Perioperative Hypotension in Infants: Insights From the GAS Study : Anesthesia & Analgesia

Secondary Logo

Journal Logo

Editorials: Editorial

Perioperative Hypotension in Infants: Insights From the GAS Study

Vutskits, Laszlo MD, PhD*; Skowno, Justin MBChB, DA, FCA, FANZCA

Author Information
doi: 10.1213/ANE.0000000000001959
  • Free

Although anesthesiologists routinely measure blood pressures in infants, and, in principle, agree that hypotension should be avoided, we do not yet have good evidence to help with the definition of what constitutes hypotension. This unanswered question is of utmost relevance where the possibility of a causal link between early life exposure to anesthesia and subsequently altered neurodevelopmental outcome has been suggested.

In this issue of the Journal, investigators of the General Anesthesia compared to Spinal anesthesia (GAS) Consortium compare blood pressure profiles and hypotension rates between infants of <60 weeks postmenstrual age undergoing inguinal herniorrhaphy under either general anesthesia (GA) or regional anesthesia (RA).1 Using threshold values of mean arterial blood pressure (MAP) < 45 mm Hg to define hypotension, they demonstrate that hypotension occurs in 87% of infants undergoing GA while the prevalence is 41% in the RA group. Importantly, if this threshold is lowered to MAP < 35 mm Hg (defined as moderate hypotension), there are still 49% and 16% of subjects, respectively, in the GA and RA groups whose blood pressures, at certain stages of the intraoperative period, are under these values. Multivariable modeling reveals a nearly 3-fold increase both in the relative risk of intraoperative hypotension and in the number of interventions to treat this hypotension in the GA compared with the RA group.

The GAS study aims to determine whether 5-year neurobehavioral outcomes are different between general and regional anesthesia and has recently published 2-year interim results.2 Although laboratory investigations strongly suggest neurotoxic potential of most currently used general anesthetics, it is also widely assumed that suboptimal anesthesia management in young children, including intraoperative hypotension, may also be a cause for perioperative developmental neuromorbidity.3 Given the limited availability of detailed anesthesia records, large-scale retrospective epidemiological studies provide us little information on how changes in any specific physiological variable during the perioperative course relate to long-term neurocognitive outcome. Prospective clinical trials with detailed intraoperative records are in contrast well suited to study such correlations. The GAS study is particularly relevant in this context since this prospective, randomized, controlled, multisite trial, aimed to assess the impact of GA in infants on subsequent neurodevelopment, provides us with an important collection of physiological data sets related to the perioperative period in this patient population.

These observations are important for several reasons. First, they provide a large-scale and so far undescribed comparison of hemodynamic profiles between infants undergoing surgery using 2 distinct anesthesia techniques. Second, the authors also identify 2 other independent risk factors, body weight at the time of surgery and minimal intraoperative temperature to predict intraoperative hypotension. These findings are of great clinical value since they allow us to identify patient populations at particular risk for hypotension and also point to the importance of actively maintaining normothermia in this fragile population. Third, they draw our attention to the large interindividual variability in terms of blood pressure values in this relatively homogenous cohort. The wide range distribution of blood pressure values in this trial is in line with recent observational studies on perioperative blood pressure profiles in children, and further questions our current non–evidence-based textbook definitions of threshold values for systemic hypotension. Fourth, as the authors themselves report, blood pressure values were missing in more than 30% of children at anesthesia induction in both arms. Although this may sound somewhat embarrassing, such observations have also been recently reported by other investigators and reflect the current status of blood pressure monitoring in pediatric populations. For example, in a retrospective analysis of anesthesia records of over a thousand neonates and infants, Weber et al4 report that blood pressure is measured in <25% of cases at anesthesia induction and up to 10% of patients may not have a blood pressure value available up to 20 minutes after anesthesia induction. Similarly, a review of more than 50,000 electronic anesthesia records found that <2% of children had blood pressure recordings available immediately before anesthesia induction.5

What defines hypotension in infants under anesthesia? Here the authors arbitrarily set specific cutoff values to gain insights into the incidence of hypotension. This decision was based on recently published data arguing that the lower limit of cerebral autoregulation in infants is around 45 mm Hg and that cerebral oxygenation does not decrease before MAP drops to 35 mm Hg.6,7 However, a closer look at those studies clearly advocates for a more nuanced interpretation of blood pressure limits, and recognition of the fact that cerebral oximetry is itself a surrogate of cerebral blood flow and oxygen delivery. Indeed, the lower limit of cerebral autoregulation per se has not been defined by Rhondali et al,6,7 and recent series of investigations seriously questions the validity of our current view on cerebral autoregulation.8 In addition, data sets from Rhondali et al6,7 suggest that cerebral oxygenation under GA can be equal or superior to awake values in MAP ranges as low as 30 mm Hg questioning thereby the physiological relevance of defining 35 mm Hg as a threshold. Most importantly, the same study demonstrates significant interindividual variability in cerebral oximetry values, questioning strongly a “one size fits all” approach. An alternative, more personalized definition of hypotension could be to define an acceptable blood pressure drop from baseline values. In line with this approach, by studying cerebral oxygenation as a function of intraoperative changes in blood pressure, Michelet et al9 found that in infants <3 months of age a decrease of <20% in systolic blood pressure is associated with <10% probability of cerebral desaturation (defined as >20% decrease from initial values). In contrast, a decrease of 45% in systolic blood pressure compared with baseline was associated with a more than 90% chance of cerebral desaturation.9 Although these observations are useful, one could also argue that, in 10% of infants as small as a 20% decrease in blood pressure could lead to impairment of cerebral oxygen delivery while as much as 45% of blood pressure drop will not affect this same parameter in another 10% of patients. Again, this uncertainty clearly underlines the importance of the need for individualized care if one wants to assure adequate oxygen supply to the brain during the perioperative period. This need for individualization of care is supported by current work in cerebral autoregulation that emphasizes the critical and complex interdependence of cerebral blood flow, blood pressure, CO2, temperature, and other variables.10 Individualized care is however difficult or impossible in the absence of clinically robust bedside measurements of cerebral blood flow or perfusion adequacy. With or without such additional tools, clinicians should perhaps focus their efforts on avoiding extremes of physiology, particularly with respect to CO2 and blood pressure, as noted elsewhere.3

If we want to understand the importance of blood pressure management in providing adequate perioperative care, one of the most important questions to determine is how perioperative blood pressure values correlate with immediate or longer-term neurocognitive outcome. Assuming the presence of a detectible neurobehavioral or neurocognitive phenotype following anesthesia exposure, the design of the GAS study is well suited to evaluate this possibility. However, despite the significantly higher incidence of so-called moderate intraoperative hypotension under GA, interim analysis of neurocognitive function 2 years after anesthesia exposure did not reveal any difference between the GA and RA groups. It is, nevertheless, important to note that case reports indirectly correlating devastating neurological outcome with intraoperative hypotension are available.11 Based on these observations, where many of the intraoperative blood pressure values were comparable with those observed in the GAS Study, it remains very difficult to determine why some children develop overt neurological deficit following anesthesia, while others remain seemingly unaffected. The answer is most probably highly complex and may also depend, among others, on individual microvascular anatomy, brain metabolic status at the moment of inadequate cerebral blood supply, as well as on the duration and numbers of hypotensive episodes. Figuring out how these factors influence each other in terms of neuromorbidity should therefore be a priority of the perioperative research agenda. In the meantime, the excellent current safety record of modern anesthesia should not obscure the potential for harm from hypotension in our youngest patients, in whom the risks are demonstrably higher. Indeed, no anesthetist would dispute that neurological damage from severe hypotension is possible; they merely dispute where the threshold is. While that threshold remains obscure, it is prudent to avoid excursions into those uncharted waters, until the GAS consortium and other researchers prepare a map.


Name: Laszlo Vutskits, MD, PhD.

Contribution: This author helped write the manuscript.

Name: Justin Skowno, MBChB, DA, FCA, FANZCA.

Contribution: This author helped write the manuscript.

This manuscript was handled by: Richard C. Prielipp, MD.


1. McCann ME, Withington DE, Arnup SJ, et alDifferences in blood pressure in infants after general anesthesia compared to awake regional anesthesia. Anesth Analg. 2017;125:837–845.
2. Davidson AJ, Disma N, de Graaff JC, et al.; GAS consortium. Neurodevelopmental outcome at 2 years of age after general anaesthesia and awake-regional anaesthesia in infancy (GAS): an international multicentre, randomised controlled trial. Lancet. 2016;387:239–250.
3. Weiss M, Vutskits L, Hansen TG, Engelhardt T. Safe anesthesia for every TOT—the SAFETOTS initiative. Curr Opin Anaesthesiol. 2015;28:302–307.
4. Weber F, Honing GH, Scoones GP. Arterial blood pressure in anesthetized neonates and infants: a retrospective analysis of 1091 cases. Paediatr Anaesth. 2016;26:815–822.
5. Sottas CE, Cumin D, Anderson BJ. Blood pressure and heart rates in neonates and preschool children: an analysis from 10 years of electronic recording. Paediatr Anaesth. 2016;26:1064–1070.
6. Rhondali O, Mahr A, Simonin-Lansiaux S, et al. Impact of sevoflurane anesthesia on cerebral blood flow in children younger than 2 years. Paediatr Anaesth. 2013;23:946–951.
7. Rhondali O, André C, Pouyau A, et al. Sevoflurane anesthesia and brain perfusion. Paediatr Anaesth. 2015;25:180–185.
8. Tzeng YC, Ainslie PN. Blood pressure regulation IX: cerebral autoregulation under blood pressure challenges. Eur J Appl Physiol. 2014;114:545–559.
9. Michelet D, Arslan O, Hilly J, et al. Intraoperative changes in blood pressure associated with cerebral desaturation in infants. Paediatr Anaesth. 2015;25:681–688.
10. Willie CK, Tzeng YC, Fisher JA, Ainslie PN. Integrative regulation of human brain blood flow. J Physiol. 2014;592:841–859.
11. McCann ME, Schouten AN, Dobija N, et al. Infantile postoperative encephalopathy: perioperative factors as a cause for concern. Pediatrics. 2014;133:e751–e757.
Copyright © 2017 International Anesthesia Research Society