Impaired Cerebral Autoregulation in Children

Abstract

Managing acute brain injury involves protecting the brain from secondary injury by addressing the mismatch between metabolic demand and cerebral perfusion. Observational studies have associated impaired cerebral autoregulation, a physiological process governing the regulation of cerebral blood flow, with unfavorable neurological outcomes in both pediatric and adult populations. We review the pathophysiology of cerebral autoregulation and discuss methods for assessing and monitoring it in children after acquired brain injury. We also examine the current research investigating the relationship between impaired cerebral autoregulation and outcomes following traumatic brain injury, cardiac arrest, cardiopulmonary bypass, and extracorporeal membrane oxygenation. Furthermore, we outline potential areas for future research in cerebral autoregulation and its clinical implications for pediatric patients with brain injuries.

Introduction

Among 220,000 children admitted annually to pediatric intensive care units, 20%-26% have a new or existing neurological diagnosis, including acute brain injury (ABI).^1-3 Children with ABI, including cardiac arrest, stroke, and traumatic brain injury (TBI), have higher morbidity and mortality rates compared with the general pediatric intensive care unit population.^4,5 Although pediatric neurocritical care has advanced significantly over the past few decades,^3,6 there is a need for newer approaches to neuroprotection to reduce morbidity and mortality in critically ill children.

The current approach to neuroprotection following primary brain injury focuses on minimizing the risk of secondary brain injury.^7-9 Secondary brain injury occurs hours to days after the initial event and is caused by hypoperfusion, worsening hypoxia and ischemia, increased intracranial pressure (ICP), and reperfusion injuries mediated by an increase in free radicals, oxidative damage, and both pro- and anti-inflammatory signaling that leads to apoptotic cell death.^10,11 This strategy aims to optimize cerebral physiology to better align metabolic demand with cerebral perfusion. Traditionally, this approach has followed a “one-size-fits-all” model, where standardized neuroprotective strategies are applied to different types of ABI based on population-derived targets, without fully considering the pathophysiologic heterogeneity and phenotypic diversity that contribute to secondary brain injury.⁹ More recently, a shift toward personalizing neuroprotection has emerged, utilizing multimodality neuromonitoring to detect, characterize, and treat the pathophysiologic processes occurring during secondary brain injury.

Cerebral autoregulation (CA) is a protective hemodynamic mechanism that maintains cerebral blood flow (CBF) in response to fluctuations in systemic blood pressure through vasoconstriction and vasodilation of cerebral vasculature.^12,13 CA is impaired following ABI, increasing the risk of hypo- or hyperperfusion, which can exacerbate secondary brain injury.^14-20 Advances in technology have facilitated the real-time, continuous evaluation of CA at the bedside,^21-24 thus presenting opportunities for CA-focused neuroprotection to mitigate these effects and improve patient outcomes.^25,26

This review discusses the physiology of CA and methods for real-time CA assessment in pediatric neurocritical care. We review research on the association between impaired CA and outcomes after TBI, cardiac arrest, cardiopulmonary bypass (CPB), and extracorporeal membrane oxygenation (ECMO).

Regulation of cerebral blood flow

Perfusion pressure in any organ is defined by the difference between inflow and outflow pressures. In the brain, inflow pressure is primarily represented by the mean arterial pressure (MAP) in the internal carotid artery and outflow pressure is reflected in the bridging veins.²⁷ Since direct measurement of these pressures is not routinely used in clinical practice, estimate measures are commonly employed. MAP is typically used as a surrogate for inflow pressure. Owing to the linear relationship between ICP and pressure in the bridging veins, ICP is used as an estimate of outflow pressure.²⁸ If the venous pressure remains normal or low, MAP can be used as an estimate of the cerebral perfusion pressure (CPP).²⁹ This simplified approach does not apply to patients with elevated ICP, in whom CPP ultimately determines cerebral perfusion.³⁰

CBF is the rate at which blood flows to the brain.³¹ CBF is determined by the CPP divided by cerebral vascular resistance.^13,31-33 The Poiseuille law for resistance describes that the resistance is inversely proportional to the radius raised to the fourth power. These principles highlight the significant influence of vessel radius on CBF and underscore the importance CA in reducing the risk of injury from hypo- or hyperperfusion.

CA can be understood as a complex process influenced by multiple factors, including myogenic, neurogenic, metabolic, chemical, and endothelial mechanisms.^13,29 These processes synergistically collaborate to ensure sufficient delivery of oxygen and nutrients to the brain. Table 1 briefly covers these mechanisms, while this review focuses solely on the myogenic mechanisms and its clinical implications in pediatric neurocritical care.

TABLE 1.

CA Mechanisms That Regulate Cerebral Blood Flow

Mechanism	Trigger	Physiologic Mechanism
Myogenic	CPP/MAP	Changes in intraluminal pressure results in mechanotransduction, depolarization of vascular muscle, and Ca++ mobilization.13,34-36
Chemical or metabolic	CO2/pH, partial pressure of O2, or O2 content	Local hypoxia induces anaerobic metabolism and release of vasodilator substances (i.e., adenosine)
Neuronal	Intrinsic neuronal and glial activation	Feedback model: metabolic by-products of brain activity including potent vasodilators such as adenosine, CO2, H+, and lactate, which could potentially initiate the flow response.37 Feedforward model: neurovascular signaling pathways resulting in the release of vasoactive by-products of synaptic activity, such as K+, nitric oxide, and prostanoids38
Endothelium-dependent responses	Physical stimuli (i.e., shear stress or hemorrhage), neurotransmitters, or cytokines.	Modulation of vascular tone through endothelial factors (i.e., thromboxane A2, endothelin-1, and nitric oxide)29

Abbreviations:

CA = Cerebral autoregulation

CPP = Cerebral perfusion pressure

MAP = Mean arterial pressure

Cerebral autoregulatory capacity refers to the range of MAP/CPP over which CBF remains constant. The lower limit of autoregulation (LLA) and upper limit of autoregulation (ULA) define the boundaries of this range. Below the LLA or above the ULA, these autoregulatory mechanisms fail, resulting in a linear relationship between CPP and CBF. This relationship may lead to either inadequate or excessive cerebral perfusion, potentially causing ischemia or hyperperfusion/cerebral edema, respectively.^35,39 The original description by Lassen³⁹ illustrated the relationship between CBF and CPP, forming an inverted “S”-shaped curve of CA (Fig 1). Although initial studies based on Lassen's work suggested that the cerebral autoregulatory capacity in adults ranged from approximately 50-150 mm Hg, more recent data have shown a narrower plateau and steeper slope than initially described. This finding highlights significant individual variability, with the LLA ranging from 43-90 mm Hg in adults during CPB.²⁹

FIGURE 1.

Cerebral autoregulation curve as described by Lassen. The cerebral autoregulatory capacity, marked with a double arrow, is the range of MAP/CPP over which CBF remains constant. The lower and upper limits of CA mark the boundaries of this range. More recent data have shown a narrower plateau and steeper slope than initially described. CBF, cerebral blood flow; CPP, cerebral perfusion pressure; LLA, lower limit of autoregulation; MAP, mean arterial blood pressure; ULA, upper limit of autoregulation.

Assessment of CA

CA can be assessed using both static and dynamic methods. Static CA involves measuring CBF once a steady state is achieved, usually over extended periods.²⁹ In contrast, dynamic assessments of CA assess the spontaneous relationship between CPP and CBF in response to changes occurring within seconds.³⁶ Dynamic investigations of CA provide additional information about the latency and magnitude of the CBF response to CPP variations.^36,40,41 Although there is a strong correlation between these methods in both animals and humans,⁴² the clinical implications require further investigation.

Cerebral autoregulation monitoring

Assessing CA requires continuous and synchronized measurements of MAP or CPP and a surrogate of CBF. A surrogate of CBF is a measurable physiological parameter that reflects changes in CBF. Some CBF surrogates described in the literature for this indication are mean flow velocity measured by transcranial Doppler ultra-sound, brain tissue oxygenation from near-infrared spectroscopy, and ICP, among others (Table 2). These techniques offer several advantages, including the ability to provide continuous monitoring, and some are widely available in clinical settings, allowing for real-time assessment of CA without the need for direct perfusion imaging. However, they differ in their degree of invasiveness and scope, with some measuring regional CA and others assessing overall cerebral perfusion. These factors influence their applicability depending on the clinical context and the resources available at each institution.

TABLE 2.

Methods for the Assessment of CBF and CA⁴³

Approach	Technique	Measurement	CA Index (Paired Physiological Measures)	Clinical Implications
Noninvasive	TCD	MCA FV	Mx (FV/CPP)	Easily affected by high ICP and low CPP; responds to temporary hemodynamic changes, allowing CPP-guided therapy.
			Mxa (FV/MAP)	Less robust than Mx, but ideal for patients without an ICP monitor.
			ARI (FV/MAP)	Dynamic or static model of CA, not used frequently in critically ill children.
	NIRS	Cerebral oxygen saturation	TOx or COx (TOI/CPP or MAP)	Easy to apply, but affected by adipose tissue, perfusion state, arterial O2 saturation, metabolic demand and anemia.
		Relative total hemoglobin	THx or HVx (THI/MAP)	Measures vascular resistance; ideal for patients without an ICP monitor.
Invasive	Microdialysis	PbtO2	ORx (PbtO2/CPP)	Highly focal measurement with conflicting literature regarding its association with outcomes
	LDF	Blood cell perfusion in the microvasculature	Lx (LDF/CPP)	Provides highly focal measurements to assess cortical autoregulation.
	ICP monitor	ICP	PRx (ICP/MAP)	Extensively reported, measures vascular resistance. Abnormal values precede refractory ICP elevation and help determine optimal CPP.
			Pax (pulse amplitude ICP/MAP)	Measures compliance of intracranial vessels, aiding in the determination of optimal CPP.

Abbreviations:

ARI = Autoregulation index

CA = Cerebral autoregulation

CBF = Cerebral blood flow

COx = Cerebral oxygenation index

CPP = Cerebral perfusion pressure

FV = Flow velocity

HVx = Hemoglobin volume reactivity index

ICP = Intracranial pressure

LDF = Laser Doppler flowmetry

Lx = LDF-derived autoregulation index

MAP = Mean arterial pressure

MCA = Middle cerebral artery

Mx = TCD-based reactivity index

Mxa = TCD-based reactivity index

NIRS = Near-infrared spectroscopy

ORx = Oxygenation reactivity index

Pax = Pressure-amplitude index

PbtO₂ = Brain tissue oxygen tension

PRx = Cerebrovascular pressure reactivity index

TCD = Transcranial Doppler

THx = Total hemoglobin index

TOx = Tissue oxygenation index

Once these measurements are obtained, a CA index can be calculated to quantify the integrity of CA. Two employed methods for analysis of these measurements either focus on how the signals change over time (e.g., moving correlation coefficients) or examine changes in signal patterns and frequencies over time (e.g., wavelet methods).^44,45 A detailed exploration of these mathematical models is beyond the scope of this review; however, it is important to recognize that each model has distinct advantages and limitations and their indices are not interchangeable. For instance, correlation coefficient-based methods are simpler to calculate and commercial software is available for this purpose. In contrast, wavelet methods may be more robust against external artifacts and exhibit reduced variability in CA indices.²² However, no commercial software currently exists for implementing wavelet-based analyses. Common CA indices used in pediatric neurocritical care include the pressure reactivity index (PRx), which is derived from the correlation between ICP and MAP, and the cerebral oxygenation index (COx), which is based on the moving correlation between cerebral oxygen saturation or tissue oxygenation index (TOI) and MAP. Various CA indexes exist depending on the surrogate measure used and the type of analysis that is performed. Other CA indexes are explained in more detail in Table 2.

When CA is impaired, both MAP or CPP and the surrogate marker of CBF change in the same direction, resulting in a positive correlation. In contrast, when CA is preserved, CBF remains stable despite changes in MAP or CPP, leading to a negative correlation. The threshold for defining impaired CA varies depending on the specific CA index; however, moving correlation coefficients above 0.3-0.5 generally indicate impaired CA.⁴⁶

To define the optimal MAP or CPP, referred to as MAP_OPT or CPP_OPT, respectively, one can plot MAP or CPP on the x-axis and an index of CA on the y-axis; this will result in a U-shaped curve (Fig 2), where the lowest point of the curve represents MAP_OPT or CPP_OPT. The points at which the curve intersects a predetermined cutoff define the lower and upper limits of CA. MAP_OPT values identified by wavelet methods compared with correlation methods have shown to have a higher association between clinical outcomes in neonates with hypoxic-ischemic encephalopathy²³ and adults with TBI.⁴⁷ Although these techniques provide valuable insight of CA function, they also introduce clinical heterogeneity and contribute to inconsistencies across different studies. There is currently no consensus regarding the gold standard for assessing CA, and the method to quantify CA remains dependent on institutional resources and personal preference.^48,49

FIGURE 2.

Calculation of optimal mean arterial blood pressure. The CA index COx represents the moving correlation coefficient between the cerebral oxygen saturation and MAP. From these data, a U-shaped curve (black line) was constructed to determine the optimal MAP. The vertical gray bars represent the distribution ( ± 1 S.D.) of COx values within each MAP bin. This curve suggests an optimal MAP of 73 mm Hg (blue dashed line). However, CA is preserved within a MAP range of 46-100 mm Hg. A horizontal red dashed line intersects at a threshold of 0.3 (although this may vary depending on the CA index),⁴⁶ indicating impaired CA and an increased risk of ischemia (blue area) or edema (red area) below and above these limits. CA, cerebral autoregulation; COx, cerebral oxygenation index; MAP, mean arterial pressure.

CA function in healthy and critically ill children

The following section will present the most recent evidence on CA function in children. Although emerging data suggest a potential association between impaired CA and poorer outcomes in a variety of clinical entities, most of this evidence is derived from small retrospective studies. There are no randomized clinical trials involving children that assess the feasibility or effectiveness of CA techniques in minimizing the risk of morbidity or mortality.

Limited data exist regarding the limits of CA in healthy children. The LLA in extremely-low-birth-weight infants is ~ 30 mm Hg.⁵⁰ Although no evidence establishes the LLA in term neonates, it is reasonable to presume that this limit increases with age. Studies examining CA in healthy children indicate that the LLA appears to be comparable to that of adults.⁵¹ To determine the LLA in healthy children, Vavilala et al. performed static CA testing by measuring the mean flow velocity in the middle cerebral artery while administering phenylephrine to children aged 6 months to 14 years undergoing general anesthesia for surgery. The LLA was very similar between older children and infants (59 ± 17 mm Hg vs 60 ± 8 mm Hg; P = 0.6). By using a tilt test, Vavilala et al. also showed that among healthy children aged between 10 and 16 years, boys demonstrated a greater CA function of the middle cerebral artery compared with girls; however, the reasons for these differences are not yet well understood.^52,53

Cardiac arrest

Studies in animals, adults, and children have demonstrated associations between impaired CA, MAP_OPT, and outcomes after cardiac arrest.^{11,17,22,54-60} In a pediatric swine model of cardiac arrest, Kirschen et al. found that a higher burden of CA impairment, quantified as the area under the cerebral blood flow reactivity index curve with a threshold of 0.3, was associated with abnormal neurological outcomes.⁵⁷ Sundgreen et al. showed that 72% of adult patients have an absent or right-shifted CA curve within the first 24 hours after return of spontaneous circulation (ROSC).⁶⁰

In children, a correlation exists between the burden of MAP < MAP_OPT and outcomes. Kirschen et al. examined the relationship between impaired CA, measured by COx, and unfavorable neurological outcomes, defined as death or a change in the Pediatric Cerebral Performance Category at hospital discharge or 30 days postcardiac arrest. Among 34 patients with a median age of 2.9 years, those with a higher burden of MAP < MAP_OPT in the first 24 hours after cardiac arrest were more likely to experience unfavorable outcomes.⁶¹ Similar findings were reported by Lee et al. Thirty-six children resuscitated from cardiac arrest underwent CA monitoring during the first 72 hours after ROSC. CA was quantified by using the hemoglobin volume index. Children who did not require ECMO and underwent tracheostomy/gastrostomy experienced longer durations with a MAP < MAP_OPT and exhibited more significant deviations from MAP_OPT during the second 24 hours following ROSC. In addition, children who died from a neurological cause (n = 19) had a greater MAP deviation from MAP_OPT during the first 48 hours after ROSC compared with children who survived or died from cardiovascular failure.⁵⁸ There are conflicting results in the literature regarding the association between the burden of impaired CA and neurological outcomes in children.^61,62

Traumatic brain injury

The most frequently used CA index in patients with TBI is PRx, which represents the correlation between ICP and MAP.^7,63 In studies involving adults, impaired CA has been observed in 49%-87% of patients with severe TBI.^64-66 Several studies have shown that PRx values > 0.2-0.3^67-69 are associated with unfavorable functional outcomes and increased mortality at six and 12 months post-TBI.^67,68,70-73 Vavilala et al. found that impaired CA is present in 42% of children after moderate to severe TBI,²⁰ especially among those younger than four years.¹⁶ In children, PRx was consistently elevated in those with poor outcomes and was associated with mortality after adjusting for CPP, ICP, Glasgow Coma Scale, and age.^74,75 Two additional pediatric studies have shown that impaired CA, measured by ARI and Mx, is also associated with poor outcomes, defined as a six-month Glasgow Outcome Scale <4.^20,76-80

The time spent with impaired CA and the percentage of time with CPP < LLA has also been associated with worse neurological outcomes, including death. Among 15 children with severe TBI, those with a Glasgow Outcome Scale of 2-3 at discharge (n = 7) spent a greater proportion of monitoring time with impaired CA (64 vs 6 hours, P < 0.01), defined as a PRx ≥0.2. Two pediatric studies have demonstrated that an increased percentage of time spent with CPP < LLA based on multiple CA indices is also correlated with poorer neurodevelopmental outcomes.^21,81 Moreover, a higher percentage of time with CPP > CPP_OPT has been associated with favorable outcomes in children with TBI.⁸² The association between CA status and patient outcomes is particularly significant in children aged 15 years and younger. In this group, a lower percentage of time with CPP < 10% below CPP_OPT has been linked to favorable outcomes.⁸³ In adults, the duration of CPP < LLA predicts mortality with significance from the third day postinjury, as evidenced by an area under the curve of 0.73 (P < 0.001).¹⁴

Cardiopulmonary bypass

Children with congenital heart disease are at high risk for neurological injury and neurodevelopmental impairment.⁸⁴ The etiology of this is often multifactorial, including genetic factors, altered CBF and oxygen delivery, as well as perioperative insults related to cardiac surgery and CPB.^84,85 In humans, most of our current knowledge of CA is based on the CBP literature.⁸⁶

Impaired CA is common during CPB.⁸⁷ In a cohort of 137 adult patients undergoing CBP during cardiac surgery, 43 individuals (34%) displayed an Mx ≥ 0.4 during the cooling phase.⁸⁸ Furthermore, 68 patients (53%) exhibited an average Mx of 0.4 or higher during rewarming, highlighting a possible modulation effect of a higher temperature on CA.⁸⁹ In addition, real-time monitoring of CA may offer a more precise method for MAP goals during CPB. In infants, Brady et al. estimated that the LLA could be assessed using a COx threshold value of 0.4, which corresponds to a MAP ±7 mm Hg, acknowledging there was a wide range of individual LLA in this cohort.⁹⁰

Studies in adults have demonstrated that CA is commonly impaired during CPB, and this impairment may be associated with worse clinical outcomes. However, there is relatively limited research examining the association between CA and clinical outcomes in children.¹⁵ A multicenter observational study involving 57 children undergoing CPB showed that higher CA indices during CPB were associated with higher levels of serum glial fibrillary acidic protein level, a serum biomarker of brain injury.⁹¹ A prospective single-center study of 80 children who underwent CPB showed that more impaired CA during the 48 hours post-CPB was associated with a higher degree of brain injury based on electroencephalography and brain magnetic resonance imaging.⁹² A single-center retrospective observational study of 28 neonates who underwent cardiac surgery showed that those patients with an acute neurological event experienced higher CA indices in the first 72 hours postoperatively, a larger percentage of time with impaired CA, and a longer duration of time with CPP < LLA.²⁴

Extracorporeal membrane oxygenation

Children on ECMO experience frequent fluctuations in MAP, partial pressure of oxygen, and carbon dioxide, all of which affect CA and CBF.^93,94 As a result, this population is particularly prone to develop cerebral ischemia. It has been proposed that the pathophysiology of impaired CA in patients on ECMO may be related to a decrease in the pulsatility of CBF^95-97 and activation of vasoactive mediators, including prostacyclin, thromboxanes, complement, and cathecolamines, which ultimately causes endothelial dysfunction.^96,98

In an animal model of venoarterial ECMO (VA-ECMO), CA curves were constructed in 14 newborn lambs under normothermia by directly measuring CBF with a microsphere technique. Intracranial hypertension was induced with an artificial infusion of CSF to identify changes in CPP and calculate the limits of CA. Compared with control subjects, lambs on ECMO had a lower LLA under conditions of normocapnia.⁹⁹

The first 24 hours following ECMO represent a critical period regarding CA in animal models and humans as most ABIs occur during this period,^100,101 the same period in which CA is impaired. Short et al. showed that CA is affected after initiation of ECMO therapy based on an animal model of venovenous ECMO (VV-ECMO). Owing to the experimental conditions CA was only assessed for one hour after ECMO initiation.^97,99 Joram et al. confirmed these findings in children a few decades later. Among 29 children (mean age = 4.8 months) supported by VA- or VV-ECMO, the mean COx and the percentage of time spent with impaired CA were significantly higher among children who sustained ABI (n = 12). The LLA and ULA were 47.2 and 65.6 mm Hg, respectively, and patients with ABI also spent more time outside this range.¹⁰⁰

Studies have shown that abnormalities in CA in children on ECMO have been associated with neuroimaging abnormalities. Tian et al. showed that among 25 children on ECMO, the degree of CA impairment as measured by wavelet methods was significantly associated with worse neuroimaging scores.¹⁰² The authors also suggested that different CA impairment profiles might be associated with specific types of injury, such as ischemic or hemorrhagic.¹⁰³ A study by Sanford et al. on 16 children on ECMO showed that the degree of CA impairment as measured by diffuse correlation spectroscopy was significantly correlated with radiologic injury.¹⁹

The site of the carotid ligation with ECMO may be an important risk factor for developing ischemia, particularly when CA is absent. Although one study found that the cannulation site was more prone to CA disruption at low ECMO flows,¹⁰⁴ the clinical implications of this finding remain uncertain, as conflicting results have been reported regarding the impact of carotid artery ligation on the risk of developing new ipsilateral ischemic or hemorrhagic cerebral lesions.^102,105-107

Clinical implications and future directions

Obtaining age- and sex-specific normative values of the LLA and ULA is a critical step to understanding CA and CBF in critically ill children and neonates.¹⁰⁸ CA-guided therapies may enable us to personalize care according to cerebral hemodynamics rather than relying on general age-based guidelines, allowing for adjustments in neuroprotection parameters following ABI.⁷ For instance, an ICP threshold of 20 mm Hg for greater than or equal to five minutes has been used by most pediatric centers to define intracranial hypertension that warrants treatment. This approach is largely based on adult thresholds and a few pediatric studies showing the association between higher ICP values and poor neurological outcomes and death.^109,110 Based on findings from studies involving healthy children, normal CPP values are closer to the LLA when compared with adults. As a result, these patients might be prone to experiencing episodes of cerebral hypoperfusion when CPP < LLA even during slight increases in ICP (Fig 1). Although an ICP threshold of 20 mm Hg is commonly employed, a lower threshold may be physiologically more appropriate for infants and young children.^109,111 Two European multicenter prospective studies are underway to better understand the optimal thresholds for PRx, CPP, and CPP_OPT, associated with improved outcomes after pediatric TBI.^112,113

Another significant clinical implication of CA-guided therapies pertains to children who have experienced cardiac arrest. The most recent scientific statement from the American Heart Association recommends maintaining a systolic blood pressure ≥5th percentile for age after cardiac arrest.¹¹⁴ Kirschen et al. found that the MAP_OPT was equivalent to the 77^th percentile for age and the difference between the LLA and ULA post–cardiac arrest was 38 mm Hg.⁶¹ These findings raise the question of whether higher MAP goals than recommended in current guidelines may improve neurological outcomes. In adults, targeting higher MAP goals improved indices of brain oxygenation¹¹⁵; however, it did not improve the extent of brain injury; levels of neuron-specific enolase, a biomarker of brain injury; or neurological outcomes.^54,56 It is still unknown if active titration of MAP to achieve an MAP_OPT is a modifiable risk factor that requires early intervention to decrease mortality and morbidity in children or adults.

Over the past decade, significant progress has been made in understanding the utility of actively targeting CA to improve outcomes. A randomized clinical trial (COGiTATE) revealed that CA-based therapies are feasible and safe in adults with TBI.²⁵ Although early results of case series and retrospective studies suggest that this approach is also feasible in children, evidence in pediatrics regarding its benefits and harms is lacking.

Unfortunately, the clinical utility of CA monitoring remains largely unknown, as most studies have focused on identifying cutoff values associated with unfavorable neurological outcomes. Not surprisingly, a panel of 25 international experts who, via a Delphi process, recently reviewed the literature on the role of CA in the management of adults with severe TBI reached no consensus on crucial statements about validity and reproducibility.^9,116 Furthermore, the European and North American guidelines differ in recommendations for the implementation of CA monitoring during the acute phase of TBI. CA monitoring is not recommended routinely for children, given the lack of high-quality evidence.^109,110 On the other hand, the Seattle International Severe Traumatic Brain Injury Consensus Group, an international group of leaders in the TBI field, recommends performing a “MAP challenge” to integrate CA monitoring to guide parameters of neuroprotection.

Although CA monitoring and the CA-guided therapies derived from it hold promise in our field, there are inherent risks associated with their implementation. The balance between improving cerebral perfusion and volume overload should be particularly assessed in patients with cardiac dysfunction.^116,117 Furthermore, MAP augmentation may theoretically lead to further cerebral edema or contusion/hematoma expansion, although this has not yet been proven in clinical studies. Finally, although the placement of invasive devices to monitor CA is considered a quick and safe procedure, the risk of complications, such as displacement, hematoma, or infection, is not negligible.¹¹⁸

This review highlights knowledge gaps on the role of impaired CA in the management of ABI in pediatrics. Observational studies in children and adults have shown that monitoring CA could impact the current practice in pediatric neurocritical care. This technique offers a unique opportunity to accelerate our understanding of the pathophysiology of different types of ABI while paving the way for a new era of precision medicine in critically ill children. Further research should focus on defining specific pediatric thresholds of impaired CA that are associated with outcomes and the applicability, standardization, and reproducibility of this technique in children. We advocate for creating a multicenter registry of cerebral and hemodynamic signals to better understand the role of CA in children with severe ABI.