Debunking the Myth: The Null Effect of Antiseizure Medications on the Functional Outcomes of Acute Symptomatic Seizures

Abstract

Antiseizure Medication Use and Outcomes After Suspected or Confirmed Acute Symptomatic Seizures

Zafar SF, Sivaraju A, Rubinos C, Ayub N, Awodutire PO, McKee Z, Chandan P, Byrnes M, Bhansali SA, Rice H, Smith-Ayala A, Haider MA, Tveter E, Erlich-Malona N, Ibanhes F, DeMarco A, Lewis S, Dhakar MB, Punia V. JAMA Neurol. 2024;81(11):1159–69. doi: 10.1001/jamaneurol.2024.3189.

Importance: Antiseizure medications (ASMs) are frequently prescribed for acute symptomatic seizures and epileptiform abnormalities (EAs; eg, periodic or rhythmic patterns). There are limited data on factors associated with ASM use and their association with outcomes. Objectives: To determine factors associated with ASM use in patients with confirmed or suspected acute symptomatic seizures undergoing continuous electroencephalography, and to explore the association of ASMs with outcomes. Design, Setting, and Participants: This multicenter cohort study was performed between July 1 and September 30, 2021, at 5 US centers of the Post Acute Symptomatic Seizure Investigation and Outcomes Network. After screening 1717 patients, the study included 1172 hospitalized adults without epilepsy who underwent continuous electroencephalography after witnessed or suspected acute symptomatic seizures. Data analysis was performed from November 14, 2023, to February 2, 2024. Exposure: ASM treatment (inpatient ASM continuation ≥48 h). Main Outcomes and Measures: Factors associated with (1) ASM treatment, (2) discharge ASM prescription, and (3) discharge and 3-month Glasgow Outcome Scale score of 4 or 5 were ascertained. Results: A total of 1172 patients (median [IQR] age, 64 [52-75] years; 528 [45%] female) were included. Among them, 285 (24%) had clinical acute symptomatic seizures, 107 (9%) had electrographic seizures, and 364 (31%) had EAs; 532 (45%) received ASM treatment. Among 922 patients alive at discharge, 288 (31%) were prescribed ASMs. The respective frequencies of inpatient ASM treatment and discharge prescription were 82% (233 of 285) and 69% (169 of 246) for patients with clinical acute symptomatic seizures, 96% (103 of 107) and 95% (61 of 64) for electrographic seizures, and 64% (233 of 364) and 48% (128 of 267) for EAs. On multivariable analysis, acute and progressive brain injuries were independently associated with increased odds of inpatient ASM treatment (odds ratio [OR], 3.86 [95% CI, 2.06-7.32] and 8.37 [95% CI, 3.48-20.80], respectively) and discharge prescription (OR, 2.26 [95% CI, 1.04-4.98] and 10.10 [95% CI, 3.94-27.00], respectively). Admission to the neurology or neurosurgery service (OR, 2.56 [95% CI, 1.08-6.18]) or to the neurological intensive care unit (OR, 7.98 [95% CI, 3.49-19.00]) was associated with increased odds of treatment. Acute symptomatic seizures and EAs were significantly associated with increased odds of ASM treatment (OR, 14.30 [95% CI, 8.52-24.90] and 2.30 [95% CI, 1.47-3.61], respectively) and discharge prescription (OR, 12.60 [95% CI, 7.37-22.00] and 1.72 [95% CI, 1.00-2.97], respectively). ASM treatment was not associated with outcomes at discharge (OR, 0.96 [95% CI, 0.61-1.52]) or at 3 months after initial presentation (OR, 1.26 [95% CI, 0.78-2.04]). Among 623 patients alive and with complete data at 3 months after discharge, 30 (5%) had post-discharge seizures, 187 (30%) were receiving ASMs, and 202 (32%) had all-cause readmissions. Conclusions and Relevance: This study suggests that etiology and electrographic findings are associated with ASM treatment for acute symptomatic seizures and EAs; ASM treatment was not associated with functional outcomes. Comparative effectiveness studies are indicated to identify which patients may benefit from ASMs and to determine the optimal treatment duration.

Commentary

An acute symptomatic seizure (ASyS) is a clinical seizure occurring shortly after an acute brain insult, such as metabolic, toxic, structural, infectious, or inflammatory causes. ¹ The timing of these seizures typically falls within the first 7 days postinsult. Unlike epilepsy, the cause of ASyS is often identifiable, and the risk of subsequent unprovoked seizures is lower. ² However, ASyS are associated with higher mortality rates, with a 30-day case fatality rate of 20%. ³ ASyS account for about 40% of all epileptic seizures and occur in approximately 3.6% of individuals over their lifetime. ² The increased use of continuous EEG (cEEG) monitoring has led to more frequent detection of electrographic ASyS and other epileptiform abnormalities (EAs) in patients with acute brain injuries. ⁴

The treatment of ASyS focuses on addressing the underlying cause of the seizures, with short-term antiseizure medications (ASMs) commonly used. ⁵ However, the optimal duration of ASM use after hospital discharge remains unclear, with 70%–90% of patients continuing ASM use for over a year post-insult. ⁶ The role of cEEG in guiding treatment decisions and the impact of ASMs on clinical outcomes in ASyS also remain uncertain. To explore these factors, Zafar and colleagues conducted a multicenter observational study from July to September 2021, involving five institutions from the Post-Acute Symptomatic Seizure Investigation and Outcomes Network (PASSION). ⁷ The study involved 1172 hospitalized adults without a prior history of epilepsy who underwent cEEG following suspected or witnessed ASyS.

The study revealed several key findings. ⁷ Clinical ASyS occurred in 285 patients (24%) before cEEG was initiated, while electrographic ASyS were identified in 107 patients (9%), with electrographic status epilepticus affecting 40 patients (3%). EAs were found in 364 patients (31%). Nearly half of the participants received ASM treatment during hospitalization, including both initiation and continuation for at least 48 h. Of these, 150 patients (24%) were treated despite no confirmed clinical or electrographic ASyS or EAs. Among the 922 patients discharged alive, 288 (31%) left the hospital with ASMs.

Patients with acute brain injuries, as well as those with progressive (eg, tumors) and subacute (eg, inflammatory or autoimmune) etiologies, were more likely to receive ASMs during hospitalization and after discharge. A significant association was also found between the presence of clinical or electrographic ASyS, EAs, and the location of initial care (neurology or neurosurgery services) and the timing of cEEG initiation. At the three-month follow-up, approximately one-third of the cohort continued ASM therapy, with only 5% experiencing a recurrence of seizures. Despite the widespread use of ASMs, their use did not correlate with favorable functional outcomes at discharge or at three months, as assessed by the Glasgow Outcome Scale (GOS).

What can we learn from these findings? First, ASMs remain widely used for ASyS, both in the acute phase and long-term, despite the lack of robust data to guide treatment decisions. Over 60% of patients were discharged on ASMs, indicating prolonged use beyond the acute phase. However, most patients did not experience seizure recurrence during the 3-month follow-up. While one might argue that the lower observed rate of seizure recurrence could be linked to the widespread use of ASMs in this cohort, prior research has consistently shown that the risk of unprovoked seizure recurrence—even in patients with structural causes of ASyS—is significantly lower. ⁸ This suggests that for most patients with ASyS, the continued use of ASMs after the acute phase may be unnecessary. The overuse or inappropriate use of ASMs could lead to adverse effects (AEs), prolonged hospital stays, and increased healthcare costs. AEs such as cognitive slowing, mood disturbances, gait instability, polypharmacy, and potential drug interactions are particularly concerning for patients recovering from acute brain injuries. Additionally, currently available ASMs are neither anti-epileptogenic nor disease-modifying and do not prevent seizure recurrence, even if initiated. ⁹

Second, the study highlights an association between the inpatient location of initial care—specifically whether patients were admitted to neurology or neurosurgery services—and ASM administration. This is likely influenced by the level of neurologist involvement, as neurologists typically conduct cEEG, and findings may impact ASM treatment decisions. There may also be an inherent bias toward diagnosing ASyS and initiating ASM treatment. Consistent with this, a recent global survey of over 500 neurological providers indicated that ASM use following ASyS is nearly universal, even in cases where cEEG findings reveal no EAs. ¹⁰

This publication also underscores the significance of not resorting to overtreatment based solely on EEG findings, particularly in cases of acute brain injury where EAs may not reliably predict the likelihood of seizure recurrence. Previous studies have shown that EEG findings in patients with electrographic ASyS or EAs often resolve in subsequent EEG tests and do not consistently indicate the risk of future seizures. ¹¹ Once again, this reinforces the notion that long-term ASM therapy is unnecessary for most patients, even for those with confirmed ASyS or EAs.

Finally, despite the extensive use of ASMs, the current study provides strong evidence that ASMs may not significantly improve functional outcomes in patients with ASyS. Previous studies have shown similar findings, with some instances where continued or escalated ASM therapy has been associated with poorer outcomes. Therefore, ASMs should primarily be used during the acute phase of treatment, with careful consideration for patients at higher risk of unprovoked seizures. Moving forward, further research is needed to determine whether a more tailored management strategy—taking into account seizure type (clinical vs electrographic), cEEG findings, disease severity, and the underlying cause—can improve patient outcomes.

Limitations of the article reviewed in this commentary include its reliance on inpatient data and outcomes measured only over a three-month period, which may underestimate the development of unprovoked seizures later on. Additionally, the study did not examine the clinical decision-making behind ASM use or factors like driving restrictions that might influence ASM prescriptions postdischarge. Future prospective studies should explore these aspects further. The study also falls short in tracking the progression of cEEG abnormalities over time, which could provide valuable insights into how follow-up EEG findings may affect treatment decisions. Importantly, this observational retrospective study does not serve as a basis for long-term recommendations or standardized guidelines concerning the management of ASMs in patients with ASyS.

In conclusion, this publication highlights the widespread use of ASMs during hospitalization and postdischarge for patients with ASyS and identifies key factors that shape current ASM practices in this population. However, it also highlights the pressing need for prospective, comparative effectiveness studies to identify which patients benefit most from ASM therapy and to determine the optimal duration of treatment. Future research should also explore the role of postdischarge EEG and brain MRI in predicting seizure recurrence and guiding ASM tapering. Until such studies are conducted, ASM treatment should be individualized based on factors like the underlying cause of ASyS, rather than following a one-size-fits-all approach. Given the increasing incidence of ASyS—especially with the rise of cEEG monitoring—enhancing our understanding of ASyS and the use of ASMs is crucial.