Pharmacology of antipsychotics

This chapter tries to address Section K2(iv) of the 2023 CICM Primary Syllabus, which expects the exam candidate to develop an " understand the pharmacology of anti-depressant and anti-psychotic drugs". No detailed exploration of this topic has ever appeared in any of the CICM exams, and the only thing that has come close were Question 21 from the second paper of 2014 (a comparison of haloperidol and diazepam), or Question 16 from the second paper of 2021 (classification of antipsychotics, but mainly about haloperidol). As such, there is probably little reason to go into a lot of depth here in the course of revision. Class characteristics are offered here for the purpose of rapid comparison and reference, in case an SAQ or viva on this topic appears at some stage.

Class	Classical antipsychotics	Atypical antipsychotics
Examples	Phenothiazines: Chlorpromazine Prochlorperazine Butyrophenones: Haloperidol Droperidol	Olanzapine Quetiapine Risperidone Aripiprazole
Routes of administration	Oral, IV, SC and IM administration	Oral only in the majority of cases
Absorption	Well absorbed orally
Bioavailability	Generally, a large first pass effect decreases bioavailability to 10-70%
Distribution	Large volume of distribution (~10-20 L/kg)
Protein binding	Most are >90% protein bound
Metabolism	Most are metabolised extensively in the liver. Many have active metabolites
Elimination	Metabolites are renally excreted Most of the newer agents and some of the older have long half-lives and require once daily dosing.
Mechanism of effect	D2 receptor blockade gives rise to the antipsychotic effect, and controls the positive symptoms of psychosis Serotonin receptor activity of the newer agents also manages the negative symptoms
Side-effects:	Sedation (antihistamine effect). Extrapyramidal side effects (dystonia, oculogyric crisis, laryngospasm, akathisia, rigidity, parkinsonism and tardive dyskinesia) Hyperprolactinaemia (dopamine blockade) Postural hypotension and sexual dysfunction (α-adrenergic receptor blockade) Anticholinergic side effects (xerostomia, urinary retention, tachycardia, constipation, blurred vision, tachycardia and delirium)  Lowered seizure threshold QT interval prolongation

If one want to refer to a single resource but needs something with more legitimacy than some online blog post, "Antipsychotics" by Sadek from his Clinician's Guide to Psychopharmacology (2021) would be an excellent start. That book in general is written in a conversational style resembling The Physiology Viva, which is to say it resembles a somewhat threatening and uncomfortable conversation, where one participant keeps asking really difficult and detailed questions. "Clinical pharmacokinetics of antipsychotics" by Javaid (1994) would also be nice, except it is paywalled. For the reader with endless patience for esoteric detail, "NeuroPsychopharmacotherapy" by Riederer et al (2022) is an immediately logical choice because it spans 4652 pages.

Chemical classes of antipsychotics

The distinction between "typical" and "atypical" antipsychotics is an entirely artificial line that was drawn for marketing purposes when clozapine became available on the market in the early 1970s. Less extrapyramidal side effects, they said. Contemporary reviews of this distinction have been unable to confirm the existence of this sharp distinction. Some "typical" antipsychotic agents have very few extrapyramidal side effects (for example, perazine), and some of the novel atypical agents (eg. risperidone) have plenty of them, making it an ineffective method of discriminating between agents. 

• Typical antipsychotics
  • Phenothiazines (chlorpromazine)
  • Butyrophenones (haloperidol, droperidol)
• Atypical antipsychotics
  • Phenylpiperaxines  (aripiprazole)
  • Dibenzodiazepine (clozapine, quetiapine)
  • Thienobenzodiazepines (olanzapine) 
  • Benzisothiazoles and benzisoxazoles (risperidone, lurasidone)

An alternative classification is possible. Here is one from Jes Gerlach (1991), which reassembles these agents into pharmacodynamic mechanism groups:

• Selective D2 receptor antagonists:
  • Sulpiride
  • Remoxipride
  • Raclopride
• Partial D2 receptor agonists:
  • Terguride
  • Roxindole
• Serotonergic agents:
  • Buspirone
• Glutamate agonists:
  • Milacemide
• Multi-receptor blockers (D2 plus others):
  • Serotonin + D2 antagonists:
    • Risperidone
  • Serotonin + D2 + α1 antagonists:
    • Amperozide
    • Sertindole
  • Serotonin + D1 + D2 + α1 antagonists:
    • Clozapine
    • Savoxepine

Unfortunately, as one might have guessed from their experience with other such mechanism-based classification systems, as soon as you start putting things into categories you discover that some things span multiple categories. An alternative, which appeals to the person prone to dumpster-diving for obsolete literature, is a timeline of development such as the one that follows,  borrowed from Saloni (2024), Shen (1999) and Cunningham Owens & Johnstone (2018). The main reason this works so well is because

Antipsychotic agents

Agent	Year of availability
Phenothiazines
Promethazine	1949
Chlorpromazine	1950
Prochlorperazine	1956
Levomepromazine	1956
Butyrphenones
Haloperidol	1958
Droperidol	1961
Zuclopenthixol	1962
"atypical" antipsychotics
Clozapine	1958
Risperidone	1993
Olanzapine	1996
Quetiapine	1997
Ziprasidone	2001
Aripiprazole	2002
Paliperidone	2007
Lurasidone	2010

The history of the development of these drugs from Cunningham et al reads similarly to the first chapter of Matthew from the New Testament, insofar as it lists the genealogy of these drugs arising from earlier, more primitive and violent molecules. And lo, 

• Coal tar begat methylene blue and phenothiazine, which begat the whole range of early classical antipsychotics (some of which were subsequently reclassified as antihistamines or antiemetics)
• Pethidine begat norpethidine, which begat haloperidol and the butyrophenones (because Janssen was looking for an antidote to amphetamines)
• Of the tribe of butyrophenones, pirenperone, an agent developed to antagonise the effects of LSD, begat risperidone and all of its ilk. 
• Risperidone, notably, directly begat paliperidone, which is its major active metabolite
• Procainamide begat metoclopramide; metoclopamide begat sulpiride; sulpiride begat amisulpiride, who like his fathers prolonged the QT interval.
• Imipramine begat clozapine, who then lay dormant until the 2010s
• Clozapine begat olanzapine through the directed pursuit of daughter molecules that might be kinder to the heart and bone marrow
• Clozapine also begat quetiapine by a similar method, where the investigators were looking for agents that did not cause agranulocytosis

Pharmacokinetics

For a family of drugs with so many members, the pharmacokinetic properties of them all are remarkably uniform, which makes it easy to describe and study them en masse. And it is fortunate that the exam candidates are not expected to study them en masse, as only haloperidol  olanzapine   and quetiapine are listed in the First Part Exam pharmacopeia.

Thus:

Drug	Haloperidol	Olanzapine	Quetiapine
Chemistry	Butyrophenone - grouped with the "classical" antipsychotics, though not really a phenothiazine	Thienobenzodiazepine - grouped with the "atypical" antipsychotics, but they are all very chemically different	Dibenzodiazepine - grouped with the "atypical" antipsychotics, but they are all very chemically different
Routes of administration	Oral, IV, IM, s/c	Oral, IV, IM	Oral only
Absorption	Rapidly and completely absorbed; oral bioavailability is about 60%	Rapidly and completely absorbed; oral bioavailability is about 60%	Rapidly and completely absorbed (faster onset than others); oral bioavailability is thought to be close to 100%
Solubility	pKa 8.66; minimally water-soluble	pKa 7.78; minimally water-soluble, highly lipophilic	pKa 6.8; minimally water-soluble, moderately lipophilic
Distribution	VOD=18L/kg; 92% protein-bound	VOD=10-20L/kg; 90% protein-bound	VOD=10L/kg; 83% protein-bound
Metabolism	Hepatic metabolism into inactive metabolites	Hepatic metabolism into inactive metabolites	Hepatic metabolism into inactive metabolites by P450 enzyme CYP3A4, into metabolites at least one of which is active.
Elimination	Inactive metabolites are renally cleared	Inactive metabolites are renally cleared (60%) and 30% via the faeces, with about 10% of the drug eliminated unchanged	Inactive metabolites are renally cleared (70%) and the rest via the faeces, with only about 1% of the drug eliminated unchanged
Time course of action	Elimination half-life of 14-36 hours	Elimination half-life of about 33 hours	Elimination half-life of about 33 hours

Of course it would be possible to summarise the properties of the rest for the purpose of reference,

For these others, in summary:

• Administration and absorption is similar for both typical and atypical groups, with the older typical agents being somewhat more versatile. They are usually available in IV and IM formulations, whereas the newer agents are usually only oral. Both groups have various conveniently long-acting versions which can sit happily in a deltoid or a buttock as a depot, slowly dissociating into the bloodstream over weeks.
• Bioavailability is pretty variable, but most of these drugs are subject to a significant first pass effect. The range is from 10% to 70%. 
• Protein-binding is predictably high, just like it is with the antidepressants. Most agents are 90-99% protein bound (for example, haloperidol is92% protein-bound)
• Distribution is vast, as these highly lipophilic protein-loving drugs tend to accumulate in tissues. For haloperidol, for example, the volume of distribution is 18L/kg.
• Elimination is invariably by hepatic mechanisms, with inactive metabolites escaping via the urine. The number of possible metabolites for each drug is vast, and many have independent CNS activity (though not all of them are predictably D2 antagonists like the parent molecule). For example, risperidone is metabolised into a metabolite so active and so long-lived that it became a marketable drug all on its own (paliperidone).

Atypical antipsychotics also have a few clinically significant pharmacokinetic peculiarities, as outlined by Sheldon Prescorn (2012), and a table of these properties can be reviewed by anybody who has that kind of time. 

Mechanism of antipsychotic action

Dopamine antagonism (specifically, the antagonism of D2 dopamine receptors in the mesolimbic system) seems to be the main mechanism of action for typical antipsychotics. At the same time, the same mechanism is responsible for the extrapyramidal side effects. The antipsychotic effect is seen when the drugs are blocking65-75% of the D₂ receptors, and disappears below this narrow window (Amato et al, 2020). The idea that this receptor blockade underpins the antipsychoticness of the drug is also supported by the finding that potency of D₂ blocking and receptor affinity seems to be related to the magnitude of the clinical antipsychotic effect seen for each agent. Obviously, the brain is more complex than that, and there are numerous caveats to this oversimplification, but for exam purposes it will have to do.

Indications for the use of antipsychotics in the ICU

Quetiapine was originally developed by Edward Warawa in the 1980s, who added various piperazine structures to the original clozapine molecule to  "identify novel compounds ... which, comparably to [clozapine], potentiated gnawing in mice while inhibiting apomorphine-induced climbing." Neither gnawing or climbing seem like particularly desirable  behaviours to impart upon ICU patients, which leads us naturally to the topic of agent selection; and specifically to the question, how does an intensivist intelligently decide on the specific agent they will use.

Leaving aside the completely reasonable scenario of managing acute psychosis (where most likely there is a psychiatrist or three somewhere, driving the choices of agent), the uses of antipsychotics in ICU are largely limited to the following scenarios:

• Management of agitation in the delirious or intoxicated patient
• Co-sedative agent to decrease the total dose of another more toxic sedative agent
• Off-label or incidental use, eg. where the drug is used as an antiemetic

None of these are the normal applications of these agents, which means one cannot use the psychiatric literature and evidence base to guide one's choices. Psychiatrists are interested in different effects, as compared to the effects that the intensivists are interested in; the latter may be less concerned with defeating the negative social and cognitive clinical features of schizophrenia, and more concerned with sedation (which is conventionally regarded to be an undesirable side effect). From this, it follows that the choice of antipsychotic agent in the ICU should be mostly based on their spectrum of side effects, of which some may be desirable and some may not. 

Mechanisms of antipsychotic side effects

So numerous and important are these, that it felt appropriate to present them in a variety of ways, to suit a variety of learning styles and revision techniques. Or perhaps just to act as a reference of sorts, as there is not easy to come by a single place where all such things are compiled.  

Unfortunately, this topic is very difficult to summarise or illustrate. As an example, some of the best papers on this subject, such as  Siafis et al (2018) and Goldstein (2000), are a showcase of some truly questionable graphical representation techniques. For example, it is difficult to know which among this table, or this pie chart, are more intimidating and incomprehensible:

Or, even worse,


However, the author is not capable of following this criticism with anything better or more sensible. The difficulty and density of the subject needs to be acknowledged. The only way forward seemed to be to get the data from the original source for these pie charts, which is Goldstein (1995), as well as numerous other sources, and create this monstrosity (because bar graphs are somehow better):

Antipsychotic Agents Ranked by Receptor Affinity
where a longer bar represents higher affinity

Drug Name	D1	D2	5-HT 2A	5-HT 1A	α-1	α-2	H1	M1	QTc*	S.T.**

Or, if you prefer to have the raw numbers,

Antipsychotic Agents Ranked by Receptor Affinity, Again;
where a larger value represents a lower affinity, because it is the K_d, an equilibrium dissociation constant that quantifies the ratio between k_off and k_on. Thus, a lower value indicates more k_on and indicates a higher affinity of the ligand for the receptor.

Drug Name	D1	D2	5-HT 2A	5-HT 1A	α-1	α-2	H1	M1	QTc*	S.T.(%)**

* QTc is the QT prolonging effect of these drugs; where the value is the hERG (IKr) block potency in nanomoles. These values are very approximate and should be regarded with suspicion. The specific number would depend on the kind of in vitro assay you used and the cloned receptors you tested. A larger value represents less QTc effects (i.e. the more nanomoles required, the less Qtc prolongation per using dose).

**  S.T is the seizure threshold effect of these drugs. It is a percentage extracted from various sources of postmarketing data, and is also extremely approximate. The value represents a reported incidence of worsening seizure disorder, i.e a higher value = a higher incidence

*** The references for this horror are listed in the lowermost part of the references section

But what does any of this mean, the frustrated reader might rightfully ask. To what extent should I care? And this would be a well-reasoned response, grounded in the knowledge that we really have no clue as to how these receptor affinities factor into the drug behaviour

• D1 inhibition  is, if anything, pro-psychotic, insofar as excessive antagonism at these receptors seems to exacerbate psychosis and mania. On the plus side, there seems to be a somewhat lower risk of extrapyramidal side effects (Hietala et al, 1990).
• D2 inhibition is thought to be the effect that mediates all the main antipsychotic activity. Affinity here is apparently correlated with clinical potency. That does not necessarily mean that affinity translates directly into utility. For example, quetiapine, with its terrible high Ki, might seem worse than all the others, but it does mean that it is useless as an antipsychotic - quite to the contrary, it is among the most often prescribed agents, but the dose at which its antipsychotic effects are manifested (eg. 500-800mg/day) is well in excess of what one might give in the ICU for agitation.
  Additionally, D2 inhibition is also associated with all the antidopaminergic side effects:
  • Hyperprolactinaemia: Dopamine is a normal regulator of prolactin secretion, and dopamine (specifically D2 receptor) block in the tuberoinfundibular tract leads to the suppression of this regulation. With the breaks disabled, the pituitary gland pumps out gallons of prolactin. Galactorrhea and breast enlargement occurs in women, and gynecomastia in men.
  • Extrapyramidal side effects are the consequences of blocking striatal D2 receptors:
    • Dystonia: involuntary muscle spasm involving usually facial muscles (tongue, neck,  eyes, etc) which can happen on the very first day of treatment. Hideously, it can involve the larynx. 
    • Oculogyric crisis is the extreme version of this, which classically also involves an arched back and eyes which have rolled back inside their sockets.
    • Akathisia, a sort of motor restlessness, often occurs within the first week of treatment
    • Rigidity and parkinsonism typically takes months to develop
    • Tardive dyskinesia is only seen with prolonged treatment
• H1 inhibition mediates the sedation effects of antipsychotic agents. This effect is not always undesirable, and in fact often in the ICU these drugs are used for this specific purpose. As such, if sedation is the goal, giving the patient promethazine may be just as good as giving them quetiapine
• α1 inhibition peripherally produces hypotension, and in overdose, the block can be sufficiently deep to produce significant vasodilation and hypotension.
• α2 inhibition centrally produces hypertension and peripherally leads to sexual dysfunction
• M1 blockade leads to anticholinergic side effects (xerostomia, urinary retention, tachycardia, constipation, blurred vision, tachycardia and delirium)

It is also possible to discuss some of their other side effects without resorting to monoamine receptors.

• Lowered seizure threshold: All antipsychotics do this to some extent, but some do it more than others. It appears the oldest of the drugs have some of the greatest risk of seizures. Sadek (2021) claims that chlorpromazine and thioridazine are the most epileptogenic, but there is no reference to back it up.
• QT interval prolongation occurs because antipsychotics block the delayed rectifier currents (I_k), which are responsible for Phase 3 of the action potential. According to Zareba & Lin (2003), this is also something that appears to be common to all of them, with a distribution ranging from minor effects (aripiprazole, olanzapine, haloperidol) to significant effects (thioridazine). The upshot is an increase in the risk of sudden cardiac death (from polymorphic VT) which is 2-3 times higher among chronic schizophrenia patients when compared to the rest of the population.