Transmural Pressure

There is one equation in this day, and it is not difficult. What is difficult is that it has two names.

One subtraction, two names

Transmural pressure is the pressure difference across the wall of the aneurysm — the thing that distends it and eventually tears it. It equals mean arterial pressure minus intracranial pressure.

Cerebral perfusion pressure is the pressure driving blood into the brain. It equals mean arterial pressure minus intracranial pressure.

They are numerically identical. Which means the two goals of this operation — minimise transmural pressure to avoid rupture and maintain perfusion pressure to avoid ischaemia — are exactly opposed. There is no setting that satisfies both. There is only a compromise, and the correct compromise depends on which risk this patient actually carries.

A perfusion pressure of 60 to 80 mm Hg is a reasonable general goal. The rest of the day is about when to sit above it and when to sit below it.

Grade determines the compromise

The clinical grade is not a formality. It predicts, in order: intracranial pressure, the state of autoregulation, cardiac dysfunction, hypovolaemia and hyponatraemia, and the risk of vasospasm. Intracranial pressure of at least 20 mm Hg is present in 36 to 67% of patients after subarachnoid haemorrhage, and it correlates with grade — generally normal in grades I and II, elevated in IV and V.

That produces the counter-intuitive rule.

Good grades (0, I, II) have normal intracranial pressure and are not ischaemic. They tolerate a bigger transient fall in blood pressure — 30 to 35%, a systolic around 100 mm Hg. And they should not be hyperventilated, because reducing cerebral blood flow reduces intracranial pressure and therefore raises transmural pressure.

Poor grades have raised intracranial pressure, low perfusion pressure and ischaemia. The elevated pressure decreases transmural pressure and partially protects the aneurysm from rerupture. They tolerate hypotension poorly, so the depth and duration of any fall is moderated, and they are the patients managed with moderate hyperventilation to improve perfusion.

The general principle for induction is a reduction of 20 to 25% below baseline with prophylaxis against the intubation response, and the general principle for everything is that the change should always be gradual, never abrupt.

A normal intracranial pressure, incidentally, does not mean normal compliance. The reserve can be gone with the number still reading 12.

Why the pressure matters more here than elsewhere

Two pieces of physiology explain the anxiety.

Autoregulation is impaired, and the lower limit has shifted right. The severity tracks the clinical grade — Nornes and colleagues found the lower limit significantly higher in grade III than in grades I and II during intracranial surgery. Impaired autoregulation in the presence of vasospasm predicts delayed ischaemic deficits and unfavourable outcome. There are well-documented cases of a new deficit appearing with a fall in blood pressure and reversing when the pressure was restored pharmacologically. You must not let the perfusion pressure fall below a limit you cannot measure and which is higher than normal.

Wall tension rises with pressure and with size. By the law of Laplace, wall tension is the product of pressure and radius divided by two — so for any given blood pressure, wall stress is higher in an aneurysmal sac than in a normal artery, and higher in a large aneurysm than in a small one. That is the classical justification for lowering blood pressure during dissection, and it is the reason giant aneurysms, over 2.5 cm, are a different operation.

Induction is the dangerous part

Rupture at induction is rare — reported at 2% in one series, probably under 1% with modern technique — and is usually precipitated by a sudden rise in blood pressure at intubation. It carries a high mortality.

It is useful to think of induction as two separate problems:

Loss of consciousness. Propofol 1.5 to 2 mg/kg with fentanyl 3 to 5 µg/kg, sufentanil 0.3 to 0.5 µg/kg, or remifentanil 0.5 to 1.5 µg/kg. Etomidate 0.3 to 0.4 mg/kg and midazolam 0.1 to 0.2 mg/kg are alternatives.

Prophylaxis against the laryngoscopy response, which is a separate drug decision requiring separate drugs: high-dose opioid (fentanyl 5 to 10 µg/kg, sufentanil 0.5 to 1.0 µg/kg, remifentanil 1 to 1.5 µg/kg), esmolol 0.5 mg/kg, labetalol 10 to 20 mg, lidocaine 1.5 to 2.0 mg/kg, a second dose of propofol, or deep inhalational anaesthesia. Intravenous adjuncts are preferred in poor grades; deep volatile suits good grades but should be avoided with raised intracranial pressure.

Two rules make this safer than any drug list. Intubate only when paralysis is complete, monitored — coughing is a transmural pressure event. And if the blood pressure begins to rise unexpectedly during laryngoscopy, stop, ventilate, deepen, and try again.

The arterial line goes in before induction where possible, transduced at the level of the base of the skull and re-zeroed with any change in position, because it is standing in for perfusion pressure and not for systemic pressure.

Relaxing the brain, at a price

Every manoeuvre here reduces one of the three intracranial compartments, and every one of them raises transmural pressure. That is today’s pitfall, so here only the numbers.

Mannitol 20%, 0.5 to 2 g/kg — usually 0.5 to 1 — given over 30 minutes. Onset 4 to 5 minutes, peak 30 to 45. Given faster than 10 minutes it drops peripheral vascular resistance and causes transient hypotension, followed by a rise in central venous pressure and cardiac output. A serum osmolarity above 320 mOsm is the conventional ceiling for further dosing (a weakly-evidenced threshold) — the same ceiling as Neuro 2 day 03. Hypertonic saline 3% is a reasonable alternative, with no difference in brain relaxation and lower volume shift.

Cerebrospinal fluid drainage by lumbar catheter. The average adult has about 150 mL of cerebrospinal fluid; 20 to 30 mL is usually drained just before dural opening, at a rate not exceeding 5 mL/min. Contraindicated with an intracerebral haematoma.

Hyperventilation. Cerebral blood flow changes 2 to 3% per mm Hg of PaCO₂ within the physiologic range; blood volume changes about 1% per mm Hg. Mild hypocapnia 30 to 35 mm Hg before the dura is open, moderate 25 to 30 after, relative normocapnia after the clip.

Position. Ten degrees of reverse Trendelenburg improves intracranial pressure while maintaining perfusion pressure.

And when the brain stays tight despite all of it, the source gives a checklist worth memorising because it is mostly not pharmacological: exclude hypoxaemia and systemic hypertension; check the head and neck position for venous obstruction; check the drain is patent; stop nitrous oxide; head-up tilt after telling the surgeon; then a test bolus of propofol 100 to 200 mg and an infusion at 150 to 200 µg/kg/min if it helps. Occasionally the answer is that there is an intracerebral haematoma.

Failure to inspect head and neck position before draping is named as a common cause of unexplained intraoperative cerebral swelling. Tape the tube rather than tying it, so a tie cannot slip and tighten around the neck.

Hypotension is out; occlusion is in

Deliberate hypotension was standard for decades and is now exceedingly rare. Impaired autoregulation after subarachnoid haemorrhage, an unpredictable cerebrovascular response, an increased risk of vasospasm and reported poorer outcomes retired it. In a patient with vasospasm it is contraindicated.

Temporary occlusion of the feeding artery replaced it, and the two differ in almost every respect:

	Temporary occlusion	Controlled hypotension
Systemic pressure	normotension or hypertension	systemic hypotension
Flow	temporarily ceased	uninterrupted
Ischaemia	regional	global
Duration	short, 10–20 min	longer
Collaterals	depends on them	independent of them
Vessel damage	possible	none
Control of bleeding	possibly complete	incomplete

The clock matters. Five to seven minutes with prompt reperfusion is usually well tolerated but is often not long enough for a difficult aneurysm. Fifteen to twenty minutes is generally taken as the upper limit of safety, and occlusion under 20 minutes is associated with better outcomes. In Samson’s series, every patient occluded longer than 31 minutes infarcted, and older patients in poorer condition tolerated occlusion less well.

Restore the blood pressure to normal before the temporary clips are applied. The technique depends on collateral flow, and collateral flow depends on pressure.

Where a broad neck makes even temporary clipping difficult, adenosine can produce a transient cardiac standstill: a dose-response established with two or three incremental doses of 6 to 18 mg, then about 30 seconds of asystole at an average of 30 to 36 mg. Recovery is spontaneous, often with rebound tachycardia and hypertension. It is contraindicated with pre-existing conduction abnormalities or severe asthma.

Protecting the brain, honestly

The theory is clean: reduce cerebral metabolic rate and the tissue distal to the clip tolerates ischaemia longer. Thiopental, etomidate and propofol have all been used, titrated to burst suppression on EEG within 2 to 3 minutes. Suzuki’s high-dose mannitol at 2 g/kg, and the “Sendai cocktail” of mannitol with vitamin E and dexamethasone, come from the same idea. Lavine and colleagues reported that brain-protective therapy extended tolerable occlusion from 12 to 19 minutes without infarction.

Four honest qualifications belong with that:

• There are no randomised trials of pharmacologic neuroprotection for temporary occlusion, and the source says plainly it does not have sufficient data to support routine use.
• Additional drug may not reach the ischaemic tissue if the collateral supply is inadequate — though what did arrive will linger, because there is no flow to wash it out.
• Etomidate may cause cerebral tissue hypoxia during temporary occlusion despite reducing global metabolic rate, on brain oxygen tension measurement.
• Burst suppression and motor evoked potentials fight each other. The depressive effect on signal quality can be so pronounced that the motor response is lost completely. You cannot maximise both the protection and the monitoring that would tell you whether it was needed.

Mild hypothermia is the cleanest result in the chapter and it is negative: IHAST found no effect on neurologic outcome at 33.5°C in good-grade patients, and — a detail worth keeping — no coagulopathy either. The manoeuvre is not dangerous. It just does not help the patients who were studied.

When it ruptures

The Cooperative Study reported a leak in about 6% and frank rupture in 13%, a combined 19%; a later series reported 3.8% frank and 7.9% overall. Batjer and Samson found the timing: 7% before dissection, 48% during dissection, 45% during clip application. In roughly 8% of the Cooperative Study’s ruptures there was frank haemorrhagic shock.

Management divides on whether volume is being lost:

• Small leak, dissection complete — the surgeon gains control with suction and clips the neck.
• Temporary occlusion possible — proximal and distal clips, which is the preferred route, with propofol or etomidate for some protection before the clip. Maintain normotension during occlusion.
• Occlusion impossible, blood loss not significant — transiently drop the mean arterial pressure to 50 mm Hg or lower to let the surgeon see.
• Frank haemorrhage — resuscitate first. Cerebroprotective agents may not be giveable because of their haemodynamic effects, and induced hypotension is not available until volume is restored.

A video monitor showing you the field is named as a genuine aid here, which is a reminder that the anaesthetic response to this event depends on watching the surgery rather than the numbers.

Vasospasm, and a reconciliation

Angiographic vasospasm is not seen before 72 hours, peaks at day 7, and is seldom seen after two weeks. Cerebral blood flow can fall 20 to 25% without symptoms, so considerable narrowing is silent — and of those who do become symptomatic, about half die or are left with serious deficit.

Two things about treatment carry beyond this day. Nimodipine works without improving the angiogram, which is in today’s synthesis. And triple-H has become hypertensive therapy, because hypervolaemia adds no benefit and may worsen pulmonary dysfunction. Blood pressure is titrated to reversal of symptoms, up to 160 to 200 mm Hg systolic if the aneurysm is clipped and only 120 to 160 if it is not.

Which raises an apparent conflict with Neuro 1 day 15 and Neuro 2 day 06, both of which treat SIADH with fluid restriction.

It is not a contradiction — it is a different context. Hyponatraemia occurs in 30 to 57% of subarachnoid haemorrhage, and intravascular volume is abnormally low in 36 to 100% of these patients. Because hypovolaemia aggravates vasospasm and is associated with cerebral ischaemia and infarction, restricting fluid here trades a sodium problem for an infarct. So in subarachnoid haemorrhage the source says normovolaemia is the goal but it is safer to err on the hypervolaemic side, and hypertonic saline is the treatment of choice for both SIADH and cerebral salt wasting. Elsewhere, with no vasospasm risk, restriction is reasonable. The difference in management follows from the difference in what else the patient is at risk of.

Waking up

The stimulus profile inverts during the case — pins, then bone flap, then almost nothing once the dura is open — so the anaesthetic must be planned around it rather than around the clock.

Grade I and II patients after an uneventful clipping should wake and be extubated in theatre. Lidocaine 1.5 mg/kg suppresses cough but lasts only 3 to 5 minutes and can be repeated. Grade III is a judgement, and the source says err towards caution. Grades IV and V stay ventilated.

Blood pressure on emergence is managed to a different rule than during the case, because the aneurysm is now secured and hypertensive therapy is what reverses delayed ischaemia: modest hypertension below 180 mm Hg systolic is not treated aggressively, while above 200 risks swelling or haemorrhage. Labetalol in 5 to 10 mg increments or esmolol in 0.1 to 0.5 mg/kg increments.

Unless there are multiple aneurysms. Then the pressure is kept within 20% of the patient’s normal, because the ones that were not clipped are still subject to the equation this day is named for.

Finally, the rules for telling a slow wake-up from a surgical catastrophe: anaesthesia causes global depression, so any new focal deficit should be assumed surgical; volatile effects should have largely gone by 30 to 60 minutes; mid-sized reactive pupils with unimpaired respiration argue against opioid excess; and new unequal pupils always suggest a surgical event. Neurological status is assessed every 15 minutes, and sometimes the answer is a scan.