What Is the Most Appropriate Antidepressant to Use in Patients with Epilepsy?

Q&A 24.6

What is the most appropriate antidepressant to use in patients with epilepsy?

Prepared by UK Medicines Information (UKMi) pharmacists for NHS healthcare professionals

Before using this Q&A, read the disclaimer at

Published:September 2014

Summary

The first consideration should always be to check the patients’ anticonvulsant regimen for potential drug-induced depression. It may be that the patient would benefit from changing the anticonvulsant to another agent with a more favourable effect on mood rather than adding in an antidepressant.
The risk of seizures with most antidepressants is low, but is probably not zero for any of them, and patients should be made aware of this when prescribing. The risk of seizures increases with increasing doses.
SSRIs are considered the first line antidepressant option in patients with epilepsy. Citalopram or sertraline may be considered the better options due to safety data and reduced interaction potential with the anticonvulsants. Fluoxetine is not the best choice due to its long half-life, a possibly greater incidence of seizures and an increased risk of drug interactions.
Moclobemide is a good alternative, as it has a low incidence of seizures but due to limited data it should be reserved as a second choice to SSRIs.
TCAs should be used cautiously in patients with epilepsy and reserved for patients who poorly respond to or are intolerant of other antidepressants. Where a TCA is needed, doxepin is possibly of lowest risk and therefore the agent of choice within this group.
Clinicians should be aware of the possibility of interactions between antidepressants and anticonvulsants and should monitor carefully patients with epilepsy who are prescribed antidepressants.
Introducing the antidepressant gradually, starting with a low dose, and not exceeding the maximum recommended doses may reduce the risk of a seizure.
If seizures occur or if the incidence of seizures increases, the antidepressant should be discontinued.
Detailed information regarding drug interactions has not been included in this Q&A: drug interactions between antidepressants and antiepileptic drugs must be checked prior to prescribing.

Background

Depression is the most common co-morbidityin people with epilepsy(PWE) compared with the general population(1;2). Rates vary from 3-9% in patients with well-controlled epilepsy to much higher (20-55%) in those with recurrent seizures and those treated at tertiary referral centres(1;3). A population-base study estimated a lifetime prevalence of 17.4% in those with epilepsy compared to 10.7% in those without epilepsy (3). The presence of depression has been associated with poorer seizure control and has a greater negative impact on quality of life than the number of prescribed antidepressants or seizure frequency (4).

The association between epilepsy and depression may be explained in part by serotonin; depletion of serotonin increases the risk of both depression and epilepsy(4;5). Abnormalities in other neurotransmitters such as noradrenaline, dopamine, and GABA are also seen in both conditions(3;4).

Depressive symptoms present in a number of ways. Peri-ictal depression is relative to seizure activity, is short-lasting and does not require treatment with antidepressants; this should be treated by optimising anticonvulsant therapy (3;5). Peri-ictal depression can be divided into three types. Pre-ictal (before a seizure) depressive symptoms manifest hours to days prior to a seizure. Ictal depression occurs minutes before a complex partial seizure. Post-ictal depression is characterised by a depressed mood developing hours to days after a seizure (1;3).

Inter-ictal depression presents as chronic depression; this is not related to seizure activity and is likely to require treatment with antidepressants (1;3;5). PWE are three times more likely to be prescribed an antidepressant than the general population and two times more likely to report suicidal thoughts (6). Many antidepressants may lower the seizure threshold, but there is also evidence that when used within their therapeutic dosage range, the risk of seizure activity is low(1).

Guidelines suggest that SSRIs are an appropriate first-line choice for treating depression in PWE(7;8). What data supports this choice, or the choice of antidepressants from other drug classes?

Answer

Use of antiepileptic drugs (AEDs) with mood stabilising properties may be appropriate for selected individuals with mild depressive symptoms(3). For example, gabapentin, carbamazepine, oxcarbazepine, valproate, and lamotrigine improve depressive symptoms in epileptic patients(1;2;9), whereas phenobarbitone, vigabatrin, primidone, topiramate and tiagabine have negative psychotropic properties and have been associated with depression(1).

There is a lack of information regarding the safety of antidepressants in PWE (8) and the choice is a common clinical dilemma. Some antidepressants, especially the tricyclic antidepressants (TCAs), are well known to lower the seizure threshold but the risk is dose related(5). There is no large study comparing seizure frequency in PWE who take antidepressants with those who do not (10). With few exceptions, all antidepressants are associated with seizures in overdose; SSRIs are less likely to cause seizures in overdose than TCAs (11). However, the perceived risks are often overestimated and rarely outweigh the risk of leaving depression untreated (6). Treatment of depression in patients with epilepsy can improve their quality of life and reduce their risk of suicide (4). It is important to check the anticonvulsant regimen of an epileptic patient with depression for potential drug-induced depression before considering adding an antidepressant.

Selective Serotonin Reuptake Inhibitors (SSRIs)

In general, the SSRIs are considered the first-line choice in patients with stable epilepsy because they have a low pro-convulsive effectand are associated with a lower incidence of seizures when compared with TCAs (1-3;5-7;12). Treatment should be initiated with a low dose and increased gradually, as necessary (8). If seizures occur, or if the incidence of seizures increases, the antidepressant should be discontinued (13-18).

The risk of seizures at therapeutic doses of SSRIs is between 0.1-0.3%, compared with 0.4 – 1.2% with TCAs(11), though this varies between individual agents. Studies ofshort and long-term SSRI treatment, (1 month to around 15 months), suggest there is no worsening in seizure frequency and in some patients improvement in seizure outcomes are seen. It should also be noted that in some patients, seizure frequency worsened but could be controlled by either increasing the dose of antiepileptic drug, or discontinuing the antidepressant (2). There is a lack of double-blind, randomised controlled trials in large patient populations; most have been small studies in highly selected patients (2).

Drug interactions with SSRIs must be taken into consideration: escitalopram, fluoxetine, fluvoxamine, paroxetine and sertraline inhibit cytochrome P450 enzymes to varying degrees and may increase the plasma levels of AEDs (4;13-18).

Fluoxetine is the most studied SSRI and has a probable seizure incidence of 0.2%(12;19). Fluoxetine should be avoided in patients with unstable seizure disorders and patients with controlled epilepsy should be monitored (15). It should also be noted that fluoxetine can cause changes in blood levels of phenytoin such that cases of phenytoin toxicity have been reported (15). The long half-life of fluoxetine should be taken into consideration when switching from fluoxetine to other antidepressants (15).

Pre-marketing studies and a retrospective survey of prescriptions (n=10,401)suggested the incidence of seizure with fluvoxamine to be 0.2% in the general population(11;20). One small-scale study (n=28) reported that fluvoxamine was not proconvulsant at doses ranging from 50 to 200mg, when given to depressed epileptic patients(21), however there have been case reports of seizures following therapeutic doses of fluvoxamine in patients with and without a previous history of seizures(22;23). Fluvoxamine should be avoided in patients with unstable epilepsy and patients with controlled epilepsy should be carefully monitored (16). Fluvoxamine inhibits CYP1A2 and, to a lesser extent, CYP2C and CYP3A4; co-administration with antiepileptics metabolised by these isoenzymes, such as phenytoin and carbamazepine, may lead to higher plasma levels of the antiepileptic and dose adjustments may be necessary(16).

Paroxetine appears to have a minimal potential for producing seizures at therapeutic doses(12)and reviews suggest paroxetine has a seizure incidence of less than 0.1%(19). Paroxetine is not known to have any effect on the pharmacokinetics and pharmacodynamics of anticonvulsants but paroxetine dose adjustment may be necessary if an enzyme inducer (such as carbamazepine, phenytoin or phenobarbitone) is initiated or discontinued and it should be used with caution (17).

Citalopram and escitalopram have not been reported to have a proconvulsive effectin post-marketingdata(12). In an open label study citalopram 20mg/day for 4 months was associated with an improvement in depressive symptoms and a reduction in seizure frequency (from 1.31 per month to 0.82, p<0.005) in 39 patients who had epilepsy and depression(24). A study of 43 patients with epilepsy showed citalopram (average dose 22.6mg) to be an effective antidepressant without affecting monthly seizure frequencyor causing de novo generalised tonic-clonic seizures when used for 2 months(25). A third study also showed citalopram (mean dose 24mg) to be an effective treatment without causing an increase in seizure frequency in 33 patients with temporal lobe epilepsy, when used for up to 30 weeks(26). However, due to the potential for all SSRIs to cause seizures, citalopram should still be used with caution in patients with controlled epilepsy (13;14). Citalopram and escitalopram should be avoided in patients with unstable epilepsy and patients with controlled epilepsy should be carefully monitored (13;14). Despite this, citalopram is considered one of the first line options due to it is perceived safety in PWE and lack of interactions with the anticonvulsant agents(5;19).

Early clinical trials with sertraline suggested that seizures occurred at a similar frequency to placebo and only in people with a history of seizures(12). The seizure frequency in 100 patients with epilepsy after being started on sertraline (mean dose 108mg) was assessed in one study(27). The mean treatment duration was 10.3 months (median duration 6 months, range 0.2 to 38 months). Six patients experienced an increase in seizure frequency but this could only be definitely linked with sertraline in one patient; sertraline was stopped in this patient. There was probable causality between seizure increase and sertraline in the other five patients; AED doses were adjusted and four continued with sertraline treatment. Data from the manufacturers suggests the incidence of seizures to be between 0.1% and 1% at therapeutic doses of sertraline. Sertraline has the benefit of having no interactions with carbamazepine but it may interact with phenytoin (reduced sertraline levels and potential inhibition of phenytoin metabolism)and should be used with caution in PWE (18).

Serotonin / Noradrenaline Reuptake Inhibitors (SNRIs)

An SNRI can be used as an alternative to an SSRI (8) but clinical experience and data regarding the safety of duloxetine and venlafaxine in PWE is limited(5).

A seizure rate of 0.2% with duloxetine therapy was reported in clinical trials (n=2418) (11). Although seizures have been reported rarely (5)duloxetine should still be used with caution in patients with a history or diagnosis of seizures (28). Cases of convulsions have been reported on discontinuation of therapy (28). Seizures have been reported in 0.3% of depressed patients (n=3082) treated with venlafaxine ≤150mg/day in clinical trials (11). Venlafaxine is considered to increase seizure risk in overdose (5;29). Venlafaxine should be introduced slowly in patients with a history of seizures; no interactions with antiepileptics is anticipated(12;29).

Tricyclic Antidepressants (TCAs)

All TCAs appear to lower the seizure threshold,with amitriptyline reputed to be the most proconvulsive and doxepin possibly of lowest risk (12). Ideally they should not be used in PWE (5). The overall incidence of seizures at therapeutic doses of TCAs is estimated at 0.4–1.2% (11), though the incidence for individual TCAs varies markedly (see Appendix 1). Most TCAs are known to potentially interact with anticonvulsants and manufacturer’s advice should be consulted for individual TCAs.

Monamine Oxidase Inhibitors (MAOIs)

MAOIs (phenelzine, isocarboxazid and trancylcypromine) have a low seizure risk but they are seldom used now due to their interactions with certain foods and drugs and the fact that there are safer alternatives available. However, they should be used with caution in patients with epilepsy (30) and can increase or decrease seizure frequency (12).

Moclobemide is a reversible inhibitor of monoamine oxidase type A (RIMA) which is less likely to interact with foods and drugs, and there have been no reports of problems in epilepsy to date(5;12;30). It is therefore considered a good choice in patients with epilepsy and is not cautioned or contraindicated in these patients(5;31).

Miscellaneous Antidepressants

Rare cases of seizures with reboxetine (a noradrenaline reuptake inhibitor) have been reported in clinical studies but it has been tested in few patients with epilepsy; therefore it should be used with care in PWE (5;32).One study showed reboxetine (mean dose 6.9mg) to be an effective treatment without causing an increase in seizure frequency in 15 patients with temporal lobe epilepsy, when used for up to 30 weeks (26). Premarketing clinical trials suggest a seizure rate of 0.2% (11). Reboxetine is metabolised by CYP3A4 and inducers, such as phenobarbitone, carbamazepine and phenytoin, may reduce reboxetine plasma levels (32).

Mirtazapine(a centrally active presynaptic alpha-2 antagonist) is also associated with a minimal incidence of seizureswith an incidence of less than 0.1% with a 30mg dose suggested(33). However, one grand mal seizure has been reported in a patient with a history of seizures receiving mirtazapine at a dose of 80mg/day during a trial(12) (maximum licensed dose in UK is 45mg/day(34)). A second patient, with no history of seizures, developed generalised tonic-clonic seizures following initiated of mirtazapine 15mg/day, which stopped spontaneously (35). One study however showed mirtazapine (mean dose 32mg) to be an effective treatment without causing an increase in seizure frequency in 27 patients with temporal lobe epilepsy, when used for up to 30 weeks (26). In another small study, mirtazapine30mg/day for 3 weeks was found to enhance cortical excitability in seven patients with epilepsy, but there was no increase in seizure frequency, duration or severity. The small numbers and short duration of the study limit the power of the study in this respect(36). Although clinical experience indicates that seizures are rare during mirtazapine therapy, caution is advised in patients with epilepsy (34). In addition, there is a significant interaction with carbamazepine and phenytoin (CPYP3A4 inducers), resulting in large reductions in mirtazapine plasma concentrations of 60% and 45% respectively; the mirtazapine dose may need to be increased(34).

Agomelatine (a melatonergic agonist and 5-HT antagonist) appears to have no cautions or contraindications with regards to use in epilepsy (37) but there are very limited data regarding its safety in PWE.

Other factors to consider

Drug Interactions

Some antidepressants may interact with anticonvulsants; this potential for interactions would need to be considered when choosing a suitable antidepressant agent for the patient. Plasma concentrations of the anticonvulsants should be carefully monitored if appropriate, particularly during the early phase of treatment with the antidepressant and doses adjusted accordingly(5). NICE recommends that generally SSRIs should be first-line treatment for depression in people with a chronic physical health problem and of the SSRIs, sertraline and citalopram probably have the lowest interaction potential(38).

Dose initiation

As a general rule, the more sedating a drug is, the more likely it is to induce seizures(5). There is a dose dependent relationship between antidepressant use and seizures. Patients should be commenced on a low dose, and this should be increased slowly until a therapeutic dose has been achieved, to reduce the risk of seizures(5). In addition, maximum recommended doses of antidepressants should not be exceeded.

Limitations

In general, very little specific information is available about the use of antidepressants inPWE. The SSRIs are the most widely studied antidepressants in PWE.
Seizure rates quoted are often based on the incidence in patients with depression rather than PWE who suffer from depression, and this latter figure would be expected to be higher. Most studies investigating the relationship between antidepressants and seizure activity have examined the effects of antidepressants in overdose in the general population, making it difficult to extrapolate the findings to PWE.
Seizure frequency often fluctuates and can increase as a result of various clinical conditions, such as fatigue or non-compliance with AEDs, and it is sometimes difficult to attribute the observed aggravation of seizures only to antidepressants in the absence of controls (10).
Many of the early estimates of seizure incidence were based on case reports where there were many influencing variables and poor definitions for what constituted a convulsive event. Patients were also often taking concomitant medication, which may have affected seizure threshold, or, through inhibition of antidepressant metabolism, caused higher antidepressant plasma levels.
Newer antidepressants have undergone clinical trials with improved methodology and with systematic reporting of adverse events. However, even these do not allow accurate comparisons to be made between the antidepressants.
There is no large study comparing seizure frequency in PWE who take antidepressants with those who do not (10). Unless a large-scale trial is conducted, which is unlikely, the best antidepressant in epilepsy will remain unknown.
Detailed information regarding drug interactions has not been included in this Q&A: drug interactions between antidepressants and antiepileptic drugs must be checked prior to prescribing.

References

(1) Ojong M, Allen SN. Treatment of depression in patients with epilepsy. U S Pharmacist 2012; 37(11):29-32.

(2) Cardamone L, Salzberg MR, O'Brien TJ et al. Antidepressant therapy in epilepsy: can treating the comorbidities affect the underlying disorder? Br J Pharmacol 2013; 168:1531-1554.

(3) Muzerengi S, Moor CC. What do we know about depression in people with epilepsy? Prog Neurol Psychiatry 2013; March/April:20-24.

(4) Kondziella D, Asztely F. Don't be afraid to treat depression in patients with epilepsy! Acta Neurol Scand 2009; 119:75-80.

(5) Taylor D, Paton C, Kapur S. Chapter 7: Use of psychotropic drugs in special patient groups. The Maudsley Prescribing Guidelines in Psychiatry. Eleventh edition. Oxford: Wiley-Blackwell, 2012: 419-423.

(6) Noe HK, Locke DEC, Sirven JI. Treatment of depression in patients with epilepsy. Curr Treat Opt Neurol 2011; 13:371-379.