Our next speaker is Laila Mohamed, discussing her work with observation of cortical spreading depolarizations after chronic subdural hematoma evacuation. Dr. Greg Hariluk will be the discussant. Good morning. My name is Laila Mohamed, and I'm a fifth-year resident at the University of New Mexico. Today, I will be presenting the data from our CSD2 trial, which stands for cortical spreading depolarizations in chronic subdural hematoma patients after evacuation. This project is a collaboration of the efforts of resident Dr. Mohamed Abbas, medical student Rostin Amadian, neuroscientist Dr. Bill Shuttleworth, and our PI, Dr. Andrew Carlson. I have no disclosures to report. It is known that chronic subdural hematomas may cause recurrent TIA-like symptoms of the involved hemisphere, including transient aphasia, hemisensory, or motor abnormalities. The duration of symptoms tends to be longer than the typical TIA. This can result in prolonged hospital stays and delayed recovery. In order to better characterize the incidence of these transient neurological deficits, we identified 220 patients over a five-year period that underwent chronic subdural hematoma evacuation at our institution. We found that 50% of these patients required an EEG, which suggested a post-operative neurological deterioration. Although this rate is much higher than expected, only 14% of these deficits were explained by seizures. This leaves a group of patients who had a delayed recovery, but no obvious neurophysiological explanation. Some of these patients may have had seizures that were unable to be detected by EEG or a metabolic dysfunction. However, we have to consider that there may be another neurophysiological explanation that are not seizures. Now, let's talk about the other CSD in the CSD-2 trial, cortical spreading depolarization. CSD is a slowly propagating wave of near-complete cellular depolarization that can result in the depression of electrocortographic activity. Recent work has documented that CSD occurs in humans and can be associated with transient neurological deficits. CSD is also responsible for the scintillating scotoma in migraine auras. Both the migraine aura and CSD propagate along the cortical surface, as seen in this drawing. This demonstrates a 20-minute progression of scintillations with a similar MRI-bold response in the occipital cortex contralateral to the visual aura. This data strongly suggests that an electrophysiological event, such as CSD, generates the aura in the human visual cortex. CSD has also been clearly linked to delayed neurological deficits in brain injury, as seen in this patient with subarachnoid hemorrhage. The patient demonstrated time-linked clinical deterioration with multiple related CSDs. She was initially oriented times four, but spontaneously developed vasospasm-like symptoms. This change in the level of consciousness corresponded with a series of four spreading depolarizations shown by the arrows. Furthermore, recent data in elective craniotomies have suggested that these events may occur outside of brain injury and migraines, but during neurosurgical procedures. Based on the observation that CSDs can occur after acute brain injury and elective neurosurgical procedures, it is plausible that CSDs may contribute to some transient neurological deficits despite adequate evacuation of their subdural hematoma. In order to answer this question, we performed a prospective observational study of patients undergoing clinically indicated evacuation of chronic subdural hematomas. The following is our methodology. Despite multiple studies attempting to observe CSD non-invasively, the capacitance of the scalp blocks low-frequency changes. Therefore, there are currently no non-invasive methods to reliably detect CSD, so it was necessary to place an intraoperative subdural electrode. In order to ensure that the electrode was over the area of the evacuation, it was placed in the same opening as the subdural drain and tunneled posteriorly for later removal. All readable ECOG times were recorded to further assess the feasibility of the study. Recordings occurred for approximately two days until the subdural drain was removed for clinical indications. CSDs were then scored using the standard criteria of slowly propagating DC shift associated with high-frequency ECOG suppression. This figure demonstrates a typical CSD. The first six recordings are the raw referential DC potential data. The red tracings show a characteristic shape of a CSD. The five tracings below are bipolar high-frequency filter data, which removes much of the artifact from the raw DC data. This image is highly time-compressed, demonstrating a 20-minute interval with one large box indicating five minutes. Therefore, when you see a propagation, it is occurring over a three-minute period. This is a characteristic example of a definite CSD with three DC shifts associated with a subsequent depression in frequency. Next, we attempted to quantify the neurological deterioration in three ways. The first was using the Mark-Walder grading system, a grading scale for evaluating the neurological condition of patients with chronic subdurals. Grade zero patients are typically with incidental findings of chronic subdural hematomas. Grade one patients are typically presenting with severe headaches, plus or minus a pronator drift. Grade two patients are by far the most common patients. They present with mild weakness or confusion. The second method was with the documented occurrence of a neurological deficit, and the third was with clinically indicated EEG. The goal was to compare the association of CSD with postoperative neurological deterioration. 30 subjects were recruited and completed data and ECOG analysis. The mean age was 71 years old, with 30% being female. All subjects had a technically successful evacuation of the subdural hematoma. With over 1,000 hours of readable data, all patients had a readable ECOG. This was a pleasant surprise, since there was concern that the patient's activity and postoperative pneumocephalus may result in a significant amount of artifact. Definite CSD occurred in four of the 30 subjects at a rate of 13%. In these subjects, the number of events varied from four to seven, with CSD events generally occurring in clusters. No patients had a clinical deterioration based on MGS. However, one patient's MGS remained a two immediately postoperatively, and declined to a three by postoperative day 10. Furthermore, clinically indicated EEG was only ordered in three of the 30 subjects at a rate of 10%, which is much less than expected, based on our predicted 50%. There were no findings of seizures or epileptiform activity. The single patient who had a clinical deterioration demonstrated a CSD at hour 19. The strip had been removed by the time the deterioration occurred. An EEG was completed, and it did not demonstrate any seizures. None of the subjects with a CSD demonstrated a time-locked clinical deterioration. The first limitation of this study is that the rate of subjects with deterioration was notably lower than what we expected, based on our retrospective data. It is possible that the enrollment in a prospective study affected other variables, such as increased attention during the surgical procedure, and more attentive postoperative care, which could potentially affect CSD occurrence and other factors leading to deterioration. Given that anti-epileptic medication was left to the discretion of the treating team, it was found to be variable in the study. A larger population would be needed to better determine the relative importance of seizures versus CSD in postoperative neurological deterioration. Finally, none of the subjects had a clinical deterioration during electrode recording. This may be related to technical reasons, such as the location of the electrode, which was generally placed over the frontal lobe for convenience. A localized CSD in one frontal lobe would not be expected to be led to a measurable deficit, unless motor or language areas were involved. While none of the subjects had a clinical deterioration during electrode recording, this occurrence of CSD is the first step to determine if there's a potential role of CSD in some cases of deterioration. We present the first observation of CSD occurring after chronic subdural hematoma evacuation at a rate of 13%. These data are preliminary, and certainly no direct causal relationship is established, but this is the first observation that CSD can occur after chronic subdural hematoma evacuation, even without parenchymal injury or ischemia. This phenomenon may be responsible for some cases of delayed clinical deterioration. Such data would potentially allow for the design of interventional studies to determine the efficacy of seizures and CSD prophylaxis after chronic subdural hematoma evacuation. While ketamine has been shown in brain injury to effectively suppress CSD, multiple migraine prophylactic medications, such as topiramate and gabapentin, have been suggested to act via an anti-CSD mechanism in mouse models. Pre-treatment with these medications may therefore be feasible in cases of subdural hematoma, where there is typically less urgency for immediate action compared to acute brain injury. Thank you. Thank you, Dr. Mohamed. Dr. Harilek will discuss the paper. Neurotrauma and neurocritical care are well behind other areas of medicine, and I believe that this is at least in part because the field has been built upon some shaky pillars. One such problem is our failure to recognize cortical spreading depolarizations. These recently rediscovered, slowly spreading waves of depolarization have been hiding in plain sight, filtered out by EEGs. They are now the accepted cause of migraine aura, and are thought to be initiated by neurons depolarizing as a result of energy failure. Evidence to date suggests that CSDs are prevalent in neurosurgical conditions, and are clinically important with at least five times the energy cost of a seizure. Moreover, they may underlie transient or permanent neurological deficits, hard to explain ischemic strokes, and alterations in auto-regulatory status. I'm sure that many of you, like me, have resisted the urge to learn about CSDs, because they are not currently easy to record, even in a university hospital. However, given growing evidence for their clinical importance and the finding that they might be treated with ketamine, I believe it is now urgent for neurosurgeons to gain familiarity with these phenomena, and to lead in relevant research efforts, as our colleagues at the University of New Mexico have done. It is striking that UNM found a 13% prevalence of CSDs in chronic subdural hematoma patients. In the context of the broader literature, this finding raises the possibility that monitoring and a targeted treatment for CSDs may ultimately be considered for a large proportion of neurosurgical patients. In this case, it is, of course, speculative that the described neurological decline was related to a CSD. They also reported the retention of an electrode tip, and that reminds us that efforts to monitor these phenomena will not come without some risk. In conclusion, it is my opinion that monitoring and treatment of CSDs is still investigational, but is very likely to become mainstream over the next five to 10 years. Please be aware that our colleagues in the CAUSBD research group have already published recommendations for how clinical CSD monitoring should be conducted. Thank you for the opportunity to discuss this important work. Thank you.

cortical spreading depolarizations

chronic subdural hematoma

neurological deficits

EEG

neurosurgical patients