The initial literature search was carried out using Google Scholar to collect available data on cerebral oedema with a focus on DKA. The lack of available literature prompted a wider exploration to all types of cerebral oedema management and basic pathophysiology. We also searched PubMed, MEDLINE, The Cochrane Library, and Scopus databases for studies published up to June 2024. Search terms included "diabetic ketoacidosis," "cerebral oedema," "children," and "pediatric, raised intracranial pressure." medical subject headings (MeSH) terms were applied.

A simple research question opted which was specific, concise and consisting to the object that ensure validity and appropriate inform of evidence-based practice. We included in this review studies that contained 3 of the 4 proper elements of population (P), intervention (I), comparative intervention (C), and outcome (O). The P here was a pediatric group with DKA who developed cerebral oedema during or prior to treatment. The I was the treatment option used in the study and was analysed relative to a C, like oxygen or neck position, to contrast the O, such as reduction of cerebral oedema or mortality [6]. Our aim was to review the management options of cerebral oedema in pediatric patients with DKA.

Inclusion criteria for the review were studies involving children (<18 years) with DKA and cerebral oedema; original clinical data; and presentation of risk factors, pathophysiology, clinical features, or treatment outcomes. Studies were excluded if they were reviews, editorials, animal studies, commentaries, or did not contain relevant data.

Titles and abstracts were screened independently by 2 reviewers. Full-text articles were assessed for eligibility. Discrepancies were resolved through discussion.

Data on study design, population characteristics, risk factors, pathophysiology, clinical presentation, and treatment outcomes were extracted. Quality assessment was performed using the Newcastle-Ottawa Scale for observational studies.

Risk of bias was evaluated using the Cochrane Risk of Bias Tool for randomized studies and ROBINS-I (Risk Of Bias In Non-randomised Studies - of Interventions) for nonrandomized studies.

A total of 941 articles was identified using the MeSH keywords. After the appropriate exclusions, we chose 29 full-text articles that focused on management strategies; the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) flow diagram is Fig. 1.

The studies included reviews, expert opinions, clinical trials and guidelines, and were conducted across nations. Most were reviews, few are expert opinions and randomized controlled trials.

The pathophysiology of cerebral oedema and related brain injury is not fully understood. Cerebral oedema is defined as excess accumulation of water within the intra-and/or extracellular spaces of the brain [7]. The process starts with release and accumulation of glutamate and free fatty acids, followed by activation of sodium and calcium channels to expel calcium intake sodium. This results in sodium accumulation within the cell, resulting in an osmotic gradient and water influx. Intracellular hypoxia and/or hypoglycaemia of any cause may lead to depleted ATP production and reduced calcium exchange [8].

In early 1980s, some retrospective studies recognised the increasing frequency of cerebral injury in patients who received rapid and higher rates of intravenous fluids [9]. A rapid decrease in intravascular osmolality results in movement of water into brain cells, resulting in increased intracranial pressure (ICP), cerebral oedema, and cerebral injury [10]. It was proposed that rapid and large fluid shifts may lead to cerebral oedema due to increased intracellular accumulation during DKA. Hence, DKA management guidelines were adjusted to recommend conservative rehydration and slower rates of intravenous isotonic fluids [11].

However, multiple findings have challenged this theory of cerebral injury caused by osmotic changes and fluid shifts, such as those of the PECARN fluid trial (2018) [12]. That study randomized intravenous fluid treatment and compared it with rapid versus slow rehydration and use of 0.45% versus 0.9% sodium chloride. The results showed no significant difference in mental status (Glasgow Coma Scale, GCS) or in rates of clinically apparent cerebral oedema/injury.

Moreover, few other studies adjusting the severity of DKA reported correlation between the rate of fluid infused and cerebral oedema [3]. Similarly, high initial glucose, rapid decline in glucose, and high osmolality have not been associated with cerebral injury [4]. However, one retrospective study showed that a decrease in serum sodium during DKA treatment was associated with risk of cerebral oedema; this was not observed in a subsequent prospective study [13]. In addition, there were no differences in children who experienced decrease in sodium level during treatment and those who did not. These data suggest that associations between decreased serum sodium and cerebral injury in previous retrospective studies may have reflected a delayed diagnosis of cerebral injury occurring before treatment. This suggests that intravenous fluid therapy cannot be solely responsible for cerebral injury and supports intrinsic DKA factors (like high blood urea nitrogen or severe acidosis), rather than treatment-related factors, as responsible for related cerebral injury.

Fluctuation of cerebral blood flow (CBF) and neuroinflammation may cause DKA-related cerebral oedema and injury. Glaser et al. [14] identified CBF to be markedly affected by dehydration and hypocapnia and to be associated with life threatening cerebral injury.

Some studies have assessed CBF during DKA treatment and shown increased CBF to be associated with increased extracellular water content, hyperaemia, cytotoxic oedema, and related vasogenic cerebral injury/oedema. Such increased CBF may persist for hours after DKA resolution [15]. However, the precise roles of cytotoxicity and neuroregulation in altered CBF and perfusion-related injury are under investigation.

Recently, high-concentration oxygen therapy has become increasingly popular, as it is very effective and can alleviate ischemic hypoxic conditions and reduce cerebral oedema. Hence, DKA patients need to be supported with oxygen therapy, preferably via face mask to avoid nasal stimulation [18].

High-concentration (100%) oxygen via face mask will increase the partial pressure of oxygen and was identified by Zhou et al. [19] to have no difference in efficacy between hyperbaric or normal application during the early phase of oedema. Hence, it was recommended for use among patients with cerebral oedema. However, the detrimental effects of high-concentration oxygen must be considered, like the free radical formation that may cause more tissue damage and ischaemia. To avoid these, a short course of 41%–90% oxygen is recommended [20,21].

DKA is a hyperosmolar state in which water accumulates within the brain cells to decrease the extracellular osmolar state, resulting in swelling [10]. Slow fluid replacement and addition of electrolytes over 48 hours are recommended rather than bolus administration. However, this has not been supported by subsequent studies [22]. Patients receiving replacement over 12 hours instead of 48 hours demonstrated no higher risk of oedema [23].

However, if a patient develops cerebral oedema, restricted fluid infusion is of utmost importance to minimize the DKA-induced effect. On the contrary, additional fluid restriction may lead to hyperviscosity, hypoperfusion, and hypotension, which may worsen the neurologic outcome [24]. Controversy persists, but most guidelines recommend restricted fluid infusion if there is cerebral oedema or risk of it.

Loop diuretics (furosemide) most commonly have been used to manage cerebral oedema, though the results are inconsistent, and are associated with acute hypovolemia, reduction of mean arterial pressure, worsening cerebral perfusion, and poor clinical outcomes [25]. Data suggest that low-dose furosemide (0. 7 mg/kg) may prolong the reversal of the osmotic gradient of the blood brain barrier established with hypertonic saline (HTS) and mannitol by primarily excreting free water [26].

There are different options for neuroprotective methods such as head elevation, where the recommended tilt is 15° to 30° [27] to enhance the CBF and impede venous drainage. Another option is midline positioning of the patient head to avoid obstruction of venous return. as well as increased ICP through processes like sneezing, coughing, mechanical suction of the nostril, and deep buccal (vagal) stimulation with suction. The most commonly recommended process is to remove excessive secretion with superficial and low-pressure suction and avoid deep suction.

Brief episodes of hyperventilation may be used for patients with acute increase in ICP (strong recommendation, very low-quality evidence) as suggested by Cook et al. [28]. This may reduce the partial pressure of CO₂, leading to intracranial vasoconstriction to temporarily reduce the ICP. Such treatment should be used limitedly because prolonged hyperventilation may reduce CBF and can lead to ischaemia. The targeted PaCO₂ should be 30–35 mmHg.

Seizure management with prophylactic antiseizure medication may be started in cases of known seizure disorder or with high suspicion of seizure development. The occurrence of barbiturate coma, as in traumatic brain injury-related oedema, has not been studied in DKA. Prophylactic anticonvulsants such as phenytoin or levetiracetam may be used in patients at risk of seizures.

The most effective and rapid mode of decreasing the tissue water content is osmotherapy, which aims to draw water from the brain based on an osmotic gradient to decrease blood viscosity [29]. An osmotic gradient is created due to a difference in diffusion perfusion from the parenchyma to the vascular space, reducing the intracranial free water volume and decreasing ICP [30].

When the glucose level is below 250–300 mg/dL (14–17 mmo1/L), 5% glucose solutions should be administered as a source of free water and as a prophylactic measure to prevent late cerebral oedema syndrome [31].

Mannitol was first introduced in 1960s, and remains part of routine medical management of cerebral oedema as a 20% solution. The mechanism of this treatment remains unclear but is thought to involve increased vasoconstriction, reduced blood and cerebral spinal fluid volumes, and enhanced cerebral perfusion to produce a protective effect against biochemical injury [8].

A dose of 0.5–1.0 g/kg IV in 10–15 minutes is sufficient as an intermittent dose to maintain serum osmolality around 300–310 mOsm/L and may be repeated after 30 minutes, action is apparent in 15 minutes. Continuous infusion is not recommended. Prolonged and repeated doses should be avoided to prevent electrolyte imbalance [27,32].

Since 1990, HTS has been a mainstay of clinical practice for the management of cerebral oedema. Such saline fluids are available in different concentrations from 2% to 23.4% based on institutional practice [33]. Injection of HTS produces an osmotic gradient in cerebral tissues in the presence of an intact blood brain barrier, drawing intracellular fluid into the intravascular space and reducing the cerebral volume. With a higher reflection coefficient of sodium chloride (1.0), HTS is more effective than mannitol, has a lower rate of rebound cerebral oedema, and does not produce osmotic diuresis or hypovolemia [34].

For patients with cerebral oedema or GCS≤8, HTS (3%) is suggested as a first-line agent, especially in children. the reason behind may be minimal osmotic diuretic effect with HTS and may reduce elevated ICP to a greater degree and may be use if initial therapy of mannitol shows minimal results, combination of both may be beneficial in severe cases [11]. A dose of 3 to 5 mL/kg over 10 to 15 minutes may be repeated if required. Continuous infusion of HTS may be used to maintain osmotic gradients and to achieve a targeted sodium level of 140–160 mmol/L [35].

Complications related to the use of HTS include metabolic derangements, volume overload, phlebitis, hypotension, and coagulopathy [33].

Anaesthetic agents have no direct impact on cerebral oedema but may induce a sedative effect and decrease intracerebral pressure through reduction of cerebral metabolic demand and intracerebral arterial volume [36]. Though several agents have been discussed in the literature, none has a clear direct pharmacological effect on cerebral oedema. In addition, there is no clear evidence to support the use of barbiturate-induced coma in the management of cerebral oedema [37].

Propofol and midazolam are also used as sedative agents in cerebral oedema but are not mainstays in its medical management [34].

Intubation in cerebral oedema should be undertaken only with extreme caution as it may cause vagal stimulation and increase ICP. The procedure should be carried out in select cases like patients with apnoea or abnormal respirations and must consider airway protection. Lidocaine may have a preventive effect on cerebral oedema but will not reduce existing oedema [38]. Aggressive hyperventilation to lower pCO₂ should be avoided as it is associated with poor outcomes [10]. The pCO₂ should not be reduced to less than 25 mmHg, as there will be a vasoconstrictive effect of hypocarbia that can cause hypoxia and ischaemic cell inflammation and oedema [38].

Though studies on trauma patients recommend neuroprotective hypothermia, hyperthermia must be avoided, as it can lead to increased risk of infections, coagulopathy, and electrolyte disturbances.

Hypothermia is a protective mechanism to reduce CBF by reducing the metabolic demand through the release of transmitters to decrease vascular permeability and oedema. Body temperature >37.5°C and blood glucose >150-mg/dL exacerbate cerebral oedema [39].

Induced hypothermia of 32°C–35°C is used for management of cerebral oedema, and its mechanism is multifactorial. However, recent clinical trials of induced hypothermia do not suggest improved clinical outcomes in use with available management [40].

Monitoring clinical status along with vital and laboratory parameters is important in management of cerebral oedema. The classical triad of bradycardia, hypertension, and irregular respiration must be kept in mind in all cases of cerebral oedema as this may develop very rapidly and may be missed. Behavioural change may be the initial clinical feature of such changes, like abnormal sleep, difficult arousal, irritability, or headache.

Mental status should be documented hourly, along with the GCS score, as an essential component of monitoring. Laboratory values like hourly glucose level, venous blood gas, and electrolyte levels also are important to monitor. Central lines are not indicated, as they may increase the risk of deep venous thrombosis [1].

Physicians managing patients with cerebral oedema related to DKA must be vigilant, monitoring clinical and behavioural effect to detect early clinical findings and to initiate management.