Broca’s area: do not touch?

A couple of years ago I’d been tasked with some international galivanting loosely associated with professional development, and this time it was for the American Association of Neurological Surgeons (AANS) conference held in New Orleans. A neurosurgeon by the name of Prof Hugues Duffau took to the podium with preposterous enthusiasm and presented a high volume of slides within a short period of time that were highly informative and way over my head. There was one statement, though, that in retrospect completely changed the trajectory of my career; ‘…you can resect Broca’s area with impunity’.

Awake brain surgery is probably a horrifying idea to the uninitiated. However in circumstances where tumours grow near areas of the brain that mediate so-called ‘eloquent’ function – such as speech, language, movement and sensation – oftentimes an awake approach permits the greatest chance at identifying and subsequently avoiding these areas to ensure they remain uncompromised during surgery, lest the patient experience debilitating neurologic consequences post-operatively.

Electrical stimulation of the brain during these procedures helps facilitate the objective identification of areas to avoid. It’s not a novel concept, nor is it a recent development. In fact, awake brain tumour surgery with mapping has probably been considered a ‘staple’ neurosurgical procedure since the 1920’s. Of course this excludes the numerous experimental reports from the late 1800’s, which were retrospectively acknowledged as scientifically important. For example, one neurosurgeon in particular elected to stimulate the exposed brain of patient showing symptoms compatible with imminent ‘early extinction of life’ and was subsequently ostracised by the American Medical Association, but has since been hailed as the progenitor of human cortical stimulation[1]. These days awake brain tumour surgeries are routinely performed and often partnered with a sophisticated, systematic and stereotactic cartographic exploration of function, where functional mapping with electrical stimulation delineates functional cortical or subcortical regions to avoid prior to resecting tumour. Though it’s far from a ‘one size fits all’ approach.

Beyond the classical brain mapping techniques popularised by Penfield and Ojemann, there are now numerous electrophysiological techniques designed to map the eloquent brain more effectively. Current approaches are capable of objectively mapping the corticospinal tract to ~4mm, prognosticating an intact corticospinal system, objectively delineating pericentral anatomy, detecting and preventing seizures associated with stimulation, identifying epileptogenic foci, as well as the real-time evaluation of language function (including multilingual mapping). These approaches are far from homogeneous in neurosurgery, and clinical utility is also subject to the patient’s presenting symptoms, the extent to which function is affected prior to surgery, their age and psychological status, as well as histopathology and tumour grading. More importantly, these techniques are most effective only in consideration with the surgical goal, often premeditated in consideration of tumour grading.

The World Health Organization characterises brain tumours according to their aggressiveness. ‘Lower grade’ tumours grow at a lesser rate than ‘higher grade’ tumours, but surgical approaches can vary given the degree of infiltration, access to appropriate surgical treatment centres, and location of lesions. When it comes to electrically mapping these areas during surgery, there are also noticeable differences in approaches, techniques, and philosophies. This is thanks in part to inter-individual variability of neuroanatomy, but also because we know and understand that the human brain and nervous system can be stimulated or modulated with various electrical stimulation parameters, akin to modern computing systems and medical devices. The complexity of how these systems are networked and organised is what we’re yet to fully (or partially) elucidate.

This isn’t surprising. The human cerebral cortex alone contains 40 billion neurons crowed into 3 square meters of surface area, and the total number of neural contacts on the surface of the brain is in order of 40 followed by 14 zeroes, a number that is as large as the number of all the stars in our galaxy[2]. In the last decade or two, research has continued to describe the ‘plastic’ functional reorganisation of the brain based on neuro-imaging studies, lesional studies and a-priori brain mapping studies conducted during awake brain surgery. Yet back in New Orleans, watching Prof Duffau take to the podium and speak of his extensive experience in awake brain mapping, it seemed considered audacious to claim that you could resect part of the brain universally recognised as being associated with speech production. However there’s slightly more to consider.

The basis of Duffau’s claim is to remove an eloquent cortical region that inevitably induces a transient deficit, in the hopes of achieving a maximal resection that contributes to progression-free survival. The statement resonated with me purely because my feeble mind couldn’t process that an area of the brain historically synonymous with ‘do not touch’ could, in fact, be resected. Whereas it was once considered disastrous to risk damage to cortical structures comprising the classical Broca-Wernicke-Lichtheim-Geshwind model of speech function, it has since been widely contested, with data indicating that resection of these very structures does not necessarily contribute to permanent injury provided subcortical white matter tracts are not compromised. Contemporary models of speech and language processing are now regarded as a dual-stream model championed by Gregory Hickok (who also presented in New Orleans), whereby the dorsal stream involves sensorimotor integration including the arcuate fasciculus and superior longitudinal fasciculus, and the ventral stream involves comprehension and includes the inferior frontal-occipital fasciculus, uncinate fasciculus and middle and inferior longitudinal fasciculi. The downside of ‘touching’ Broca’s in this instance is that it would result in a temporary inability to speak, sometimes up to twelve weeks.

But what if we could ‘relocate’ the eloquent area prior to surgery and avoid a deficit altogether? That’s what we’re trying to explore at the moment.

As the wider literature continues to increase reports of brain mapping techniques used in the treatment of lower grade glioma, combined with improved neuro-imaging methodologies relating to the assessment of neural function, there has been increasing interest in commandeering the potential neuromodulatory effects of emerging clinical technologies to induce functional organization under circumstances where eloquent regions infiltrated by tumour are considered ‘inoperable’ (the definition of which, I’m told, is widely heterogeneous). And no, I’m not talking about Neuralink. Here, the rationale is to regularly electrically modulate targeted functional regions or networks involving the tumour partnered with relevant neuropsychological or motor function tasks in order to catalyse cortical plasticity over time.

Resecting Broca’s area is one thing. Relocating Broca’s area might be another.

Current models of therapeutic neuromodulation (invasive or non-invasive) exist to support stroke-related motor or speech recovery, the amelioration of symptoms in medication-resistant depression patients, and management of neuropathic pain, in addition to various applications currently under investigation. The idea is to target the affected or homologous network via stimulation to promote a clinical response. The mechanisms at play are suspected to be associated with modulation of molecular pathways of the brain involving brain-derived neurotrophic factor, dopamine, gamma aminobutyric acid, glutamate, serotonin, cortisol, or endogenous opioids. Whereas with cortical prehabilitation, stimulation contributes to the synchronization of affected neurons and GABAergic inhibition, contributing to temporary brain disruption and lesion effect, which over time may contribute to a “virtual lesion” resulting in diaschisis and subsequent compensation.

At present there are very few reports of attempts at cortical prehabilitation mainly involving perisylvian cortex relating to the ‘relocation’ of speech. Whether this would be favourable in comparison to a temporary inability to speak following brain tumour surgery remains to be seen, especially since neuro-oncological strategies for the management of lower grade glioma favours early surgical intervention as the primary treatment option to facilitate opportunities of progression-free survival. This presents a number of inherent ethical considerations relating to the time necessary to catalyse cortical prehabilitation versus the earliest opportunity for surgical intervention to increase opportunities of progression-free survival. Of course, there are also economic factors to consider; the resources necessary for prehab versus those necessary for rehab, for example.

From a technical perspective, clinical approaches and optimal stimulation paradigms for prehab are yet to be systematically investigated[3]. For example, surgeons can perform an awake craniotomy and map the eloquent regions of interest then proceed to implant an electrode intended for continuous extraoperative stimulation to catalyse plasticity (procedurally similar to epilepsy surgery). Of course, any surgical procedure presents risk to the patient, relating to anaesthesia, infection, mechanical insult etc. The alternative is to use non-invasive approaches, such as transcranial magnetic stimulation, where there is compelling data supporting the effective modulation of neural networks over time. Distinguishing whether an invasive or non-invasive approach is therapeutically effective in reorganising eloquent function to avoid damage remains to be seen, but probably warrants continued investigation. Or it won’t work as well as anticipated, which is also a distinct possibility.

Two years after I watched Duffau encourage an audience of surgeons to ‘resect Broca’s area with impunity’ I was fortunate to speak with him as part of a profile for the Intraoperative neurophysiology Society of Asia-Pacific[4]. When I asked him about this claim, he maintained his position. Based on his research, though, plasticity following such a resection is best achieved only under certain circumstances; lower-grade lesions, younger patients with less preoperative infiltration, and always partnered with extensive neuropsychological and physical rehabilitation. When I inquired as to whether or not other centres should be endorsing his philosophy, Duffau remarked; “This is not science fiction. We are really capable of doing this. The postoperative possibilities of neurosurgery are just amazing, particularly if the patient is doing well before surgery…plasticity exists in patients, but sometimes it’s very difficult to find plasticity in researchers and medical doctors”.

1. Patra, D. P., Hess, R. A., Abi-Aad, K. R., Muzyka, I. M., & Bendok, B. R. (2019). Roberts Bartholow: the progenitor of human cortical stimulation and his contentious experiment. Neurosurgical focus, 47(3), E6. https://doi.org/10.3171/2019.6.FOCUS19349

2. Professor Marsel Mesulam, writing the foreword for Duffau’s Brain mapping: from neural basis of cognition to surgical applications.

3. Hamer, R. P., & Yeo, T. T. (2022). Current Status of Neuromodulation-Induced Cortical Prehabilitation and Considerations for Treatment Pathways in Lower-Grade Glioma Surgery. Life, 12(4), 466. https://doi.org/10.3390/life12040466

4. https://ryanhamer.medium.com/this-is-not-science-fiction-a656c30b1485

Written by Ryan Hamer

Clinical lecturer with the Faculty of Medicine and Health at the University of Sydney, honorary fellow (Department of Surgery) at St Vincent’s Hospital / the University of Melbourne, Chair of the Intraoperative Neurophysiology Society of Asia-Pacific (INSA) and Managing Director of Neuroclast.