J Psychopharmacol OnlineFirst, published on February 8, 2010 as doi:10.1177/0269881109359098
Review
CNS toxicity involving methylene blue: the exemplar for understanding and predicting drug interactions that precipitate serotonin toxicity
Journal of Psychopharmacology 0(00) 1–8 ! The Author(s) 2010 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0269881109359098 jop.sagepub.com
P Ken Gillman
Abstract Methylene blue has only recently been noted to cause severe central nervous system toxicity. Methylene blue is used for various conditions, including, intravenously, in methemoglobinemia, vasoplegia and as an aid to parathyroidectomy (at doses of 1–7.5 mg kg1). This review of the current evidence concludes that 13 of 14 of the reported cases of CNS toxicity were serotonin toxicity that met the Hunter Serotonin Toxicity Criteria. That has important preventative and treatment implications. Serotonin toxicity is precipitated by the monoamine oxidase inhibitor (MAOI) property of methylene blue interacting with serotonin reuptake inhibitors. Serotonin toxicity is reviewed, using the lessons inherent in the methylene blue story and experience, to illustrate how the mechanisms and potency of serotonergic drugs interact to determine severity. Recent human data showed that an intravenous dose of only 0.75 mg kg1 of methylene blue produced a peak plasma concentration of 500 ng ml1 (1.6 mM), indicating that the concentration in the central nervous system reaches a level that inhibits monoamine oxidase A. That is consonant with the actual occurrence of severe serotonin toxicity in humans at the dose of only 1 mg kg1. It seems that all proposed uses of methylene blue entail levels that block monoamine oxidase, so cessation of serotonin reuptake inhibitors should be very carefully considered before using methylene blue.
Keywords methylene blue, monoamine oxidase inhibitor, serotonin reuptake inhibitor, serotonin syndrome, serotonin toxicity
Introduction Methylene blue (MB), now called methylthioninium chloride, has recently been noted to cause central nervous system (CNS) toxicity, variously described as encephalopathy, CNS toxicity, serotonin syndrome (SS) or serotonin toxicity (ST), but seemingly only when given to patients already on (selective) serotonin reuptake inhibitors ((S)SRIs). See Table 1. The extent of the use of MB is unknown, much being ‘off-label’. There are reviews of its intravenous (i.v.) use in methemoglobinemia (McRobb and Holt, 2008), parathyroidectomy (Dudley, 1971; Han et al., 2007) and vasoplegia (Del Duca et al., 2009; Stawicki et al., 2008). Also, there are (March 2009) 13 clinical trials involving MB registered at ClinicalTrials.gov. These involve, inter alia, dementia, bipolar disorder and malaria. A warning has been issued by the UK Medicines and Healthcare Products Regulatory Agency (MHRA) (MHRA, 2008) that refers to CNS toxicity. It does not discuss either ST or MB’s monoamine oxidase inhibitor (MAOI) potency. As a result of observations and predictions concerning MB toxicity (Gillman, 2006d) work was initiated which has demonstrated it to be a very potent MAOI in vitro. MB’s Ki for MAO-A is 27 nM (Ramsay et al., 2007). Recent evidence indicates tissue levels are sufficient to produce full MAO-A inhibition in humans at even the lowest recommended dose (1–2 mg kg1 intravenously for methemoglobinemia). Recent reviews of the pharmacology of MB discuss the evidence in relation to ST (Stanford et al., 2009) and other aspects of its pharmacology
(Oz et al., 2009), including inhibition of inducible and constitutive nitric oxide synthase and the formation of L-citrulline.
The intricacies of serotonin toxicity Serotonin toxicity is a moderately complex topic. This review uses the history of the emergence of MB toxicity to illustrate how to understand ST, and the probability of the degree of severity with various drug combinations, which are represented in the figures and tables. These set out in detail the interaction relationships between the relevant classes of drugs involved (Figure 1), and illustrate why weakly serotonergic drugs like mirtazapine are not a problem for precipitating ST (Table 1). A key point is the spectrum concept of ST, which highlights that it is a synaptic 5-HT concentration related phenomenon; this spectrum concept has been elaborated, combining animal and human data, to show the extent to which various classes of drugs, and their various combinations, elevate serotonin (Gillman, 2006a). For other aspects of ST readers are referred to selected reviews, especially because there are reviews that perpetuate outdated information, and many poor case reports (Gillman, 2006c). There have been
PsychoTropical Research, Bucasia, Queensland, Australia. Corresponding author: Dr P Ken Gillman, PO Box 86, Bucasia, Queensland 4750, Australia. Email:
[email protected]
2
Journal of Psychopharmacology 0(00)
significant developments in the quantification of the frequency and severity of ST risk with different classes of drugs when taken in therapeutic doses or overdose (Gillman, 1998, 2005a, 2006a,b; Isbister et al., 2003; Whyte and Dawson, 2002) and also in the definition of the toxidrome of ST in both animals and humans (Dunkley et al., 2003; Isbister and Buckley, 2005; Whyte, 2004). Good descriptions of the pathophysiology (Isbister and Buckley, 2005), clinical presentation (Boyer and Shannon, 2005), management and treatment (Gillman, 1999; Isbister et al., 2007; Table 1. Clinically active serotonin reuptake inhibitors* Paroxetine sertraline fluoxetine fluvoxamine citalopram escitalopram Venlafaxine desvenlafaxine milnacipran duloxetine sibutramine Clomipramine imipramine (but not other TCAs) Tramadol pethidine methadone dextromethorphan dextropropoxyphene pentazocine (but not other opioids) Chlorpheniramine brompheniramine (but not other anti-histamines) *This list is derived and updated from tables and data on the author’s website, and analysed and discussed in (Gillman, 1998, 2005a, 2006a, 2006b); consult references for the rationale of exclusion of particular drugs, for example, mirtazapine and most TCAs, and inclusion of others, for example, selected opioid analgesics. In the context of this paper the terms serotonin reuptake inhibitor (SRI) and selective serotonin reuptake inhibitor (SSRI) are effectively synonymous; but the range of drugs that have significant SRI properties is structurally diverse, encompassing: clomipramine (a TCA), the SSRIs, venlafaxine (a serotonin and noradrenaline reuptake inhibitor (SNRI)), sibutramine and some opioid analgesics. For all these drugs it is only their SRI potency that matters in relation to the propensity to precipitate ST.
Sun-Edelstein et al., 2008) are available. It is also important to appreciate that overdose of particular classes of drugs often produces a clinical picture of signs and symptoms that are quite characteristic. When overdoses have been ingested this allows toxicologists to distinguish the offending substance, sometimes with a high degree of reliability. Such clinical presentations have become designated as toxidromes (a contraction of toxicity and syndrome), and are now well described for a number of CNS drugs, for example anti-muscarinics, sympathomimetics, sedatives and SSRIs. Serotonin toxicity is a specific well delineated toxidrome that can be diagnosed with accuracy and confidence. It is not, as has frequently been stated, a diagnosis of exclusion that can only be made when other pathologies have been ruled out (Gillman, 2005b). That notion has caused much confusion, especially in relation to the differential diagnosis between NMS and ST, which is usually straightforward, because the two syndromes are quite different (Whyte, 1999). An incorrect diagnosis of NMS may be made because prior ingestion of an anti-psychotic is thought to ‘trump’ a diagnosis of ST: prior administration of anti-psychotics is not a reason to suspend a diagnosis of ST, if serotonergic drugs have been ingested and characteristic symptoms are present (Gillman, 2005b). Figures 1 and 2 represent the classes of drugs (and combinations) which are capable of precipitating ST. In usual treatment settings involving therapeutic drugs, only an SRI plus an MAOI is likely to produce severe ST (see Figures 1 and 3,
MAOI-B (3) Selegiline, and Furazolidone?
ST reaction Increasing severity
MAOI-A (2) Moclobemide, Linezolid, MB
MAOI-A+B (1) Tranylcypromine, Phenelzine MAOIs
5-HT Releasers
SRIs
(1) MDMA, Fenfluramine
(1) SSRIs, Venlafaxine, Clomipramine (2) Imipramine, Tramadol, Pethidine (3) Fentanyl, Chlorpheniramine
SRIs
Attenuated effect No ST
REL
(2) Amphetamine, Methamphetamine* (3) Phentermine
Figure 1. The serotonin toxicity triangle. The diagram represents interactions that precipitate serious ST (based on ‘usual’ clinical doses). Potency is indicated as (1) strong, serious interactions probable, (2) medium, interactions possible, (3) weak, interactions unlikely. A supra-therapeutic dose of a level 2 drug may elevate the risk to level 1. *Serious reactions and deaths do occur from hypertension, but probably not from ST. Source: modified from Gillman, http://www.psychotropical.com/
3
Gillman and Table 1 for a list of (S)SRIs). SSRIs by themselves, even if taken in overdose, do not usually precipitate severe ST (Figure 3). This has been confirmed by analysis of large prospective case series of thousands of overdoses (Whyte, 2004; Whyte et al., 2003) and the side-effects encountered in trials, and case reports (Gillman, 2006a). Rare case reports exist suggesting more severe ST with SSRIs. The spectrum concept of ST also conforms with animal and human data concerning the degree to which drugs elevate serotonin. It correctly predicts those drugs that are unable to precipitate ST, such as those TCAs that are insufficiently potent as SRIs to cause serotonergic side-effects or serotonergic interactions with MAOIs (Gillman, 2006a). A similar lack of serotonergic effects (at either therapeutic doses or after overdose, or on interaction with MAOIs) is found with trazodone, nefazodone, mianserin and mirtazapine (Biswas et al., 2003; Gillman, 2003; 2004; Isbister and Hackett, 2003; Isbister and Whyte, 2003; LoVecchio et al., 2008; Schaper et al., 2002; Waring et al., 2007). A clear understanding of these interactions is heuristically valuable in revealing the mechanism of action, and potency, of many psychotropic drugs and allows confident predictions about them to be made (for details and discussion, see Gillman, 2006a): in this instance the prediction, now confirmed, that MB is an MAOI. The occurrence of severe ST in patients on an SSRI (see Figure 3) indicates that an MAOI is likely to have been co-ingested (Gillman, 2006a,d). The occurrence of definite and severe ST with an SRI and a putative MAOI may be considered as an in vivo assay of the CNS activity and
potency of the MAOI. This interaction is informative because extrapolation from animal work on MB, or any other drug like linezolid (which has some MAOI potency, see Figure 1), cannot generate a precise prediction for the dose range for effective MAO inhibition in humans, because there are a number of often imprecisely known intervening variables (e.g. volume of distribution, in vivo metabolism, brain penetrance etc.). There are few, if any, other practical and reliable ways of determining an MAOI’s in vivo CNS potency, other than establishing the occurrence of definite ST. The figures illustrate why the initial report by Rosenbaum, and the original report by Stanford, immediately raised the suspicion that an unidentified MAOI might be involved. Firstly, the clinical features allowed a confident diagnosis of ST; secondly, ST of that severity is extremely improbable with an SSRI alone; thirdly, there is no other known class of therapeutic drug that is capable of inducing that degree of severity of ST; only a combination of MAOI and SRI will do so.
Aims This review seeks to: (a) consolidate the evidence about the nature of the interaction between SRIs and MB by applying established diagnostic criteria to all known reports, because many cases of CNS toxicity with MB exhibit features suggesting ST; and (b) use this example to explain and illustrate the key aspects of ST and the drug–drug interactions of consequence that relate to ST. If the ‘CNS toxicity’ with MB is ST, that has important predictive, preventative and treatment implications.
If a potent serotonergic agent has been taken (see fig. 1, STT), ST is established at a high confidence level (see HSTC for sensitivity/specificity) if any one ofthe signs and symptoms 1–5 below are present:
(1) Spontaneous clonus
(3) Ocular clonus and either agitation or diaphoresis (4) Inducible clonus and either agitation or diaphoresis
Serotonin toxicity
(2) Tremor and hyperreflexia
(5) Hypertonia, temperature >38°C, and either ocular clonus or inducible clonus
Figure 2. Algorithm for diagnosis of serotonin toxicity (HSTC). This algorithm is referred to as the Hunter Serotonin Toxicity Criteria and is derived from (Dunkley et al., 2003) and (Isbister et al., 2007) where the sensitivity and specificity are elaborated. In clinical practice the neuromuscular signs of generalized, or ocular, clonus, tremor and lower limb hyper-reflexia are strongly indicative of the ST: their occurrence in the setting of serotonergic drug overdose/combinations establishes the diagnosis. Note: in severe cases (almost all of which involve an MAOI plus an SRI) hyperthermia is often severe (>39 C) and muscle rigidity is so intense it overrides other neuromuscular signs. Source: modified from Gillman, http://www.psychotropical.com/
4
Journal of Psychopharmacology 0(00)
Method All published reports of all forms of toxicity with MB that were located by searching the Medline database, and all cases cited in bibliographies of reports were checked for indications of CNS toxicity of any nature. All cases with signs suggestive of ST are summarized and analysed and rated for diagnostic certainty as either definite, probable, or possible, and also for severity, using the Hunter Serotonin Toxicity Criteria (HSTC) (Dunkley et al., 2003; Isbister et al., 2007). These criteria have been shown to be of high sensitivity (84%) and specificity (97%) for ST. The reported cases have all involved patients who were on SRIs prior to being administered MB. No reports of CNS toxicity, or possible ST, with MB alone were located by this author, nor have any been suggested or reported by others (except Stanford, see below). This is as predicted by the spectrum concept of ST, because even overdoses of selective reversible MAO-A inhibitors (RIMAs) do not precipitate ST (see Figure 3), as exemplified moclobemide (Isbister et al., 2003). Therefore, also, even large i.v. doses of MB would not, in the absence of SRIs, be expected to
precipitate ST. Using the HSTC is a stringent definition and test of caseness as applied by experienced toxicologists (see Figure 2 for HSTC diagnostic flow chart). In the context of a general surgery setting, and where anesthesia is known to suppress key ST symptoms, false negatives are inevitable.
Results Currently known cases of methylene blue related central nervous system toxicity, or possible serotonin toxicity Rosenbaum first suggested that the symptoms observed in a patient could be ST resulting from an interaction between MB and an SSRI, in a report posted on the internet in early 2006 (Rosenbaum, 2006). He also noted the similarities of that case to a previous report (Martindale and Stedeford, 2003) that did not mention ST or serotonin, but which also involved MB and an SSRI. That information suggested that MB was an MAOI because the cases appeared to be ST, and there was no other candidate as the putative MAOI except MB.
Severity of ST with classes of drugs & combinations
Clinical signs
Temperature 39°C
Sustained clonus
Agitation Restlessness Diaphoresis Mood dysphoria
Hyperreflexia Tremor
Nausea, diarrhoea, insomnia, nervousness
MAOI-AB MAOI-A + + SRI SRI
MAOIAB alone Overdose
MAOI-A alone
REL
SRIs
Serotonergic side effects
Mydriasis
Inducible clonus
ST moderate
5-HT levels and symptoms severity
Confusion
ST severe
Rigidity
Mirt
High therapeutic dose
Figure 3. Degree of serotonin toxicity typical with drugs. The bars depict ‘overdose’ and ‘high therapeutic dose’, which reflects the full effect of the drug as used for refractory illness. MAO-AB: non-selective irreversible inhibitors, for example, tranylcypromine and phenelzine, MAO-A: moclobemide, REL: releasers like MDMA (the only one in therapeutic use, fenfluramine, has been withdrawn), SRI: all serotonin reuptake inhibitors, selective and non-selective and SNRIs, MIRT: mirtazapine. Amphetamine is a releaser, mainly of dopamine not 5-HT. The typical sequence of signs at their approximate level of initial appearance is shown. Note the clear ‘ceiling’ effects: for example, SRIs alone rarely or never induce temperatures of >38.5 C. Moclobemide overdose does not induce ST of even moderate degree, mirtazapine hardly has serotonergic side effects, and no serotonergic toxicity. The data on which this figure is based has been checked with the clinical signs recorded in thousands of cases documented in the HATS database, with the assistance of Professor Ian Whyte. Source: modified from Gillman, http://www.psychotropical.com/
5
Gillman A review of these three cases was published (Gillman, 2006d) and followed up by the initiation of the in vitro monoamine oxidase (MAO) assay of MB; that work yielded the data unequivocally demonstrating that MB is a potent selective reversible inhibitor of monoamine oxidase A (MAO-A) at nanomolar concentrations (Ramsay et al., 2007), confirming several older publications suggesting MB’s antidepressant activity and ability to raise serotonin (Aeschlimann et al., 1996; Naylor et al., 1987; Wegener et al., 2000). The MHRA subsequently commented on further published reports of MB/SRI and CNS toxicity, and cases on their files (MHRA, 2008, 2009), although there was no discussion
of its MAOI potency, or that some of the cases were typical ST (and one was fatal). The 14 reports currently identified (as of March 2009) are: (1) Stanford and Stanford (1999); (2) Martindale and Stedeford (2003); (3) Bach et al. (2004); (4) Majithia and Stearns (2006); (5) Mathew et al. (2006); (6) Rosenbaum (2006); (7) Khan et al. (2007); (8) Mihai et al. (2007); (9) Ng et al. (2008); (10) Shanmugam et al. (2008); (11) Khavandi et al. (2008); (12) Schwiebert et al. (2009); (13) Rowley et al. (2009); and (14) Pollack et al. (2009). These are tabulated and evaluated below (Table 2): this list includes Stanford’s case: the original report (Stanford and
Table 2. Cases of possible ST: ratings of certainty of diagnosis and severity. All reports are of i.v. use of MB for parathyroidectomy, except (10) cardiac, and (12) ureteric surgery. MB is usually given over about one hour pre/peri-operatively, but the details usually not reported. Clonus and hyper-reflexia are strongly indicative of ST; see Figure 2 for key ST symptoms. *This is reflected by ‘probable/definite’ designation for ‘Certainty of diagnosis’ (e. g. cases 3, 4, 7, 9), that is, if the symptom, as described, is accepted as representing clonus then the rating would be definite ST. For example, (7) ‘Jerky movement of all four limbs’ probably represents spontaneous clonus. A ‘clinical diagnosis’ would mean all probable and definite cases would be treated as ST: that is, 12/13. Note: temperatures were reported in five cases (5, 6, 10, 11, 13). Case (8) Mihai is only justifiable as a ‘possible’ because it is post-operative and there is a probability of suppression of typical ST symptoms by anesthesia, and there is no other likely etiology. The i.v. MB dose was 5–7.5 mg kg1 in all cases except Mihai (1.75 mg kg1), Shanmugam (2 mg kg1) and Schweibert (1 mg kg1) Certainty of diagnosis of ST
Severity of clinical state
(1) Stanford (Stanford and Stanford 1999; Stanford et al., 2009)
Definite
Severe
(2) Martindale (Martindale and Stedeford, 2003)
Definite
Severe
(3) Bach (Bach et al., 2004)
Probable
Moderate
(4) Majithia (Majithia and Stearns, 2006) (5) Mathew (Mathew et al., 2006) (6) Rosenbaum (Rosenbaum, 2006)
Probable/ definite Definite Definite
Severe
(7) Khan (Khan et al., 2007)
Probable/ definite Possible
Case reference (chronological order)
(8) Mihai (Mihai et al., 2007)
Severe Severe Severe N/A
(9) Ng (Ng et al., 2008)
Probable/ definite
Severe
(10) Shanmugam (Shanmugam et al., 2008)
Definite
Severe
(11) Khavandi (Khavandi et al., 2008) (12) Schweibert (Schwiebert et al., 2009) (13) Rowley (Rowley et al., 2009)
Probable/ definite Definite
Moderate Severe
Definite
Severe
(14) (Pollack et al., 2009)
Definite
Moderate
Source: modified from Gillman http://www.psychotropical.com/
Reported symptoms and signs. Comments. SRI before surgery MB used, but not mentioned in original report. ‘agitation, confusion, brisk reflexes, ankle clonus, hypertension, temperature 38 C.’ Paroxetine. ‘Rotational nystagmus’ (represents ocular clonus). Rigid, jerky movements and increased tone of all four limbs, very agitated, sweating profusely. Fluoxetine. ‘A suggestion of clonus, with forced dorsiflexion of feet.’* Paroxetine. ‘Nystagmus’, was very probably ocular clonus.* Paroxetine. Venlafaxine. MH queried, tremor, agitation, temperature 40 C. Citalopram. The first recognized case. ‘Agitated, tachycardic and diaphoretic. . .lower extremity rigidity.’ Temperature 38.3 C. Sertraline. ‘Confused, agitated, jerky movement of all four limbs.’* Clomipramine. ‘Agitated and restless. . .unable to speak no response to verbal command. . .no limb weakness. . . no focal neurological signs.’ Paroxetine. ‘Agitated, moving all limbs purposelessly. . .increased tone in all limbs’ and ‘rapid, fluid eye movements’ represents probable ocular clonus.* Paroxetine. ‘Confused and agitated. . .temperature 40 C, fine tremors, shivering, hyperactive reflexes, hypertonicity.’ Fluvoxamine and paroxetine. ‘Agitated and restless’, ‘myoclonic movements of the lower limbs, brisk reflexes’, Temperature 37.5 C. Citalopram. ‘Confusion, agitation, aphasia, ocular clonus, mydriasis, hyperreflexia and arterial hypertension.’ MB dose 1 mg kg1. Paroxetine. ‘End-gaze nystagmus (probably ocular clonus), hypertonia, diaphoresis, mildly agitated. . .reflexes were symmetric and normal’. 38.1 C’ (rectal, not warmed intra-operatively). Duloxetine. Agitation, diaphoresis, lower extremity hyperreflexia, opsoclonus. Citalopram.
6 Stanford, 1999) did not mention MB but it was subsequently established that it had been used; a correction has been published (Stanford et al., 2009). This case is another example of the predictive power of the spectrum concept of ST, because the unexpectedly severe symptoms, in the apparent absence of an MAOI, made the case so exceptional that a re-checking of the original case report data was initiated (as part of this review), leading to the post hoc discovery that MB had been used. This number of probable and definite cases of serious ST (Table 2), with not a single similar instance in the absence of SRIs, makes it probable that it is a pharmacodynamic interaction between SRIs and MAOIs.
Surgical retrospective case series There are two retrospective case series from units where MB was always used for parathyroid surgery that have reported on ‘CNS toxicity’: Kartha (Kartha et al., 2006) and Sweet (Sweet and Standiford, 2007). These two series totaled 325 patients who all had MB; 45/325 were on SSRIs pre-operatively: of those, 17/45 showed CNS toxicity. No case exhibited toxicity who had not been taking SSRIs pre-operatively (N ¼ 280). There are insufficient clinical details reported to allow a diagnosis to be reached that is more precise than CNS toxicity.
Discussion The surgical retrospective case series are helpful only insofar as they indicate the frequent occurrence of ‘CNS toxicity’ exclusively with the combination of MB and SRIs, but not in a single instance in the 280 cases administered MB but without previous SRI treatment. The difference between 17/ 45 (MB + SRI) vs. 0/280 (MB alone) is highly significant (Genstat, two-sample, two-tailed binomial test of two proportions, p ¼ 0.001). The individual case reports all contain histories with sufficient detail to permit a probable or definite diagnosis of ST. It is thus highly probable that these cases represent a pharmacodynamic interaction between MB and SRIs. Nine of the 14 case reports rated as definite using the HSTC, of which 7/9 were severe; and a further 5/14 rated as probable ST. Only 1/13 was only a possible ST. That confirms, beyond reasonable doubt, that this combination of MB and an SRI, as predicted, does precipitate severe ST. This is important because that fact constitutes very strong evidential corroboration that MB does indeed attain sufficient concentration in the brain to inhibit MAO-A, at doses at least as low as 1 mg kg1. Reviews of recent data on animal and human plasma and tissue levels of MB accord with the above (Burhenne et al., 2008; Oz et al., 2009; Stanford et al., 2009; Walter-Sack et al., 2009). This indicates that warnings (like the one issued by the MHRA (2009)) that imply that smaller doses are safer are probably inappropriate. MB is undergoing something of a renaissance and is currently being used for various conditions in 13 registered trials (see ClincalTrials.gov) in: for example, malaria (Meissner et al., 2006; Zoungrana et al., 2008), bipolar disorder and dementia (Wischik et al., 2008). The MAO-A inhibition may partly or wholly explain some of MB’s putative benefits
Journal of Psychopharmacology 0(00) in bipolar disorder and dementia. The use of MB needs to take account of its most potent property of MAO inhibition, especially because of widespread use of (S)SRIs, because the resultant drug–drug interaction risks have already been predicted and quantified. The failure by the UK MHRA (MHRA, 2008, 2009) to mention the conclusive evidence that MB is an effective MAOI, in relevant doses in humans, and that the clinical picture is frequently typical ST (not non-specific ‘CNS toxicity’) denies doctors, and patients, information they need to know. It is also helpful to understand, from the explanations elaborated herein, why the MHRA’s inclusion of, inter alia, mirtazapine and bupropion (MHRA, 2008) as having significant interaction potential is erroneous. The contradictory conclusions of the USA FDA and the UK MHRA in their respective assessments of ST with the triptans, and ST with methylene blue, is striking (Gillman, 2009). The MHRA fail to warn of ST despite strong evidence, whereas the FDA do warn of fatal toxicity where no substantive evidence exists, even of mild toxicity. This contrast may be accounted for by a difficulty in comprehending drug–drug interaction data in the explanatory context of the spectrum concept of ST. The pharmacology of MB is discussed in more detail by Stanford et al. (2009) and Oz et al. (2009). The most recent human data (Walter-Sack et al., 2009) established MB plasma concentrations after an i.v. dose of only 0.75 mg kg1 (usual therapeutic doses are 1–7 mg kg1) as being a peak of 500–600 ng ml1 (1.6 mM) in plasma (Burhenne et al., 2008). The mean T1/2 was 13.6 3.7 h. No other therapeutically used drugs are able to cause this interaction of severe ST, other than SRIs combined with MAOIs (Figures 1 and 2): the only class of drugs other than SRIs that produces severe ST reactions with MAOIs is ‘releasers’, that is, indirectly acting agonists that release neuro-transmitter from the pre-synaptic stores, for example, MDMA, ecstasy (3,4-methylenedioxymethamphetamine) and fenfluramine, which is no longer used. SRIs attenuate the effect of releasers by blocking their uptake into the pre-synaptic neuron, so there is no potential for ST with that combination (Mechan et al., 2002; Rothman and Baumann, 2003). The high probability that MB is a clinically potent MAOI at i.v. doses as low as 1 mg kg1 will assist decisions that need to be made prior to surgery, and other procedural and treatment uses of MB. If SRIs are being taken careful consideration needs to be given to ceasing them prior to MB use, following the accepted and recommended drug washout intervals already established for the MAO-A inhibitor moclobemide. The history of events and discoveries surrounding MB, documented in this review, demonstrates both the predictive capacity of the spectrum concept of ST, and the need and usefulness of an accurate diagnosis of ST. This is facilitated by a clear knowledge of the specificity of the toxidrome of ST, and the diagnostic significance of key signs, especially of clonus. It is imperative to test for its presence carefully. MB may be thought of as a dye, rather than a drug, which may have resulted in its role in interactions being overlooked. There are other implications for prevention and treatment of ST, detailed in the recommended references provided in the introduction, the further elaboration of which is beyond the scope of this review.
Gillman Acknowledgements I acknowledge the expertise, support and help of my wife Isobel, who maintained the indispensable computers and programs, and Professor Ian Whyte for assistance, especially with verifying the accuracy of Figure 3 from the data on several thousand cases in the HATS database.
References Aeschlimann C, Cerny T, Kupfer A (1996) Inhibition of (mono)amine oxidase activity and prevention of ifosfamide encephalopathy by methylene blue. Drug Metab Dispos 24: 1336–1339. Bach KK, Lindsay FW, Berg LS, Howard RS (2004) Prolonged postoperative disorientation after methylene blue infusion during parathyroidectomy. Anesth Analg 99: 1573–1574. Biswas PN, Wilton LV, Shakir SA (2003) The pharmacovigilance of mirtazapine: results of a prescription event monitoring study on 13554 patients in England. J Psychopharmacol 17: 121–126. Boyer EW, Shannon M (2005) The serotonin syndrome. N Engl J Med 352: 1112–1120. Burhenne J, Riedel KD, Rengelshausen J, et al. (2008) Quantification of cationic anti-malaria agent methylene blue in different human biological matrices using cation exchange chromatography coupled to tandem mass spectrometry. J Chromatogr B Biomed Sci App 863: 273–282. Del Duca D, Sheth SS, Clarke AE, Lachapelle KJ, Ergina PL (2009) Use of methylene blue for catecholamine-refractory vasoplegia from protamine and aprotinin. Ann Thorac Surg 87: 640–642. Dudley NE (1971) Methylene blue for rapid identification of the parathyroids. Br Med J 3: 680–681. Dunkley EJC, Isbister GK, Sibbritt D, Dawson AH, Whyte IM (2003) Hunter Serotonin Toxicity Criteria: a simple and accurate diagnostic decision rule for serotonin toxicity. Q J Med 96: 635–642. Gillman PK (1998) Serotonin syndrome: history and risk. Fundam Clin Pharmacol 12: 482–491. Gillman PK (1999) The serotonin syndrome and its treatment. J Psychopharmacol 13: 100–109. Gillman PK (2003) Mirtazapine: unable to induce serotonin toxicity? Clin Neuropharmacol 26: 288–289. Gillman PK (2004) Mirtazapine: not a dual action antidepressant? Aust NZ J Psychiatry 38: 266–267. Gillman PK (2005a) Monoamine oxidase inhibitors, opioid analgesics and serotonin toxicity. Br J Anaesth 95: 434–441. Gillman PK (2005b) Understanding toxidromes: serotonin toxicity: a commentary on Montanes-Rada et al. J Clin Psychopharmacol 25: 625–626. Gillman PK (2006a) A review of serotonin toxicity data: implications for the mechanisms of antidepressant drug action. Biol Psychiatry 59: 1046–1051. Gillman PK (2006b) A systematic review of the serotonergic effects of mirtazapine in humans: implications for its dual action status. Hum Psychopharmacol 21: 117–125. Gillman PK (2006c) Extracting value from case reports: lessons from serotonin toxicity. Anaesthesia 61: 419–422. Gillman PK (2006d) Methylene blue implicated in potentially fatal serotonin toxicity. Anaesthesia 61: 1013–1014. Gillman PK (2009) Triptans, serotonin agonists, and serotonin toxicity. Headache [Epub ahead of print]. Available at http:// www.ncbi.nlm.nih.gov/pubmed/19925619?dopt=Citation (accessed November 17, 2009). Han N, Bumpous JM, Goldstein RE, Fleming MM, Flynn MB (2007) Intra-operative parathyroid identification using methylene blue in parathyroid surgery. Am Surg 73: 820–823.
7 Isbister GK, Buckley NA (2005) The pathophysiology of serotonin toxicity in animals and humans: implications for diagnosis and treatment. Clin Neuropharmacol 28: 205–214. Isbister GK, Hackett LP (2003) Nefazodone poisoning: toxicokinetics and toxicodynamics using continuous data collection. J Toxicol Clin Toxicol 41: 167–173. Isbister GK, Whyte IM (2003) Adverse reactions to mirtazapine are unlikely to be serotonin toxicity. Clin Neuropharmacol 26: 287–288. Isbister GK, Buckley NA, Whyte IM (2007) Serotonin toxicity: a practical approach to diagnosis and treatment. Med J Aust 187: 361–365. Isbister GK, Hackett LP, Dawson AH, Whyte IM, Smith AJ (2003) Moclobemide poisoning: toxicokinetics and occurrence of serotonin toxicity. Br J Clin Pharmacol 56: 441–450. Kartha SS, Chacko CE, Bumpous JM, Fleming M, Lentsch EJ, Flynn MB (2006) Toxic metabolic encephalopathy after parathyroidectomy with methylene blue localization. Otolaryngol Head Neck Surg 135: 765–768. Khan MA, North AP, Chadwick DR (2007) Prolonged postoperative altered mental status after methylene blue infusion during parathyroidectomy: a case report and review of the literature. Ann R Coll Surg Engl 89: W9–11. Khavandi A, Whitaker J, Gonna H (2008) Serotonin toxicity precipitated by concomitant use of citalopram and methylene blue. Med J Aust 189: 534–535. LoVecchio F, Riley B, Pizon A, Brown M (2008) Outcomes after isolated mirtazapine (Remeron) supratherapeutic ingestions. J Emerg Med 34: 77–78. McRobb CM, Holt DW (2008) Methylene blue-induced methemoglobinemia during cardiopulmonary bypass? A case report and literature review. J Extra Corpor Technol 40: 206–214. Majithia A, Stearns MP (2006) Methylene blue toxicity following infusion to localize parathyroid adenoma. J Laryngol Otol 120: 138–140. Martindale SJ, Stedeford JC (2003) Neurological sequelae following methylene blue injection for parathyroidectomy. Anaesthesia 58: 1041–1042. Mathew S, Linhartova L, Raghuraman G (2006) Hyperpyrexia and prolonged postoperative disorientation following methylene blue infusion during parathyroidectomy. Anaesthesia 61: 580–583. Mechan AO, Esteban B, O’Shea E, Elliott JM, Colado MI, Green AR (2002) The pharmacology of the acute hyperthermic response that follows administration of 3,4-methylenedioxymethamphetamine (MDMA, ‘ecstasy’) to rats. Br J Pharmacol 135: 170–180. Meissner PE, Mandi G, Coulibaly B, et al. (2006) Methylene blue for malaria in Africa: results from a dose-finding study in combination with chloroquine. Malar J 5: 84. MHRA (2008) Methylthioninium chloride (methylene blue): CNS toxicity with serotonergic drugs. Drug Safety Update 1: 5. MHRA (2009) Methylthioninium chloride (methylene blue): update on CNS toxicity with serotonergic drugs. Drug Safety Update 2: 3. Mihai R, Mitchell EW, Warwick J (2007) Dose-response and postoperative confusion following methylene blue infusion during parathyroidectomy. Can J Anaesth 54: 79–81. Naylor GJ, Smith AH, Connelly P (1987) A controlled trial of methylene blue in severe depressive illness. Biol Psychiatry 22: 657–659. Ng BK, Cameron AJ, Liang R, Rahman H (2008) Serotonin syndrome following methylene blue infusion during parathyroidectomy: a case report and literature review: [Syndrome serotoninergique suite a une perfusion de bleu de methylene pendant une parathyroidectomie: presentation de cas et analyse bibliographique]. Can J Anaesth 55: 36–41.
8 Oz M, Lorke DE, Petroianu GA (2009) Methylene blue and Alzheimer’s disease. Biochem Pharmacol: doi:10.1016/ j.bcp.2009.1004.1034. Pollack G, Pollack A, Delfiner J, Fernandez J (2009) Parathyroid surgery and methylene blue: a review with guidelines for safe intraoperative use. Laryngoscope 119: 1941–1946. Ramsay RR, Dunford C, Gillman PK (2007) Methylene blue and serotonin toxicity: inhibition of monoamine oxidase A (MAO A) confirms a theoretical prediction. Br J Pharmacol 152: 946–951. Rosenbaum H (2006) January 2006 Case of the Month. http://medical.mhaus.org/index.cfm/fuseaction/Content.Display/PagePK/ CaseOfTheMonth2.cfm (accessed March 2009) Rothman RB, Baumann MH (2003) Monoamine transporters and psychostimulant drugs. Eur J Pharmacol 479: 23–40. Rowley M, Riutort K, Shapiro D, Casler J, Festic E, Freeman WD (2009) Methylene blue-associated serotonin syndrome: a ‘green’ encephalopathy after parathyroidectomy. Neurocrit Care 11: 88–93. Schaper E, Farber E, Ebbecke M, Desel H (2002) Suicide with mirtazapine –hardly possible. J Toxicol Clin Toxicol 40: 343–344. Schwiebert C, Irving C, Gillman PK (2009) Small doses of methylene blue, previously considered safe, can precipitate serotonin toxicity. Anaesthesia 64: 924–924. Shanmugam G, Kent B, Kirby S, Baskett R (2008) Serotonin syndrome following cardiac surgery. Interact Cardiovasc Thorac Surg 7: 656–657. Stanford BJ, Stanford SC (1999) Postoperative delirium indicating an adverse drug interaction involving the selective serotonin reuptake inhibitor, paroxetine? J Psychopharmacol 13: 313–317. Stanford SC, Stanford BJ, Gillman PK (2009) Risk of severe serotonin toxicity following co-administration of methylene blue and serotonin reuptake inhibitors: an update on a case report of post-operative delirium. J Psychopharmacol: doi:10.1177/ 0269881109105450).
Journal of Psychopharmacology 0(00) Stawicki SP, Sims C, Sarani B, Grossman MD, Gracias VH (2008) Methylene blue and vasoplegia: who, when, and how? Mini Rev Med Chem 8: 472–490. Sun-Edelstein C, Tepper SJ, Shapiro RE (2008) Drug-induced serotonin syndrome: a review. Expert Opin Drug Saf 7: 587–596. Sweet G, Standiford SB (2007) Methylene-blue-associated encephalopathy. J Am Coll Surg 204: 454–458. Walter-Sack I, Rengelshausen J, Oberwittler H, et al. (2009) High absolute bioavailability of methylene blue given as an aqueous oral formulation. Eur J Clin Pharmacol 65: 179–189. Waring WS, Good AM, Bateman DN (2007) Lack of significant toxicity after mirtazapine overdose: a five-year review of cases admitted to a regional toxicology unit. Clin Toxicol (Phila) 45: 45–50. Wegener G, Volke V, Rosenberg R (2000) Endogenous nitric oxide decreases hippocampal levels of serotonin and dopamine in vivo. Br J Pharmacol 130: 575–580. Whyte IM (1999) Serotonin syndrome complicating the treatment of recurrent depression. Curr Ther 40: 6–7. Whyte IM (2004) Serotonin toxicity (syndrome). In: Dart RC (ed.) Medical toxicology. Baltimore: Lippincott Williams & Wilkins, 103–106. Whyte IM, Dawson AH (2002) Redefining the serotonin syndrome. J Toxicol Clin Toxicol 40: 668–669. Whyte IM, Dawson AH, Buckley NA (2003) Relative toxicity of venlafaxine and selective serotonin reuptake inhibitors in overdose compared to tricyclic antidepressants. Q J Med 96: 369–374. Wischik CM, Bentham P, Wischik DJ, Seng KM (2008) Tau aggregation inhibitor (TAI) therapy with remberTM arrests disease progression in mild and moderate Alzheimer’s disease over 50 weeks. Alzheimers Dement 4: T167. Zoungrana A, Coulibaly B, Sie A, et al. (2008) Safety and efficacy of methylene blue combined with artesunate or amodiaquine for uncomplicated falciparum malaria: a randomized controlled trial from Burkina Faso. PLoS ONE 3: e1630.
