A drug cleared faster than the liver is perfused
— and the emulsion it had to arrive in.

• Clearance exceeds hepatic blood flow. Extrahepatic metabolism must exist, and it does — mainly renal, partly pulmonary. Everything good about propofol follows from this.
• The pressure falls mainly because resistance falls, with reduced contractility and a blunted baroreflex on top — and the reflex tachycardia that should compensate usually does not appear.
• It obtunds airway reflexes better than any other induction agent, which is why a supraglottic airway goes in without a relaxant, and why apnoea follows a quarter to a third of inductions.
• Marsh scales with weight; Schnider fixes V₁ and uses age and height. That single difference is most of what separates them in a real patient.
• Propofol infusion syndrome is a mitochondrial lesion, and rising triglycerides may be its first sign — before anything happens to the heart.

A 1% preparation is 10 mg/mL of propofol in an emulsion of 10% soybean oil, 1.2% purified egg phospholipid as the emulsifier and 2.25% glycerol to make the emulsion isotonic, with sodium hydroxide adjusting the pH to about 7. An antimicrobial — EDTA in the commonest formulation, sodium metabisulfite or benzyl alcohol in others — was added after concerns about bacterial growth in the emulsion.

The generic formulations differ in their excipients, and the difference is not cosmetic. All contain soybean oil, egg lecithin and glycerol; some substitute part of the long-chain triglyceride with medium-chain triglycerides, which increases the proportion of propofol held inside the lipid droplet rather than free in the aqueous phase — and correspondingly reduces both the free concentration and the pain on injection.

The emulsion is stable at room temperature, not light sensitive, and may be diluted with 5% dextrose. It is also an excellent bacterial growth medium, which is the reason for the preservative and for the strict limits on how long a drawn-up syringe may be kept.

Distribution. Propofol is 98% bound — to albumin and to erythrocytes — leaving 1 to 2% free in plasma, and it crosses the blood–brain barrier readily. Its volume of distribution is the largest of the induction agents at about 4 L/kg, a direct consequence of its lipid solubility. Miller reports a volume of distribution at steady state of 2 to 10 L/kg and a central compartment of 6 to 40 L. After a bolus the initial distribution half-life is 2 to 8 minutes and the slow distribution half-life 30 to 70 minutes.

Metabolism. About 70% is conjugated directly to propofol glucuronide. The remaining 30% undergoes a phase I hydroxylation to 4-hydroxypropofol, mediated by CYP2B6 with a contribution from CYP2C9, and that metabolite is then conjugated to 4-hydroxypropofol sulphate and glucuronide, and to 1-hydroxypropofol glucuronide. All the conjugates are excreted renally. Less than 1% of the dose appears unchanged in the urine.

Clearance is the number that matters. Propofol’s hepatic extraction ratio is 0.79 to 0.92 — the liver removes nearly everything presented to it — so its clearance is flow-limited, and reducing hepatic blood flow reduces its metabolism. Total clearance is 1.5 to 2.2 L/min. Hepatic blood flow is about 1.5 L/min. A clearance at or above that figure cannot be produced by the liver alone.

The terminal half-life is a trap. It is quoted as 5 to 12 hours, but if sampling continues beyond 24 hours the figure approaches 60 hours, which reflects slow release from fat and describes nothing a patient experiences. The number that describes what the patient experiences is the context-sensitive half-time, under 40 minutes even after eight hours of infusion. The short distribution half-life is a third number again, and it describes something different from either: it is what terminates a single bolus, and it says nothing about what happens at the end of a steady infusion. The three are routinely conflated, and they answer three separate questions.

Propofol impairs its own clearance. Because it has a high extraction ratio, and because it reduces cardiac output and hence hepatic blood flow, doubling the dose can more than double the resulting concentration. The converse holds: a sympathomimetic that raises cardiac output lowers the propofol concentration. In haemorrhagic shock, concentrations rise about 20% until decompensation, at which point they rise steeply.

Age and size. The elderly have a smaller central compartment and reduced clearance, both largely from reduced cardiac output, and are more sensitive to a given concentration — so a patient over 80 needs roughly half the dose of a 20-year-old. Children have a central compartment about 50% larger and clearance about 25% faster, so they need more per kilogram: the ED95 in children is 2 to 3 mg/kg, and the maintenance requirement is correspondingly higher.

Hypnosis comes from enhancement of GABA-induced chloride current at the β subunit of the GABA-A receptor, as lesson 1 sets out. Propofol also inhibits acetylcholine release in the hippocampus and prefrontal cortex, inhibits the NMDA receptor through modulation of sodium-channel gating, and acts on GABA-A and glycine receptors in the spinal dorsal horn.

Naming the effects — pressure down, output down, resistance down — describes what happens without explaining any of it, and it leaves out the two findings that are most characteristic of propofol: that the compensatory tachycardia does not appear, and that profound bradycardia and even asystole occur. The table below gives the mechanism of each limb, because the mechanisms are what let you predict the size of the fall in a patient you have not met.

• Apnoea follows 25 to 30% of induction doses, and its incidence and duration depend on dose, speed of injection and premedication. Apnoea lasting more than 30 seconds becomes considerably more likely when an opioid is given as premedication or immediately before induction.
• A maintenance infusion of 100 μg/kg/min reduces tidal volume by about 40% and increases respiratory rate by about 20%, so the change in minute ventilation is unpredictable. Doubling the infusion rate deepens the depression further.
• Airway reflexes are obtunded more than by any other induction agent. Cough and laryngospasm are rare after propofol. This is why a supraglottic airway can be placed under propofol alone without a neuromuscular blocking drug, and it is the single most useful respiratory fact about the agent.
• Note what propofol is not. It is not a bronchodilator, not a muscle relaxant, and not an analgesic, and it is not a treatment for laryngospasm. It obtunds the reflex that produces laryngospasm, which is why laryngospasm is uncommon under propofol; that is a different claim from relieving one that has already occurred.

Propofol is antiemetic at sub-hypnotic concentrations. The mechanism is a reduction in serotonin concentration in the area postrema, probably through its action at GABA receptors there; antagonism at the dopamine D₂ receptor has also been proposed. It is effective after induction, after maintenance, and given as a small post-operative dose. A meta-analysis of recovery after propofol maintenance against the newer volatiles found only minor differences in time to recovery goals but significantly less nausea and vomiting after propofol.

The pain comes from free propofol in the aqueous phase of the emulsion in contact with the venous endothelium and the nociceptive nerve endings around it — not from the lipid, and not from the pH, which is neutral. That single mechanistic statement explains every effective intervention, because they all reduce either the aqueous-phase concentration or the exposure.

• Use a large vein and avoid the dorsum of the hand. Greater flow dilutes the aqueous-phase drug faster and there are fewer nociceptors per unit of endothelium.
• Add lidocaine to the propofol, or give it into the same vein first with venous occlusion. This is the intervention with the most consistent evidence.
• Change the formulation. Medium-chain triglyceride preparations hold a greater proportion of the drug in the lipid phase, reducing the free concentration and the pain with it — the mechanism made into a product.
• Elevate the limb to speed venous drainage.
• Pretreatment with a small dose of propofol, an opioid, an NSAID, ketamine, esmolol or metoprolol, magnesium, clonidine with ephedrine, dexamethasone and metoclopramide have all been tried, with variable efficacy.

Total intravenous anaesthesia means inducing and maintaining anaesthesia with intravenous drugs alone, usually propofol with an opioid. Its advantages are largely the absence of the volatiles’ disadvantages — no fluoride ions, markedly less post-operative nausea and vomiting, no distension of air-filled spaces, no diffusion hypoxaemia, no malignant hyperthermia trigger, no theatre pollution and no occupational exposure — together with two positive indications: it does not suppress evoked potentials the way the volatiles do, so it is preferred for spinal surgery with motor and somatosensory evoked potential monitoring, and it permits a shared airway for bronchoscopy.

What makes propofol the agent for it is the material of section 03: high clearance from multiple organs, so plasma concentration falls quickly when the infusion stops; a context-sensitive half-time that rises only modestly with infusion duration; metabolites that contribute little or nothing; and enough kinetic predictability that a model can be built.

What a target-controlled pump is doing

A TCI pump holds a three-compartment pharmacokinetic model of the drug and solves it forward in time. Given the patient’s covariates, the drug concentration in the syringe and the target you set, it computes the infusion rate needed to reach and hold that concentration — a bolus to fill the central compartment, then a decreasing infusion that replaces both what is being eliminated and what is being distributed peripherally. It is calculating a concentration; it never measures one.

Studies suggest the displayed plasma concentration is generally within 25% of the actual concentration, which is useful and is not the same as accurate. The infusion must still be titrated to effect, exactly as a vaporiser setting is.

Marsh against Schnider

One row does most of the work. Marsh describes V₁ as 0.228 L/kg, so it scales directly with the weight entered: at 70 kg V₁ is 15.96 L, and at 140 kg it is 31.92 L. Double the weight, double the initial dose. For a normally proportioned patient that is reasonable. Where the extra weight is adipose it is not: V₁ becomes an overestimate and the patient is overdosed.

Schnider fixes V₁ at 4.27 L — nearly four times smaller than Marsh’s value for a 70 kg patient — and uses age for V₂ and the intercompartmental rate constants, and weight with height for clearance. Because V₁ is fixed, the induction bolus barely changes with weight, and if Schnider’s usual effect-site targets were used in plasma-targeted mode the dose would be too small to induce anaesthesia reliably. The two models are therefore built for different targeting modes: Marsh for plasma targeting, Schnider for effect-site targeting, and that is ordinarily how they are used.

Typical maintenance targets when propofol and remifentanil are used together are 2.5–4 μg/mL and 3–6 ng/mL respectively, and the two are synergistic — raise one target and the other can come down for the same clinical effect. For a long case, priority usually goes to keeping the propofol target low and the remifentanil target higher, because remifentanil’s context-sensitive half-time is short and almost independent of duration. Many pumps display a decrement time: the predicted interval until the concentration falls to a value at which the patient is expected to wake, commonly set at 1.2 μg/mL for propofol. Opioids and other hypnotics prolong it beyond the prediction.

Depth of anaesthesia monitoring

Volatile anaesthesia lets you measure the expired concentration, which is a real-time proxy for the delivered dose; a disconnected breathing circuit is detected. Total intravenous anaesthesia has no equivalent — a disconnected infusion line, an extravasated cannula or a pump programmed with the wrong drug or concentration all produce a patient receiving nothing, with nothing to show it. That, and not any property of propofol, is why depth-of-anaesthesia monitoring is recommended where TIVA is used together with a neuromuscular blocking drug. The processed EEG indices themselves — how a bispectral index is derived and what its limitations are — are taught in EEG and evoked potentials.

Infusion rates for sedation are half or less of those for general anaesthesia — 25 to 75 μg/kg/min, or 30 to 60 μg/kg/min to supplement a regional technique in a healthy patient. In patients over 65 and in the sick the requirement falls by up to another 50%. Amnesia is generally reliable at rates above about 30 μg/kg/min, and tolerance develops: in 20 to 40% of intensive-care patients the dose has to be adjusted upward repeatedly to maintain the same effect. Sedation holidays and the smallest effective dose are both part of a long-term regimen, and the reason is the previous section.

The sections above were organised by system, which is how the pharmacology is conventionally set out. They are not, however, independent of one another. Almost all of propofol’s clinical identity descends from two structural facts, and following the chain is a better way to hold the drug in memory than the systems are:

1. Two isopropyl groups flank the phenol. The molecule is therefore intensely lipid soluble and will not dissolve in water.
2. So it must be given as an emulsion — and the emulsion supplies the pain on injection, the microbial risk, the calorie load, the allergy question, and the lipid substrate at the far end of the propofol infusion syndrome chain.
3. Lipid solubility also gives it a large volume of distribution and instant access to the brain, so it meets the one-arm–brain specification and wears off by redistribution in five to ten minutes.
4. It is cleared faster than the liver is perfused. Extrahepatic metabolism must therefore exist, and does — renal and pulmonary. Hepatic disease needs no dose adjustment for the same reason.
5. That clearance is what flattens the context-sensitive half-time. Under 40 minutes at eight hours, because elimination keeps pace with the drug returning from the periphery.
6. A flat context-sensitive half-time is what makes TIVA possible, and a predictable kinetic model is what makes TCI possible on top of it.
7. And running it for days is what reaches the ceiling. Propofol infusion syndrome is the price of the technique that step 5 bought, which is why the dose and duration limits are not arbitrary.

The cardiovascular effects sit outside that chain and are the counterweight to it. They are the reason propofol is not the answer to every question in this module, and lesson 4 is where that argument is settled.