EDITORIAL
Resection of gliomas around language areas Can fMRI contribute?
Larry Junck, MD Shawn L. Hervey-Jumper, MD Oren Sagher, MD
Correspondence to Dr. Junck: [email protected] Neurology® 2015;84:550–551
Approximately 19,000 gliomas are diagnosed in the United States each year, most ultimately leading to death.1 Resection improves outcome, especially for patients with low-grade gliomas.2,3 These tumors commonly involve eloquent areas, and the risk of injuring eloquent brain regions limits the extent of resection in many patients. The challenge is greatest for tumors near the perisylvian language network due to the wide variability in functional anatomy of language sites and changes in functional anatomy induced by the gliomas. The best established technique for assessing eloquent cortex to guide resection is direct cortical stimulation (DCS).4,5 With this approach, the patient is generally anesthetized during opening, allowed to awaken during DCS and for most of the resection, then anesthetized again for closing. With DCS, lowcurrent stimulation of the brain creates a transient localized lesion, and testing of language function during DCS can help assess the site’s importance in language function. Thus, DCS identifies and helps to preserve important sites of language function. DCS is a well-established technique for guiding resection.4 It has high face validity and is uniformly agreed to be of substantial benefit. However, it has important limitations. (1) Some patients cannot tolerate it due to anxiety or other factors. Even the most cooperative patients will tire, limiting the number of sites and language modalities assessed. (2) Usefulness may be limited with moderate or severe baseline language impairment, yet preservation of partial language function is important in these patients. (3) There is a risk of inducing seizures. Although this risk can be mitigated by antiepileptic drugs and by direct cortical application of cold solutions, seizures can interfere with further DCS testing. (4) Results may differ depending on currents used. Low currents (e.g., ,3 mA) can lead to false-negative results; high currents (.5–8 mA) increase risk of seizures. New techniques provide different approaches to cortical mapping, including functional MRI (fMRI), PET, transcranial magnetic stimulation, and diffusion
tensor imaging (DTI). Of these, the greatest amount of experience is with fMRI. In this issue of Neurology®, Kuchcinski et al.6 retrospectively compared results of DCS and fMRI for cortical mapping in 40 patients undergoing resection of gliomas (24, low-grade; 16, high-grade). fMRI was performed at 3T using 3 language tasks: letter word generation, category word generation, and semantic association. Results of DCS and fMRI were compared, first by mapping DCS sites to a 1 3 1 cm virtual cortical grid and then assessing fMRI activation, using DCS as the gold standard, within squares of the grid. Of 2,114 stimulated sites, 103 were positive for language function during DCS. Adding the 3 language tasks together, sensitivity of fMRI was disappointingly low at 37%, while specificity was 83%. Using only DCS and not fMRI to guide resections, the authors found worse neurologic function postoperatively in 78% of patients, a rate higher than reported in many other published series,4,5 but function returned to baseline within 6 months in nearly all. Based on these and other results, fMRI is not ready to use as a stand-alone technique for language mapping to guide resections. Methods are needed to improve its sensitivity. In this study, low sensitivity was related to the stringent statistical threshold (p , 0.001). The authors found that a more lenient threshold (p , 0.05) was associated with higher sensitivity of 63% but lower specificity of 66%. However, considering use of fMRI to guide resection, sensitivity is the more important measure, as high sensitivity for detection of essential language areas is necessary to protect language function. Low specificity could lead the neurosurgeon to avoid resecting areas that might otherwise have been safely resected, arguably a more acceptable limitation. Other studies have found higher sensitivity,7,8 but as Kuchcinski et al. point out, a factor contributing to this difference is that some of these studies used a less-stringent requirement for spatial match-up between DCS and fMRI. As mentioned by the authors, fMRI may face a theoretical limit in its use for guiding surgical resection insofar as it demonstrates areas that participate
See page 560 From the Department of Neurology (L.J.), Comprehensive Cancer Center (L.J., S.L.H.-J., O.S.) and the Department of Neurosurgery (S.L.H.-J., O.S.), University of Michigan, Ann Arbor. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the editorial. 550
© 2015 American Academy of Neurology
ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
in language tasks, not those essential to performing these tasks. While this may be an advantage in studying language networks, it seems a disadvantage for guiding resections. Considering their results further, it is likely that fMRI results are limited by tumor-induced partial uncoupling of vascular reactivity (brain oxygen level– dependent effect) from neuronal activity.9,10 Differences in the degree of uncoupling may account for the authors’ findings of marginally higher falsenegative fMRI results with high-grade tumors, also of lower false-positive fMRI with high tumor regional cerebral blood volume (rCBV), low cortical rCBV, diagnosis other than astrocytoma, and distance to tumor ,1 cm. How could results using fMRI be improved? The main goal would not be lower incidence of language deficits, which are already at a fairly low and acceptable rate with DCS-guided resections. Rather, the goal would be development of a technique that circumvents the limitations of DCS, mentioned above. Possibilities include the following: (1) fMRI with stronger magnets and more uniform magnetic fields. (2) Different approaches to fMRI statistical analysis. Receiver operating characteristic analysis of wholebrain statistical thresholds between 0.05 and 0.001 might help to find the best threshold. (3) Exploration of other language stimulation paradigms. Optimal tasks may differ for different brain regions, e.g., posterior and anterior language areas. (4) Logically, activation by multiple tasks may help to define an area as essential. In addition to considering an area to be positive for language function if activation using any of several tasks meets a high threshold (“or” approach), as done by Kuchcinski et al., one could consider an area positive for language function if activation using more than one task meets a lower threshold (“and” approach). The present article by Kuchcinski et al. is a valuable contribution, but the results confirm that fMRI is not ready for prime time in guiding glioma resection.
STUDY FUNDING No targeted funding reported.
DISCLOSURE L. Junck receives clinical trial support from Celldex Therapeutics. S.L. Hervey-Jumper and O. Sagher report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.
REFERENCES 1. Ostrom QT, Gittleman H, Farah P, et al. CBTRUS statistical report: primary brain and central nervous tumors diagnosed in the United States 2006–2010. Neuro Oncol 2013;15:ii1–ii56. 2. Capelle L, Fontaine D, Mandonnet E, et al. Spontaneous and therapeutic prognostic factors in adult hemispheric World Health Organization Grade II gliomas: a series of 1,097 cases: clinical article. J Neurosurg 2013;118: 1157–1168. 3. 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;26:1338–1345. 4. De Witt Hamer PC, Robles SG, Zwinderman AH, Duffau H, Berger MS. Impact of intraoperative stimulation brain mapping on glioma surgery outcome: a metaanalysis. J Clin Oncol 2012;30:2559–2565. 5. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after language mapping for glioma resection. N Engl J Med 2008;358:18–27. 6. Kuchcinski G, Mellerio C, Pallud J, et al. Three-tesla functional MR language mapping: comparison with direct cortical stimulation in gliomas. Neurology 2015;84: 560–568. 7. Bizzi A, Blasi V, Falini A, et al. Presurgical functional MR imaging of language and motor functions: validation with intraoperative electrocortical mapping. Radiology 2008; 248:579–589. 8. FitzGerald DB, Cosgrove GR, Ronner S, et al. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. AJNR Am J Neuroradiol 1997;18:1529–1539. 9. Schreiber A, Hubbe U, Ziyeh S, Hennig J. The influence of gliomas and nonglial space-occupying lesions on bloodoxygen-level-dependent contrast enhancement. AJNR Am J Neuroradiol 2000;21:1055–1063. 10. Wang L, Chen D, Olson J, Ali S, Fan T, Mao H. Reexamine tumor-induced alterations in hemodynamic responses of BOLD fMRI: implications in presurgical brain mapping. Acta Radiol 2012;53:802–811.
Neurology 84
February 10, 2015
551
ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Resection of gliomas around language areas: Can fMRI contribute? Larry Junck, Shawn L. Hervey-Jumper and Oren Sagher Neurology 2015;84;550-551 Published Online before print January 14, 2015 DOI 10.1212/WNL.0000000000001241 This information is current as of January 14, 2015 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/84/6/550.full.html
References
This article cites 10 articles, 6 of which you can access for free at: http://www.neurology.org/content/84/6/550.full.html##ref-list-1
Subspecialty Collections
This article, along with others on similar topics, appears in the following collection(s): fMRI http://www.neurology.org//cgi/collection/fmri Primary brain tumor http://www.neurology.org//cgi/collection/primary_brain_tumor
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 2015 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
Resection of gliomas around language areas Can fMRI contribute?
Larry Junck, MD Shawn L. Hervey-Jumper, MD Oren Sagher, MD
Correspondence to Dr. Junck: [email protected] Neurology® 2015;84:550–551
Approximately 19,000 gliomas are diagnosed in the United States each year, most ultimately leading to death.1 Resection improves outcome, especially for patients with low-grade gliomas.2,3 These tumors commonly involve eloquent areas, and the risk of injuring eloquent brain regions limits the extent of resection in many patients. The challenge is greatest for tumors near the perisylvian language network due to the wide variability in functional anatomy of language sites and changes in functional anatomy induced by the gliomas. The best established technique for assessing eloquent cortex to guide resection is direct cortical stimulation (DCS).4,5 With this approach, the patient is generally anesthetized during opening, allowed to awaken during DCS and for most of the resection, then anesthetized again for closing. With DCS, lowcurrent stimulation of the brain creates a transient localized lesion, and testing of language function during DCS can help assess the site’s importance in language function. Thus, DCS identifies and helps to preserve important sites of language function. DCS is a well-established technique for guiding resection.4 It has high face validity and is uniformly agreed to be of substantial benefit. However, it has important limitations. (1) Some patients cannot tolerate it due to anxiety or other factors. Even the most cooperative patients will tire, limiting the number of sites and language modalities assessed. (2) Usefulness may be limited with moderate or severe baseline language impairment, yet preservation of partial language function is important in these patients. (3) There is a risk of inducing seizures. Although this risk can be mitigated by antiepileptic drugs and by direct cortical application of cold solutions, seizures can interfere with further DCS testing. (4) Results may differ depending on currents used. Low currents (e.g., ,3 mA) can lead to false-negative results; high currents (.5–8 mA) increase risk of seizures. New techniques provide different approaches to cortical mapping, including functional MRI (fMRI), PET, transcranial magnetic stimulation, and diffusion
tensor imaging (DTI). Of these, the greatest amount of experience is with fMRI. In this issue of Neurology®, Kuchcinski et al.6 retrospectively compared results of DCS and fMRI for cortical mapping in 40 patients undergoing resection of gliomas (24, low-grade; 16, high-grade). fMRI was performed at 3T using 3 language tasks: letter word generation, category word generation, and semantic association. Results of DCS and fMRI were compared, first by mapping DCS sites to a 1 3 1 cm virtual cortical grid and then assessing fMRI activation, using DCS as the gold standard, within squares of the grid. Of 2,114 stimulated sites, 103 were positive for language function during DCS. Adding the 3 language tasks together, sensitivity of fMRI was disappointingly low at 37%, while specificity was 83%. Using only DCS and not fMRI to guide resections, the authors found worse neurologic function postoperatively in 78% of patients, a rate higher than reported in many other published series,4,5 but function returned to baseline within 6 months in nearly all. Based on these and other results, fMRI is not ready to use as a stand-alone technique for language mapping to guide resections. Methods are needed to improve its sensitivity. In this study, low sensitivity was related to the stringent statistical threshold (p , 0.001). The authors found that a more lenient threshold (p , 0.05) was associated with higher sensitivity of 63% but lower specificity of 66%. However, considering use of fMRI to guide resection, sensitivity is the more important measure, as high sensitivity for detection of essential language areas is necessary to protect language function. Low specificity could lead the neurosurgeon to avoid resecting areas that might otherwise have been safely resected, arguably a more acceptable limitation. Other studies have found higher sensitivity,7,8 but as Kuchcinski et al. point out, a factor contributing to this difference is that some of these studies used a less-stringent requirement for spatial match-up between DCS and fMRI. As mentioned by the authors, fMRI may face a theoretical limit in its use for guiding surgical resection insofar as it demonstrates areas that participate
See page 560 From the Department of Neurology (L.J.), Comprehensive Cancer Center (L.J., S.L.H.-J., O.S.) and the Department of Neurosurgery (S.L.H.-J., O.S.), University of Michigan, Ann Arbor. Go to Neurology.org for full disclosures. Funding information and disclosures deemed relevant by the authors, if any, are provided at the end of the editorial. 550
© 2015 American Academy of Neurology
ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
in language tasks, not those essential to performing these tasks. While this may be an advantage in studying language networks, it seems a disadvantage for guiding resections. Considering their results further, it is likely that fMRI results are limited by tumor-induced partial uncoupling of vascular reactivity (brain oxygen level– dependent effect) from neuronal activity.9,10 Differences in the degree of uncoupling may account for the authors’ findings of marginally higher falsenegative fMRI results with high-grade tumors, also of lower false-positive fMRI with high tumor regional cerebral blood volume (rCBV), low cortical rCBV, diagnosis other than astrocytoma, and distance to tumor ,1 cm. How could results using fMRI be improved? The main goal would not be lower incidence of language deficits, which are already at a fairly low and acceptable rate with DCS-guided resections. Rather, the goal would be development of a technique that circumvents the limitations of DCS, mentioned above. Possibilities include the following: (1) fMRI with stronger magnets and more uniform magnetic fields. (2) Different approaches to fMRI statistical analysis. Receiver operating characteristic analysis of wholebrain statistical thresholds between 0.05 and 0.001 might help to find the best threshold. (3) Exploration of other language stimulation paradigms. Optimal tasks may differ for different brain regions, e.g., posterior and anterior language areas. (4) Logically, activation by multiple tasks may help to define an area as essential. In addition to considering an area to be positive for language function if activation using any of several tasks meets a high threshold (“or” approach), as done by Kuchcinski et al., one could consider an area positive for language function if activation using more than one task meets a lower threshold (“and” approach). The present article by Kuchcinski et al. is a valuable contribution, but the results confirm that fMRI is not ready for prime time in guiding glioma resection.
STUDY FUNDING No targeted funding reported.
DISCLOSURE L. Junck receives clinical trial support from Celldex Therapeutics. S.L. Hervey-Jumper and O. Sagher report no disclosures relevant to the manuscript. Go to Neurology.org for full disclosures.
REFERENCES 1. Ostrom QT, Gittleman H, Farah P, et al. CBTRUS statistical report: primary brain and central nervous tumors diagnosed in the United States 2006–2010. Neuro Oncol 2013;15:ii1–ii56. 2. Capelle L, Fontaine D, Mandonnet E, et al. Spontaneous and therapeutic prognostic factors in adult hemispheric World Health Organization Grade II gliomas: a series of 1,097 cases: clinical article. J Neurosurg 2013;118: 1157–1168. 3. 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;26:1338–1345. 4. De Witt Hamer PC, Robles SG, Zwinderman AH, Duffau H, Berger MS. Impact of intraoperative stimulation brain mapping on glioma surgery outcome: a metaanalysis. J Clin Oncol 2012;30:2559–2565. 5. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after language mapping for glioma resection. N Engl J Med 2008;358:18–27. 6. Kuchcinski G, Mellerio C, Pallud J, et al. Three-tesla functional MR language mapping: comparison with direct cortical stimulation in gliomas. Neurology 2015;84: 560–568. 7. Bizzi A, Blasi V, Falini A, et al. Presurgical functional MR imaging of language and motor functions: validation with intraoperative electrocortical mapping. Radiology 2008; 248:579–589. 8. FitzGerald DB, Cosgrove GR, Ronner S, et al. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. AJNR Am J Neuroradiol 1997;18:1529–1539. 9. Schreiber A, Hubbe U, Ziyeh S, Hennig J. The influence of gliomas and nonglial space-occupying lesions on bloodoxygen-level-dependent contrast enhancement. AJNR Am J Neuroradiol 2000;21:1055–1063. 10. Wang L, Chen D, Olson J, Ali S, Fan T, Mao H. Reexamine tumor-induced alterations in hemodynamic responses of BOLD fMRI: implications in presurgical brain mapping. Acta Radiol 2012;53:802–811.
Neurology 84
February 10, 2015
551
ª 2015 American Academy of Neurology. Unauthorized reproduction of this article is prohibited.
Resection of gliomas around language areas: Can fMRI contribute? Larry Junck, Shawn L. Hervey-Jumper and Oren Sagher Neurology 2015;84;550-551 Published Online before print January 14, 2015 DOI 10.1212/WNL.0000000000001241 This information is current as of January 14, 2015 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/84/6/550.full.html
References
This article cites 10 articles, 6 of which you can access for free at: http://www.neurology.org/content/84/6/550.full.html##ref-list-1
Subspecialty Collections
This article, along with others on similar topics, appears in the following collection(s): fMRI http://www.neurology.org//cgi/collection/fmri Primary brain tumor http://www.neurology.org//cgi/collection/primary_brain_tumor
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 2015 American Academy of Neurology. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
