cerebral angiography
D. Indocyanine green (ICG) injection
Electrocorticography; Extent of resection; Glioblastoma; Low-grade glioma; Residual tumor volume
Figure. No available...Image Tools
Copyright © by the Congress of Neurological Surgeons
Highlight selected keywords in the article text.