Driving Into the Unknown: ASM Reductions After Intracranial EEG

Abstract

Characterization of Anti-Seizure Medication Reduction and Discontinuation Rates Following Epilepsy Surgery

Jha R, Chua MMJ, Liu DD, Cosgrove GR, Tobochnik S, Rolston JD. Epilepsy Behav. 2024;158:109944. doi: 10.1016/j.yebeh.2024.109944. Epub 2024 Jul 13. PMID: 39002278.

Objective: Many patients pursue epilepsy surgery with the hope of reducing or stopping antiseizure medications (ASMs), in addition to reducing their seizure frequency and severity. While ASM decrease is primarily driven by surgical outcomes and patient preferences, preoperative estimates of meaningful ASM reduction or discontinuation are uncertain, especially when accounting for the various forking paths possible following intracranial EEG (iEEG), including resection, neuromodulation, or even the absence of further surgery. Here, we characterize in detail the ASM reduction in a large cohort of patients who underwent iEEG, facilitating proactive, early counseling for a complicated cohort considering surgical treatment. Methods: We identified a multi-institutional cohort of patients who underwent iEEG between 2001 and 2022, with a minimum of 2 years’ follow-up. The total number of ASMs prescribed immediately prior to surgery, choice of investigation modality, and subsequent surgical treatment were extracted for each patient. Primary endpoints included decreases in ASM counts from preoperative baseline to various follow-up intervals. Results: A total of 284 patients were followed for a median of 6.0 (range 2–22) years after iEEG surgery. Patients undergoing resection saw an average reduction of ∼0.5 ASMs. Patients undergoing neuromodulation saw no decrease and trended towards requiring increased ASM usage during long-term follow-up. Only patients undergoing resection were likely to completely discontinue all ASMs, with an increasing probability over time approaching ∼10%. Up to half of resection patients saw ASM decreases, which was largely stable during long-term follow-up, whereas only a quarter of neuromodulation patients saw a reduction, though their ASM reduction decreased over time. Conclusions: With the increasing use of stereotactic EEG and non-curative neuromodulation procedures, realistic estimates of ASM reduction and discontinuation should be considered preoperatively. Almost half of patients undergoing resective surgery can expect to reduce their ASMs, though only a tenth can expect to discontinue ASMs completely. If reduction is not seen early, it likely does not occur later during long-term follow-up. Less than a third of patients undergoing neuromodulation can expect ASM reduction, and instead, most may require increased usage during long-term follow-up.

Commentary

A young man with pharmacoresistant focal epilepsy and multiple bilateral periventricular nodular heterotopia presented with disabling seizures in his early twenties, ending his career in construction due to the inability to drive. He nonetheless managed to start a family, but could not hold his baby alone.

Stereo-EEG (SEEG) determined that a heterotopion in the occipital horn of the right lateral ventricle was most involved early in the seizures, with a complex network spreading to the right temporal lobe, causing the disabling semiology. Laser interstitial thermal therapy (LITT) was performed on the heterotopion to reduce seizure frequency and severity, ¹ with limited expectation that seizures would stop. At the time of surgery, he was on 4 antiseizure medications (ASMs) and was advised that ASM freedom was unlikely. Two years post-surgery, he is remarkably seizure-free, working, driving, and holding his second baby. One seizure would ruin a lot. He would like to be on fewer ASMs. Is it time to wean them?

This clinical scenario highlights the broader challenge of managing ASM use after epilepsy surgery—a question for which guidelines and quality evidence are lacking. The decision he makes will be based on his neurologist's understanding of his epilepsy, the personal stakes for the patient, and whatever relevant scientific evidence is available. A recent American Academy of Neurology (AAN) Guideline Subcommittee report on ASM withdrawal identified only 4 studies meeting inclusion criteria that estimated the risk of seizure recurrence after stopping ASMs in adults with stable epilepsy. ² The results are variable and difficult to interpret. The situation is worse for adults after epilepsy surgery, with only a single paper addressing the question in this population included in the Guideline. ³

Other studies, which either did not meet inclusion criteria for the AAN Guideline report or were released subsequent to it, have also contributed to our understanding of ASM withdrawal after surgery, but the problem of individualized risk stratification remains largely unresolved. For example, a recent study presented a predictive model developed from an observational cohort study based on 850 patients from 9 different centers. ⁴ According to this model, certain factors that could be known before surgery were associated with a higher adjusted hazard ratio of seizure recurrence after withdrawal of medication: a history of focal to bilateral tonic-clonic seizures before surgery (1.24) and a greater number of ASMs before surgery (1.60). Two other factors, from the period between surgery and ASM withdrawal, were associated with a modified adjusted hazard ratio of seizure recurrence: time between surgery and ASM withdrawal (0.90) and focal non-motor aware seizures between surgery and ASM withdrawal (5.56). Only this last factor really stands out as having a major impact on risk calculation and leads to a common-sense conclusion: patients who are already having seizure recurrences after surgery, even if awareness is preserved, are at high risk of more clinically significant seizures if ASMs are reduced from that point. An online clinical tool was also derived from this model and can be used to estimate seizure recurrence risk, but only a small proportion of imaginable clinical factors can be inputted.

As SEEG allows for more complex cases to be considered for surgery and, one hopes, eventual consideration of ASM withdrawal, the present retrospective cohort study characterizing ASM use in 284 adult patients after intracranial EEG (iEEG) at 2 Boston epilepsy centers is more than welcome. ⁵ After iEEG, 71.1% of patients underwent resective surgery or LITT, with the balance either undergoing neuromodulation or no surgical intervention at all. Neither of these 2 latter groups showed a reduction on ASMs in subsequent years. For the resective surgery group, 95% of whom were on multiple ASMs, there was a reduction of 0.5 ASMs between the pre-operative period and subsequent time points 2 years and beyond. On average, this reduction occurred within the first 2 years, with a slight tendency for the ASM number to then increase in the following years. Likelihood of complete ASM discontinuation was low even within this group, reaching only 12.5% at 15 years.

We must keep in mind that a limitation of this study is that these results are highly dependent on the clinical practices of the neurologists and the decisions of the patients who were treated at these epilepsy centers, which may be quite different from those of other centers, and these would have had a dramatic impact on the results. Furthermore, the retrospective nature of the study will have affected the quality of the data and, perhaps most importantly, does not allow us to understand the factors that drove decision-making around ASMs since this information was not systematically categorized in the charts.

Nevertheless, a reasonable conclusion is that patients should not undergo iEEG if their main objective is to stop taking ASMs, as the odds of accomplishing this are slim. If the outcomes at these centers are typical, they will be helpful in counseling patients before they decide to undergo iEEG. In their discussion, the authors rightly point out the important gap between the high patient expectations of ASM discontinuation after surgery and the reality, which persists even despite in-depth discussions.

When making decisions about withdrawal of ASMs after iEEG and surgery, however, we remain still largely unable to estimate seizure recurrence risk after ASM withdrawal or know when or how to attempt it. Here, the study's value is to highlight the need for more research in this area, moving beyond retrospective analyses to prospective trials that can offer more precise guidance for patients with specific clinical profiles.

The decision to withdraw ASMs requires a careful balance between the limited scientific evidence, clinical judgment, and the patient's unique circumstances. Ultimately, the discussion I had with my patient was simple. Once he understood that I could not guarantee he would be able to drive after weaning an ASM, he declined any reduction, as he needed his driver's license to feed his family. As so often happens, the decision to continue ASMs was based less on evidence than on the patient's priorities. Interestingly, his decision was also consistent with the study data showing that iEEG rarely leads to ASM discontinuation, reflecting a cautious but understandable human desire to avoid risk and maintain gains. Indeed, this impulse, often shared between patients and their physicians, may well prove to be a barrier to future prospective trials. In an ideal world, we would have access to more targeted studies providing us with actionable data, and perhaps someday we will. Until that (faraway) time, and even beyond it, a personalized approach will always be needed. Epilepsy care is as much about understanding the person as it is about understanding the condition.