Abstract
In his thoughtful editorial “Early discontinuation of antiseizure medication in neonatal seizures - Pro-ceed with caution,” Dr. Volpe discusses his concerns that the recommendation for “routine discontinuation” of anti-seizure medication (ASM) after resolution of acute provoked neonatal seizures is too broad. We recommend this practice based on evidence from a large, multi-center comparative effectiveness study from the nine centers of the Neonatal Seizure Registry (NCT02789176) [1, 2]. We agree with Dr. Volpe’s recommendation to proceed with caution and also stand by our conclusions that routine discontinuation of ASM after resolution of acute provoked neonatal seizures and prior to hospital discharge is safe and warranted in most cases.
Dr. Volpe’s considerations focus on four key points, which we clarify below
Establishing the cause of seizures
We agree with Dr. Volpe that the treating clinician must always consider the neurological disorder causing neonatal seizures. Our study included only neonates with acute, symptomatic neonatal seizures, as diagnosed using modern neurocritical care tools. These included continuous video-EEG monitoring to confirm electrographic seizures and their resolution, magnetic resonance imaging to confirm an acute symptomatic seizure etiology and exclude a congenital malformation, and testing to identify genetic neonatal onset epilepsy. Children with neonatal onset epilepsy require a longer treatment course than infants with acute provoked seizures; the recommendation for routine discontinuation of ASM prior to hospital discharge only applies to neonates with resolved seizures due to an identified acute cause. As Dr. Volpe points out, most neonates with acute, provoked seizures will not develop subsequent epilepsy, making ongoing ASM treatment after hospital discharge unnecessary for the majority in this specific group.
The importance of acute treatment to prevent the potential consequences of ongoing electrographic seizures
We wholeheartedly agree that early, rapid treatment of acute EEG seizures is important to optimize short and long term outcomes, as long duration of acute seizures has been linked to adverse outcomes [3, 4]. We encourage centers to improve resources and treatment pathways toward this end, and our recent publications provide potential strategies for improving the speed of treatment [5, 6]. We also agree that the benefits of a brief treatment course with phenobarbital or fosphenytoin outweigh the potential risks. However, current evidence, including ours, indicates that the 1–3 months of extended ASM treatment suggested by Dr. Volpe is likely unnecessary. Furthermore, if ASM is continued beyond hospital discharge into early infancy, there are no evidence-based criteria to decide at what point to maintain or discontinue ASM. We show that early follow-up EEG at three months does not refine predictions of the risk of epilepsy [7]. In our study, close to 20%of children with acute provoked seizures remained on ASM at 24 months (although just 13%overall developed post-neonatal epilepsy). This suggests that once a child is sent home on ASM, the pediatrician or parent may be reluctant to stop the medication, even in the absence of epilepsy. Continuing ASM beyond hospital discharge with the thought that it should be discontinued within a few months assumes that an infant will attend timely follow up with a clinician knowledgeable, experienced, and comfortable with ASM management in young infants, which unfortunately is not always the case.
How does the bedside clinician know when acute provoked seizures have resolved (and is 24 hours of EEG seizure freedom sufficient to determine resolution)?
Dr. Volpe cites World Health Organization (WHO) guidelines that recommend 72 hours of seizure freedom before discontinuing ASMs [8]. We are not opposed to this approach, which aligns with our recommendation to discontinue ASM prior to discharge home from the hospital. We also speculate that it is unlikely that most centers in the USA, and less so internationally, have the resources to continue EEG monitoring for a full 72 hours after the last seizure. Prolonged monitoring is not without risk; there is increased risk of skin breakdown with > 72 hours of EEG monitoring. We strongly assert that a minimum of 24 hours EEG monitoring to confirm seizure resolution is needed. This aligns with the ACNS guidelines [9]. cited by Dr. Volpe. We further highlight conditions under which we would maintain EEG for longer. For example, extended monitoring is warranted when the EEG background remains extremely abnormal (e.g., burst suppression, or severe continuous suppression) or is progressively worsening. Additionally, if there are ongoing runs of epileptiform discharges that fall on the ictal-interictal spectrum, or if there is ongoing risk of brain injury (for example in children with CNS vascular malformations), then continued EEG monitoring may also be warranted. The exact duration of EEG to determine “resolution” of acute seizures is not known and may depend somewhat on the etiology and clinical condition as described in example scenarios above.
A critical distinction is that the duration of EEG required to confirm resolution of acute neonatal seizures is not identical to the time at which we recommend routine discontinuation of ASM. While EEG monitoring is required for at least 24 hours to confirm seizure resolution as above, this does not require ASM to be stopped 24 hours after the last seizure. Our recommendation is simply that discontinuation of ASM occur routinely after resolution of acute symptomatic neonatal seizures and prior to hospital discharge. In our study, this approach led to a median of six days of ASM treatment (interquartile range 3, 11 days) for phenobarbital [1].
With respect to convalescent or serial EEG, we consider this to be clinically useful in two key scenarios: (1) to help with differential diagnosis in cases where it is not known whether acute provoked seizures, transient seizures, or genetic epilepsy is the cause, and (2) to determine treatment effect in children with early onset epilepsy and subtle or subclinical seizures. In those scenarios, a period of EEG monitoring after several days of scalp rest could reduce the risk of skin breakdown associated with prolonged EEG monitoring. We do not advocate universal EEG prior to discharge for all children with acute provoked seizures in the absence of these factors, as the natural history of provoked seizures is to resolve within days [10, 11].
The known risk of post-neonatal epilepsy in the setting of particular brain injuries, especially injuries that manifest with an early abnormal neurological examination
Dr. Volpe outlines the known risk of post-neonatal epilepsy in the setting of certain brain injuries, e.g., those involving the deep grey nuclei and/or large cortical territory injuries, as well as the importance of the neurological examination at discharge. Again, we agree with Dr. Volpe and add our own data showing that injury to the deep grey nuclei or brainstem and abnormal tone at discharge portend a high risk for infantile spasms [12]. However, even in the setting of high risk injuries, there is no evidence to suggest that maintaining ASMs such as phenobarbital or levetiracetam (the two most commonly used ASMs for neonatal seizures [3] and early-life epilepsy [13]) will alter the risk of epilepsy. In particular, we know that these medications will not be effective for treating infantile spasms [14]. Early and effective treatment for infantile spasms provides the best chance for favorable outcomes [15, 16]. Therefore, we suggest that rather than routinely continuing for months a medication that is not expected to prevent epilepsy, families would benefit from education such that they recognize seizures, especially infantile spasms, at their earliest onset and know how to reach the appropriate clinicians for efficient evaluation and treatment. Given the lack of evidence for benefit of maintaining ASMs and the neurodevelopmental and epilepsy safety data from our study, [1] we conclude that the advantages of ASM discontinuation prior to hospital discharge outweigh the potential risks, an approach that was endorsed by our Parent Advisory Panel of 12 parents of children with neonatal seizures, including representatives from key parent organizations [2].
In sum, we agree with Dr. Volpe that the treating neonatologist and neurologist must carefully consider the cause of neonatal seizures and the anticipated natural history of the underlying condition. We all must “think twice” about discontinuing ASM. We suggest that clinicians ask themselves these three simple questions to help determine whether ASMs can safely be discontinued prior to discharge from the neonatal seizure admission: Have the EEG seizures resolved for at least 24 hours? Has epilepsy been excluded as the cause of neonatal seizures (i.e., has an acute provoked or transient etiology been confirmed)? Have the parents been educated about what to do in case of suspected post-neonatal seizures and how to recognize infantile spasms?
If the answer to each of these questions is “yes,” we assert that the clinician has “proceed[ed] with caution” and appropriately considered whether or not to stop ASM. This approach should lead to discontinuing ASM routinely prior to hospital discharge for most neonates with acute provoked seizures.
Footnotes
Acknowledgements
The authors would like to thank Drs. Donna Ferriero, Janet Soul, and Courtney Wusthoff for their thoughtful contributions.