Should thalamic recording be standard practice or institutional review board-approved research in stereoEEG?

In this issue of Brain we ask a group of international epilepsy experts to give their opinions on a rapidly growing practice in stereoEEG (SEEG): thalamic implantation. SEEG is an invasive diagnostic procedure performed in patients who are candidates for epilepsy surgery. Intracerebral electrodes are implanted in brain to map seizure onset and spread using passive recording and electrical stimulation. This information is used to create a plan for definitive surgical management—resection, ablation, neuromodulation with implantable devices, or to continue with medications alone. Here, a group of respected experts in the field address a single question: should implanting electrodes and recording from the thalamus be considered the clinical standard of care in epilepsy or should it be performed only as institutional review board (IRB)-approved, hypothesis-driven research? The commentators were chosen for their extensive experience and for their passionate, disparate opinions. The authors are neurosurgeons, neurologists and neurophysiologists, and together they represent leading institutions in the USA, Canada and Europe. Their practices are grounded in the long, storied teachings of Walter Penfield in Montreal and Jean Bancaud in Paris, dove-tailing with neurosurgical practice that has aggressively adapted SEEG to centres in the USA over the past 10 years.1 The pieces agree on the importance of several points in the discussion: foremost, whether the approach directly benefits patients, on the requirement for patient safety, the need for testable hypotheses, and for ethical considerations. However, the groups differ on a crucial point: how strong is the evidence that thalamic recordings benefit patients? To answer this question, our authors draw from expert opinion, single centre case series and retrospective cohorts—but not yet from blinded, controlled clinical trials—a point of contention. Proponents of thalamic SEEG as standard of care highlight its safety,2-4 regulatory compliance and historical precedent. They question whether holding thalamic SEEG to a standard greater than traditional SEEG is fair. These practitioners emphasize the value of thalamic recordings for several clinical applications, among them: Lateralizing and helping to localize the epileptogenic zone (EZ),3-6 mitigating the risk of missing it with standard SEEG implants. To test candidate sites for optimal therapeutic neuromodulation (i.e. deep brain or responsive stimulation devices). This technique uses anatomical and functional network mapping, electrical stimulation and interictal biomarkers to guide targeted device therapy.7,8 Proponents of the traditional SEEG view emphasize that the basic tenant of SEEG is to record from structures that can generate focal seizures, which does not include the thalamus. They cite the lack of evidence that thalamic electrodes provide clinical information superior to standard SEEG, and point to the need for randomized, blinded clinical trials demonstrating that thalamic recording leads to improved patient outcome. Without this evidence, these experts state that thalamic interventions remain ‘research'. Which group is correct? What are the consequences of adopting one point of view or the other? How should centres decide upon a strategy? These are the challenging questions our discussants address. Proponents of standard of care thalamic SEEG argue that it enables clinicians to test specific network hypotheses, which they feel already benefits patients. However, expertise, targeting precision, patient selection and protocols vary widely across institutions, which may make it difficult to interpret results from these procedures, as does the lack of controls. Making SEEG standard of care might also induce more variability in implantations, degrading our ability to learn from these procedures. Collaborative platforms like Epilepsy.Science, whose mission is to share data and analytical tools across existing databases and repositories, may facilitate pooling data across institutions for analysis, which may help. Similarly, new natural language processing tools to extract clinical metadata from the electronic medical record at scale could facilitate this approach. Categorizing thalamic SEEG as research may have significant benefits to offset restricting its application in individual cases. IRB approved investigation requires focusing thalamic studies to answer specific questions in protocols with clearly defined quantifiable outcome measures. This does not obviate using information from specific patient studies in clinical decision making, as long as it is within the confines of the protocol. Randomized, prospective clinical trials ultimately provide the very best quality data, though linking thalamic SEEG studies to outcomes from treatment, including neuromodulation, may take years to get results. Ethical considerations, including risk, are central to all the authors’ positions. Several of the experts quote large series emphasizing the relative safety of SEEG.2,3 Others note that the haemorrhage risk increases as a function of the number of intracranial electrodes implanted,9 and additional implants not essential to care increase patient risk. A rising trend in some US centres towards implanting multiple thalamic targets bilaterally adds to this concern. On the other hand, it is increasingly common practice to implant more than a single subcortical target for deep brain stimulation applications, and techniques for avoiding complications are continuing to improve. Finally, the opinions below weigh the risks of restricting thalamic SEEG to specific protocols, potentially depriving the clinical team of information they feel is important, versus protecting against practices and therapies that may be ineffective or potentially harmful. The role of controls in invasive antiepileptic devices is the subject of a recent study,10 which is stimulating further discussion in this area. The quandary we present is not unique to epilepsy surgery. There are many examples where a beneficial medical practice became accepted without class I evidence, such as early adoption of pacemakers or laparoscopic surgery. Requiring randomized trials could have had substantial negative impact on care by delaying their availability. Conversely, there are examples where randomized controlled clinical trials significantly changed established medical practice, such as in coronary artery bypass surgery and arthroscopic knee surgery for osteoarthritis. In these cases, one could argue that earlier clinical trials might have had huge societal and economic benefit and prevented many procedures that did not ultimately benefit patients. What is the correct path? We hope that the articles in this issue of Brain will help the reader form their own opinion. The fact that well-informed and intentioned practitioners can reasonably disagree on the issue attests to its difficulty. For now, individual clinicians and centres are looking to leaders in the field for clarity. Our intention in presenting opinions from such leaders is twofold: (i) to consider salient points of the discussion and ensure that each centre can choose their approach in an informed way, until more data are available; and (ii) to stimulate the field to fill this knowledge gap, through collaborative research, specifically rigorous, transparent data sharing, controlled clinical trials, or ideally both. The authors report no competing interests.