30 

What Is the Best Algorithm for Treatin Intracranial Hypertension?

Jürgen Meixensberger and Matthias Jaeger

BRIEF ANSWER

Background

Although various options are available for the treatment of elevated intracranial pressure (ICP), no consensus exists regarding the optimal therapeutic regime for patients with intracranial hypertension. The available options are usually applied as required in an escalating step-by-step fashion. The sequence of their application, however, may vary considerably at different neurotrauma centers.

The ” optimal“ treatment algorithm (according to the principles of evidence-based medicine) still needs to be defined. Ideally, prospective randomized trials generating class I evidence and level I recommendations should validate and compare the available treatment options in terms of their effectiveness for controlling ICP and improving patient outcome. A large amount of existing data does provide some guidance in choosing among different therapies. Unfortunately, little solid evidence is available to guide neurotraumatologists in the many ICP treatment decisions that they must make on a daily basis.

From a Medline search using ” intracranial pressure AND treatment,“ we extracted those results that provided the best data for developing an evidence-based algorithm for the treatment of elevated ICP.

Literature Review

Basic Measures

SEDATION AND ARTIFICIAL VENTILATION

No studies have evaluated the effects on outcome of sedation and artificial ventilation by randomizing some patients with severe traumatic brain injury (TBI) and intracranial hypertension to control groups in which these interventions were not used. Nevertheless, expert opinion and widespread recognition of the ICP-lowering effects of these measures prompted the European Brain Injury Consortium to advocate their use as a level III recommendation.¹ Regarding the choice of sedative drugs, propofol and midazolam were both declared to be safe and effective for use in severely injured patients (some of whom had TBI) in class III studies,^2,3 but direct effects on ICP were not seen, possibly because all patients in these studies received some form of sedation. For analgesia, the opioids alfentanil, fentanyl, and sufentanil were all found to produce transient increases in ICP (8 mmHg for ~15 minutes) when administered by bolus injection, presumably reflecting an autoregulatory response to an associated decrease in mean arterial pressure; however, no cerebral ischemia was detected in measurements of jugular venous lactate-oxygen index (class II data).⁴ The ultrashort-acting opioid remifentanil was determined to be safe for analgesia in the neurosurgical intensive care unit (ICU), partly because it did not produce alterations in ICP or in blood pressure (class III data).5 Despite the fact that ketamine has been largely abandoned for a long time in terms of analgesia in TBI patients, Kolenda et al⁶ conducted a class II study in which the combination of midazolam plus ketamine had more beneficial effects than the combination of midazolam plus fentanyl in terms of vasopressor requirements and enteral food intake, but no major differences were seen in ICP or outcome.

In patients requiring artificial ventilation, use of positive end-expiratory pressure (PEEP) up to 10 cm H₂O was considered safe by Cooper et al⁷ based on their study of 33 TBI patients (class II data). Schneider et al⁸ observed that hyperventilation (PaCO₂<30 mmHg) produced decreases of brain tissue oxygen tension; in one of their 15 severe TBI patients, hyperventilation had to be stopped because the decrease in brain tissue oxygen tension was potentially of sufficient magnitude to cause cerebral tissue hypoxia (class II data). In a randomized trial reported by Muizelaar et al⁹ (class II data), prolonged hyperventilation was associated with a trend toward worse outcomes in severe TBI patients with an initial Glasgow Coma Scale motor score of 4 to 5.

PATIENT POSITION

Correct patient positioning was the focus of a class II study of 25 patients by Schneider et al.¹⁰ They investigated ICP and SjvO₂ at different head elevations ranging from 0 to 45 degrees. With 45 degrees of head elevation, an ICP reduction of almost 50% could be obtained without significant compromise of cerebral oxygenation or cerebral perfusion pressure (CPP), despite the potential for reduction in CPP upon head elevation. Although Schneider et al found that average CPP and SjvO₂ did not change significantly as head position varied, CPP responses of individual patients were unpredictable. Other investigators have reported that flexing the neck or turning the head to either side produces modest but statistically significant increases in ICP (class II data).¹¹

Extended Measures

HYPEROSMOTIC SOLUTIONS

Hyperosmotic solutions, such as mannitol, are still the most common ” extended“ treatment for posttraumatic ICP control. A randomized study reported by Schwartz et al¹² provided class I evidence supporting the use of mannitol prior to the use of barbiturates for the treatment of elevated ICP in head-injured patients. Additional class II studies have demonstrated that bolus doses of mannitol are effective for ICP treatment.^13,14 Infusable agents other than mannitol have also been found to reduce ICP effectively. Bolus infusions of hypertonic saline in concentrations ranging from 7 to 23% decreased mean ICP by at least 40% in several studies, without significant side effects (class II and class III data).^15–17 Furthermore, tromethamine (THAM) administered as a continuous infusion for 5 days after TBI was found to successfully reduce ICP in a multicenter class I trial reported by Wolf et al.¹⁸ Beneficial effects on outcome, however, were not observed. THAM is also effective in lowering ICP when given as bolus infusion, with a reported reduction in ICP of 33% (class II data).¹⁹

BARBITURATES

The use of barbiturates for ICP control has been studied extensively. Ward et al²⁰ investigated prophylactic barbiturate coma in 53 severely head-injured patients (class I data). They found no difference between groups in outcome or in ICP control, but the barbiturate-treated group was much more likely to experience arterial hypotension (54% versus 7%). Thus, the authors did not recommend prophylactic barbiturate coma for patients with severe TBI. On the other hand, Cormio et al²¹ described a class II study that identified subgroups of patients who experienced considerable benefit from barbiturates. In particular, those patients in whom ICP decreased after the initial loading dose of barbiturates might benefit from the barbiturate-induced reduction of cerebral metabolism. As discussed above, however, Schwartz et al’s¹² class I study comparing barbiturates and mannitol for ICP control found barbiturates to be less successful than mannitol for ICP reduction, at least at the earlier stage of intervention at which barbiturates were used in the Schwartz study.