Propofol‐Related Infusion Syndrome in Paediatrics—Cautious Consideration or Considered Caution? Case Report

Propofol-related infusion syndrome (PRIS) is a rare but potentially fatal clinical entity that was first described in children in 1990 and in adults in 1998 and continues to pose challenges in critically ill patients in intensive care [1, 2]. These challenges are especially important in the paediatric setting, as the utility of propofol as a common and effective drug in paediatric neurocritical care must be balanced with the increased risk of PRIS in this vulnerable patient cohort [2]. We present a case of an adolescent patient who was admitted to intensive care with a traumatic brain injury (TBI) and refractory intracranial hypertension who subsequently developed PRIS, leading to her death. We urge clinicians caring for children with TBI to use propofol with caution. An 11-year-old female patient was admitted to our tertiary children's hospital following a high-speed motor vehicle accident. There was evidence of a closed head injury with a Glasgow Coma Score of 4 at the scene. She was intubated prior to transport and received hyperosmolar therapy (3% sodium chloride) when she developed anisocoria during transfer to the trauma centre. She required a red blood cell transfusion and intravenous tranexamic acid en route to the hospital. The patient was stabilised in the emergency department and taken for an urgent CT scan, which revealed traumatic subarachnoid haemorrhage in the left sylvian fissure with effacement of the supratentorial ventricles and sulcal spaces in keeping with severe diffuse TBI (Figure 1). Other injuries included an undisplaced left mandibular fracture, an open and completely displaced right humeral fracture and cardiac contusion, with no other solid or hollow organ injury identified in the thorax, abdomen or pelvis. The patient went emergently to the operating theatre for insertion of an external ventricular drain and fixation of her humerus fracture, after which she was admitted to the paediatric intensive care unit (PICU). Of note, the intracranial pressure (ICP) was noted to be normal on insertion of the EVD. Neuroprotective measures were instituted according to TBI guidelines [3]. The ICP became elevated (> 20 mmHg) on Day 1 of PICU admission, requiring escalation of medical management, including increased levels of sedation (morphine, midazolam and propofol), targeted temperature management (36.5°C) and hyperosmolar therapy with boluses of hypertonic saline. Of note, propofol had been commenced in the operating theatre and was continued following admission to the PICU at a dose range of 3–5 mg/kg/h, with intermittent boluses used to manage spikes in ICP. The patient also required infusions of catecholamines (noradrenaline and adrenaline), as well as vasopressin, to maintain her cerebral perfusion pressure in the context of her high sedation requirements and high ICPs. ICP and CPP trends, alongside a table of propofol dosing over time, are presented in Figure 2. Due to persistent intracranial hypertension over the first 48 h of admission, a sodium thiopental infusion was commenced, which initially was successful in decreasing her ICPs to < 20 mmHg. Although propofol was initially ceased following commencement of thiopental, it was recommenced in addition to thiopental on Day 2 due to ongoing ICP spikes. Despite some initial stability following commencement of thiopental, there was a subsequent deterioration with worsening intracranial hypertension, and thiopental was unable to be weaned. During this period, the patient underwent two CT scans which showed stable imaging findings. An MRI brain on Day 4 of admission showed findings consistent with Grade 1 diffuse axonal injury, left parieto-temporal contusion with post-traumatic oedema, left putamen traumatic focal infarction, and effaced extra-axial CSF spaces and distended optic nerve sheaths suggestive of raised ICP. An EEG was isoelectric; however, this was in the context of a thiopental coma. Given the refractory intracranial hypertension, the patient's family was counselled about the need for further surgical management with decompressive craniectomy, as well as the increasingly high likelihood of a poor neurological outcome. The patient proceeded to have a bicoronal decompressive craniectomy on Day 6 of admission. Following return to the PICU and over the next 24 h, the patient developed worsening acidosis, new ECG changes showing widening of the QRS complex, prolonged QTc and new left bundle branch block (Figure 3) and severe dyselectrolytaemia with hypokalaemia (2.3 mmol/L), hypophosphataemia (0.7 mmol/L) and hypomagnesaemia (0.6 mmol/L). A transthoracic echocardiogram showed normal cardiac structure and function. Refeeding syndrome was considered as a differential diagnosis given the recent commencement of total parenteral nutrition (TPN); however, raised creatine kinase (peaking at > 10 000 U/L) and triglyceride levels (peaking at 5.3 mmol/L) led us to conclude that PRIS was the most likely diagnosis. Despite cessation of the propofol on Day 7, the patient continued to deteriorate with worsening haemodynamic instability and hyperkalaemia (peaking at 6.6 mmol/L) over the following 24 h. Aggressive medical management was unable to prevent progression to cardiac arrest on Day 8 of admission. ECLS was not felt indicated. Resuscitative efforts were unsuccessful, and the patient passed away in the PICU surrounded by her family. A postmortem examination was limited to a CT scan and external examination. However, the findings suggested that despite a propofol concentration within the typical range for therapeutic use (1.0 mg/L), PRIS likely complicated treatment of the primary cause of death (blunt force trauma to the head). The first case of PRIS was reported in a Danish child in 1990 [2]. Subsequent case series by Parke et al. in 1992 and Bray et al. in 1998 led to recommendations that propofol be avoided for long-term sedation in children [1, 4]. Since that time, cases of PRIS have been increasingly reported in the literature, including in adults [2]. Likely caused by propofol impairing mitochondrial energy production, the syndrome is characterised by severe metabolic acidosis, often accompanied by rhabdomyolysis, renal failure, hepatomegaly, hypertriglyceridaemia, hyperkalaemia, cardiac arrhythmias and significant cardiorespiratory depression [5]. The primary risk factors are younger age (< 18 years), dosing ≥ 4 mg/kg/h and treatment duration > 48 h [5]. While high and extended duration of dosing increases the risk of PRIS, cases have been reported with short-term use at relatively low doses [5, 6]. Further risk factors include low carbohydrate intake, neurological injury and concomitant catecholamine and corticosteroid use [5]. Our patient had several of these risk factors, including young age, high catecholamine requirements, prolonged and high propofol dosing, low carbohydrate intake, neurological injury and corticosteroid use. In addition to the high infusion rate, the patient received boluses in response to intermittent ICP spikes, with the administered propofol dose over the ~7 days of treatment amounting to between 17 and 18 g of propofol (500–514 mg/kg based on a weight of 34 kg). Our patient had refractory intracranial hypertension which was not responsive to escalation of standard multimodal therapy, and intravenous propofol was continued as a component of the treatment strategy. The diagnosis of PRIS was made based on the presence of a metabolic acidosis with ECG changes, evidence of rhabdomyolysis and hypertriglyceridaemia, with the exclusion of other potential and likely causes. A range of conduction defects and arrhythmias have been described in PRIS, including bundle branch block, AV block, bradycardia, asystole, Brugada-like changes on ECG (right bundle branch block with convex-curved ST elevation in anterior chest leads), atrial fibrillation, supraventricular tachycardia, ventricular tachycardia and ventricular fibrillation [1, 6]. Our patient's ECGs were reviewed by our cardiology service and felt to be consistent with PRIS. The patient did not experience the idiosyncratic ‘green urine’ sometimes erroneously associated with propofol toxicity, and clinicians should be cautioned about relying on the absence of green urine to discount PRIS as a likely diagnosis [7]. PRIS is often fatal, with a mortality rate up to 52%, and treatment relies on supportive management and discontinuation of propofol given the absence of a pharmacological antidote [2]. Due to the bleeding risks with anticoagulation and the very poor neurological prognosis of our patient related to her TBI, extracorporeal support, such as ECMO or CRRT, was not initiated. When performed, postmortem findings have included fatty infiltration of the liver and myocytolysis without evidence of myocarditis [4]. Our patient had a limited postmortem examination without histological specimens. Differential diagnosis of PRIS is often difficult given its non-specific presentation in the context of critically ill patients. The hypokalaemia and high urine output observed in our patient on Days 6 and 7 of admission were not consistent with the acute kidney injury and hyperkalaemia which are often associated with PRIS [2]. This prompted some consideration of thiopental toxicity as a potential diagnosis, as barbiturates are more likely to cause significant hypokalaemia and then rebound hyperkalaemia than propofol [8]. The pronounced hyperkalaemia—in the absence of renal injury—and increasing inotropic support on Day 7 of admission may suggest that simultaneous drug toxicities occurred. Although there is limited evidence evaluating the simultaneous use of propofol and thiopental, it is possible that the extended duration of both agents caused an additive or synergistic pharmacodynamic response, which amplified adverse effects (specifically, haemodynamic instability and dyselectrolytaemia) [9]. Similarly, propofol and thiopental have comparable pharmacokinetics: both are highly lipophilic and protein-bound compounds and may theoretically compete for binding sites when co-administered over an extended period [9]. This could have influenced their respective volumes of distribution, plasma concentration and therefore associated toxicity. This hypothesis notwithstanding, it is not standard practice for these agents to be used concomitantly, and this should be avoided in future patients. Finally, targeted temperature management in the context of neuroprotection (targeting 36.5 degrees Celsius in this case) or hypothermia is known to reduce the metabolism of propofol due to slowing hepatic blood flow and cytochrome P450 activity [10]. Extra caution should be applied to propofol dosing in cases where therapeutic cooling is required. Although the dangers of PRIS are highlighted by our patient's case, the utility of propofol as a sedative agent in the intensive care setting must also be acknowledged. Commonly used for short-term and procedural sedation or anaesthesia, propofol has a rapid onset and offset and a relatively low risk of delirium and facilitates control of ICP. This case reinforces the importance of balancing these benefits against the potential risks of propofol, consistent with existing literature recommending cautious use in children. Clinicians should remain aware of the risk of PRIS and relevant patient-specific risk factors and use propofol judiciously, ideally limiting infusion rates to < 4 mg/kg/h and duration to less than 24–48 h. This case report is limited by its retrospective nature and that the postmortem examination did not include histological examination of the liver. Additionally, propofol concentrations during the period of propofol therapy were not available to the treating clinicians and may have informed the diagnosis. Without specific biomarkers, PRIS remains a clinical diagnosis. Open access publishing facilitated by The University of Sydney, as part of the Wiley - The University of Sydney agreement via the Council of Australasian University Librarians. The authors have nothing to report. The authors declare that the research presented in this manuscript adheres to the ethical principles outlined by Sydney Children's Hospitals Network Human Research Ethics Committee (Project Number: CCR2025/48). All procedures involving human participants were conducted in accordance with the ethical standards of the Sydney Children's Hospitals Network and the Declaration of Helsinki (1964), as revised in 2013. Written consent for publication of anonymised data and any potentially identifiable information was obtained from the relevant legal guardian. The authors declare no conflicts of interest. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.