Rajesh C. Sachdeo MD*
Roopal M. Karia MD**
* Clinical Professor of Neurology, Department of Neurology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School; and Director, Comprehensive Epilepsy Center, Robert Wood University Hospital, New Brunswick, New Jersey
** Fellow and Instructor, Department of Neurology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, New Jersey
New antiepileptic drugs (AEDs) are approved on the basis of controlled clinical trials in a limited number of patients. Although these types of trials do enable us to see the most common emerging side effects of the new medication, it is really with postmarketing surveillance, when the drug has had a large patient population exposure, that we can begin to appreciate its full adverse potential.
Most second-generation anticonvulsants are fairly well tolerated; nonetheless, there are certain well-recognized side effects. Topiramate, currently in use in 750,000 patients worldwide, is no exception. As is the case with most AEDs, most adverse events are related to the central nervous system, although the spectrum of potential toxicities associated with topiramate use is quite broad. Other systems affected include renal, metabolic, and endocrine.
CENTRAL NERVOUS SYSTEM
The most common and most important central nervous system side effect is cognitive slowing. A prospective study designed to evaluate cognition was performed in a group of healthy adults who were randomized to one of three drug groups: topiramate, lamotrigine, and gabapentin (1). The subjects were assessed for attention, psychomotor speed, memory, language, mood, concentration, and cognition. After single doses, topiramate (5.7 mg/kg) decreased performance in attention and word fluency tests, whereas lamotrigine (3.5 mg/kg) and gabapentin (35 mg/kg) had no effect. After 4 weeks of multiple doses, topiramate (5.7 mg/kg) was still associated with impairment in verbal memory and psychomotor speed, whereas lamotrigine (7.1 mg/kg) and gabapentin (35 mg/kg) did not cause any impairment. In an add-on study conducted in patients receiving carbamazepine comedication, topiramate was also titrated rapidly by 50 mg/day, at weekly intervals up to 400 mg/day and compared with valproate, which was titrated up to 2, 250 mg/day (2). In this study, the topiramate-treated group had slightly more psychomotor slowing, speech problems, confusion, and mood problems in comparison with the valproate-treated group, but after 20 weeks, the effects of the two drugs did not differ significantly in most patients. Another study with a more gradual dose-escalation (25 mg weekly) paradigm demonstrated a higher dropout rate in the topiramate group compared to the valproate group because of side effects. At the end of the maintenance phase of this trial, changes in one of 10 cognitive function variables (short-term verbal memory) were significantly worse for the topiramate-treated group (3).
Patients taking topiramate may have word-finding difficulty as well as difficulties using words, and they tend to repeat themselves. In a meta-analysis on the safety of topiramate based on double-blind controlled studies, many of which used target doses higher than those currently recommended, “abnormal thinking” was noted in 25% to 33% of patients (4). This term encompassed slowed thoughts, difficulty in calculating, dulled thinking, difficulty with calculation, slowness in response, and blunted mental reaction. Burton et al. (5) studied topiramate's effects specifically on attention, which was assessed by way of digit span. Subjects were tested at weekly intervals over a 3-month period, and results were compared with those of the control group. These investigators found that four of nine patients had problems with attention, especially at higher drug doses (5). Most of these side effects are seen during the first 2 months of initial titration period, they are often transient, or they abate after dosage reduction (6). Although their incidence can be reduced with a low starting dose and a slower dose-escalation rate, these effects continue to be seen in everyday practice, and their importance cannot be overemphasized.
Other side effects of topiramate include somnolence, fatigue, and slurred speech (7). These effects were notably seen in a double-blind, randomized trial studying the use of
topiramate in Lennox-Gastaut syndrome (8). In the initial randomized placebo-controlled trials (9, 10, 11, 12, 13) in adults with refractory partial-onset seizures, topiramate doses of 100 mg/day were added to the other AEDs and were titrated at 100 to 200 mg/day on a weekly basis, which is faster than currently recommended. The most common adverse effects seen with a 10% higher rate than in the placebo-treated group were dizziness, somnolence, psychomotor slowing, ataxia, and memory difficulty. Treatment withdrawals were more commonly seen in the higher-dosage groups. In a pooled analysis of data from 527 patients treated with topiramate and 216 patients taking placebo, the central nervous system side effects associated with topiramate were mild to moderate and included speech problems, memory difficulty, somnolence, fatigue, and dizziness (14).
The use of topiramate in children is also associated with significant side effects. In a 3-year retrospective analysis of 51 children and adolescents ranging in age from 3 to 16 years, most of the patients had the drug withdrawn because of adverse effects. Of the 51 patients, 57% experienced side effects including cognitive difficulties (15). Other adverse events noted in children include dizziness, fatigue, ataxia, and confusion (16). Moreland et. al. followed-up 49 children with intractable epilepsy. In these patients, the most efficacious dose of topiramate was between 2.5 and 7.5 mg/kg/day. More then 50% of the children experienced side effects that could interfere with their learning ability at school. Hence, the authors recommended that although topiramate may be efficacious, patients taking the drug should be closely watched for a possible decline in cognitive function (17).
Adverse events emerging when topiramate is added to other AEDs are not likely to reflect pharmacokinetic interactions because topiramate does not significantly alter plasma levels of concomitant AEDs. With topiramate, pharmacodynamic interactions appear to be a major contributor to treatment-related adverse events. This is demonstrated by the observation of a lower incidence of side effects with topiramate monotherapy (18,19).
Because topiramate is a carbonic anhydrase inhibitor, this mechanism may be the explanation for the occurrence of paresthesias, which have been reported in ≤15% of the patients who participated in the controlled clinical trials. The symptoms are mild and are often transient (20). However, in a monotherapy study involving 48 patients with partial seizures who were randomized to topiramate at 100 mg/day, topiramate at 1,000 mg/day, or placebo, paresthesias were present in 63% of the 100 mg/day group and 58% in the 1,000 mg/day group (21). In a another monotherapy study comparing topiramate with standard AEDs such as carbamazepine and valproate, the incidence of paresthesia was 25% in the topiramate-treated group (20).
Development of hemiparesis was reported in two patients during the first few weeks of topiramate therapy (22). The condition cleared after drug withdrawal. Both patients had preexisting cerebral damage, which may have facilitated the appearance of this unusual side effect.
Alterations in mood and behavior are common in patients with epilepsy, and AEDs may play a significant role in the pathogenesis of these disorders. In the already discussed study in health volunteers in whom topiramate was compared with lamotrigine and gabapentin, subjects allocated to topiramate had a higher incidence of depression 4 weeks after starting the drug (1). The topiramate-treated group also had more anger and hostility when these patients were evaluated for mood profile compared with the lamotrigine group. Overall analysis showed that the patients treated with topiramate had higher scores for depression, anger, and hostility at 4 weeks of starting treatment in comparison to baseline. As discussed earlier, however, the applicability of these data to the therapeutic situation is uncertain.
Psychotic episodes are not uncommon in patients with refractory epilepsy (23,24). The incidence of psychosis in patients taking topiramate was 0.8% in randomized controlled trials. In a separate cohort of ≤1,000 patients treated for ≤5.3 years, an incidence of 3% was reported (25,26). A retrospective chart review of 80 patients taking topiramate (50 to 400 mg/day) showed a 6% incidence of acute psychotic symptoms (25). A postmarketing study of 787 patients revealed an incidence of 1.3% (27). A prior history of psychotic symptoms was a risk factor in some, but certainly not all, of the patients. Symptoms of psychosis in most instances were thought to be related to topiramate, because rapid resolution occurred once topiramate was discontinued.
Renal calculi are a well-recognized concern with the use of topiramate. The reported incidence is about 1.5% (21), representing a two- to fourfold increase over the estimated occurrence in the general population. Most patients faced with this problem do not need surgery and go on to continue treatment with topiramate once the stone is passed (21). Renal calculi may be more common in men (28). The proposed mechanism for the calculus formation is the inherent action of topiramate as a carbonic anhydrase inhibitor. The result is reduced urinary citrate excretion, which, in turn, raises the urinary pH. It is proposed that the risk of stone formation from this process can be reduced with proper hydration. The concomitant administration of other carbonic anhydrase inhibitors such as acetazolamide and the use of diuretics may increase the risk.
As a carbonic anhydrase inhibitor, topiramate reduces serum bicarbonate levels. This, in turn, can lead to
reversible metabolic acidosis in some patients (29, 30, 31, 32, 33). The clinical expression of this in a pediatric population has been described as the central hyperventilation syndrome (33). Other reports of dyspnea on exertion were described in the context of topiramate use as migraine prophylaxis (34). In these otherwise healthy migraineurs, respiratory symptoms resolved with discontinuation of the drug, and these patients were not found to have any cardiopulmonary problems to account for the symptoms. Another clinical manifestation of reversible metabolic acidosis is encephalopathy. Such encephalopathies have been reported in patients taking topiramate and also receiving adjunctive therapy with valproate (35,36). Once again, the symptoms resolve when topiramate or valproate is discontinued.
That patients taking topiramate actually lose weight rather than experiencing weight gain, as with some other AEDs, has been widely observed. The average body weight decreases 2% to 7% (1.6 to 6.5 kg) (14). The mechanism by which this phenomenon happens is unclear, although an attempt has been made to investigate it through the use of animal models. Rats given topiramate showed decreased body fat as well as acutely reduced food intake and an increased metabolic rate. These animals also had decreased levels of total insulin, leptin, and corticosterone (37). The precise mechanism for all these changes remains to be determined.
Other investigators suggest that topiramate inhibits fat deposition. The activity of lipoprotein lipase is reduced in adipose tissue in topiramate-treated rats (38).
The reasons for the observed weight loss may, in fact, be mulifactorial and remain to be properly characterized. In patients taking topiramate, weight loss occurs early in the treatment and is maximal by 15 to 18 months. It appears to be the greatest in patients who are heavier at the onset and is most commonly seen in female patients (39).
As with most AEDs, topiramate may reduce the effects of oral contraceptive medications. Rosenfeld et al. (40) gave topiramate to 12 women who were taking valproate and a contraceptive containing a combination of norethindrone 1.0 mg/ethinyl estradiol 35 µg. Serum norethindrone, ethinyl estradiol, and progesterone levels were measured first with the patient taking valproate and the oral contraceptive pill alone (cycle 1). Once topiramate was added at three different doses (200, 400, and 800 mg/day over cycles 2, 3, and 4, respectively), the same tests were repeated. Ethinyl estradiol concentration decreased in a dose-dependent fashion during topiramate administration, with a 30% reduction seen at the highest topiramate dose (800 mg/day). Hence, in women taking topiramate who are also taking oral contraceptive medication, an oral contraceptive medication with a higher estrogen content should be preferably used (41).
PREGNANCY AND BREAST-FEEDING
It is estimated that approximately 1 million women with epilepsy are of childbearing age in the United States (42). In the postmarketing experience, cases of hypospadias have been seen in male infants exposed to topiramate in utero. This has been observed in monotherapy as well as with the use of other concomitant anticonvulsants; however, a causal relationship has yet to be determined, and overall available data remain insufficient to assess potential risks for the human embryo (43). Teratogenicity has been described in experimental animals. Problems include craniofacial malformations, low fetal body weight, and limb malformations such as ectrodactyly, micromelia, and amelia. Rats who were exposed to topiramate in late pregnancy and during lactation had offspring with poor physical development. Although AEDs have been implicated as a major cause of teratogenesis, the clinician should be aware that uncontrolled epilepsy is also a risk to the mother as well as to the infant (44).
In animal studies, topiramate was found to be secreted in breast milk (45). Most authorities advise weighing the risk:benefit ratio in patients taking topiramate who wish to breast-feed their infants.
Topiramate binds to erythrocytes more so than to plasma proteins, even though binding to red blood cells has been shown to be saturable (46). Nonetheless, there have not been any well-described changes in hematologic parameters with the use of this AED (47).
HYPERSENSITIVITY AND IDIOSYNCRATIC REACTIONS
The so-called anticonvulsant hypersensitivity syndrome can manifest as a rash, lymphadenopathy, fever, hepatitis, and/or eosinophilia. To date, this kind of a reaction has not been observed with the use of topiramate (48). In controlled clinical trials, the incidence of rash was similar in topiramate-treated groups when compared with placebo-treated groups.
In controlled trials, no clinically significant changes in the mean values of clinical laboratory tests, including liver and renal function, were observed. In addition, there were
no treatment-related changes in electrocardiographic, ophthalmologic, or audiometric parameters.
Since then, a few case reports have described problems in clinical use. A brief case report (49) described the development of liver failure requiring a transplant in a female patient receiving topiramate (300 mg/day) who was also receiving carbamazepine (800 mg/day). More data and observation will be needed to ascertain a causal relationship.
Thus far, the most important adverse effect of topiramate is cognitive dysfunction, as evidenced by mental slowing and word-finding difficulty, which can be quite significant. The other significant side effects are renal calculi, weight loss, paresthesias, and a pharmacokinetic interaction with steroid oral contraceptives. Significant effects on cardiovascular function, as well as effects on bone density, bone marrow cells, and thyroid function, have not been reported so far. With the exception of the rather high incidence of troublesome central nervous system problems, topiramate appears otherwise to be a relatively well-tolerated and reasonably safe addition to our antiepileptic armamentarium.