Serotonin syndrome in the perioperative period

Learning objectives.

By reading this article, you should be able to:

• •Explain the Sternbach and Hunter criteria for diagnosing serotonin syndrome.
• •Describe the wide variety of signs and symptoms of serotonin syndrome.
• •Distinguish the pathophysiology of serotonin syndrome.
• •Summarise the principles of management of a patient with serotonin syndrome.

Key points.

• •Serotonin syndrome has been described as a triad of changes in mental status, neuromuscular abnormalities, and autonomic hyperactivity.
• •It is difficult to diagnose serotonin syndrome in a patient under general anaesthesia because the signs and symptoms are similar to other perioperative conditions.
• •Concurrent use of antidepressants has been strongly associated with the occurrence of serotonin syndrome.
• •The Hunter criteria for predicting serotonin toxicity are both more sensitive and more specific than the Sternbach criteria.
• •Early diagnosis and aggressive supportive care are the mainstays of management.

The serotonin syndrome (SS) is a potentially life-threatening drug interaction caused by excessive serotoninergic activity in the CNS. It can arise from therapeutic drug use, drug interactions, or intentional overdose of medications that affect the serotonergic system. The clinical features of SS have been described as a triad of changes in mental status, neuromuscular abnormalities, and autonomic hyperactivity.¹ Clinical presentation is highly variable, and ranges from mild symptoms to a severe life-threatening condition.

The first human case reports of serotonin-related adverse effects of psychotropic drugs were published in the 1950s. The term serotonin syndrome was first used in the early 1980s to describe the presence of hyperthermia and behavioural changes in patients receiving medications that have serotoninergic activity.² Diagnosis during the perioperative period is particularly challenging because many confounding factors can obscure the clinical picture. In addition, other conditions, such as malignant hyperthermia and neuroleptic malignant syndrome, have similar features to SS, and are part of the differential diagnosis of delirium and neuromuscular abnormalities during the perioperative period. Many agents with potential serotoninergic activity are given during relatively short periods of time in the perioperative period. This increases the likelihood for drug interactions and their consequences, such as the SS.

In this article, we review the epidemiology, diagnostic challenges, and management of perioperative SS. In addition, we summarise the characteristics of the cases of perioperative SS that have been reported in the literature.

Epidemiology

Prescriptions for antidepressant drugs in England have doubled in the past decade. This has been linked to the increasing awareness of mental health illness, improvement in diagnosis of mental health conditions, and a greater willingness of patients to seek help.³ The annual report of the American Association of Poison Control Centers lists antidepressants as the third most common cause of reported toxic substance poisoning, accounting for 9% of the total exposures in adults. Many of these reports are of selective serotonin reuptake inhibitors (SSRIs) and associated SS.⁴

Significant numbers of patients presenting for surgery are taking SSRIs. A study using data from a large database of 530,416 patients undergoing major surgery found that 72,540 (13.7%) of patients received an SSRI during the perioperative period.⁵ The variability in clinical presentation of SS makes the true incidence of this condition difficult to ascertain, with many cases probably going underdiagnosed.

Diagnosis

The diagnosis of SS is based on the presence of the triad of change in mental status, neuromuscular irritation signs, and autonomic hyperactivity in a patient with current or previous exposure to a medication with serotoninergic activity. The onset of SS typically occurs all of a sudden within 24–48 h of exposure to the triggering agents, and usually resolves quickly after the triggering agent is discontinued. However, the ‘washout’ period after discontinuation of psychotropic drugs is highly variable because of wide variations in their elimination half-lives, which range from 15 h for paroxetine and sertraline to 7 days with fluoxetine.⁶ Therefore, SS can occur even when serotonergic drugs have been discontinued several days before exposure to a trigger.

Sternbach described the most commonly encountered signs and symptoms in 12 case reports, including 37 patients suffering from SS.⁷ Mental status changes were the most frequently reported symptoms, followed by restlessness, myoclonus, hyperreflexia, diaphoresis, shivering, tremor, and diarrhoea in decreasing order of frequency. Based on these clinical features, Sternbach suggested that the presence of three or more of these clinical manifestations combined with previous or new exposure to a serotoninergic agent was suggestive of SS.⁷

The analysis of a large toxicology database (Hunter Area Toxicology Service) containing data from 2222 patients after an overdose of serotoninergic drugs was used to derive a model that was predictive of serotonin toxicity based on clinical features. This Hunter model contains seven clinical features and decision rules that, when applied to the database, were highly predictive of serotonin toxicity. The Hunter criteria were found to be more sensitive (84% vs 75%) and more specific (97% vs 96%) than Sternbach's criteria.⁸ Application of Sternbach's criteria alone may rule out mild or subacute presentations of SS. However, the Hunter criteria should be used to improve diagnostic accuracy (Fig. 1).

Fig 1.

Algorithm using the Hunter criteria for the diagnosis of SS in the perioperative period.

Serotonin syndrome can present in a clinical spectrum ranging from mild symptoms to severe life-threatening toxicity. There are no laboratory or imaging tests specific to SS, and it remains a diagnosis made on the bases of clinical presentation. Laboratory abnormalities, such as metabolic acidosis, increased liver enzymes, increased serum creatinine, and leucocytosis are frequent and likely reflect severity of associated organ dysfunction. Increased serum creatinine phosphokinase can occur in cases where rhabdomyolysis is present. There is no correlation between serum serotonin concentrations and the severity of SS.¹

The reliance on clinical signs for the detection of SS makes its diagnosis particularly challenging in the perioperative period. The detection of all three elements of the diagnostic triad can be obscured by common circumstances or conditions encountered in the perioperative period. Observation of abnormal movements, pupillary changes, and abnormal muscle tone can be very difficult in a patient who is anaesthetised and receiving neuromuscular blocking agents (NMBAs), and who is covered by surgical drapes. Confusion and agitation after surgery and anaesthesia are not uncommon and have a number of causes. Alterations in body temperature can be related to an underlying pathology or to the inflammatory response that accompanies surgery.

Perioperative manifestations of SS can overlap with other conditions, such as malignant hyperthermia, neuroleptic malignant syndrome, anticholinergic syndrome and opioid toxicity (Table 1). However, it is very important to remember that a combination of some of the aforementioned signs and symptoms (e.g. hyperthermia, severe haemodynamic fluctuations, rhabdomyolysis, acute kidney injury, disseminated intravascular coagulation, severe agitation and seizures) could point towards life-threatening SS, and the clinician should be vigilant for these.

Table 1.

Differential diagnosis of SS in the perioperative period.

	History	Clinical
Malignant hyperthermia	Family history adverse reactions associated with anaesthesia; exposure to halogenated anaesthetics or depolarising NMBAs	(i) Hyperthermia >41°C (ii) Rigidity (iii) Hyporeflexia (iv) No myoclonus
Anticholinergic syndrome	History of exposure to an anticholinergic agent (e.g. neostigmine and pyridostigmine)	(i) Miosis (ii) Bradycardia (iii) Bronchoconstriction (iv) Salivation (v) Diarrhoea (vi) Cramps
Opioid toxicity	Exposure to opioids	(i) Miosis (ii) Hypothermia (iii) Respiratory depression (iv) Hyporeflexia
Neuroleptic malignant syndrome	Associated with the use of an antipsychotic (neuroleptic agents)	(i) Gradual onset (ii) Lead-pipe rigidity, fever, and dysautonomia (iii) No gastrointestinal symptoms
Perioperative delirium	Acute onset and fluctuating course; multifactorial origin	(i) Inattention, disorganised thinking, and altered level of arousal (ii) No neuromuscular activation signs (e.g. tremor, clonus, or hyperreflexia)
Sedative–hypnotic withdrawal	History of chronic intake and abrupt cessation of sedatives–hypnotics (e.g. benzodiazepines, baclofen, ethanol, and barbiturates)	(i) Confusion and agitation (ii) Autonomic dysfunction may be present (iii) No inducible clonus

Pathophysiology

Hyperstimulation of postsynaptic serotonin receptors is the main mechanism postulated to be responsible for SS. Therefore, an understanding for serotonin physiology is necessary to explain the effects of excess serotonin in the body. Serotonin or 5-hydroxytryptamine (5-HT), first isolated by Rapport in 1948, is a neurotransmitter produced by the hydroxylation and decarboxylation of the essential amino acid L-tryptophan.

Serotonin is synthesised in the CNS in the brainstem raphe nuclei and concentrated in the serotonergic pathways in the hypothalamus, thalamus, limbic system, cerebellum, spinal cord, and retina. Serotonin can also be found peripherally in the enterochromaffin cells of the enteral nervous system, platelets, and mast cells.⁹ The role of serotonin centrally is regulation of mood, wakefulness, appetite, and sexual behaviour. Serotonin is also involved in the pathways that control emesis, core body temperature, and pain signals.⁹ Enteral motility, secretions, and vasomotor tone are influenced by serotoninergic activity with the gastrointestinal tract being the largest producer of serotonin in the body. Serotonin also plays an important role in platelet function and coagulation homeostasis.

Serotonin is stored in presynaptic vesicles, and upon release, its effects are mediated by 5-HT receptors. 5-HT receptors are a family of G-protein-coupled receptors, with the exception of 5-HT_3, a ligand-gated Na⁺ and K⁺ channel, that exert excitatory or inhibitory postsynaptic responses through activation of intracellular second messenger cascade. There are seven classes of 5-HT receptors (5HT_1–7) with further subdivisions in each class of receptor (e.g. 5-HT_1a to 5-HT_1f). The classification was based upon their pharmacological and molecular biology properties.¹⁰ Multiple receptors are responsible for the clinical manifestations of SS. The 5-HT_2a receptors have been found to be of particular relevance in animal models or serotonin toxicity. Animal studies have also found high concentrations of noradrenaline (norepinephrine) in the hypothalamus of rats with drug-induced serotonin toxicity, suggesting that noradrenaline also plays an important role in the clinical manifestations of SS.¹¹

Multiple pharmacological agents have been associated with SS. In essence, any drug that alters the synthesis, release, metabolism, or reuptake of serotonin has the potential to cause toxicity (Table 2). Selective serotonin reuptake inhibitors are also of subtypes of cytochrome P450 enzymes, which can lead to the accumulation of pro-serotoninergic drugs (e.g. tramadol, venlafaxine, oxycodone, risperidone, and dextromethorphan). Therefore, the pro-serotoninergic effect can be potentiated when combinations of these medications occur.

Table 2.

Mechanisms and pharmacological agents associated with perioperative SS. MAO,monoamineoxidase; MDMA,3,4methylenedioxymethamphetamine; SNRI, serotonin noradrenaline reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor.

Serotonin synthesis	L-tryptophan
Serotonin metabolism	MAO inhibitor antidepressants (phenelzine, moclobemide, and selegiline); drugs with MAO inhibition activity (methylene blue, hydralazine, and linezolid)
Enhanced release	MDMA (Ecstasy), cocaine, fenfluramine, phenylpiperidine opioids (fentanyl and meperidine), oxycodone, and tramadol
Reuptake inhibition	SSRIs (paroxetine, sertraline, citalopram, and fluoxetine); SNRIs (venlafaxine and duloxetine); tricyclic antidepressants (amitriptyline, imipramine, clomipramine, and desipramine); phenylpiperidine opioids and tramadol; other antidepressants (trazodone); ondansetron and granisetron
Direct serotonin agonist	Triptans (sumatriptan, rizatriptan, and zolmitriptan); ergot alkaloids

Individual variability and genetic differences may explain why some patients develop SS with particular doses and combinations of serotoninergic drugs, whereas others do not. Knock-out mice with disruption in the serotonin transporter (SERT) exhibit a marked susceptibility to developing SS when exposed to serotoninergic agents.¹² Polymorphism in cytochrome isoforms CYP2D6 and CYP3A4 may also explain individual susceptibility to develop SS.¹³

Our literature search found that methylene blue and phenylpiperidine opioid derivatives (e.g. fentanyl and meperidine) were reported in numerous case reports as being associated with the development of SS in the perioperative period.

Methylene blue is a phenothiazine derivative, which can be administered for both therapeutic and diagnostic purposes. Methylene blue has been used for the treatment of severe vasodilatation and methaemoglobinaemia. It is also a medical dye used to enhance the surgical visualisation of anatomical structures, such as the parathyroid glands. Methylene blue is a reversible inhibitor of both isoforms of the monoamine oxidase (MAO) enzyme with high selectivity for MAO-A, which is the isoform responsible for the metabolism of serotonin. Doses as low as 1 mg kg⁻¹ are sufficient to produce inhibition of the MAO-A, and have been reported to be associated with severe serotonin toxicity.¹⁴

Severe serotonin toxicity occurring with overdose of SSRIs or reversible MAO inhibitors, such as methylene blue, as single agents has not been reported in the literature. Nevertheless, the additive effect of an increase in serotonin release and a reduction in extracellular clearance that occurs with the co-administration of SSRIs and reversible MAO inhibitors has resulted in severe life-threatening toxicity.¹⁴ Multiple mechanisms have been postulated to explain the serotonergic effects of opioids and their relationship with SS. Mechanisms include weak serotonin uptake inhibition and increased release of serotonin through stimulation of the 5HT-1_a receptor (e.g. fentanyl and meperidine) and increase of intra-synaptic concentrations of serotonin (morphine, oxycodone, and tramadol). Similar to methylene blue, opioids alone have not been described in cases of SS, but in combination with other serotonergic drugs, such as SSRIs.¹⁵

Ondansetron is a known trigger of SS, as it antagonises the 5HT₃ receptor with a resulting increase in serotonin availability at the 5HT₁ and 5HT₂ receptors, and needs to be considered as a contributor, despite not being commonly implicated in the development of SS by itself.

Characteristics of reported cases of perioperative SS

We conducted a review of case reports that recorded occurrences of SS in the postoperative period to further characterise the clinical presentation of perioperative SS (see Supplementary Table 1). The reports covered a period of time spanning from 1994 to 2018. We reviewed 29 case reports that included 31 patients reported to have suffered SS in the perioperative period.

The median age of patients was 58 with 61% (19 cases) of the cases reported in female and 39% (12 cases) in male patients. Agitation was the most commonly reported clinical presentation, followed by delayed awakening from anaesthesia and confusion. Nystagmus, clonus, and myoclonic jerks were the most commonly reported neuromuscular manifestations. Hyperthermia, tachycardia, and diaphoresis were frequently reported as signs of autonomic instability. The tree map in Figure 2 graphically displays the reported symptoms and signs with reported frequency.

Fig 2.

Tree map graph illustrating the frequency of reporting of signs and symptoms of perioperative SS in case reports.

Cardiac surgery was the surgical specialty with the most reported cases of SS, followed by orthopaedic and general surgery, each with the same number of cases. This could be attributable to the high dose of fentanyl used during cardiac surgery. Head and neck surgery followed in fourth place with parathyroidectomy as the only procedure reported from this group. We reviewed the preoperative medication with serotoninergic activity that patients were receiving, and also the likely triggers of SS reported by the authors of the case reports.

The great majority of patients (65%) were reported as taking an SSRI before their procedure and the development of SS. Other medications in order of frequency were atypical antipsychotics (7%); tricyclic antidepressants (7%); gamma-aminobutyric acid analogues (pregabalin) (5%); amphetamines (5%); and with smaller percentages for MAO inhibitors, ergot alkaloids, sympathomimetic amines (ephedrine), and tramadol.

The most common potential trigger of serotonin toxicity identified by the authors of the case reports was methylene blue, accounting for 52% of reported triggers. Fentanyl was reported as a potential contributor in five cases, accounting for 15% of triggers. Other phenylpiperidine derivatives (meperidine, remifentanil, and dextromethorphan) were also reported as potential triggers. Ondansetron, cyclobenzaprine, oxycodone, and lidocaine were also associated with cases of perioperative SS.

There was only one reported death among the case reports reviewed.¹⁶ The great majority of cases reported full recovery with no lasting sequelae, with only two reports disclosing that patients were discharged to a rehabilitation facility.^17,18 One study reported a patient requiring prolonged ICU stay.¹⁹ It was unclear whether the occurrence of SS was directly responsible for the need for a rehabilitation facility or prolonged ICU stay.

Management

The type and intensity of management depends on the clinical manifestations of SS. Patients manifesting more than mild symptoms should be closely monitored in a high-dependency care environment (Level 2 or 3). Discontinuation of serotonergic agents; supportive treatment; and management of agitation, hyperthermia, and autonomic instability are the mainstays of treatment for all presentations of SS. Supportive care should aim to restore normal physiology in all cases. Aggressive hydration is particularly important in the presence of rhabdomyolysis to avoid acute kidney injury.

Agitation and anxiety are usually managed with benzodiazepines. The use of different agents, including lorazepam, midazolam, and diazepam, have been described. In an animal model of SS, diazepam attenuated the development of hyperthermia.²⁰ However, there is no clinical evidence that favours the administration of any particular benzodiazepine over the others.

Autonomic instability can be severe, and can present with rapid swings in arterial BP and HR. Hypotension should be treated with i.v. fluids, and in more severe cases, directly acting sympathomimetic amines (noradrenaline, adrenaline [epinephrine], or phenylephrine) may be required. Indirectly acting catecholamines, such as dopamine and ephedrine, should be avoided especially when agents that inhibit the activity of MAO are involved. In the presence of MAO inhibition, the effects of indirect amines can be unpredictable.¹

Hypertension and tachycardia appear to be far more prevalent than hypotension in reported cases of perioperative SS. Because of the unpredictable nature of changes in arterial BP and HR in SS, only short-acting agents, such as esmolol and glyceryl trinitrate, are recommended for the treatment of hypertension and tachycardia. Hydralazine should be avoided, as it has significant inhibitory effects on MAO and may exacerbate serotonin toxicity.²¹

Hyperthermia in SS is mostly mediated by muscle hyperactivity, which renders paracetamol ineffective as a treatment. Mild hyperthermia can be treated with topical cooling and benzodiazepine sedation to decrease muscle activity. Severe uncontrolled hyperthermia (temperature >41.1°C) can lead to severe complications, such as rhabdomyolysis, metabolic acidosis, and disseminated intravascular coagulation. Aggressive treatment with sedation, ventilation, and neuromuscular block with non-depolarising agents is necessary to reduce muscle activity and control body temperature. Depolarising NMBAs, such as suxamethonium, should be avoided, as they can result in life-threatening hyperkalaemia brought about by rhabdomyolysis and metabolic acidosis.

Dantrolene, used for the management of malignant hyperthermia, has not been proved to be effective for the management of hyperthermia associated with SS in animal or clinical studies.²²

The literature suggests that specific serotonin antagonists should be given when supportive care and sedation fail to control symptoms, and in severe presentations of SS.¹ Cyproheptadine is an H₁ receptor antagonist that also has 5-HT_1A and 5HT_2A receptor antagonist activity. Case reports have found cyproheptadine to be useful in providing symptomatic relief in cases of serotonin toxicity.^23,24 Our review of the literature also found multiple cases where cyproheptadine was administered for the management of SS occurring during the perioperative period. The recommended regimen is an initial dose of 12 mg with further 2 mg h⁻¹ doses until clinical improvement is observed.¹ A potential limitation for the use of cyproheptadine in the perioperative period is that it is only available in oral tablet form. However, the tablet can be crushed and administered via a nasogastric tube.

Antipsychotic agents with serotonin 5-HT_2A receptor blockade, such as chlorpromazine and olanzapine, have been used as alternative therapies for SS, but there are little data regarding their effectiveness and safety in the perioperative period.

Dexmedetomidine is an α₂-adrenergic receptor agonist used for sedation in anaesthesia and intensive care. Several of its pharmacological features make dexmedetomidine attractive for the treatment of SS. Alpha_-2 receptor stimulation in serotonergic neurones inhibits the release of serotonin, and animal studies suggest that this mechanism explains the superiority of dexmedetomidine over midazolam for the treatment of serotonin toxicity in animal models.²⁵

Dexmedetomidine produces effective sedation and anxiolysis without significant respiratory depression. The autonomic instability that characterises more severe forms of SS is mediated by excess serotonergic and noradrenergic activity. Dexmedetomidine decreases central sympathetic output, and can effectively control hypertension and tachycardia.

A recent case series described the effective use of dexmedetomidine for the management of severe SS refractory to conventional therapies. The use of dexmedetomidine was associated with effective stabilisation of the autonomic system with improvement in tachycardia, hyperthermia, and myoclonus.²⁶ Figure 3 contains an algorithm of suggested treatment for perioperative SS.

Fig 3.

Suggested treatment algorithm for perioperative SS.

Prevention

The increasing prevalence of patients receiving serotonergic medication requires preoperative screening of patients at risk of serious drug interactions during the perioperative period. Discontinuation of serotonergic agents before surgery is not always practical or free from risk. In the case of SSRIs, serotonin–noradrenaline reuptake inhibitors, and MAO inhibitors, abrupt cessation can be associated with unpleasant withdrawal symptoms and exacerbation of psychiatric conditions.²⁷ Therefore, the risks and benefits of continuing psychotropic medications during the perioperative period should be carefully considered. Perioperative SS is a relatively rare condition; therefore, psychotropic agents should not be routinely discontinued. Instead, prevention should focus on avoidance of known triggers and early identification of signs or symptoms of serotonin toxicity.

There must be a clear plan for managing patients at risk of adverse drug interactions presenting for surgery, where exposure to methylene blue is likely for diagnostic or therapeutic purposes. A discussion between the patient, surgeon, anaesthetist, and primary care or psychiatry team should take place. Planned cessation of psychotropic medication may be considered. Alternatively, other medical dyes, such as indigo carmine or indocyanine green, may be used instead.

The risk of end-organ damage from severe vasoplegia should be weighed against the risk of SS when methylene blue is administered for vasodilatory shock in patients receiving serotonergic medications. Finally, the importance of raising patients’ awareness and education about the risk of perioperative drug interactions cannot be overstated.

Declaration of interest

The authors declare that they have no conflicts of interest.

MCQs

The associated MCQs (to support CME/CPD activity) will be accessible at www.bjaed.org/cme/home by subscribers to BJA Education.

Biographies

Ashish Bartakke FRCA EDAIC is a senior clinical fellow in cardiothoracic anaesthesia and critical care who has special interests in cardiothoracic anaesthesia, cardiopulmonary exercise testing, and pain management.

Carlos Corredor FRCA FFICM is a consultant in cardiothoracic anaesthesia and intensive care whose interests include anaesthesia for major aortovascular surgery, mechanical circulatory support, and perioperative echocardiography.

Adriaan van Rensburg FCA (SA) MMED (ANES) MD FRCPC is an associate professor in anaesthesiology at the University of Toronto. Dr van Rensburg has clinical and research interests in cardiovascular and transplant anaesthesia and perioperative ultrasound. He is the Chair for the annual meeting of the Canadian Anaesthesiologists' Society.

Matrix codes: 1A02, 2A06, 3I00

Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1