Elucidating the Neural Correlates of Anesthesia-induced Unconsciousness in Neonates

Overview

This study will try to elucidate the neural responses of neonatal patients to surgery-level propofol infusions during induction and maintenance of anesthesia. The main objective is to elucidate the EEG patterns of neural activity that can reliably reflect the level of hypnotic depth in full term neonates undergoing propofol anesthesia based on the classification of EEG into states and the characterization of discrete neonatal EEG events.

Appropriated dosage of anesthetic drugs is essential to reduce the incidence of negative outcomes during and after surgery. In neonates, this is of particular importance because of their intrinsic vulnerability. Administration of an insufficient dose of hypnotic drugs like propofol to a neonate can result in episodes of intraoperative awareness and strong pain responses that cause unnecessary physiological stress, which may be hard to detect intraoperatively because of muscle relaxants. On the other hand, excessive dosage of drugs like propofol for long periods of time can drastically disrupt ongoing spontaneous activity, which is essential for the normal development of synaptic connections within the cerebral cortex, especially around the time of birth. Also, it has been suggested that these drugs may have a neurotoxic effect in neonates. This highlights some of the reasons why the highest degree of precision is required when administering hypnotic anesthetic drugs to neonates. In adults, intraoperative anesthesia monitoring provides an objective measure of the magnitude of the anesthetic effect achieved by directly evaluating the patient's brain activity in real time. With this information, anesthesiologists can continuously titrate the doses of hypnotic drugs required to produce an optimal effect. This monitoring is done by continuously recording electroencephalographic (EEG) activity and comparing that activity with known markers of hypnotic depth like oscillatory activity of different frequencies. Importantly, this is possible because we know to a great extent what patterns of brain activity are related to insufficient, optimal and excessive dosages of propofol in the adult brain. However, we do not currently have this information about neonates' brain during anesthesia. Neonatal EEG activity is notoriously different to adults' EEG, both during anesthesia and without it. The EEG activity of adults under propofol anesthesia is dominated by either alpha (8 - 12 Hz, which is not present in neonates) and/or delta (1- 4Hz) oscillations that are relatively consistent in time. This means that the amplitude and frequency of these oscillations do not vary significantly in time but remain mostly constant unless there are changes in the patient's hypnotic depth because of a change in dosage. In contrast, neonatal EEG activity during anesthesia is notoriously discontinuous; it transitions from low to high amplitude activity, and is filled by discrete events including sharp waves, bursts and delta brushes. This makes the adult's EEG markers of anesthesia unusable in neonates. The specific characteristics of these discontinuities and discrete events have been mainly described by visual inspection or rough historical criteria, but not using modern data-driven EEG analysis techniques. In this work the investigators will directly record EEG activity from neonates before anesthesia (baseline), during propofol induction and during surgery-level anesthesia. In this way the investigators will be able to evaluate how do the putative neural signatures of hypnotic depth behave, and how closely they follow propofol concentration and propofol effect (i.e. loss of consciousness/response). Because of this, the current study will be a prospective observational analytic cohort study. The study will include full term neonates (37 - 42 weeks of gestational age) that require general anesthesia/deep sedation during their first 28 days of life because they require a surgery or a specific procedure. This will include patients from cardiac and non-cardiac surgeries like hernias, superficial abdominal surgery, radiological procedures, procedures related to airway assessment, among others. Exclusion criteria will include patients scheduled for neurological surgeries, with perinatal asphyxia, with evidence of neurological injury or pathology, suspected brain malformations, and patients with metabolic or hemodynamic instability at the time of the surgery or procedure. A considerable number of neonatal surgeries and procedures are conducted in patients that are already partially sedated because their pathology or general medical condition requires it. This implies that EEG activity during baseline will depend on whether the patient was already sedated or not. To manage this, the investigators will separate patients into two groups: 1) a previously sedated group and 2) a not previously sedated group. Although we know that it will be harder to evaluate neural signatures of anesthesia in patients that are already sedated during baseline, we chose to include them in our study because we want our results to reflect the clinical reality of many neonatal patients. Also, at least in our institution, patients that require surgery or a procedure with anesthesia are more likely to be already sedated. We will record the drugs and doses administered to each patient in the previous 24 hours including vasoactive and sedatives. Separating neonates based on their sedation state before the surgery or procedure will, at the same time, separate patients based on their level of severity of illness. The researchers will directly measure this possible confounding effect to include it in our models. They will quantify severity of illness with two indices, the "Neonatal therapeutic intervention scoring system" (NTISS) and the "Neonatal sequential organ failure assessment" score (nSOFA). These consider dozens of factors like if the patient requires supplemental oxygen, if they have required transfusions of any blood products or if they have a central venous line, among many others. Each one of these criteria is considered for the final score. In this way, we will have two objective measures of patient's severity of illness to be used as covariates in data analyses. They will help to better understand if a given EEG property is a marker of hypnotic depth or just a marker of severity of illness.