Because depression is common in the general population and occurs even more often in patients with epilepsy (up to 50%, depending on the screening tool used39), and because treatment of depression usually requires long-term drug treatment, the choice of antidepressant medication must be made carefully, taking into account the possible effect on seizure threshold. The choice of antidepressant medication should be based on the needs of the individual patient, however, not the small risk of a seizure or fear of drug interactions.
Evidence that several antiepileptic drugs (AEDs) also function as mood stabilizers means that epilepsy and depression in some patients can be managed using one drug. AEDs such as carbamazepine or valproic acid, however, may trigger depression or mania when discontinued, and barbiturates can initiate or exacerbate depression.
The risk of seizures in patients taking antidepressant drugs depends on three factors:41–44
The complex neurotransmitter effects of antidepressant drugs make it impossible to offer simplistic assumptions about their proconvulsant effects. Recent experimental studies of AEDs used to treat depression lead to the conclusion that it is unlikely that alterations in serotonin and norepinephrine levels are related to an increased risk of seizures.42 In fact, some studies suggest that fluoxetine40 and doxepine41 may occasionally have anticonvulsant properties.
By analyzing which transmitter systems are shared by proconvulsant drugs like local anesthetics, cocaine, phenothiazines, and tricyclic antidepressants, and then eliminating those with unique properties, Dailey and Naritoku concluded that it is the antihistaminic, antimuscarinic, and local anesthetic properties of antidepressants that are most likely responsible for any increased susceptibility to seizures.42 Even simple side effects of medication, however, such as drowsiness, can lower the seizure threshold in susceptible patients.
The overall incidence of seizures in patients taking antidepressants at reasonable doses and presumed therapeutic ranges is 0.1–4.0%; this can be compared to an incident seizure risk of 0.073–0.086% in the general population of Rochester, Minnesota.41 By class of drug, the highest risk of seizures has been reported with:
The tricyclic antidepressants (TCAs) have the next highest seizure risk (which is increased by toxic levels). The newer selective serotonin reuptake inhibitors (SSRIs)—fluoxetine, sertraline, fluvoxamine, citalopram, and paroxetine—and serotonin receptor modulators—trazodone and nefazodone—are intermediate in risk, as is venlafaxine, a combined norepinephrine and serotonin reuptake inhibitor. The monoamine oxidase inhibitors (MAOIs) carry the least risk.
The incidence is 0.2% for fluoxetine, paroxetine, fluvoxamine, venlafaxine, and sertraline (at the higher doses generally required for obsessive-compulsive disorder); less than 0.1% for imipramine and amitriptyline; and 0.04% for mirtazapine, an adrenergic antagonist that releases norepinephrine and serotonin and blocks some serotonin receptors.41 The newest agents have not been used long enough to make realistic estimates of their epileptogenic potential. Recent estimates of seizure incidence in patients taking antidepressant drugs, based on large series of patients, have declined, implying that figures from early case series are unreliable.39
Some seizures in patients using SSRIs may be caused by hyponatremia due to the syndrome of inappropriate secretion of antidiuretic hormone (SIADH).44,46 SIADH has also been reported with one of the newest antidepressants, reboxetine, a norepinephrine reuptake inhibitor.47
Blood levels can be obtained for most antidepressant drugs, and these should be used to establish the steady-state baseline concentrations associated with optimal treatment effects and, in the event of a seizure, to document the associated blood level.7 If a toxic level is found (or at least a level higher than baseline), the antidepressant drug can generally be continued after lowering the dose.50
Dose-effect relationships can be informative. For example, bupropion (now used more often as an aid to smoking cessation than as treatment for depression) has a risk of seizures as high as 2.2% in doses higher than 450 mg per day. The risk is as low as 0.4% with lower doses, and the slow-release formulation lowers seizure risk further still.39 Seizure rates with older antidepressants, such as maprotiline or amoxapine, are now much lower than earlier estimates,39 but their use has also decreased substantially because of the availability of safer alternatives. When clomipramine is used to treat obsessive-compulsive disorder, somewhat higher dosages are used than are typically used for depression. At doses above 350 mg per day, the risk of seizures is 2.1% but only 0.48% if the dose is kept below 250 mg per day.41,44
Seizures occur with severe overdoses of TCAs, just before cardiac arrest. Treatment should include alkalinization to avoid malignant arrhythmias,4 but seizures in such circumstances may be prolonged or recurrent, despite appropriate use of AEDs.8 At San Francisco’s poison control center, overdoses of imipramine and amitriptyline accounted for 29% of the patients who had seizures.5 Massive overdoses of newer antidepressant drugs, such as fluvoxamine in doses above 1500 mg, have also been associated with seizures.45
Seizures are a neurotoxic effect of lithium and can occur even with plasma concentrations within the usual therapeutic range. EEG abnormalities are predictive of seizures.48
The risk of seizures is increased in elderly people, probably owing to age-related reduction in metabolism and clearance that leads to drug accumulation, despite dosages well tolerated in younger patients.7
Toxic amounts of antidepressant drugs also can accumulate in younger people. The active drug concentration actually within the brain is a product of many overlapping factors:
A newly emphasized property of drugs is chirality: the beneficial or toxic effects of a drug depend on its chemical “handedness.” The racemic (L or D) form of a drug affects its metabolism and reuptake, which, in turn, results in variable effects on antidepressant efficacy and on seizure propensity.49
Enzyme-inhibiting antidepressant drugs, such as the SSRIs and TCAs, may require downward adjustment in the dosage of concurrently administered AEDs that are also metabolized by the cytochrome P-450 (CYP450) microsomal system in the liver. In other cases, phenytoin and carbamazepine inhibit the metabolism of antidepressant drugs, like fluoxetine or the TCAs, which leads to unintended toxicity from the antidepressant.40,41
Reference to drug interaction charts is necessary to anticipate the effects of one drug on another, as there are several isoenzymes of both the CYP450 and GT systems. In general, antidepressant and antipsychotic drugs are metabolized by CYP2D6 and CYP1A2, the enzyme systems that also metabolize carbamazepine (See Table 2: Antidepressants and their metabolizing enzymes). Most benzodiazepines and newer antipsychotic agents, as well as most other AEDs except phenytoin, use CYP3A4. Phenytoin and a few benzodiazepines use CYP2C9. Valproic acid is also metabolized by the glucuronyl transferase system of enzymes, as are the tricyclic drugs. Chlorpromazine and clozapine use CYP1A3; lamotrigine, tricyclics, and clozapine use CYP1A4; and lorazepam and most narcotic drugs use CYP2B7. Narcotics are also metabolized by the GT1A1 system.
Drugs can induce or inhibit individual enzymes; such information is generally readily available in the Physicians’ Desk Reference or other medication guides.51–54 Although it is tempting to use gabapentin and citalopram solely because of their “clean” metabolic profiles (they have no effect on hepatic enzyme systems),54 they are not free of either risks or side effects.44,54,55
When an AED is added to an antidepressant, induction of hepatic enzymes may decrease the antidepressant’s effectiveness.56 Sedating antidepressants may trigger seizures in patients whose seizure threshold is lowered by drowsiness.42
Because most drug interactions are predictable, they can usually be anticipated, especially with chronic use. The choice of antidepressant medication should be based on the individual's needs rather than the fear of drug interactions or the small risk of a seizure.
Although isolated case reports have described spontaneous seizures after electroconvulsive therapy (ECT),57 an increased incidence has not been confirmed by larger series.58 A seizure disorder is no longer a contraindication to ECT. In patients with epilepsy who require ECT, maintenance AEDs should be withheld the day of treatment to obtain the best response.44
Newer technologies for treating depression, such as transcranial magnetic stimulation and vagus nerve stimulation (VNS), have not been associated with any increased risk of seizures.59
Reviewed and revised March 2004 by Steven C. Schachter, MD, epilepsy.com Editorial Board.
© 2014 Epilepsy.com. All rights reserved.