In a brain with no compensatory reserve left, a fall in perfusion pressure raises intracranial pressure — and the loop runs both ways. The other half of the problem arrives twenty minutes later, when the clot comes out and the pressure the patient was generating for themselves stops.
A decrease in cerebral perfusion pressure causes cerebral vasodilation, which raises cerebral blood volume, which raises intracranial pressure, which lowers perfusion pressure again. In a brain whose compliance is spent that loop closes, and a modest fall in blood pressure at induction becomes a herniation. The loop also runs in reverse, which is the treatment. And when the clot is finally evacuated the Cushing response is no longer needed and stops — so the same case can kill the patient from too little pressure, then too much, then too little again.
Neuro 1 day 04 was the elective craniotomy — a brain with reserve, a schedule you could write down, and countermeasures you could give in advance. This is the same operation without any of that. The mass arrived quickly, the compensatory mechanisms are spent, and the patient is holding their own perfusion pressure up by a reflex you are about to remove.
The effect of an intracranial mass on neurological status and pressure depends on how fast it arrived. A chronic subdural haematoma produces subtle signs and small pressure rises, and can be anaesthetised much like a tumour. An acute epidural or intracerebral haematoma arises far more rapidly, with massive neurological impairment and acutely life-threatening pressure.
The reason is compliance. For rapidly expanding masses the pressure–volume curve shifts markedly to the left. Compensation comes first from extracranial shift of intracranial blood, then from the larger-capacity displacement of CSF — which is ineffective if CSF flow is obstructed. Once those are exhausted, pressure rises steeply, the cerebral circulation is impaired, and herniation follows as the end stage of compensation.
Here is the mechanism that makes a routine-looking hypotension dangerous.
A decrease in cerebral perfusion pressure produces cerebral vasodilation, which increases cerebral blood volume, which increases intracranial pressure — which decreases cerebral perfusion pressure. In a brain with reserve the extra volume is absorbed. In this brain it is not, and the loop closes.
The same relationship run backwards is the treatment: an increase in perfusion pressure produces vasoconstriction, decreased blood volume and decreased pressure. That is why blood pressure here is not merely a number to keep above a line — it is an intervention on intracranial volume.
The Cushing triad — systemic hypertension, bradycardia, and an irregular breathing pattern — is described as a weak but significant indicator of life-threatening brain injury requiring immediate action, because at this point even small increases in intracranial volume produce disproportionate pressure rises and swelling.
Weak but significant is the right phrase to carry. Its absence does not reassure you. Its presence should move you.
Oxygenation is paramount. Secure the airway, then mild hyperventilation with 100% oxygen. In trauma, exclude cervical spine fracture where possible. Laryngoscopy and mask ventilation should be done by experienced people, and the patient should be in a deep plane of anaesthesia at laryngoscopy to avoid coughing and sympathetic stimulation — conditions that are hard to achieve in a patient with multiple injuries and shock, which is precisely when they are most needed.
Agent choice follows the haemodynamics:
Rapid sequence is usually required for aspiration risk, and the best way to do it in a neurosurgical emergency remains controversial. A modified rapid sequence, bag-mask ventilating during the time succinylcholine takes to reach peak effect, helps keep CO₂ in an acceptable range. High-dose rocuronium permits intubation within a minute and is appropriate only if sugammadex reversal is immediately available.
Once intubated and ventilated, the brain relaxation measures continue.
Mannitol 0.5 to 1 g/kg — 150 to 400 mL of 20% — split between a rapid precraniotomy dose and a slower infusion until dissection is complete. The effect is prompt, removes about 90 mL of brain water at peak, and lasts 2 to 3 hours. A serum osmolality above ~320 mOsm/kg is the conventional ceiling for further mannitol; the evidence for a hard threshold is weak, and the osmolar gap tracks nephrotoxicity better than serum osmolality does, so treat 320 as a prompt to reassess rather than a proven cutoff. Replace urinary losses with isotonic crystalloid, and watch for hypernatraemia, hypokalaemia and acute hypervolaemia.
CSF drainage by direct ventricular puncture, or by a lumbar catheter placed preoperatively — the latter effective only if there is no caudal obstruction to outflow. Because of the risk of causing acute herniation, lumbar drainage is used cautiously and only when the dura is open, with at least mild hyperventilation while draining. Removing 10 to 20 mL is very effective; up to 50 mL can be drained if necessary.
Position, because it is reversible and free. Head-down tilt and extreme neck rotation impair venous drainage and raise brain volume. When the brain swells after the dura is opened for no obvious reason, repositioning the head to relieve rotation and jugular compression may be the answer.
Volatile agents and nitrous oxide should be avoided. Neuromuscular blockade is indicated even in comatose patients, because muscle tone and shivering raise CO₂ production and undermine ventilatory control of pressure.
And corticosteroids are not given. They have not been shown to improve outcomes in intracranial haematoma, traumatic or primary, and have been shown to raise mortality in brain injury.
Both hypertension and hypotension are associated with poor outcome, which is a frustrating sentence until you separate the phases.
Before decompression, high pressure is often doing necessary work. Intracranial pressure can rise to 50 mm Hg with haematomas over 30 mL, and in a patient with impending herniation a high blood pressure is maintained deliberately to preserve perfusion pressure. In less severe patients, guidelines recommend a target systolic pressure of 140 mm Hg for hypertensive patients. If the pressure does need lowering, the order is analgesia first — opioids — then increased depth of anaesthesia with propofol, barbiturate or etomidate.
At decompression, the problem inverts. That is today’s pitfall.
An arterial line is indicated, and preferably placed before induction when the patient is haemodynamically unstable — except with impending herniation, where every minute counts and the line waits. An intracranial pressure monitor, if needed, is generally placed at the end of the procedure.
Blood gases, glucose and coagulation are all followed. Damaged brain releases large quantities of thromboplastin into the circulation, which disturbs coagulation, so clotting is checked regularly rather than assumed.
Patients with acute haematomas have significant brain injury and significant swelling, and should undergo slow weaning and delayed extubation in a neurointensive care unit. Chronic subdural haematoma patients, who arrived with minimal impairment, can usually be extubated immediately.
The distinction is the same one the day opened with. How fast the mass arrived determines almost everything, including how the case ends.
No randomized trial addresses this.
The cascade is established physiology. A decrease in cerebral perfusion pressure produces cerebral vasodilation with a subsequent increase in cerebral blood volume and intracranial pressure; an increase in perfusion pressure produces vasoconstriction with a decrease in blood volume and pressure. Nothing about that is contested, and it is the mechanism behind both the deterioration and the rescue.
The compliance argument is equally settled. For rapidly expanding masses the pressure–volume curve shifts markedly to the left. Compensation comes first from extracranial shift of intracranial blood, then from the larger-capacity displacement of CSF — which is ineffective if CSF flow is obstructed. Once those are exhausted, pressure rises steeply and the circulation fails.
The doses and the sequence are consensus. Mannitol at 0.5 to 1 g/kg, the induction agent choices, the volumes of CSF that can safely be drained, the blood pressure targets — these are accumulated neuroanaesthetic practice rather than trial results.
One thing here is a trial result, and it is negative. Corticosteroids have not been shown to improve outcomes for intracranial haematomas, whether traumatic or primary, and have been shown to raise mortality in brain injury. That is Neuro 1 day 09's CRASH result arriving in a different chapter, and it is the reason steroids are absent from this day's checklist.
The trap in this case is not the induction. It is twenty minutes later, and it catches people because it is the opposite of the problem they have spent the case solving.
A patient with a large acute haematoma arrives hypertensive. That hypertension is not incidental and it is not primary hypertension — it is a Cushing response, the circulation generating the perfusion pressure needed to push blood into a skull whose internal pressure may be as high as 50 mm Hg in haematomas over 30 mL. The patient is doing work on their own behalf.
Then the surgeon evacuates the clot, the intracranial pressure falls, and the reason for the response disappears. Rapid and severe hypotension can follow. The source is explicit that vasopressors should be immediately available during evacuation for exactly this, and that small adrenaline boluses of 0.1 mg may be necessary in extreme cases.
Two things follow, and they are both about timing rather than knowledge. Draw the vasopressor up before the bone flap comes off, not when the pressure drops — the interval between the surgeon saying "we're in" and the pressure falling is short. And if you are asked to treat the hypertension before decompression, treat it in the right order: improved analgesia with opioid first, then increased depth of anaesthesia with propofol, barbiturate or etomidate. That sequence exists because the hypertension is partly a response to stimulation and partly a response to the pressure, and only one of those should be blunted directly.
The wider lesson is worth naming. This case asks you to defend the blood pressure, then to lower it, then to defend it again, in about an hour. The only way to get all three right is to know which mechanism you are looking at each time.
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.
A patient with a large acute subdural haematoma becomes hypotensive after induction. Why is this more dangerous here than in a patient with a normal brain?
You are anaesthetising a patient for evacuation of a large acute subdural haematoma. They arrived hypertensive and bradycardic. What should be prepared before the dura is opened?
Which statement about osmotic therapy during emergency craniotomy is correct?
No returning items before day 4. The retrieval window opens once there is something far enough back to be worth retrieving.