Emmanuel Mandonnet1,2,3*- 1Hôpital Lariboisière, Paris, France
- 2INSERM U1127 Institut du Cerveau et de la Moelle épinière (ICM), Paris, France
- 3Université de Paris, Paris, France
In a recent opinion (1), Herbet thoughtfully questioned the conceptual relevance of mapping complex cognitive functions with direct electrical stimulation during awake surgery.
His argumentation followed two lines of reasonings.
First, it might be that it is methodologically impossible to map of these functions intraoperatively, due to both electrophysiological and behavioral limitations. On an electrophysiological point of view, a single, focal, low-intensity input gate into the brain circuits (2) might be insufficient to cause an imbalance of the whole brain dynamics, up to the point that it would cause a behavioral impairment of complex cognitive functions. This view is grounded in the meta-networking theory of these complex cognitive functions: the goal-directed, flexible, context-sensitive behavior of humans would implicate a short-lived (metastable) spatio-temporal reconfiguration/cooperation between several domain-specific networks (3). Apart from the methodological inadequacy of using low-intensity focal electrical stimulation to disturb meta-networks, Herbet also argues that it could be impossible to design a behavioral task that would enable to monitor such functions, given the intraoperative constraints. In particular, electrical stimulation should be applied no longer than a few seconds, and such a short period of time considerably limits the possibilities. Finally, the well-known interindividual anatomical variability of modular and domain-specific functional areas could be even greater for the dynamics of meta-networking, and such a variability would make the interpretation and reproducibility of stimulation-induced disturbances very difficult.
Second, Herbet suggests that there is maybe no need to pay attention to these complex cognitive functions, because they would be much more resilient to damage than are modular or domain-specific functions. Indeed, as they emerge from a widespread meta-networking brain, these functional systems would exhibit a high degree of plasticity, and no long-term dysfunction would be observed after a brain tumor resection.
In this commentary, I would like to both complement and nuance this opinion.
The hypothesis put forward by Herbet is somehow supported by a recent study (4), demonstrating the rarity of self-reported effects when stimulating areas belonging to hetero-modal and trans-modal areas, strongly contrasting with unimodal areas: the gradient in the frequency of self-reported effects clearly follows the principal gradient of the cortical hierarchy found by Margulies et al. (5). It is true that a single entry point of stimulation might not be the optimal way to unveil areas at the distal end of the cortical hierarchy. However, it should be noted that, in a recent study, multisite stimulation unmasked language sites that were not detected by mono-site stimulation (6). Such scheme of stimulation could thus be used to desynchronize a set of areas, helping to identify a meta-network. More generally, mathematical tools of network dynamics, like network controllability (7), could guide the elaboration of even more sophisticated stimulation protocols, possibly better suited to induce meta-network dysfunction.
The issue of the task design is a serious one, as it is absolutely right that the limited timing (a few seconds) puts a very strong constraint. However, in the aforementioned study (4), self-reported effects were also noted, albeit rarely, after stimulation of hetero/trans-modal areas. When these rare effects occurred, they were quite complex (for example, “Patient describes a negative emotional feeling, seemingly localized in the chest”), suggesting that they might have been indeed generated from a meta-networking dysfunction. Hence, assessing detailed introspection could provide a mean to overcome the constraints imposed by the intraoperative setting when designing a task targeting the meta-networking activity. This is also in line with the recent proposal to monitor patient's self-evaluation of his performance after each task trial, in addition to the performance itself (8). This additional monitoring is relatively easy to implement, but, admittedly, its value is currently not well-established.
We fully concur with the author that the high interindividual variability of meta-networks anatomical implementation might greatly complicate the reproducible identification of “complex cognitive areas.” And in the same perspective, the well-known non-reproducibility on a trial-by-trial basis of mono-site stimulation (9) might also fit with the idea that the dynamics of meta-networking is highly variable, even at the scale of a single brain: from one trial to another, the brain dynamics might have followed a different trajectory, explaining why two consecutive stimulations of the same site do not always produce the same behavioral dysfunction, even for such a simple task as picture naming.
We thus claim that it is possible to envision solutions that would enable to interfere intra-operatively with the meta-networks of complex cognition. But, as argued by Herbet in the second part of his argumentation, the most important practical clinical question is rather the following: assuming we are able to identify intraoperatively a complex cognitive function site, should the surgeon stop the resection at this level or should he push further the resection, betting on late recovery thanks to the virtue of plasticity?
We fully agree with the idea that there exists a gradient of plasticity from unimodal/modular areas to modality-specific or domain-specific networks. Let me consider the example of executive functions, although the same would apply to social cognition (10–12). Recently, several groups independently reported that immediate postoperative deficits in executive functions are commonly observed in low-grade glioma patients in whom the resection was performed without intraoperative monitoring of these executive functions (13–20). Importantly, these series also demonstrated that only a relatively small fraction of patients remained definitely impaired. This means that if executive functions would have been systematically mapped and preserved, the resection would have been stopped prematurely in a large proportion of patients, with a loss of chance regarding oncological benefit. This contrasts with modular input-output areas, for which we are 100 % sure that their resection would result in permanent deficits, and thus should be preserved systematically. To address this issue, two broad categories of solutions have been implemented. Skrap et al. proposed an “ad hoc” pragmatic approach: “The limit used to stop surgery temporarily or definitively coincided with a decrease in performance to 70% of baseline levels. The 30% threshold was established by examining a preliminary group of patients in whom we tested the feasibility of our technique before actually starting the present study. Thus, this threshold was correlated to follow-up performance” (21). Other authors preferred to follow a “mechanistic” approach: they attempted first to retrospectively determine the precise networks of the brain the resection of which induced deficits, and restricted the use of intraoperative executive functions tasks to these locations (14, 18, 22, 23). In the future, there is a strong call for developing new methods (24), relying on machine learning algorithms, that would enable to predict, from a set of preoperative personalized predictors reflecting individual brain functioning, which function would be at risk in an individual patient and would thus deserve to be mapped intraoperatively.
In fact, the degree of plasticity of a modality-specific/domain-specific network cannot be defined in an absolute manner. The measurement of the plasticity of a network is always relative to the task measuring its function. For example, when Herbet argues that “despite the central role of the anterior temporal structures within the semantic memory network, unilateral damage of this region in various pathophysiological conditions (including glioma) does not result in the severe impairments of semantic representations it might be expected,” this statement implicitly depends on which task is used to assess semantic knowledge. While it is true that a basic non-verbal semantic association task like the pyramid-palm-tree test fails to detect any impairment after anterior temporal resections (25), long-term deficits in proper names retrieval have been reported (26). This is likely because it is only from the left side that the lexical store can access the bilateral representations of conceptual entities, shared by the left and right anterior temporal structures. Similarly, should we measure semantic knowledge by metrics estimated on full semantic graphs (27), new kind of deficits would be likely unveiled.
To which extent plasticity is even more efficient for complex cognition remains unknown. As acknowledged in the conclusion by the author, “no works have currently assessed in a longitudinal manner if complex and flexible cognitions/behaviors are impaired following neurosurgical procedures (knowing that they are not probed with routine neuropsychological tasks) and if these possible impairments are disabling in daily life.”
To go further, it would have been informative to illustrate a little bit more what is exactly meant by “complex and flexible cognition/behavior.” The unique example given by Herbet is learning abilities, raising the question whether patients do suffer (or not) postoperatively from learning problems. And indeed, to the best of my knowledge, the literature on this question is virtually non-existent. Thinking about other examples, should we include in this category the processes allowing to understand metaphors? This ability necessitates the coordination of several networks: semantic knowledge, language, attention, working memory, executive system. Importantly, plasticity does not fully protect metaphors comprehension from surgical damage: studies have reported postoperative deficits after resection in the right hemisphere (21). And interestingly, it seems feasible to test intraoperatively the comprehension of metaphors (21).
Last but not least, it should be kept in mind that, even if plasticity allows patients to recover complex cognition performances close to normal, the resulting suboptimal brain dynamics comes with a price to pay: patients are both slower and more fatigable. These two downsides should not be overlooked: it has been demonstrated that both slowness and fatigue correlate with employment outcomes (28, 29). Hence, in the context of surgery for incidentally discovered glioma (30)—and even so in the context of a screening policy (31)—, great care should be devoted to preserve as much as possible a fully normal functioning of meta-networks.
In summary, Herbet opened a very interesting debate regarding what could be the limitations of cognitive monitoring in an awake patient for preserving the integrity of its complex cognitive functions. Such a topic might seem a little bit premature, considering that we still have a lot to learn regarding modality-specific/domain-specific functions so that we could better predict in which patient and at which location which function should be mapped intraoperatively. Nevertheless, efforts should be devoted to design new tasks in order to better explore complex cognitive functions extra-operatively, to evaluate how they might be impacted by surgical resection, and to imagine new pre- and intra- operative methods to offer the patients the maximal chances to preserve an efficient functioning of their meta-networks.
Author Contributions
The author confirms being the sole contributor of this work and has approved it for publication.
Funding
EM was supported by INSERM, contract Interface 2018.
Conflict of Interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's Note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1. Herbet G. Should complex cognitive functions be mapped with direct electrostimulation in wide-awake surgery? A network perspective. Front Neurol. (2021) 12:635439. doi: 10.3389/fneur.2021.635439
2. Mandonnet E, Winkler PA, Duffau H. Direct electrical stimulation as an input gate into brain functional networks: principles, advantages and limitations. Acta Neurochir. (2010) 152:185–93. doi: 10.1007/s00701-009-0469-0
3. Herbet G, Duffau H. Revisiting the functional anatomy of the human brain: toward a meta-networking theory of cerebral functions. Physiol Rev. (2020) 100:1181–228. doi: 10.1152/physrev.00033.2019
4. Fox KCR, Shi L, Baek S, Raccah O, Foster BL, Saha S, et al. Intrinsic network architecture predicts the effects elicited by intracranial electrical stimulation of the human brain. Nat Hum Behav. (2020) 4:1039–52. doi: 10.1038/s41562-020-0910-1
5. Margulies DS, Ghosh SS, Goulas A, Falkiewicz M, Huntenburg JM, Langs G, et al. Situating the default-mode network along a principal gradient of macroscale cortical organization. Proc Natl Acad Sci USA. (2016) 113:12574–9. doi: 10.1073/pnas.1608282113
6. Gonen T, Gazit T, Korn A, Kirschner A, Perry D, Hendler T, et al. Intra-operative multi-site stimulation: Expanding methodology for cortical brain mapping of language functions. PLoS ONE. (2017) 12:e180740. doi: 10.1371/journal.pone.0180740
7. Stiso J, Khambhati AN, Menara T, Kahn AE, Stein JM, Das SR, et al. White matter network architecture guides direct electrical stimulation through optimal state transitions. Cell Rep. (2019) 28:2554–66.e7. doi: 10.1016/j.celrep.2019.08.008
8. Ng S, Herbet G, Lemaitre A-L, Moritz-Gasser S, Duffau H. Disrupting self-evaluative processing with electrostimulation mapping during awake brain surgery. Sci Rep. (2021) 11:9386. doi: 10.1038/s41598-021-88916-y
9. Roux F-E, Djidjeli I, Quéhan R, Réhault E, Giussani C, Durand J-B. Intraoperative electrostimulation for awake brain mapping: how many positive interference responses are required for reliability? J Neurosurg. (2019) 1–11. doi: 10.3171/2019.6.JNS19925
10. Herbet G, Lafargue G, Bonnetblanc F, Moritz-Gasser S, Duffau H. Is the right frontal cortex really crucial in the mentalizing network? A longitudinal study in patients with a slow-growing lesion. Cortex. (2013) 49:2711–27. doi: 10.1016/j.cortex.2013.08.003
11. Campanella F, Shallice T, Ius T, Fabbro F, Skrap M. Impact of brain tumour location on emotion and personality: a voxel-based lesion-symptom mapping study on mentalization processes. Brain. (2014) 137:2532–45. doi: 10.1093/brain/awu183
12. Nakajima R, Kinoshita M, Okita H, Yahata T, Matsui M, Nakada M. Neural networks mediating high-level mentalizing in patients with right cerebral hemispheric gliomas. Front Behav Neurosci. (2018) 12:33. doi: 10.3389/fnbeh.2018.00033
13. Mandonnet E, De Witt Hamer P, Poisson I, Whittle I, Bernat A-L, Bresson D, et al. Initial experience using awake surgery for glioma: oncological, functional, and employment outcomes in a consecutive series of 25 cases. Neurosurgery. (2015) 76:382–9; discussion: 389. doi: 10.1227/NEU.0000000000000644
14. Mandonnet E, Cerliani L, Siuda-Krzywicka K, Poisson I, Zhi N, Volle E, et al. A network-level approach of cognitive flexibility impairment after surgery of a right temporo-parietal glioma. Neurochirurgie. (2017) 63:308–13. doi: 10.1016/j.neuchi.2017.03.003
15. Campanella F, Palese A, Del Missier F, Moreale R, Ius T, Shallice T, et al. Long-term cognitive functioning and psychological well-being in surgically treated patients with low-grade glioma. World Neurosurg. (2017) 103:799–808.e9. doi: 10.1016/j.wneu.2017.04.006
16. Hendriks EJ, Habets EJJ, Taphoorn MJB, Douw L, Zwinderman AH, Vandertop WP, et al. Linking late cognitive outcome with glioma surgery location using resection cavity maps. Hum Brain Mapp. (2018) 39:2064–74. doi: 10.1002/hbm.23986
17. Mandonnet E. Transopercular resection of IDH-mutated insular glioma: a critical appraisal of an initial experience. World Neurosurg. (2019) 132:e563–76. doi: 10.1016/j.wneu.2019.08.071
18. Puglisi G, Howells H, Sciortino T, Leonetti A, Rossi M, Conti Nibali M, Gabriel Gay L, Fornia L, Bellacicca A, Viganò L, et al. Frontal pathways in cognitive control: direct evidence from intraoperative stimulation and diffusion tractography. Brain (2019) 142:2451-2465.
19. Rijnen SJM, Kaya G, Gehring K, Verheul JB, Wallis OC, Sitskoorn MM, et al. Cognitive functioning in patients with low-grade glioma: effects of hemispheric tumor location and surgical procedure. J Neurosurg. (2019) 1–12. doi: 10.3171/2019.8.JNS191667
20. Cochereau J, Lemaitre A-L, Wager M, Moritz-Gasser S, Duffau H, Herbet G. Network-behavior mapping of lasting executive impairments after low-grade glioma surgery. Brain Struct Funct. (2020) 225:2415–29. doi: 10.1007/s00429-020-02131-5
21. Skrap M, Marin D, Ius T, Fabbro F, Tomasino B. Brain mapping: a novel intraoperative neuropsychological approach. J Neurosurg. (2016) 125:877–87. doi: 10.3171/2015.10.JNS15740
22. Mandonnet E, Vincent M, Valero-Cabré A, Facque V, Barberis M, Bonnetblanc F, et al. Network-level causal analysis of set-shifting during trail making test part B: a multimodal analysis of a glioma surgery case. Cortex. (2020) 132:238–49. doi: 10.1016/j.cortex.2020.08.021
23. Puglisi G, Sciortino T, Rossi M, Leonetti A, Fornia L, Conti Nibali M, et al. Preserving executive functions in nondominant frontal lobe glioma surgery: an intraoperative tool. J Neurosurg. (2018) 131:474–80. doi: 10.3171/2018.4.JNS18393
24. Mandonnet E, Thirion B. Tackling the complexity of lesion-symptoms mapping: how to bridge the gap between data-scientists and clinicians? In: Machine Learning in Clinical Neuroscience: Foundations and Applications (Springer Nature Publishing).
25. Mandonnet E, Nouet A, Gatignol P, Capelle L, Duffau H. Does the left inferior longitudinal fasciculus play a role in language? A brain stimulation study. Brain. (2007) 130:623–9. doi: 10.1093/brain/awl361
26. Papagno C, Casarotti A, Comi A, Pisoni A, Lucchelli F, Bizzi A, Riva M, Bello L. Long-term proper name anomia after removal of the uncinate fasciculus. Brain Struct Funct. (2016) 221:687–94. doi: 10.1007/s00429-014-0920-8
27. Kenett YN, Faust M. A semantic network cartography of the creative mind. Trends Cogn Sci. (2019) 23:271–4. doi: 10.1016/j.tics.2019.01.007
28. Moritz-Gasser S, Herbet G, Maldonado IL, Duffau H. Lexical access speed is significantly correlated with the return to professional activities after awake surgery for low-grade gliomas. J Neurooncol. (2012) 107:633–41. doi: 10.1007/s11060-011-0789-9
29. van Loon EMP, Ernens W, Heijenbrok-Kal MH, Horemans HLD, Ribbers GM, van den Bent MJ. Long-term employment status and the association with fatigue in patients with grade II glioma. J Rehabil Med. (2021) 53:jrm00198. doi: 10.2340/16501977-2837
30. Ng S, Herbet G, Lemaitre A-L, Cochereau J, Moritz-Gasser S, Duffau H. Neuropsychological assessments before and after awake surgery for incidental low-grade gliomas. J Neurosurg. (2020) 1–10. doi: 10.3171/2020.7.JNS201507
Keywords: meta-networks, glioma surgery, awake surgery, brain mapping, cognitive outcome
Citation: Mandonnet E (2021) Should Complex Cognitive Functions Be Mapped With Direct Electrostimulation in Wide-Awake Surgery? A Commentary. Front. Neurol. 12:721038. doi: 10.3389/fneur.2021.721038
Received: 05 June 2021; Accepted: 15 July 2021;
Published: 26 August 2021.
Edited by:
Sandro M. Krieg, Technical University of Munich, GermanyReviewed by:
Carlos Velásquez, Marqués de Valdecilla University Hospital, SpainEmanuele La Corte, University of Bologna, Italy
Copyright © 2021 Mandonnet. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Emmanuel Mandonnet, mandonnet@mac.com