Abstract
The case report by Falsaperla et al. (2022) describes a neonate that required extensive anticonvulsive treatment, including continuous intravenous high-dose midazolam with subsequent suppression of electrocortical activity in amplitude-integrated EEG (aEEG) during therapeutic hypothermia. In the context of hypoxemic ischemic encephalopathy, burst suppression and flat trace are associated with adverse outcomes in neonates, children, and adults (Bourgoin et al., 2020; Goeral et al., 2017; Sugiyama et al., 2016). However, deep sedation can cause burst suppression or flat trace and is sometimes deliberately induced to control refractory epileptic state (Singh et al., 2020). In this case report, the attending practitioners were unaware of the interaction of midazolam with multiple sedative agents and, therefore, misjudged the infant's clinical condition.
This report represents an important piece of work because it points out the necessity to consider all accompanying factors that potentially affect electrocortical activity. In neonatal and pediatric intensive care, the lack of knowledge on the effect of sedative and anticonvulsant agents poses serious challenges for aEEG interpretation. I congratulate the authors for their courage to publish this self-critical report, thereby increasing awareness in the neonatal and pediatric intensive care community for potentially unexpected effects of sedatives and anticonvulsants on electrocortical activity.
Our group has previously mentioned the case of a 5-month-old toddler with benzodiazepine intoxication (Bruns et al., 2021) and later-diagnosed genetic epilepsy. This infant received oxcarbazepine, vigabatrin, and repetitive administration of phenobarbital to interrupt seizures (serum level at admission 78μg/mL). Administration of a 16-fold lorazepam overdose during a seizure induced coma and hypopnea, requiring transfer to our hospital. At pediatric intensive care unit admission, the infant was unresponsive to painful stimuli but had preserved defensive and peripheral reflexes. The aEEG and raw EEG displayed a burst suppression pattern unresponsive to two antagonization trials with flumazenil (Fig. 1A).
Drug-induced changes of aEEG background.
Intubation was omitted in favor of nasal high-flow therapy to bridge the time until sufficient spontaneous breathing. Further symptoms during the intoxication included mild arterial hypotension, temperature instability, and generalized edema. Over the course of more than a week, the aEEG background recovered to a discontinuous and finally continuous pattern (Fig. 1B), and the infant became gradually responsive until achieving full consciousness after ∼1.5 weeks.
The case report by Falsaperla et al. and our case highlight that high-dose benzodiazepines carry the potential to induce coma and burst suppression, potentially enhanced by interaction with other sedative or neuroactive drugs such as barbiturates and/or opioids. However, I would like to put emphasis on the fact that the effect of sedatives is not limited to amplitude depression. For example, ketamine can cause amplitude elevation (Fig. 1C). The combination of midazolam and propofol can induce either amplitude elevation (Fig. 1D) or depression depending on the dose (Fig. 1E). Further factors that potentially affect aEEG amplitudes are intracranial hemorrhage or fluid accumulation, trapped air after neurosurgery, and a missing skullcap after decompressive craniectomy. Thus, any deviation of amplitudes from age-adjusted normal values in critically ill or sedated patients requires consideration of accompanying neuroactive drugs and underlying medical conditions.
Footnotes
Author's Contribution
N.B. wrote the letter and composed the figure.