By the time the surgeon asks whether the brain is slack, every decision that determines the answer has already been made. The pre-dural hour is a sequence of anticipatable stimuli, each with a countermeasure that has to be given before it, not after.
Laryngoscopy, skull pins, scalp incision, skull opening and dural incision are the most stimulating steps of a craniotomy, and every one of them is on the schedule. That is the whole opportunity: each is a pressure event you can see coming, and each has a specific countermeasure that only works if it is already circulating. By the time the dura is opened, nothing you give works fast enough — the brain is either slack or it is not, and it was decided upstream.
There is a moment in every craniotomy when the surgeon looks up and asks whether anything can be done about the brain. It is a fair question and it arrives too late. Almost everything that determines the answer happened in the preceding hour, and most of it happened on a schedule you could have written down in advance.
That is the useful way to hold this day. Laryngoscopy and intubation, skull pinning, scalp incision, skull opening and dural incision are the most stimulating steps of the operation, and every one of them is on the list before the case starts.
Sedation before a craniotomy is a genuine trade. Hypercapnia, hypoxaemia and partial upper airway obstruction are all poorly tolerated when intracranial compliance is low, and all three are ordinary consequences of a sedative given in a corridor. Against that, stress raises cerebral metabolism and flow, and hypertension in a brain with impaired autoregulation can worsen vasogenic oedema.
The resolution is procedural rather than pharmacological. If premedication is indicated in a patient with raised ICP, give it in the operating room, in a fully monitored patient, with respiratory support available: midazolam 0.5 to 2 mg, fentanyl 25 to 100 µg, or sufentanil 5 to 20 µg.
One consequence is worth carrying separately, because it is the kind of thing that turns a routine morning into an emergency. Small doses of benzodiazepines or opioids can unmask or worsen a pre-existing neurologic deficit. It is usually temporary. It is also very difficult to distinguish from a rapid rise in intracranial pressure. The only thing that resolves that ambiguity is a documented baseline examination taken before anything was given.
The goal is stated simply enough: lose consciousness while keeping haemodynamic stability and avoiding a rise in ICP. In practice that is three separate jobs. Achieve enough depth and analgesia to tolerate laryngoscopy. Keep the head positioned so venous outflow is not obstructed. Control ventilation so that neither hypercapnia nor hypoxaemia adds vasodilation to the problem.
Propofol 1 to 3 mg/kg and thiopental 2 to 4 mg/kg are the preferred induction agents for their effects on metabolism, flow, blood volume and CO₂ responsiveness. Both cause hypotension, so the dose is a judgement about the patient rather than a number. In the older or frailer patient etomidate 0.2 to 0.3 mg/kg is an alternative. Ketamine at an induction dose of 1 to 2 mg/kg raises metabolism, flow and pressure and should be used with caution, if at all.
Opioids go in before laryngoscopy, not with it: fentanyl 1 to 3 µg/kg, sufentanil 0.1 to 0.3 µg/kg, alfentanil 5 to 10 µg/kg, or remifentanil 0.5 to 1 µg/kg. Lidocaine 1 to 1.5 mg/kg or esmolol 0.5 to 1 mg/kg may be added to blunt the sympathetic response — and note the units on the esmolol, which is the subject of today’s pitfall.
Non-depolarizing relaxants have no direct effect on metabolism, flow or pressure. Succinylcholine causes fasciculation and a transient rise in ICP, so it is reserved for a difficult airway or an unavoidable rapid sequence; high-dose rocuronium at 1 to 1.5 mg/kg is the usual alternative.
Skull clamp placement is the most under-anticipated stimulus in the case, because it happens after induction, in a quiet room, to a patient who looks settled. It is intense, and it arrives at a moment you choose.
Anticipate it with a bolus — remifentanil 0.5 to 1 µg/kg, propofol 0.5 to 1 mg/kg, or esmolol 0.5 to 1 mg/kg — and infiltrate the pin sites with local anaesthetic, which further limits sympathetic activation. A scalp block does the same job and improves the haemodynamic and stress response to pinning.
Attention to positioning matters here more than in most operations, because the cases are long and the positions are unusual, and because the head is about to be fixed in a frame that removes the option of adjusting it.
The rule to remember is four centimetres, or two finger breadths, between chin and sternum when the neck is flexed. It exists to prevent airway swelling and acute myelopathy, and it happens to be the same clearance that keeps the jugular veins draining. Extreme rotation and flexion of the neck raise brain volume by obstructing venous outflow, which makes head and neck position an intracranial intervention rather than a comfort measure.
While you are there: arms abducted less than 90 degrees, knees mildly flexed, every pressure point padded, eyelids taped shut immediately after induction, tube secured and accessible for the whole case. And confirm tube position and cuff pressure after the patient has been positioned, not before.
There is a long-standing argument about intravenous versus volatile maintenance for intracranial surgery, and the evidence is less dramatic than the argument. Both classes have been used successfully and both appear safe and effective for elective craniotomy. Volatile agents at up to 1 MAC cause cerebral vasodilation with a small rise in flow and pressure, which moderate hyperventilation generally offsets.
The meta-analytic picture deserves reading precisely. Across 14 randomized trials and 1,819 patients, brain relaxation scores after dural opening were similar. ICP was lower with propofol by a weighted mean difference of 5.2 mm Hg, and CPP was higher with propofol by 16.3 mm Hg. Recovery profiles were similar apart from less nausea and vomiting with propofol, and there were inadequate data to compare neurological morbidity or mortality. A second, methodologically weaker meta-analysis reports more brain oedema with sevoflurane than with propofol, so the literature is not perfectly consistent on that point.
What follows is narrower than either camp usually claims. For elective surgery in a compliant brain, the choice is not decisive. For a brain that has no compliance left, a 5 mm Hg difference in ICP and a 16 mm Hg difference in CPP are not academic. Hence the specific recommendations: avoid inhalational maintenance whenever ICP is abnormally elevated, particularly in emergency craniotomy; prefer propofol in severe cerebrovascular disease, where volatiles may cause intracerebral steal; prefer it when motor evoked potentials are being recorded, since even low concentrations of volatile attenuate the signal; and stop the volatile and switch to an intravenous technique any time measures to control brain bulk are not working.
Intracranial compliance is how much the pressure moves for a given change in volume. Everything done to improve it falls into three moves, and it helps to name them that way:
If the brain protrudes, the order is dictated by speed: deepen with an intravenous agent, hyperventilate, drain CSF, elevate the head where the position allows it. Osmotic agents are efficient and slower, which is why they are started early rather than reached for in the moment. In the extreme, decompressive craniectomy expands the available volume.
Retractor injury occurs in an estimated 5 to 10% of craniotomies. That number is the reason a slack brain is a safety outcome and not a courtesy to the surgeon.
No randomized trial addresses this.
Most of this day is a consensus sequence built on established physiology, and it is worth being clear about which parts are which.
Settled. That hypercapnia, hypoxaemia, obstructed venous outflow, light anaesthesia at a noxious stimulus, and extreme neck positions all raise cerebral blood volume is not in dispute. Nor is the Monro–Kellie logic that makes intracranial compliance — how much the pressure moves for a given change in volume — the property you are managing. The interventions follow from it directly: reduce blood volume by maximizing venous return and by transient hyperventilation, subtract cerebrospinal fluid by careful drainage, or shrink the parenchyma with osmotic agents and corticosteroids.
Consensus, not trial. The specific doses, the order of the checklist, the four-centimetre chin-to-sternum rule, the decision to infiltrate the pin sites — these are the accumulated practice of neuroanaesthetic centres. They are sensible and widely agreed, and none of them has been randomized.
Where there is trial evidence, it is largely negative, and it is worth reading carefully. In a meta-analysis of 14 randomized trials and 1,819 patients undergoing elective craniotomy, brain relaxation scores after dural opening were similar between propofol and volatile maintenance — but ICP was lower with propofol by a weighted mean difference of 5.2 mm Hg (95% CI 3.6–6.81) and CPP was higher by 16.3 mm Hg (95% CI 12.2–20.46). Recovery profiles were similar apart from less postoperative nausea and vomiting with propofol, and the authors state plainly that there are inadequate data to compare neurological morbidity or mortality.
So the numbers differ and the surgeon's field does not. For an elective craniotomy in a compliant brain, either technique is defensible. When compliance is already spent, that 5 mm Hg and that 16 mm Hg stop being academic — which is exactly why the standing advice is to abandon the volatile and switch to an intravenous technique any time measures to control brain bulk are failing.
Esmolol for the pressor response is 0.5 to 1 mg/kg. You will see it written as 0.5 to 1 mcg/kg — in course handouts, in review articles, in departmental protocols, sometimes repeatedly in the same document. At microgram doses the drug does nothing at all, which is the quiet part of the problem: nobody notices a beta blocker failing to work, because the laryngoscopy response is attributed to light anaesthesia and covered with more propofol.
That substitution is the error that follows the typo, and it is the one that reaches the patient. More induction agent does blunt the response, and it does so by dropping the perfusion pressure in someone whose autoregulation may already be impaired and whose compliance is the reason you are here. Check the units on any dose you are handed. On this drug the gap between right and wrong is three orders of magnitude.
The wider version of the same failure is treating brain relaxation as something to be produced on request. Osmotic agents are effective and slow. Venous drainage, ventilation, agent choice and positioning all take effect over minutes and are set long before the field is exposed. When the surgeon says the brain is tight, the useful questions are about what has already been done — is the neck free, what is the arterial CO₂, is a volatile still running, has the head been raised, is there CSF to drain — and not about what can be given now.
Full text · Preston goes through LibKey to whatever Preston holds — the PDF where there is one — and works on a phone. The doi link always reaches the publisher's record, whether or not we hold a subscription. If you are reading on a computer, LibKey Nomad adds the same one-click PDF button to PubMed and publisher pages everywhere else; it is a desktop browser extension, so it does nothing on a phone or tablet.
You plan to blunt the sympathetic response to skull pin placement in a 70 kg patient with a large frontal meningioma. Which dose of esmolol is correct?
Forty minutes into an elective craniotomy for a parietal glioma, the dura is opened and the surgeon reports that the brain is tight. Sevoflurane at about 1 MAC is running with a remifentanil infusion. What is the best sequence of action?
A meta-analysis of 14 randomized trials in 1,819 patients undergoing elective craniotomy found brain relaxation scores after dural opening to be similar between propofol and volatile maintenance, while ICP was lower by about 5 mm Hg and CPP higher by about 16 mm Hg with propofol. What is the most defensible conclusion?
Bilateral proximal, arm-worse-than-leg weakness is the classic border-zone presentation, sitting on the seam between the anterior and middle cerebral territories where flow is pressure-passive once vasodilation is exhausted. Reviewing the actual pressure trace is the first useful step. Note what this does not do: it does not exclude embolism, because a bed with no reserve is also a bed that cannot wash an embolus through.
Under normal conditions pial flow does not cross into the adjacent territory because the pressures on either side of the boundary are equal. Change one side — a stenosis, a fall in pressure, an embolization — and the boundary moves. That is why the anatomic location varies considerably between people and can change during the course of treatment.