Wade Mueller

John L. Ulmer

Surgical excision of tumors is important to prolong survival, reduce steroid dependence, improve neurologic function (in up to 50% of glioma patients), and establish histologic diagnoses that determine postsurgical treatment algorithms. Advanced functional imaging MR-based techniques, such as fMRI and DTI, have shown applications in establishing operative risks and guiding operative strategies for patients with brain tumors and other lesions. Presurgical mapping can be useful in determining operability, guiding surgical trajectory, and establishing functional resection boundaries. The overriding goal of the application is to minimize operative neurologic complications, reported to range from 7–26% without the benefits of presurgical mapping.^1–9 Limitations of blood oxygen level–dependent fMRI and DTI notwithstanding, a neurosurgeon’s goal is to find ways to use presurgical mapping to optimize postoperative outcomes, despite all of the technique vulnerabilities discussed in this News Digest.

The limitations of intraoperative functional localization have paved the way for the clinical translation of presurgical fMRI and DTI. Intraoperative functional assessments can be obtained only in patients with sufficiently preserved network function and with a willingness to undergo such procedures. Even when intended, electrocortical mapping may induce seizures in perilesional hyperexcitable cortex, rendering functional assessments nondiagnostic. Also, cortex lining deep sulci may not be accessible to intraoperative stimulation. Presurgical fMRI offers localization data that may be especially important in patients where eloquent cortex cannot be fully mapped intraoperatively. The lack of gross anatomic features distinguishing functional white matter networks is even more problematic in establishing functional resection boundaries. Some surgeons will use intraoperative electrical stimulation to map white matter structures, but this requires significant expertise. At the Medical College of Wisconsin (MCW) and Froedtert Hospital, an ultrasonic aspirator is used to dissect, fragment, and aspirate tumor tissue. The device has an unintended but useful effect of stunning neurologic tissue within a few millimeters from the dissection plane, resulting in a transient deficit of adjacent functional brain tissue. This effect provides a mechanism to establish functional resection boundaries of white matter networks intraoperatively. The advantage of DTI is the ability to establish lesion border-risk designations that can guide intraoperative testing.

The combined use of DTI, fMRI, and anatomic imaging can provide preoperative estimates of functional networks. A quality assurance review of 70 patients with presurgical mapping and surgery at our institution revealed a positive predictive value (PPV) of 55–60% for surgically-induced transient postoperative language and motor deficits related to postoperative edema. This PPV was, of course, reduced by a prior knowledge provided to the surgeons. However, the negative predictive value (NPV) for eloquent functional systems at risk was 100% in this population of patients. Additionally, the combined use of fMRI and DTI provided a superior receiver operating characteristic curve than did fMRI alone. During the initial translation at our institution of combined fMRI/DTI presurgical network mapping, postoperative neurologic outcomes were studied as part of our translational quality review. This included 33 left-dominant, high-risk posterior frontal lobe tumors that were resected. The postoperative neurologic outcomes of 18 consecutive patients resected just prior to presurgical fMRI/DTI translation and 15 patients resected just after presurgical fMRI/DTI translation were compared. New speech or motor deficits occurred in 44% of the former group and 47% of the latter group. However, permanent deficits (at 1 month postoperative) were present in 39% of those not mapped with combined fMRI/DTI, but in only 7% of those with presurgical fMRI/DTI. This translated to a significantly (P < .05) better recovery in neurologic function with preoperative fMRI and DTI in patients with left frontal lobe tumors, compared with age-, gender-, histology-, tumor size-, and location-matched controls using identical neurosurgical techniques with the same neurosurgeon. Since that initial translation of presurgical mapping, with refinements of our approach, postoperative neurologic complication rates have dropped even further.

Because localization data are imperfect, integrating information from all available sources, including fMRI and DTI, is critical. There are 5 complementary preoperative and perioperative localization sources necessary to establish risks and to preserve neurologic functions. These are:

1. Clinical deficits: A presenting clinical deficit indicates lesion proximity to functional networks and greater risk of postoperative deficits, as well as the potential for functional improvements. An awareness of the clinical presentation improves the interpretation of imperfect presurgical mapping data.
2. Functional anatomy at standard MRI: Anatomic localization empowers functional localization, especially when fMRI and/or DTI are compromised by pathology. Anatomic analysis alone may even obviate the need for presurgical mapping.
3. Presurgical mapping (ie, fMRI, DTI, MEG, WADA, TMS, etc): The combined use of fMRI and DTI is superior to either modality used alone in establishing lesion border-risk designations, providing a better estimation of functional networks.¹⁰ Also, fMRI commonly informs the significance of white matter border-risk designations shown at DTI. For example, hemispheric language dominance at fMRI can determine the significance of perilesional association of white matter bundles (ie, SLF, IFOF, UF). Likewise, retinotopic vision cortex mapping can help to identify the location of central visual field optic radiations.
4. Intraoperative electrocortical and white matter stimulation
5. Intraoperative functional white matter testing: Intraoperative mapping techniques provide crucial alternative strategies to localize functional brain networks, and help to validate preoperative assessments. Conversely, presurgical mapping commonly guides intraoperative assessments, especially DTI, in white matter functional testing. Presurgical and intraoperative functional mapping techniques are indeed complementary.

In summary, each functional localization strategy, including fMRI and DTI, can help to compensate for the limitations of others. Congruence of functional data increases the accuracy of surgical neurologic risk assessments. Functional data discordance requires confirmation of functional relationships with alternative techniques. Integrating functional localization parameters, including but not limited to fMRI and DTI, has facilitated successful clinical translation of these techniques. Our experience with presurgical mapping and intraoperative utilization of mapping data in over 800 patients indicates that the approach can significantly influence neurosurgical decision-making and improve patient outcomes.

References:

1. Chang S, Parney IF, McDermott M, et al. Perioperative complications and neurological outcome of first versus second craniotomy among patients enrolled in the glioma outcomes project. J Neurosurg 2003;98:1175–81
2. Ciric I, Ammirati M, Vick N, et al. Supratentorial gliomas: surgical considerations and immediate postoperative results. Gross total resection versus partial resection. Neurosurgery 1987;21:21–26
3. Fadul C, Wood J, Thaler H, et al. Morbidity and mortality of craniotomy for excision of supratentorial gliomas. Neurology 1988;38:1374–79, 10.1212/WNL.38.9.1374
4. Sawaya R, Hammoud M, Schoppa D, et al. Neurosurgical outcomes in a modern series of 400 craniotomies for treatment of parenchymal tumors. Neurosurgery 1998;42:1044–55
5. Brell M, Ibanez J, Caral L, et al. Factors influencing surgical complications of intra-axial brain tumors. Acta Neurochir (Wien) 2000;142:739–50, 10.1007/s007010070088
6. Deveaux BC, O’Fallon JR, Kelly PR. Resection, biopsy, and survival in malignant glial neoplasms: a retrospective study of clinical parameters, therapy, and outcome. J Neurosurg 1993;78:767–75
7. Vorster SJ, Barnett GH. A proposed preoperative grading scheme to assess risk for surgical resection of primary and secondary intraaxial brain tumors. Neurosurg Focus 1998;4:Article 2
8. Taylor MD, Berstein M. Awake craniotomy with brain mapping as the routine surgical approach to treating patients with supratentorial intraaxial tumors: a prospective trial of 200 cases. J Neurosurg 1999;90:35–41
9. Bohinski RJ, Kokkino AK, Warnick RE, et al. Glioma resection in a shared resource operating room after optimal image-guided frameless stereotactic resection. Neurosurgery 2001;48:731–42
10. Ulmer JL, Salvan CV, Mueller WM, et al. The role of diffusion tensor imaging in establishing the proximity of tumor borders to functional brain systems: implications for preoperative risk assessments and postoperative outcomes. Technol Cancer Res Treat 2004;3:567–76