Effect of xenon and dexmedetomidine as adjuncts for general anesthesia on postoperative emergence delirium after elective cardiac catheterization in children: study protocol for a randomized, controlled, pilot trial

Emergence delirium (ED) is a manifestation of acute postoperative brain dysfunction that occurs with a relatively high frequency after pediatric anesthesia. The incidence varies depending on the diagnostic criteria used and the combination of administered anesthetic drugs [1]. It can be associated with either short-term complications, such as an increased risk of injury in children, parental dissatisfaction, increased costs, and length of hospital stay, or potential long-term consequences, including persistent psychological and behavior problems [2]. While the exact pathophysiological mechanisms underlying ED remain unknown, the use of sevoflurane has been identified as one of the most important risk factors [3, 4] There is a continuous search for anesthetics that ameliorate or avoid the occurrence of emergence agitation [2].

In children with congenital heart disease (CHD), therapeutic decisions are often based on the hemodynamic parameters measured during cardiac catheterization. The ideal anesthetic technique for this procedure should maintain hemodynamics that reflect awake baseline conditions to provide the best therapeutic option. Unfortunately, most anesthetics affect myocardial function and the vascular tone, and the ideal anesthetic regimen for these procedures remains a subject of debate.

In adults, xenon anesthesia preserves baseline blood pressure (BP) better than volatile anesthetics or propofol [5]. Moreover, xenon has been repeatedly demonstrated to offer neuroprotection in various animal models and models of neuronal injury [6]. More specifically, xenon has been demonstrated to attenuate isoflurane-induced neurotoxicity in rats [7]. Whereas xenon can be used as a mono-anesthetic in adults, it can possibly only be used as an additive to other anesthetics in children. The minimum alveolar concentration (MAC) of xenon in children is unknown but, concurring with other anesthetics, is expected to be higher than in adults (MAC 63.1%) [8]. In a meta-analysis, the MAC of xenon for children at the age of 1 year was calculated to be 92% [9]. As a consequence, xenon can only be used in sub-anesthetic concentrations for children to guarantee an inspiratory oxygen concentration of at least 30%. While sub-anesthetic concentrations of xenon were demonstrated to be neuroprotective [10], they need to be supplemented by another anesthetic or sedative drug to achieve sufficient depth of anesthesia.

We previously demonstrated in children with CHD that the addition of 60% xenon to sevoflurane resulted in a 60% reduction of the mean expiratory sevoflurane concentration of sevoflurane to achieve a comparable depth of anesthesia [11, 12]. Moreover, we did find a decreased incidence of ED when xenon was added to sevoflurane anesthesia [11]. However, we could not demonstrate increased hemodynamic stability. Most probably, the hemodynamic advantages known from adults receiving mono-xenon anesthesia were masked by the necessary addition of sevoflurane to achieve an acceptable depth of anesthesia in children receiving xenon.

The ideal drug to add to xenon for balanced anesthesia should potentiate the anesthetic effects of xenon without jeopardizing hemodynamic stability.

Due to its sedative, anxiolytic, sympatholytic, and analgesic properties, dexmedetomidine has recently been introduced in different areas of pediatric anesthesia as an adjunct for balanced anesthesia, for premedication or as part of sedation techniques. Moreover, a recent systematic review and meta-analysis in pediatric patients undergoing congenital cardiac surgery found the administration of dexmedetomidine to result in more stable intraoperative hemodynamics, a lower incidence of ED, and an attenuated stress response [13‐15]. A sparing effect of the MAC for sevoflurane has been demonstrated when an initial bolus of dexmedetomidine was followed by an infusion [16]. Moreover, the infusion of dexmedetomidine produced a dose-dependent decrease in the amount of sevoflurane required to produce 50% excellent tracheal intubation conditions in children [17]. Notably, like xenon, dexmedetomidine was also found to have neuroprotective properties in preclinical research. More specifically, it was observed that dexmedetomidine protects against anesthesia-induced neurotoxicity [18].

Altogether we can conclude that both xenon and dexmedetomidine possess numerous desirable properties with respect to the treatment of pediatric patients with CHD, including a benign hemodynamic profile [5, 14, 15], a reduced catecholamine release, anesthetic-sparing effects, a decreased incidence of ED, and neuroprotective effects [13]. This innovative combination of xenon and dexmedetomidine could be of particular interest for pediatric anesthesia, as combining these anesthetics has never been described as a balanced-anesthesia technique in children. Moreover, pediatric data on both drugs are scarce and, therefore, the FDA had no evidence to include these products in the “list of general anaesthetic and sedation drugs” for which a label change was requested indicating that “the repeated or lengthy use during surgeries or procedures in children younger than three years or in pregnant women during their third trimester may affect the development of children’s brains” [19].

ED is a frequently observed condition following pediatric anesthesia as a consequence of which a child experiences emotional suffering, restlessness, and psychomotor disturbances. It can last up to 45 min, interferes with recovery, is associated with prolonged length of stay in the PACU and the hospital, and parental dissatisfaction. Additionally, ED can result in long-term maladaptive and behavioral disturbances (e.g. insomnia, fear, bed wetting, etc.) [2]. The diagnosis of ED remains notoriously difficult in young children. For several reasons, we opted in this study (as we did in several previous ones) for the Watcha scale as an instrument to assess the incidence of ED in this young population [11, 12, 28]. In fact, the frequently used PAED scale was derived and validated in a cohort of children with a mean age of 3.7 years [23]. In clinical practice, several items of the PAED (e.g. eye contact with the caregiver or awareness of the surrounding) appear very difficult to assess in neonates and infants. Interestingly, the latter items have recently been identified to be indicative for pain rather than for delirium [29]. Moreover, in a direct comparison of different scaling systems, the Watcha scale was found to be a more practical tool to assess ED in children after surgery and to have the highest overall sensitivity and specificity [22]. Note that the results of testing with the PAED scale will also be reported to allow comparison between these two instruments, an approach that is now increasingly be performed in the literature [4].

We specifically choose this patient population because: (1) the brains of these children are probably vulnerable to the adverse neuropsychiatric effects of sevoflurane; (2) younger age is a risk factor for ED; and (3) we observed a high incidence of ED in a population of children of the same age undergoing the same procedure [4, 11, 30].

Assessing the depth of anesthesia in children remains controversial. Moreover, the validity of the BIS monitor has never been formally tested in children undergoing combined anesthesia with xenon and sevoflurane. Available evidence suggests, however, that the BIS monitor is a suitable instrument to assess the depth of xenon anesthesia. In one study, during balanced xenon anesthesia, BIS values correlated well with the clinical assessment of hypnotic depth and were within the recommended range of adequate anesthesia [31]. Moreover, Stoppe et al. demonstrated that the BIS showed a similar response to xenon as to sevoflurane-anesthesia. In the latter study, the depth of anesthesia was additionally confirmed by the measurement of auditory evoked potentials [32]. Lastly, EEG changes observed during xenon anesthesia resembled closely those instigated by propofol [33]. Furthermore, although age-related changes in BIS values in children were observed, it has been suggested that the BIS is a reliable monitor of depth of anesthesia at least in children aged > 1 year, showing better performance with increasing age [34, 35]. Note that in our study, the adequacy of depth of anesthesia will primarily be assessed by physiological signs (absence of HR or BP increases and involuntarily movements) and the BIS monitor is only an additional tool to quantify depth of anesthesia.

We will assess the peri-interventional release of S100β as a secondary endpoint. The astroglial receptor protein S100β is involved in several cellular processes such as neuronal differentiation, axonal growth, and calcium homeostasis [36]. Increased levels of S100β in the peripheral blood have been suggested to indicate an increased permeability of the blood–brain barrier and found to be associated with neurotrauma, cerebral ischemia, cardiac arrest, and cardiac surgery [37]. In neonates suffering from CHD, levels of S100β were described to be inversely related to cerebral blood flow and to mortality [38]. In children undergoing general anesthesia for cardiac catheterization, we recently reported early neurocognitive deficits that were paralleled by an increase in S100β blood levels [12]. In adult patients undergoing cardiac surgery, high levels of S100β were observed to be correlated with adverse neuropsychological and psychiatric outcomes [39] and to predict the occurrence of postoperative delirium [36]. The exact mechanisms resulting in a perioperative/peri-interventional increase in S100β remain still unknown but might involve neuro-inflammatory pathways with astrocytes being stimulated by cytokines originating from systemic inflammation to release S100β. Therefore, we will also assess the peri-interventional release of the pro- and anti-inflammatory cytokines IL-6 and IL-10 in our population. We acknowledge that the findings will be exploratory, solely describe associations, and should not be used to assess causation.

The current standard of care for general anesthesia in childhood overall, and for cardiac catheterization more specifically, is sevoflurane. The use of sevoflurane has been linked to an increased incidence of ED [4]. Moreover, sevoflurane was listed by the FDA as potentially neurotoxic when used in children aged < 3 years (during the episode of rapid brain development). In contrast, dexmedetomidine and xenon have no proven neurotoxicity in preclinical research and even have the potential to reduce the incidence of ED [40]. The combination of both drugs has never been described as a balanced-anesthesia technique in children.

The exposure of young children to the combination of dexmedetomidine and xenon will provide the first human data of a pediatric anesthesia regimen without the use of any drugs that were indicated by the FDA to potentially impair neurodevelopmental outcome when repeatedly or lengthily used for general anesthesia and sedation during surgeries or procedures in children aged < 3 years or in pregnant women during their third trimester of pregnancy.

We acknowledge that our study is subject to several limitations. First, the attending anesthesiologist cannot be blinded for the interventional treatment. However, the primary outcome will be assessed by an investigator unaware of the treatment allocation (“observer blinded”). Second, it is not defensible to extrapolate the feasibility of xenon and dexmedetomidine as adjuncts for general anesthesia in young children undergoing cardiac catheterization to surgical settings with severe painful stimulation. Third, this randomized controlled trial is not designed to prove the potential superiority of xenon/dexmedetomidine versus sevoflurane anesthesia with regard to the incidence of ED. To perform a trial with this purpose, a precise idea of the expected ED incidences is necessary for appropriate sample size estimation. The current study will deliver data allowing to design an adequately powered randomized controlled trial addressing this research question. Note that the current study is also not sufficiently powered for all secondary endpoints, which should, therefore, be considered as purely explorative and hypothesis-generating.

In conclusion, this pilot trial will be performed to investigate the incidence of ED in an anesthesia regimen that does not involve sevoflurane. Moreover, we will study the feasibility and the safety of combined dexmedetomidine-xenon anesthesia in children.