Antiepileptic drugs - Clinical Tree

See also individual agents

General information

Some of the drugs that are used to treat epilepsy can be grouped into classes (carbamazepine and its analogue oxcarbazepine, barbiturates, hydantoins, benzodiazepines, and succinimides), while others (for example valproate, felbamate, gabapentin, levetiracetam, lamotrigine, tiagabine, topiramate, vigabatrin, and zonisamide) stand on their own. Individual drugs differ in their spectra of activity and in adverse effects profiles. For adverse effects that occur with all drugs, frequency and severity vary from one agent to another: for example, sedation is more common with barbiturates and benzodiazepines, ataxia and diplopia are more common with phenytoin and carbamazepine, and aplastic anemia is more common with felbamate. Certain adverse effects are related to specific properties shared only by certain drugs: for example, renal stones can occur with drugs causing carbonic anhydrase inhibition (acetazolamide, topiramate, zonisamide), whereas reduced efficacy of oral contraceptives can occur with inducers of isoenzymes that metabolize these steroids (carbamazepine, oxcarbazepine, phenytoin, barbiturates, felbamate, and topiramate).

The clinical pharmacology and adverse effects of some new antiepileptic drugs (ganaxolone, levetiracetam, losigamone, pregabalin, remacemide, rufinamide, stiripentol, and zonisamide) have been reviewed [ ].

The uses and adverse effects of antiepileptic drugs in the treatment of painful peripheral neuropathy have been reviewed [ ].

Although many believe that some modern antiepileptic drugs are better tolerated than older ones, the authors of the ILAE Treatment Guidelines have suggested that statistically this has been very hard to show, except in a few studies [ ]. Many of these studies were designed to support marketing strategies, and some of the methods used in these trials can skew the results in favor of the sponsor’s product [ ]. For example, choice of inclusion and exclusion criteria, choice of comparator drug and formulation (modified-release or not), dosing intervals, titration rates, and end-points can influence outcome [ ].

Hypersusceptibility adverse reactions to antiepileptic drugs have been reviewed and their mechanisms of action and risk factors have been described [ ]. These reactions include (1) immune-mediated hypersensitivity reactions, (2) reactions involving unusual non-immune-mediated individual hypersusceptibility, often related to abnormal production or defective detoxification of reactive cytotoxic metabolites and (3) so-called “off-target” pharmacology, whereby a drug interacts directly with a system other than that for which it is intended (reactions that are classified as collateral adverse reactions in the DoTS framework [ ].

Comparisons of different antiepileptic drugs

Quality of life

The effects of carbamazepine and lamotrigine on health-related quality of life have been compared for 1 year in 260 patients with newly diagnosed epilepsy randomized to 48 weeks of treatment [ ]. Patients taking carbamazepine had significantly worse quality of life at week 4 but not later. They also had more cognitive adverse effects in general and more changes in energy and affect during the first 4 weeks of treatment.

Cost-effectiveness

The cost-effectiveness of four antiepileptic drugs used to treat newly diagnosed adult epilepsy has been studied by cost minimization analysis in 12 European countries [ ]. The analysis took account of each drug’s adverse effects and tolerability profiles. Lamotrigine incurred higher costs than carbamazepine, phenytoin, and valproate, whose costs were similar.

Withdrawal of therapy

Gabapentin, lamotrigine, topiramate, and vigabatrin have been compared using Kaplan–Meier survival analysis in 61 patients to see how long they chose to keep taking each drug and, if they stopped, why they stopped [ ]. The results are shown in Table 1 . Lamotrigine seemed to be the best tolerated of the four drugs and topiramate the least. These results have been mirrored by those of two larger retrospective studies [ , ].

Table 1

Persistence with therapy with different antiepileptic drugs in different studies

	Gabapentin	Lamotrigine	Tiagabine	Topiramate	Vigabatrin
Number of subjects [ ]	36	37	–	28	26
Median time to 50% drop out (months)	13	> 43	–	9.5	29
Withdrawn owing to lack of efficacy (%)	58	24	–	25	62
Number of subjects [ ]	146	122	88	70	37
Withdrawn owing to lack of efficacy (%)	25	16	30	30	46
Withdrawn owing to adverse effects (%)	16	15	26	42	16
Number of subjects [ ]	158	424	–	393	–
Withdrawn owing to lack of efficacy (%)	39	34	–	19	–
Withdrawn owing to adverse effects (%)	37	22	–	40	–

Gabapentin and vigabatrin as first-line add-on treatments have been compared in 102 patients with partial epilepsy [ ]. The improvement rate was 48% with gabapentin and 56% with vigabatrin. There were seven withdrawals in each group because of adverse events. Of the serious adverse events only one was thought to be drug-related—depression and weight gain in a patient taking vigabatrin.

In a comparison of carbamazepine and lamotrigine for trigeminal neuralgia in 18 patients with multiple sclerosis, lamotrigine was more effective [ ]. After withdrawal of carbamazepine, drowsiness resolved in 16 patients; cerebellar signs improved partially in five patients and completely in two; brainstem signs improved partially in four patients and completely in 3; ambulation improved in 11. In one patient taking lamotrigine a skin rash forced withdrawal.

General adverse effects and adverse reactions

The most important adverse effects of antiepileptic drugs affect the central nervous system and include sedation, fatigue, dizziness, cognitive dysfunction, ataxia, dysarthria, nystagmus, and headache. These effects are often dose-related; they are more prominent with multiple drug therapy and they are usually reversible after dosage adjustment. Behavioral disturbances are relatively common, especially in children and in patients with preexisting mental handicap. Exacerbation of seizures and psychiatric reactions are not uncommon. Hepatotoxic reactions have especially been reported with felbamate, valproate, carbamazepine, and phenytoin. Endocrine and metabolic changes occur with most drugs, but their clinical relevance is usually limited. Carbamazepine, phenytoin, barbiturates, and to a lesser extent felbamate, topiramate, and oxcarbazepine, are enzyme inducers, whereas felbamate and valproate are enzyme inhibitors. These effects cause significant drug interactions. Most anticonvulsants precipitate attacks in patients with acute intermittent porphyria.

The use of antiepileptic drugs (gabapentin, lamotrigine, and topiramate) as mood stabilizers has been reviewed [ ]. The authors concluded that the benefit to harm balances of these drugs have not been well enough established for their routine use in bipolar disorder.

Hypersusceptibility reactions

Hypersusceptibility reactions can involve any system, but they most often affect the skin, leading to drug withdrawal in up to 20% of patients. Aplastic anemia and hepatotoxicity have drastically curtailed the use of felbamate.

Tumorigenicity

Pseudolymphoma and a condition resembling malignant lymphoma occur very rarely with phenytoin. There is no evidence of a significant increase in the incidence of other tumors.

Second-generation effects

The use of older antiepileptic drugs in pregnancy is associated with a two- to three-fold increase in the risk of fetal malformations, including facial clefts and cardiac defects. Neural tube defects, including spina bifida, are seen in 2–3% of offspring exposed to valproate and in 1% of those exposed to carbamazepine. There is some evidence that fetal exposure to barbiturates and possibly phenytoin can cause impaired postnatal mental development, but in most studies it has been difficult to discriminate the effects of drugs from those of genetic and environmental factors. There are insufficient data to assess fetal risks after exposure to the newer drugs.

Drug studies

Comparative studies

The long-term efficacy and tolerability of antiepileptic drugs in patients with newly diagnosed epilepsy needs to be evaluated in comparative studies. Two randomized, unblinded, long-term studies have been published. In the first study carbamazepine, gabapentin, lamotrigine, oxcarbazepine, and topiramate were compared in 1721 patients with epilepsy for whom carbamazepine was deemed to be standard treatment (patients with partial epilepsies) [ ]. In the second study, 716 patients for whom valproate was considered to be standard treatment were randomized to valproate, lamotrigine, or topiramate [ ]. One of two primary outcomes was time to treatment failure, which is a mixed measure of efficacy and tolerability.

In the first study, time to treatment failure was significantly better for lamotrigine than for carbamazepine (HR = 0.78; 95% CI = 0.63, 0.97), gabapentin (0.65; 0.52, 0.80), and topiramate (0.64; 0.52, 0.79), and there was a non-significant advantage compared with oxcarbazepine (1.15; 0.86, 1.54). Adverse events occurred in 45–53% (lamotrigine 45%, topiramate 53%). The most common adverse effect associated with treatment failure was rash (7% of patients taking carbamazepine, 6% of those taking oxcarbazepine, and 3% of those taking lamotrigine).

In the second study valproate was significantly better than topiramate (HR = 1.57; 95% CI = 1.19, 2.08), but there was no significant difference between valproate and lamotrigine (1.25; 0.94, 1.68). The adverse events associated with treatment failure were most commonly psychiatric symptoms, cognitive symptoms, tiredness, and fatigue, all of which were more common with topiramate. For lamotrigine, rash was the most common symptom associated with treatment failure (4% of patients randomized), whereas for valproate weight gain was the most common symptom (4% of patients randomized).

In a multicenter, randomized, double-blind comparison of diazepam (0.15 mg/kg followed by phenytoin 18 mg/kg), lorazepam (0.1 mg/kg), phenobarbital (15 mg/kg), and phenytoin (18 mg/kg) in 518 patients with generalized convulsive status epilepticus, lorazepam was more effective than phenytoin and at least as effective as Phenobarbital or diazepam plus phenytoin [ ]. Drug-related adverse effects did not differ significantly among treatments and included hypoventilation (up to 17%), hypotension (up to 59%), and cardiac rhythm disturbances (up to 9%).

Placebo-controlled studies

Efficacy and tolerability data from double-blind, placebo-controlled add-on trials of new antiepileptic drugs in patients with refractory partial epilepsy have been reviewed [ ]. Although there were differences in adverse events profiles among the various drugs, the review identified major methodological problems, which hamper comparisons across studies and drugs. These included variability in the use of COSTART terminology, marked differences in the occurrence of specific adverse events in the placebo groups (an indication of heterogeneous evaluation procedures), and the use of non-optimal dosages or non-optimal titration schedules in many trials [ ].

Systematic reviews

In a meta-analysis of the most frequent treatment-related central nervous system adverse effects of new antiepileptic drugs from double-blind, add-on, placebo-controlled studies in adults with epilepsy 36 suitable studies were identified [ ]. No meta-analysis was possible for oxcarbazepine and tiagabine. Gabapentin was significantly associated with somnolence and dizziness; lamotrigine with dizziness, ataxia, and diplopia; levetiracetam with somnolence; pregabalin with somnolence, dizziness, ataxia, and fatigue; topiramate with somnolence, dizziness, cognitive impairment, and fatigue; zonisamide with somnolence and dizziness.

Organs and systems

Cardiovascular

Cardiac dysrhythmias induced by anticonvulsants are rare and occur mainly in patients other than those known to be at high risk of sudden death [ ]. Phenytoin has been rarely associated with bradydysrhythmias, almost exclusively after intravenous dosing, and some of these have been fatal. Hypotension can also complicate intravenous phenytoin. Carbamazepine can depress cardiac conduction, mostly in elderly or otherwise predisposed patients. Third-degree atrioventricular block occurred in one patient with pre-existing right bundle branch block treated with topiramate, but a cause-and-effect relation was uncertain [ ].

Respiratory

Respiratory adverse effects are extremely rare, apart from respiratory depression associated with high-dose benzodiazepines or drug overdose.

Nervous system

Most major anticonvulsants can cause cerebellovestibular and oculomotor symptoms (ataxia, dysarthria, dizziness, fatigue, tremor, diplopia, blurred vision, and nystagmus), alterations in cognitive function, and disorders of mood and behavior. Less common effects include parkinsonism (almost exclusively with valproate), exacerbation of seizures, headache, dyskinesias, and dystonias. Neurophysiological evidence of peripheral neuropathy may be common, but neuropathic symptoms are relatively rare. Monoplegia, Babinski reflexes, restless legs syndrome, and retinal/optic nerve disorders are very rare (except for vigabatrin-induced asymptomatic visual field defects, which are relatively common). Neurological adverse effects are usually dose-dependent and more prominent in patients on multiple drug therapy, although it has been suggested that neurotoxicity relates more to total drug load (in terms of sum of defined daily doses for each drug) than to the actual number of drugs taken [ ].

In some cases, seizure exacerbation occurs as a manifestation of drug intoxication, and is reversible on dosage reduction or elimination of unnecessary polypharmacy [ ]. In other cases, seizure exacerbation reflects an adverse reaction to a given drug in specific seizure types or syndromes. Carbamazepine in particular can precipitate or exacerbate a variety of seizures, most notably absence, atonic, or myoclonic seizures, especially in children with generalized epilepsies characterized by bursts of diffuse and bilaterally synchronous spike-and-wave EEG activity. Aggravation of seizures has also been reported with phenytoin and vigabatrin, particularly in children with generalized epilepsies. Gabapentin has been implicated in precipitating myoclonic jerks, while benzodiazepines occasionally trigger tonic seizures, particularly when they are given intravenously to patients with Lennox–Gastaut syndrome. Evidence that ethosuximide predisposes to tonic–clonic seizures remains inconclusive.

Experimental or clinical evidence of polyneuropathy, sometimes with paresthesia, has been found in up to 50% of patients treated chronically with carbamazepine, phenytoin, phenobarbital, and/or valproate [ ], but it is usually not associated with troublesome symptoms.

The risk of aggravating juvenile myoclonic epilepsy with carbamazepine and phenytoin has been assessed in a retrospective study of 170 patients, of whom 40 had taken carbamazepine or phenytoin [ ]. There was aggravation of seizures in 23 patients, 6 benefited, and there was no effect in the other 11. Of the 28 patients who used carbamazepine, 19 had aggravated symptoms, including myoclonic status in 2. Of the 16 patients who used phenytoin, 6 had aggravated symptoms, including one in association with phenobarbital. Vigabatrin was given in only one case, in association with carbamazepine, and provoked mixed absence and myoclonic status.

Antiepileptic drugs have sometimes been associated with a paradoxical increase in seizures. The evidence for this comes from isolated reports and clinical impressions. Somerville asked five pharmaceutical companies responsible for the development of new antiepileptic drugs to provide data concerning increases in seizure frequency during randomized, placebo-controlled, add-on trials in patients with uncontrolled partial seizures [ ]. Seizure frequency in individual patients taking the active drug or placebo was compared with the baseline pretreatment seizure frequency. More than 40% of the patients in trials of tiagabine, topiramate, and levetiracetam had an increase in seizures while taking a placebo. Increased seizure frequency was no more likely to occur when they were taking any of the three drugs than when they were taking placebo. A doubling or more of seizure frequency was significantly less likely to occur with topiramate or levetiracetam than with placebo, but more likely with tiagabine. There was some evidence of a dose–response effect with tiagabine, but a negative effect with topiramate (aggravation less likely with increasing dose). Unfortunately, the author did not obtain data on gabapentin and lamotrigine. Thus, aggravation of seizures in patients using some of the new antiepileptic drugs occurs no more often than with placebo and probably represents spontaneous fluctuation of seizure frequency.

Retrospective studies have suggested that antiepileptic drugs can be associated with peripheral nerve dysfunction. This has been studied prospectively in 81 patients (aged 13–67 years) without polyneuropathy who took sodium valproate (n = 44) or carbamazepine (n = 37) as monotherapy in standard daily doses [ ]. After 2 years one patient had clinical signs of polyneuropathy and six patients had symptoms of polyneuropathy, but electrophysiology did not show significant changes or trends. Only one patient had abnormal electrophysiological findings, which were only subclinical, and eight patients had abnormal values at two subsequent visits. There were no consistent patterns, and the data were unaffected when the drugs were examined separately or when patients were grouped according to whether or not they had symptoms of polyneuropathy. The authors concluded that previously untreated young to middle-aged patients who take valproic acid or carbamazepine for 2 years are not at risk of polyneuropathy.

Sensory systems

Visual field defects associated with various antiepileptic drugs (carbamazepine, diazepam, gabapentin, phenytoin, tiagabine, and vigabatrin) have been reviewed [ ]. The true frequency is unknown, but in a retrospective study in 158 patients with partial epilepsy visual field defects were detected in 21 (13%); 13 patients had concentric visual field constriction without subjective spontaneous manifestations. Of these 13 patients, 9 were taking vigabatrin.

Visual-evoked potentials and brainstem auditory-evoked potentials have been measured in 58 children and adolescents taking carbamazepine, phenobarbital, or sodium valproate monotherapy and 50 sex- and age-matched controls [ ]. After 1 year the patients taking carbamazepine had significantly prolonged visual-evoked P100 latencies compared with both baseline and control values; they also had significantly prolonged peak latencies of auditory waves I–III–V and interpeak interval I–V. Those taking sodium valproate had significantly prolonged visual-evoked P100 latencies. In contrast, children taking phenobarbital had no changes.

In 100 epileptic patients aged 8–18 years taking carbamazepine or valproate in modified-release formulations either alone or with added vigabatrin interpeak latencies of I–III and III–V of brainstem-evoked potentials were significantly delayed and N75/P100 and P100/N145 amplitudes in the visual-evoked potentials were reduced [ ]. However, the addition of vigabatrin did not worsen the effects caused by the other two drugs alone.

Psychological, psychiatric

Behavioral and psychiatric disturbances are not uncommon [ ]. Although epilepsy is itself associated with an increased risk of such disturbances, drugs play an important role. Phenobarbital-induced behavioral disturbances, especially hyperkinesia, are especially common in children, with an incidence of 20–50% and need for drug withdrawal in 20–30% of cases, whereas it is unclear whether and to what extent adults are affected.

Psychological effects

Clinically important cognitive effects of anticonvulsants have been investigated in a double-blind, parallel group, randomized, placebo-controlled study of anticonvulsant drug withdrawal in subjects with completely controlled seizures taking a single anticonvulsant [ ]. Drug withdrawal was associated with significant improvement in performance on the Controlled Oral Word Association Test and the Stroop Colour-Word Interference Test.

Psychiatric effects

Among older drugs, valproic acid and carbamazepine are least likely to cause adverse psychiatric effects, though valproate rarely causes encephalopathy and reversible pseudodementia. Phenytoin has been implicated in psychiatric adverse effects with or without other signs of toxicity, and at serum concentrations above or below the upper limit of the target range, but the actual incidence of these reactions is unknown. Benzodiazepines can cause paradoxical excitation, particularly in children and in anxious patients, and several other psychiatric symptoms can complicate the benzodiazepine withdrawal syndrome. Psychiatric or behavioral disorders have been reported with ethosuximide, but the lack of systematic studies prevents assessment of incidence and cause-and-effect relation. Among newer drugs, vigabatrin has been implicated most commonly in psychiatric adverse effects. With gabapentin, lamotrigine, and levetiracetam aggressiveness or hyperactivity can occur, especially in patients with previous behavioral problems or learning disability. Adverse psychiatric reactions to lamotrigine are uncommon, whereas with topiramate, felbamate, and other new drugs information is still insufficient.

Overall, the problem of drug-induced psychiatric disorders can be minimized by avoiding unnecessarily large dosages and drug combinations and by careful monitoring of the clinical response. In patients with a previous history of psychiatric disorders, carbamazepine and valproate are the first-line drugs, and are least likely to cause behavioral disturbances. The ideal management of such disturbances is withdrawal of the offending agent. When continuation of treatment is necessary for seizure control, psychosocial intervention and psychotropic medication can be useful.

In a retrospective study of 89 patients who developed psychiatric symptoms during treatment with tiagabine, topiramate, or vigabatrin, the psychiatric problem was either an affective or a psychotic disorder (not including affective psychoses) [ ]. All but one of the patients had complex partial seizures with or without secondary generalization. More than half were taking polytherapy. Nearly two-thirds had a previous psychiatric history, and there was a strong association between the type of previous psychiatric illness and the type of emerging psychiatric problem. Patients taking vigabatrin had an earlier onset of epilepsy and more neurological abnormalities than those taking topiramate.

Patients with chronic epilepsy have a higher likelihood of psychosis than the healthy population [ , ]. Psychosis is especially frequent in patients with temporal lobe epilepsy [ ]. Antiepileptic drugs have been reported to precipitate psychosis, although the literature is confounded by the inclusion of affective and confusional psychoses in this category. Moreover, the purported association has mostly been made through isolated case reports or small non-controlled case series. In fact, most antiepileptic drugs have been associated with psychosis: phenytoin and phenobarbital [ ], carbamazepine and valproate [ ], felbamate [ ], gabapentin [ ], levetiracetam [ ], topiramate [ ], vigabatrin [ ], and zonisamide [ ]. There have been no reports of psychosis associated with lamotrigine.

A retrospective chart review of 44 consecutive patients with epilepsy who had psychotic symptoms with clear consciousness has shown the difficulties in associating psychosis with drug effects [ ]. These patients were divided into two groups based on the presence or absence of changes in their drug regimen before the onset of the first episode of psychosis. In 27 patients the first episode of psychosis was unrelated to changes in their antiepileptic drug regimen, and in 23 of them the psychosis was temporally related to changes in seizure frequency. In 17 patients the first episode of psychosis developed in association with changes in their antiepileptic drug treatment, and in 12 of them the psychosis was temporally related to seizure attenuation or aggravation. This study therefore highlights the fact that psychosis can occur in relation to changes in seizure frequency, sometimes due to lack of effect of the new medication or to concomitant withdrawal of an efficacious medication.

Withdrawal of anticonvulsants with favorable mood stabilization properties, such as carbamazepine, has often been associated with acute psychosis [ , ]. Moreover, the phenomenon of “forced normalization,” by which complete seizure freedom in a patient with previous refractory epilepsy can lead to a psychotic state, may also contribute to the apparent association between drugs and psychosis [ ].

Information from double-blind studies of psychosis as an adverse event is relatively scarce. A double-blind, randomized, add-on, placebo-controlled trial with carbamazepine showed that there was no increase in chronic psychotic symptoms in patients with suspected temporal lobe seizures [ ].

The relation between psychosis and tiagabine has been assessed in an analysis of data from two multicenter, double-blind, randomized, placebo-controlled trials of add-on tiagabine therapy (32 or 56 mg/day) in 554 adolescents and adults with complex partial seizures during 8–12 weeks [ ]. There were psychotic symptoms (hallucinations) in 3 (0.8%) of 356 patients taking tiagabine and none of the 198 taking placebo, a non-significant difference. Thus, it appears that tiagabine does not increase the risk of psychosis, but the result is inconclusive.

An analysis of double-blind, placebo-controlled trials of vigabatrin as add-on therapy for treatment-refractory partial epilepsy showed that compared with placebo patients taking vigabatrin had a significantly higher incidence of events coded as psychosis (2.5% versus 0.3%) [ ]. There were no significant differences between treatment groups for aggressive reaction, manic symptoms, agitation, emotional lability, anxiety, or suicide attempts. In an open trial of topiramate, psychosis was seen in 30/1001 (3%) of the patients, and was severe enough to require withdrawal in eight [ ].

Should certain antiepileptic drugs be contraindicated in patients with active psychosis? Unfortunately there is not enough solid information to answer this question. Undoubtedly, anticonvulsants that are less likely to cause psychosis (lamotrigine, carbamazepine, oxcarbazepine, valproate) should be preferred [ , ]. However, patients with psychoses have been successfully treated even with drugs that are believed to be associated with psychosis, such as vigabatrin. For example, in a prospective study in 10 patients with psychosis and epilepsy to whom vigabatrin was added, there was no aggravation of the psychiatric disorder [ ].

The association of Alzheimer’s disease and all types of dementia with epilepsy and the use of antiepileptic drugs has been investigated in 5376 elderly people (aged 65 years or older) with no prior evidence of dementia, defined as a Modified Mini-Mental State score of at least 78 [ ]. Those who took antiepileptic drugs had a significantly higher risk of developing dementia but not Alzheimer’s disease. The association remained significant in those who took only phenytoin. Further investigation is warranted to determine whether it is indeed antiepileptic drug therapy or some underlying confounding pathology that is associated with the development of dementia in these patients.

The adverse effects of antiepileptic drugs on mood have been reviewed [ ]. The barbiturates, vigabatrin, and topiramate are more likely than other antiepileptic drugs to be associated with depressive symptoms, which are present in up to 10% of patients taking these drugs. Tiagabine, levetiracetam, and felbamate present an intermediate risk, with a prevalence of depression of about 4% or less. For zonisamide, the data are less clear, but it seems that mood disorders may occur in up to 7% of patients, even though in most cases slow titration can significantly reduce the risk. Phenytoin, ethosuximide, carbamazepine, oxcarbazepine, gabapentin, sodium valproate, pregabalin, and lamotrigine are all associated with low risks of depression (< 1%), and several of these have a positive effect on mood. Several mechanisms may be involved in this effect of antiepileptic drugs, including potentiation of GABA neurotransmission, folate deficiency, pharmacodynamic interactions, and forced normalization.

Epidemiological data show that the rate of suicide among patients with epilepsy is five-fold higher than in the general population, while in temporal lobe epilepsy it is even higher. It has been hypothesized that since disturbances of serotonin metabolism are involved in the pathogenesis of suicidal behavior, they also might be a common link between depression, suicidality, and epilepsy [ ]. The authors suggested that anticonvulsants with certain serotonergic properties should reduce the risk of suicidality, whereas drugs that lack serotonergic mechanisms would not be effective in preventing it. According to this hypothesis, phenobarbital and phenytoin should be drugs with a proven risk while carbamazepine, oxcarbazepine, valproate, and lamotrigine could be regarded as drugs with antisuicidal properties, because they all improve mood in epilepsy and have serotonergic mechanisms of action. Topiramate, tiagabine, vigabatrin, levetiracetam, and zonisamide all have negative effects on mood and cognition. However, zonisamide has negative psychotropic effects, even though it has serotonergic properties, whereas gabapentin has positive psychotropic effects on mood but is devoid of serotonergic properties.

Cognitive effects

The appropriate methods and timing in assessing cognitive and behavioral adverse events during drug development programs have been thoroughly reviewed [ ].

The authors of a critical review focusing on pediatric data concluded that adverse effects on learning and behavior may have been over-rated [ ]. Because of methodological flaws, many early studies could not discriminate between effects of drugs and the influence of heredity, brain damage, seizures, and psychosocial factors. In fact, the majority of children taking antiepileptic drugs do not experience major cognitive or behavioral effects from these medications. In some patients, however, drugs do produce detrimental effects, barbiturates and benzodiazepines being among those most commonly implicated [ ]. At least with some agents, such as gabapentin, behavioral adverse effects occur mainly in children with pre-existing learning disability. As to phenytoin, carbamazepine, and valproate which are the drugs most commonly recommended for first-line use, recent investigations have failed to show major differences in cognitive effects between these agents, although in some studies, patients taking phenytoin tended to have lower motor and information processing speeds [ , ].

Endocrine

Although most anticonvulsants interfere with endocrine function, epilepsy may do so itself, and it is difficult to differentiate the effects of drugs from those of the disease. In any case, symptoms of endocrine dysfunction are less common than biochemical abnormalities.

In women with epilepsy taking antiepileptic drug monotherapy, valproate monotherapy is associated with more obesity and metabolic syndrome while lamotrigine and topiramate may be safer in patients with a high risk of cardiovascular disease [ ].

Growth hormone

Normal growth hormone concentrations are found in carbamazepine-treated patients [ ]. Growth hormone secretion in response to levodopa stimulation was not affected by carbamazepine or phenobarbital, whereas phenytoin and anticonvulsant polytherapy caused an increase in growth hormone concentration, and valproate a fall at varying times after the administration of levodopa [ ]. Pubertal growth arrest has been described in a 12-year-old girl who had taken valproate for 18 months [ ].

Adrenal–pituitary axis

Neither phenytoin, valproate, carbamazepine, nor phenobarbital altered the circadian ACTH/cortisol rhythm in epileptic patients [ ]. In some studies, phenytoin and carbamazepine were associated with increased serum concentrations of unbound cortisol [ ], but cortisol concentrations were unaffected in other studies [ ] and Cushing’s syndrome has not been described with these drugs. Valproate can depress ACTH concentrations by inhibiting corticotrophin-releasing factor, and it has been used to treat Nelson’s syndrome [ ]. Serum concentrations of progesterone and cortisol and the excretion of 17-hydroxycorticosteroids have been reported to be lower in untreated patients with epilepsy and to be further reduced by phenytoin [ ].

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here