Treament of seizures requires antiepileptic drug (AED) treatment for the large majority of patients and may be accompanied by unwanted effects on cognitive function. Although the magnitude of such 'cognitive side effects' is generally considered to be mild to moderate for most of the AEDs, their impact may be substan tial in some patients when critical functions are involved such as learning in children (Aldenkamp et al., 1995«) or driving capacities in adults (often requiring milliseconds precision); or when functions are impaired that are already vulnerable, such as memory function in the elderly (Trimble, 1987). Moreover, as the cognitive side effects represent the long-term outcome of AEDs, the effects may increase with prolonged therapy, which contributes to the impact on daily life functioning in refractory epilepsies (Committee on Drugs, 1985). The following topics are relevant to clinical practice: The combined effects of seizures and AEDs on cognitive function Undoubtedly many controversies concern the relative contribution of AED therapy, compared to the effect of seizure activity on cognition. Improved seizure control (when for example a new AED is added into the existing drug regime) may cause cognitive improvement that itself may camouflage genuine cognitive side effects of the new drug. In many situations it will thus be impossible to separate seizure effects from 'genuine' AED effects. Subjective patient complaints may enlarge this problem, as patients often believe that their cognitive problems are caused by 'external' factors, such as the drugs they have to take instead of by 'internal' factors, such as their own seizures. Habituation
In most drugs 'early' side effects may occur, fortunately only for a short period, i.e. during the first few days or weeks of drug exposure. After this period normalization occurs, possibly due to the development of so-called positive tolerance or habituation (Kulig and Meinardi, 1977). Although little is known about how tolerance to the cognitive effects of AEDs develops, a failure to take this factor into account may lead to overestimation of the negative effects of drugs on cognition. In clinical assessment we should only conclude 'cognitive side effects' if these persist during long-term treatment. Subjective patient complaints
Most of the studies use formal neuropsychological tests to assess the cognitive side effects of AEDs, although in clinical practice the opportunities for such assessment are usually very limited. Patient complaints, suggestive of problems in cognitive behaviour often represent the only available evidence of possible cognitive dysfunction. Nevertheless, it would be unwise to take all patient complaints that suggest cognitive dysfunction at face value. While some patients may have a clear insight into their own failures, others may have a poor understanding of their performance. Thus, although subjective cognitive complaints are an important factor to be considered, their use may generate discrepancies between patient complaints and the outcome of cognitive tests, a situation that, in our experience, occurs fairly regularly in clinical practice and is a frequent cause for controversies. When changing to another drug, patients may be 'doing better but feeling worse', and vice versa, a topic that has been discussed since it was raised in 1890. A possible solution is to use standardized scales to assess the subjective impressions of patients. Several of these scales with proven validity now exist (e.g. the ABNAS neurotoxicity scale; Aldenkamp and Baker, 1997; Aldenkamp et al., 1995b).
The specificity of cognitive side effects of AEDs came back into debate recently when several studies failed to find earlier reported cognitive side effects of AEDs when serum concentrations were sufficiently controlled (Meador et al., 1990). When Dodrill and Troupin initiated their research on cognitive side effects of AEDs in 1977 (Dodrill and Troupin, 1977), they reported significant drug-induced impairment in patients using phenytoin. However, no impairment remained, when they removed all patients with phenytoin serum levels exceeding the upper limit of the therapeutic reference interval (30 g/ml). This was done in a reanalysis of their original work and published about 15 years later (Dodrill and Troupin, 1991).
For other types of AEDs the effects of higher serum levels seem to be milder, but were nonetheless also found for carbamazepine and for valproate. The implication of these findings is that some of the reported drug-induced cognitive impairments and differences between drugs actually may have been due to differences in drug concentrations during the study, rather than representing specific effects on cognition. Although the value of routine serum level monitoring in clinical treatment is criticized it may be helpful to control the serum levels when using high dosing if patients have cognitive complaints.
In addition to the effect of dose, the pharmacokinetic properties of antiepi-leptic drugs may have an effect on cognitive functioning. In carbamazepine, transient cognitive deficits have been detected in relation to high peak serum levels (Aldenkamp et al., 1987). The pharmacokinetic profile of this drug, characterized by rapid and marked fluctuations in serum levels, may differentially affect test performance across short periods during the day. In drugs with large differences between trough and peak levels, a part of the cognitive assessment procedure should be performed repeatedly during the day, allowing a comparison of performance at peak levels with other periods. Generics and branded formulations of one drug
It is sometimes suggested that different formulations of the same drug may have different effects on cognitive function (Crawford et al., 1996). This may especially be true for carbamazepine, due to differences in absorption rate and phar-macokinetics among the different formulations. In a recent study, we did not obtain significant differences in cognitive performance when patients were switched from the conventional branded form of carbamazepine to several generic formulations (Aldenkamp et al., 1998). Nonetheless, patients' performance should be monitored when frequent swiches are made between different formulations of an antiepileptic drug.
The specific effects of the AEDs on cognitive function have been assessed in a meta-analysis that evaluated 25 years of studies (Vermeulen and Aldenkamp, 1995):
Polypharmacy shows most impact on cognitive function as opposed to monotherapy, irrespective of the type of AEDs included. Two drugs with mild cognitive effects may show potentiation of tolerability problems and thus yield serious cognitive impairment, when used in combination (Trimble, 1987).
All established AEDs have 'absolute' cognitive side effects, i.e. all the investigated drugs have effects when compared to no treatment. These effects are larger for phenobarbitone (PHB) than for phenytoin (PHT), carbamazepine (CBZ) or valproate (VPA). But even the latter drugs, that are generally considered to be drugs with a safe cognitive profile, have cognitive effects, mostly resulting in a mild general psychomotor slowing (Aldenkamp et al., 1993).
The respective differences between the four most investigated AEDs: PHB, PHT, CBZ and VPA, can be considered as small, when studied within a normal therapeutic dose, with the exception of the cognitive effects of PHB that has a less favourable cognitive profile when compared to PHT, VPA and CBZ. Possibly the most marked findings were that all AEDs have some cognitive effects and that the effects of PHT are more moderate than has been suggested previously (Meador et al., 1990).
The cognitive side effects of each antiepileptic drug. Phenobarbitone is the only AED with specific absolute effects: when compared with no treatment memory functions are affected. In long-term treatment even impairment or delay of psychological development is reported, leading to intellectual deterioration. No information is available on dose-effects.
Phenytoin has a much milder impact on cognitive function than had hitherto been suspected. When compared with 'no treatment' PHT-induced attentional deficit and mental slowing is observed but these are not markedly different from similar effects of CBZ and VPA. No correlation with dose was found.
Carbamazepine has been compared with several drugs and some differences have been found in respect of VPA and PHT. There is some evidence that dose-effects may occur and that controlled-release formulations have a more favourable cognitive profile than conventional forms. No differences were found between different generic forms of CBZ.
Valproate has mild psychomotor slowing as an absolute effect and has a similar cognitive profile when compared with CBZ and PHT. In contrast with CBZ, dose-relationships have not been obtained and there is no effect when patients are switched from conventional to controlled-release formulations.
For the newer antiepileptic drugs there is a suggestion of equivalence with CBZ and VPA for oxcarbazepine, gabapentin and vigabatrin (although mood effects have been reported for vigabatrin). For the other newer drugs, lamotrigine and tiagabine, no data from controlled studies have been published yet. Clinical anecdotal information suggests mood effects during titration of tiaga-bine and attentional-deficits in relation to treatment with lamotrigine in children. A recent study (Aldenkamp et al., 2000) showed cognitive impairment in topiramate related to dose escalation speed.
Was this article helpful?