MAOIs: swapping and combining

by | Last updated Dec 2, 2019 | Published on Jan 23, 2017 | Anti-Depressants, MAOIs

This commentary covers all aspects of the subject, including swapping from one MAOI to another MAOI, it inevitably overlaps with various other commentaries, but that is inevitable.  Also, that overlap consolidates knowledge and understanding.

Myth: There are many drugs that are ‘dangerous’ if combined with MAOIs — Untrue.

Interactions, swapping and combining involving MAOIs is not difficult.

There are no significant pharmaco-kinetic interactions that involve MAOIs {Gillman, 2011 #10056;Dixon Clarke, 2011 #11984}. There are however many with the SSRIs and the SNRIs, so in this respect MAOIs are superior and easier to use.

Learning about interactions is interesting (and, I confess, I think it is fun) and helps one to master the principles of clinical pharmacology that are of wide applicability.

Once interactions are understood then swapping and combining drugs is simply the practical application of that knowledge. There is quite a lot of misinformation in the psychiatric literature on this subject, largely because psychiatrists’ knowledge of clinical pharmacology is not good.

Note: this is primarily about the non-selective irreversible inhibitors tranylcypromine and phenelzine and isocarboxazid. Other MAOIs such as rasagiline, selegiline, and moclobemide have even less interactions (see below).

Not too hard

There are no significant pharmaco-kinetic interactions and only two possible pharmaco-dynamic interactions: 1) ST and 2) pressor response.  That is it: nothing else. Not too hard is it?

1) ST is the only serious interaction involving MAOIs that is ever likely to be encountered these days.

2) The pressor response of increased blood pressure (BP) is the other interaction that caused much consternation in the past. It involved the ‘releasers’ that in the past were found in over-the-counter drugs such as cough and cold remedies. These drugs have largely been withdrawn from the market and almost all cough and cold remedies now contain oxymetazoline (an α1 adrenergic receptor agonist, which is quite safe), but not ephedrine-like drugs: e.g. ‘Sudafed’ used to be pseudo-ephedrine but is now actually ‘oxymetazoline’, now badged as ‘Sudafed OM’. Where ephedrine-like drugs are occasionally still available they should be avoided, but ordinary therapeutic doses are highly unlikely to lead to serious harm: it is only overdoses such drugs that are likely to be seriously harmful. As would be expected, the same applies to amphetamine.

Serotonin toxicity (ST) is caused only by the co-ingestion of serotonin reuptake inhibitors (SRIs). So, avoid ‘SRIs’ (with appropriate ‘washout’ intervals), i.e. both SSRIs and SNRIs, and you are 99% of the way there in terms of avoiding problems. There are just one or two other drugs that are not generally regarded as SRIs, but which do in fact have sufficient SRI potency to be problematic (see below, and other commentaries on my site for details).

Contrary to the opinions expressed in many texts, various other so-called ‘serotonergic’*  drugs are not significant SRIs — such as trazodone, mirtazapine, lithium, buspirone, tryptans etc., see below for full list — and are not a risk for ST interactions: these references contain detailed evidence relating to these issues {Gillman, 2009 #9452;Isbister, 2007 #13353;Gillman, 2006 #7176;Buckley, 2014 #19217}.

*‘Serotonergic’ is a poor and misunderstood term: strictly speaking it means effecting serotonin neurones. However, it is usually (mis)used to mean serotonin-enhancing (as opposed to serotonin-blocking). Add that to the pharmacological confusion and you have chaos.  So, yes, in a sense mirtazapine is a ‘serotonergic’ drug, but an anti-serotonergic one that lessens manifestations of ST, not exacerbates them.

The first general principle to keep in mind is that ST (when caused by MAOI/SRI combinations) is much more likely to lead to serious harm than is the pressor response to tyramine — or other releasers.

Even a single therapeutic dose of a potent SRI is likely to lead to a serious (i.e. putting someone in hospital) or even fatal ST reaction.

Reminder – ST is not an idiosyncratic reaction but a predictable dose-related interaction. That is why for years it has been my contention that it is nonsensical to speak about ST being ‘rare’. Strychnine poisoning is rare, except in people who have taken strychnine.

In contrast, a single therapeutic dose of releaser is extremely unlikely to lead to a seriously harmful outcome due to a pressor response.  In the vast majority of cases of releaser-induced pressor responses nothing will be observed at all, and occasionally a brief episode of hypertension will result which will not do any harm.  So, there is a great difference in the risks produced by these two different reactions.

The other major difference between the two reactions (ST & pressor) is that potentially dangerous SRI antidepressants are frequently going to be encountered in patients who might be considered for treatment with MAOIs (like leftover supplies of SRI-type medication in the bath-room cabinet), whereas the need and frequency of use of drugs that are potent releasers is rare in modern practice.  Therefore, the releaser/pressor-response scenario is both very infrequent and unlikely to lead to serious trouble.

Swapping and combining

It is commonly thought and stated that it is problematic to change from one drug to another when one of them is an MAOI.  In practice this is simply not a difficulty.

If you know how to combine drugs, then you know how to swap safely.

Combining TCAs and MAOIs serves as a good teaching example to illustrate some key points.  These two classes can be safely co-administered, except for clomipramine and imipramine which have significant potency as SRIs.  It is only SRIs that are a problem — nothing else.

Incidentally, I must mention here that there is no sound evidence that the order in which MAOIs/TCAs are given, if used in combined treatment, makes any difference to side-effects or safety.  That notion was a speculation that was contained in one of the case reports in the early 1960s.  This has become established in the literature to a degree that facts do not support.

There continues to be a great deal of misinformation about combining drugs in the psychiatric literature, and as above, it is often without a factual or even a theoretical basis.  Some recent examples of one or two basic errors are in a paper by Palaniyappan et al. that was specifically directed at professional education for specialists {Palaniyappan, 2009 #9794}.  I published a response to that {Gillman, 2010 #9925}.  Anyone interested can read the relevant material, including the authors’ response to my criticisms, but I do not particularly recommend that as a good use of your time. [link to free pdfs in the references belo]).

It is not always individual psychiatrists who are solely to blame for their lack of knowledge, what they are presented with for educational reading, even by independent and supposedly authoritative sources, such as the UK Royal College (Palaniyappan {Palaniyappan, 2009 #9794}, above), is not always of an impressive standard.  That reflects the problem I often allude to — the poor pharmacological knowledge of most psychiatrists.

The example of imipramine

Imipramine is a good example to understand because throughout the history of psycho-pharmacology it has been a focus of confusion (likewise pethidine, aka meperidine).  Some authorities (e.g. in Germany) still express the view that it is not contraindicated to combine imipramine with MAOIs.

The confusion about this arose in the early 1960s because a serious interaction occurred only infrequently, and it was not understood or distinguished from the pressor response to tyramine***.  The ST reaction only occurs sometimes because imipramine is a relatively weak SRI and therefore in smaller doses, or in people who have low blood levels, it does not cause excessive serotonin-mediated side effects or toxicity.  However, the blood level only has to go up a little and then symptoms of increasing serotonin-mediated side-effects or overt serotonin toxicity rapidly develop.  An overdose of imipramine combined with an MAOI will frequently lead to serious ST and may well prove fatal.  A recent example of a fatal case is provided in the case reported by Otte et al. {Otte, 2003 #14007}.  Note: this patient would almost certainly not have died had they known how to diagnose and treat ST (which, frankly, they should have known, especially since the incident was at a ‘teaching’ hospital).

*** For the record, as it were, here are refs to some of the old cases of ST in the 1960s involving MAOIs with imipramine: {Loveless, 1965 #623;Stanley, 1964 #547;Kane, 1963 #749;Brachfeld, 1963 #543;Luby, 1961 #588;Lee, 1961 #529;Singh, 1960 #665;Davies, 1960 #544},

and MAOIs with pethidine: {Churchill-Davidson, 1965 #12684;Vigran, 1964 #584;Spencer, 1963 #14417;Taylor, 1962 #763;Pells Cocks, 1962 #14057;Denton, 1962 #578;Cocks, 1962 #657;Brown, 1962 #12558;Palmer, 1960 #579;Mitchell, 1955 #409}.

So, imipramine is ‘on the cusp’ SRI-potency-wise.  Drugs more potent than imipramine as SRIs is are highly likely to precipitate ST and drugs that are significantly less potent do not do so (there is more detailed consideration of this in various of my peer-reviewed papers).

Because the relative SRI potency of these different drugs (TCAs and opioids like pethidine etc.) was not recognized back in the 1960s, and those reactions were confused with the pressor response, this lead to uncertainty and a blanket prohibition about combining these two classes of drugs (which is still repeated in many sources), and about analgesics and anaesthesia.  Modern knowledge resolves these uncertainties even if some current texts have not quite caught up yet.

The simple requirement is to learn and remember which drugs are in fact significantly potent SRIs.  This would be very easy if drugs were labelled correctly according to their pharmacological properties, but that is not so because they are labelled more according to marketing imperatives, not pharmacology.  Nevertheless, it is not that difficult to sort the sheep from the goats (see below).

Obviously, all the ‘SSRIs’ and the ‘SNRIs’ are ruled out.

There are a few drugs on the market in the last few decades which have been stated to be ‘serotonergic’ but which do not significantly elevate serotonin and are therefore quite safe. That is marketing for you. I have written extensively about that and the proto-typical example is mirtazapine. It is not a ‘serotonergic’ drug and is perfectly safe with MAOIs.

Swapping and ‘bridging’ candidates

When swapping drugs, it may sometimes be appropriate to stop one before starting the next one, with appropriate washout intervals. However, this is rarely necessary and it is often preferable to bridge or overlap the two, decreasing one whilst increasingly the other. I have done this in hundreds of cases without any problems whatsoever.

It must be remembered that when both drugs are non-MAOI drugs then pharmaco-kinetic interactions may need to be taken into consideration. That is one of the reasons why nortriptyline is a good candidate, because it is less likely than most other drugs to have a significant pharmaco-kinetic interaction.

The only time when ‘overlapping’ cannot be done is when one wishes to swap from an MAOI to an SRI (and vice versa) and that is when a ‘bridging’ strategy is especially useful.

The TCA nortriptyline is, pharmacologically speaking, a good and flexible candidate to fulfil a ‘bridging’ role. For instance, if one is changing between an MAOI and (des)venlafaxine, which can have quite marked withdrawal symptoms, co-administering nortriptyline*** prior to reducing or ceasing the venlafaxine, may both reduce withdrawal symptoms and act as a bridge (to reduce the chance and severity of any relapse) prior to the initiation of the MAOI. The same process works in reverse where nortriptyline can be added to a pre-existing MAOI, the MAOI is then stopped and most subsequent treatments can then be initiated with ease.

***One could equally well use mirtazapine, doxepin, amitriptyline, quetiapine, (but not ziprasidone), mood stabilizers: in short, anything you favour that is not an SRI. And all with no worries about pharmaco-kinetic interactions. How difficult is that!

The strategy I used frequently was to add sertraline*** to the nortriptyline, once the MAOI was out of the system, to get a good SNRI effect.

***Why sertraline? Because, like nortriptyline, it has the most favourable pharmacological and pharmaco-kinetic characteristics of all the drugs in its group. That is something I have been saying for a very long time, and I notice now that a number of more recent reviews put forward the same argument.

One should note here, that using two separate drugs (e.g. Sert + NTP, instead of an ‘SNRI’ like venlafaxine) makes all ‘stopping and swapping’ easier because one can raise and lower or cease each element separately. That may also assist in minimising side effects.

Some people seem to prefer using a single drug, such as (des)venlafaxine, milnacipran or duloxetine to get a dual effect, but I suspect that prejudice has got a lot to do with habit, and unfamiliarity with pharmaco-kinetic considerations concerning interactions.

I always found using two different drugs, the individual doses of which could be adjusted to optimise efficacy vs side-effects, was usually a better strategy and often more successful (and often much cheaper). Perhaps effects, other than just the NRI effect, may contribute to the benefits conferred by nortriptyline. I have certainly seen many patients who have failed to respond to venlafaxine who have subsequently responded to the combination of sertraline and nortriptyline. But rarely the other way round.

It was postulated some time ago, on a sound pharmacological basis, that TCAs, by virtue of their NRI potency, attenuate the ‘cheese effect’ {Freyschuss, 1970 #17622;Pare, 1985 #8353;Pare, 1982 #8321}.

I assembled and explained the evidence for this in my TCA review {Gillman, 2007 #8496} which concluded that the NRIs with the high affinity for the NAT (viz. reboxetine, atomoxetine, desipramine, oxyprotaline, protriptyline and nortriptyline) have all been demonstrated to block the pressor response to tyramine {Chalon, 2003 #3133;Rudnick, 1997 #14232;Bevan, 1978 #19871;Ghose, 1976 #10043;Freyschuss, 1970 #17622;Reimann, 1993 #8189;Graefe, 1999 #9840}, even when it has been potentiated in the presence of MAOIs {Dostert, 1994 #12842;Doggrell, 1977 #8299;Burkard, 1992 #8201}. The early demonstrations of NRI attenuation of the pressor response to tyramine go back a long way, past learning seems to have been lost for a long time {Brodie, 1968 #8356;Matsumoto, 1964 #19721}. NB Both those last 2 references are from the lab of the famous pharmacologist Bernard Brodie whose early papers are still worth studying.

This leads to the confident conclusion about an old and bitter-sweet irony: combinations of (non-serotonergic) TCAs or NRIs with MAOIs are not risky, they actually make MAOIs safer, not more dangerous, by attenuating the pressor response to tyramine, or any other NA releaser (ISA).

If a patient is more than usually tyramine sensitive, they will become less so if a potent NRI like nortriptyline is added to the regime: the greater the dose the greater the attenuation of the pressor response. That is not a ‘theory’, it is pharmacological fact.

Ceasing and half-lives

On ceasing any SRI-type antidepressant to start an MAOI, washout intervals varying between one and five weeks may be required***. No washout is needed for non-SRI-type drugs because it is safe to co-administer them (as above). The rule of thumb is to allow 5 half-lives to elapse*, which is about one week for many of these drugs.

***Fluoxetine (via its metabolite norfluoxetine) has an elimination half-life in some people of up to two weeks (so it can take up to 10 weeks to get out of the system). See any good pharmacology reference book for a table of half-lives, or on the net look at: http://www.health.harvard.edu/diseases-and-conditions/going-off-antidepressants

In practice 5 half-lives is a conservative approach. Most drugs will have lost sufficient SRI activity after two half-lives to allow cautious introduction of an MAOI — providing the patient can be observed for early signs of serotonergic over-activity.

Such signs are: 1) specific; tremor, hyperreflexia and clonus. Less specific; GI overactivity, mydriasis, sweating, anxiety, restlessness.

Should such symptoms be apparent it is simple enough to withhold the MAOI for a few more days and then try again. ST is a dose related phenomenon and the emergence of serotonin-mediated signs or symptoms is a cause for caution, not panic.

Other drugs with significant SRI potency***

  1. Vortioxetine, vilazodone, ziprasidone
  2. Chlorpheniramine (aka chlorphenamine) Brompheniramine.
  3. Pethidine, tramadol, dextromethorphan, dextropropoxyphene (fentanyl is safe).

Other safe drugs (with insignificant SRI potency)

  1. Mirtazapine, mianserin, doxepin (now re-badged — correctly — as an anti-histamine-type ‘hypnotic’), trazodone, nefazodone, buspirone, bupropion, reboxetine, atomoxetine, amoxapine, agomelatine, tianeptine.
  2. Lurasidone, quetiapine, aripiprazole, olanzapine, asenapine, and all other neuroleptics.
  3. Methylphenidate (it is not a 5-HT releaser or uptake inhibitor), amphetamine (it is a releaser, but may be used cautiously, ‘start low, go slow’ is the mantra to remember), the new preparation ‘lisdexamfetamine’ may be an advantage owing to its better pharmaco-kinetic profile.
  4. Lithium, carbamezepine, valproate, lamotrigine.
  5. L-tryptophan

*** Other than: all ‘anti-depressant’ SRIs (viz. SSRIs and SNRIs). See more detailed info in other commentaries on website

Swapping from one MAOI to another MAOI

The requirement or desire to swap from one MAOI to another MAOI is something that will be an uncommon occurrence.  Furthermore, it will be an urgent need even more rarely.  It may be indicated, for instance because of the excessive weight gain, sexual dysfunction, or oedema, that occur with phenelzine.  Because opinion exists in the literature suggesting dove-tailing, or a direct swap, is a potentially risky thing to do, some discussion about this is educative.

Insofar as I have been able to trace the original texts which postulate these sorts of dangers, I have one comment to make.  Some are written by clinicians who have an incomplete understanding of pharmacology, or toxicology, and they contain errors of fact.  Such texts are often books, which may not be peer-reviewed.

First, there is no known mechanism of interaction that would cause a problem.  Second, neither is there a basis for postulating a pharmaco-kinetic interaction. 

Third, nor is there a basis for a major pharmaco-dynamic interaction. 

That leaves a mystery, or possibly a phantom?

My extensive experience analysing hundreds of case reports of ST illustrates clearly that the overwhelming majority of case reports are misleading*, and they often involve supposed interactions that have no known basis, in fact or theory.  To make decisions based on such reports has repeatedly proved to be inappropriate and the resultant actions have in some cases had serious negative consequences.

*This is why many reputable journals decline to publish case reports {Gillman, 2006 #7650}.

The literature is replete with inappropriate and groundless injunctions against a host of perfectly safe drug combinations (see my other papers for a detailed analysis of this topic in relation to ST) and various ‘official’ bodies like the WHO and the FDA have been repeatedly guilty of issuing such scientifically groundless injunctions: recent examples are warnings about ‘serotonergic’ drugs precipitating ST if combined with the anti-emetic 5-HT3 antagonists, and about ST with Triptans.  Such ‘cry-wolf’ warnings are time-wasting and confuse and misdirect practicing clinicians.  So, adopting the careful conservative approach is not always the ‘win-win’ scenario that such cautious analysts suppose it to be.

In routine clinical practice the maxim ‘start low and go slow’ is wisely adhered to.  When a changeover is being considered, or is indicated, in a patient who is severely ill, and in danger because of that, a degree of risk, imagined or real, is acceptable.  In less urgent circumstances the patient and clinician are usually well-advised to opt for a cautious approach.

The history

In more than 60 years of MAOI use there are only a few reports of supposed difficulties: these are non-specific and not indicative of a cause-effect relationship involving an MAOI-MAOI interaction. 

As of 2019, all accumulated references

{Bazire, 1986 #19675;Gelenberg, 1985 #19717;Gelenberg, 1985 #19717;Jefferson, 1998 #19668;Mattes, 1998 #3253;Polnak, 2018 #21771;Safferman, 1992 #19714;Schrire, 1963 #19852;Torre, 2009 #14551;True, 1985 #19676;True, 1986 #19677}

 

 

It is notable that of these incomplete and unconvincing reports, one suggests subarachnoid haemorrhage and another serotonin toxicity.  It is highly unlikely that both of these represent a cause-effect interaction, because the required mechanism is quite different for each of them.  The most parsimonious explanation is that neither of them represent a cause-effect relationship between the changing drug regime and the outcome reported.  As with so many case-reports these ones also contain insufficient information to draw reliable conclusions.

It is well recognised that abrupt cessation of antihypertensive treatment can cause rebound hypertension; indeed, this is a not uncommon presentation in emergency departments.  It is forgotten that MAOIs are anti-hypertensive drugs, and, as I have reviewed elsewhere, were indeed used for the treatment of hypertension in the 1960s and 1970s.  I have seen patients who have developed high blood pressure when well-meaning primary-care physicians have stopped the ‘dangerous-old-antidepressants-we-don’t-use-anymore’ that the (doddery old) specialist has been giving because he does not know about the wonderful new drugs we now have (that the young lady drug rep in the short skirt who took me to lunch told me all about).

Furthermore, I have seen patients on long-term antidepressant MAOI treatment who have clearly developed idiopathic hypertension during the course of that treatment, which was ‘disguised’ by the MAOI.  These patients then had substantial rises in blood pressure on cessation of the MAOI.  Indeed, one of them actually had a small CVA whilst waiting for an appointment to see me, to decide on future treatment.  The primary care doctor had already (unilaterally) instructed her to cease the tablets prior to the appointment with me.  If that patient had already restarted another MAOI then where would the blame for her CVA have been laid?

The following are the relevant contributions that I am aware of, in the published literature relating to direct MAOI to MAOI change-overs {Chandler, 1987 #19715;Torre, 2009 #14551;Mattes, 1998 #3253;Jefferson, 1998 #19668;Bazire, 1986 #19675;True, 1985 #19676;True, 1986 #19677;Safferman, 1992 #19714;Gelenberg, 1984 #19716;Gelenberg, 1985 #19717;Schrire, 1963 #19852}: re Torre et al. see ‘case 2’.  None of them are very helpful, nor do they provide a substantive basis for prohibition of a direct swap-over of one MAOI to another MAOI.  One paper reports a small series of 8 cases where it was accomplished without a problem {Szuba, 1997 #1171}.  I have personally done it without any problem, but only on a couple of occasions, as have associates and various people who have been in contact via my web site.

Opinion: if there is good reason to swap rapidly***, do it, because there is; 1) no theoretical basis to suggest it might be contra-indicated, and 2) existing reports do not constitute evidence to the contrary, and 3) it has clearly been done many times without any problem.

***For example, I remember a patient who had failed to improve with a long course of ECT, which had involved a long stay in hospital 1000 km away from home, whom had started phenelzine and was experiencing rapid weight gain and massively swollen the legs, remained suicidal, but dreaded a return to hospital.  A direct swap to Parnate was successful and uneventful and led to great improvement.

Rasagiline, selegiline, and moclobemide

These have less interactions than the non-selective irreversible MAOIs and are discussed in more detail in other commentaries. As far as rasagiline and selegiline are concerned it is safe to combine them with SSRIs providing the dose remains in the range where they are selective for MAO-B. Moclobemide is not safe to combine in therapeutic doses with SRIs. Even at therapeutic doses serious ST has (predictably) occurred, and also fatalities. It only takes a small change in dose to go from no problems to fatal toxicity, such as might occur with accidental double dosing of a drug by a patient, or alteration of blood levels because of a pharmacokinetic interaction with some other drug given by another doctor. This note has further information:

http://www.psychotropical.com/mociobemide-s-sris

Herbals and dietary intake

A recent review looked at ‘natural’ MAOIs, from various sources, and concluded that ‘a healthy diet contains amounts of quercetin that might give sufficient amounts in brain to induce, by monoamine oxidase A inhibition, a small decrease in neurotransmitter breakdown’ {Dixon Clarke, 2011 #11984}. This was by written by my co-author Rona Ramsay (& Dixon Clarke), who did the key work that proved methylene blue was a potent MAOI {Ramsay, 2007 #8427}.

They concluded that neither diet nor herbal remedies such as St John’s Wort (Hypericum perforatum, active ingredient, hypericin) significantly inhibit MAO (the flavonol fraction of St John’s Wort produces slight (39%) inhibition of MAO {Bladt, 1994 #19894}), but that this was insufficient to give an anti-depressant effect, or a risk of ST interactions with SRIs.

In summary, herbal nostrums do not produce significant MAO inhibition. There are certainly no convincing reported cases of ST, or of pressor reactions, implicating them.

Email
Skype
BeanPod
Reddit
RSS
LINKEDIN
Google Scholar
Research Gate