Topiramate in Patients with Learning Disability and Refractory Epilepsy


Address correspondence and reprint requests to Professor M.J. Brodie at Epilepsy Unit, Department of Medicine and Therapeutics, Western Infirmary, Glasgow G11 6NT, Scotland. E-mail:


Summary:  Purpose: Management of seizures in learning disabled people is challenging. This prospective study explored the efficacy and tolerability of adjunctive topiramate (TPM) in patients with learning disability and refractory epilepsy attending a single centre.

Methods: Sixty-four patients (36 men, 28 women, aged 16–65 years) were begun on adjunctive TPM after a 3-month prospective baseline on unchanged medication. Efficacy end points were reached when a consistent response was achieved over a 6-month period at optimal TPM dosing. These were seizure freedom or ≥50% seizure reduction (responder). Appetite, behaviour, alertness, and sleep were assessed by caregivers throughout the study.

Results: Sixteen (25%) patients became seizure free with adjunctive TPM. There were 29 (45%) responders. A further 10 (16%) patients experiencing a more modest improvement in seizure control continued on treatment at the behest of their family and/or caregivers. TPM was discontinued in the remaining nine (14%) patients, mainly because of side effects. Final TPM doses and plasma concentrations varied widely among the efficacy outcome groups. Many patients responding well to adjunctive TPM did so on ≤200 mg daily. Mean carer scores did not worsen with TPM therapy.

Conclusions: TPM was effective as add-on therapy in learning-disabled people with difficult-to-control epilepsy. Seizure freedom is a realistic goal in this population.

Epilepsy is common in people with learning disability (mental retardation, intellectual disability), affecting 15–75% of this vulnerable population, depending on the extent of the brain disorder (1). When the two problems coexist, management of seizures can be challenging. Difficulties can arise in diagnosis, investigation and classification. In some patients, severe epilepsy develops, incorporating a range of seizure types (2). Concomitant medical and psychiatric conditions often complicate management (3). In particular, antipsychotic medication can lower seizure threshold (4).

Many people with learning disability have seizures that appear refractory to antiepileptic drug (AED) treatment (5). There are few outcome data to guide the pharmacotherapy of epilepsy in these patients (6). The scope has now widened, however, with the introduction of a range of new AEDs over the last decade (7). One such agent, topiramate (TPM), licensed as adjunctive therapy for partial and generalised seizures, has efficacy in drug-resistant Lennox–Gastaut syndrome (8). For this reason, a prospective outcome evaluation was undertaken with TPM in learning-disabled patients with other types of difficult-to-control epilepsy.



Sixty-four patients (36 men, 28 women; aged 16–65 years) with learning disability and refractory epilepsy were started on adjunctive TPM at the epilepsy clinic at the Western Infirmary in Glasgow, Scotland. Each had had a diagnosis of learning disability made by a specialist in that discipline. The extent of impairment was classified according to intelligence quotient (IQ) as mild (IQ, 50–70), moderate (IQ, 35–49), severe (IQ, 20–34), or profound (IQ, <20) (9).

Before the first clinic visit, the patient's family and/or caregivers were posted a standard proforma requesting information about birth and developmental history, onset and type of seizures, current and previous medical disorders, and AED and other drug treatment. Patients were reviewed within 4 weeks of referral and seen every 4–6 weeks thereafter according to clinical need. Education was provided to caregivers and, where possible, patients on seizure recording and lifestyle issues. Expectation regarding anticipated outcome was discussed with the family and/or caregivers.

Patients were categorised as having refractory epilepsy if they had experienced monthly seizures for ≥5 years despite treatment with at least four AED monotherapy or combination regimens (10). Fifty patients had partial seizures with or without secondary generalisation. The remaining 14 patients had primary generalized epilepsy (10 tonic–clonic, one myoclonic, one atonic, one tonic–clonic/absence/myoclonic, one tonic–clonic/absence/atonic).

Twenty-three patients were receiving treatment with one AED. A further 34 were treated with two, and the remaining seven patients took three drugs. Before starting TPM, seizures were monitored over a 3-month prospective baseline period on unchanged AED dosing by using standard frequency and description charts with which family and/or caregivers were familiar. The median monthly seizure count during baseline was 13 (range, one to 70). TPM was introduced according to the following schedule: weeks 1 and 2, 25 mg once daily; weeks 3 and 4, 25 mg twice daily; weeks 5 and 6, 25 mg in the morning and 50 mg at night; and weeks 7 and 8, 50 mg twice daily. Doses were adjusted by 25- to 50-mg increments or decrements every 4 weeks thereafter, according to clinical response. If the patient remained seizure free on a particular TPM dose, no adjustment was made until further seizures occurred. In those who were not seizure free, the dose was increased gradually to the limit of tolerability. An effort was made to discontinue other AEDs in patients reporting side effects.


The family and/or caregivers provided a written recording of all seizures at each clinic visit. An assessment score of 1 (poor), 2 (reasonable) or 3 (good) also was requested regarding the patient's appetite, behaviour, alertness, and sleep since the last visit. Weight was recorded, and inquiry made about adverse events. Patients continued in the study until a consistent response was obtained in terms of seizure control for 6 months compared with baseline on the highest tolerated and/or most effective TPM dose, or until TPM was discontinued because of side effects, lack of efficacy, or both.


The χ2 test was used for comparison of categoric data. All statistical tests were two-tailed.


When permitted by the patient, 20 ml of venous blood was taken at the same time after dosing at each clinic visit. The sample was drawn into heparinised tubes and immediately centrifuged before storage at –20°C for batch analysis. TPM concentrations were determined by fluorescence polarisation immunoassay with an OPUS Diagnostics Innofluor system (Biostat Diagnostics, Stockport, U.K.) on an Abbott TDx instrument (Abbott Diagnostics, Maidenhead, U.K.). The accuracy, precision, specificity, and sensitivity of the procedure were consistent with those reported on the kit insert.


Seizure freedom for ≥6 months was attained in 16 (25%) patients, with a further 29 (45%) having a ≥50% reduction (responder) in seizure frequency at optimal TPM dosage over a 6-month period compared with baseline. Marginal improvements in control (<50% seizure reduction compared with baseline) occurred in 10 (16%) patients, all of whose family and/or caregivers wished them to remain on TPM therapy. Response was not related to the extent of learning disability (Table 1). Compared with baseline values, mean carer scores for appetite, behaviour, alertness, and sleep did not worsen with the addition of TPM in those patients who continued taking the drug (Fig. 1).

Table 1.  Outcomes in patients with different degrees of learning disability
CategoryMild (IQ 50–70)Moderate (IQ 35–49)Severe (IQ 20–34)Profound (IQ <20)Total
  1. Responder, ≥50% seizure reduction; marginal effect, <50% seizure reduction.

Seizure free366116
Marginal effect424010
Figure 1.

Mean caregivers' scores for patients taking topiramate (n = 55).

Fourteen (28%) of the 50 patients with localisation-related seizures became seizure free. There were 20 (40%) responders and seven (14%) reporting marginal benefit with TPM. Two (14%) of the 14 patients with primary generalized epilepsy became seizure free, with a further nine (65%) being classified as responders. In three (21%), TPM provided only marginal benefit. One each of the two patients with absence or myoclonic seizures became seizure free, with the other two reporting a ≥50% reduction in seizure frequency. Both patients with atonic seizures became free of these.

TPM was discontinued in nine (14%) patients, all of whom had localisation-related seizures. Side effects occurred at low dosage in seven (Table 2), and worsening of seizures in the remaining two. Three patients came off TPM because of anorexia and weight loss (mean, 4.7 kg; range, 3.5–6.8 kg). Weight loss occurred in a further 22 patients (mean, 5.7 kg; range, 0.5–16 kg), who continued to take TPM. Ten patients, who benefited from the drug, gained weight (mean, 4.83 kg; range, 0.1–14 kg). Weight remained stable in the other patients. No relation was found between weight loss and extent of learning disability, TPM dose, or plasma concentration. Side effects were reported by the families and/or caregivers of nine patients who continued to take TPM (Table 2). Overall, four patients developed worsening of preexisting hemiparesis, which resolved on TPM discontinuation or reduction in dosage.

Table 2.  Side effects reported by patient, family, and/or caregivers
Topiramate discontinuedTopiramate continued
3 anorexia, weight loss2 worsening hemiparesis
2 confusion1 anorexia, stupor, worsening
1 slurred speech, ataxia1 fatigue, impaired concentration
1 fatigue1 fatigue, poor appetite
1 aggression1 nausea
1 worsening hemiparesis1 ataxia
 1 tremor
 1 acne

There was wide variation in TPM doses in the efficacy outcome groups (Fig. 2). Mean doses were 217 mg (range, 50–600 mg) in seizure-free patients, 222 mg (range, 25–800 mg) in responders, and 265 mg (range, 25–700 mg) in those experiencing a marginal reduction in seizure numbers. The mean dose was just 56 mg (range, 25–100 mg) in the nine patients in whom TPM was discontinued. More than 50% of patients with a good outcome responded to ≤200 mg TPM daily (10 of 16 seizure-free; 19 of 29 responders). Mean plasma concentrations were 8.2 mg/L (range, 1.8–25 mg/L) for seizure-free patients, 6.8 mg/L (range, 1.2–23.7 mg/L) for those with ≥50% reduction in seizure frequency, 10.3 mg/L (range, 2.6–16.9 mg/L) for those with a marginal response, and 2.2 mg/L (range, 0.2–4.4 mg/L) for patients in whom TPM was discontinued.

Figure 2.

Final daily topiramate doses in different outcome groups.

Other AEDs were discontinued in 10 patients. Phenytoin (PHT) was discontinued in four patients, lamotrigine (LTG) in two, and gabapentin (GBP), vigabatrin (VGB), and potassium bromide, respectively, in three others. LTG and GBP were discontinued in the tenth patient. Two of these patients became seizure free, with the other eight being classified as responders. Relation of outcome to individual AED combinations is shown in Table 3.

Table 3.  Final antiepileptic drug combinations in outcome groups
Antiepileptic drugsSeizure free (n = 16)Responder (n = 29)Marginal (n = 10)Discontinuation (n = 9)Seizure free/responder (%)
  1. TPM, topiramate; CBZ, carbamazepine; LTG, lamotrigine; VPA, sodium valproate; PHT, phenytoin; PRM, primidone; AEDs, antiepileptic drugs.

TPM + CBZ465110 (63)
TPM + LTG34107 (88)
TPM + VPA02012 (67)
TPM + PHT01011 (50)
TPM + VPA + LTG552010 (83)
TPM + CBZ + VPA12123 (50)
TPM + PRM + CBZ02102 (67)
TPM + 2 other AEDs34047 (64)
TPM + 3 other AEDs03003 (100)


There have been few studies of the newer AEDs in patients with learning disability. Adjunctive GBP and LTG were evaluated in adults in a randomised, open-label study (11). Half the patients taking GBP had a ≥50% reduction in seizure frequency compared with 48.6% of those taking LTG. Improved seizure control also has been claimed with LTG in observational studies (12–14). VGB reduced seizures by ≥50% in 36 patients with learning disability with a gradual reduction to 22% of patients after 5 years' follow-up (15). This drug is more often associated with behavioural problems than are LTG and GBP in this population (16).

One fourth of our patients with learning disability became seizure free with adjunctive TPM, with a further 45% experiencing at least halving of their seizure frequency. These figures were similar to those reported in our previous TPM study in patients with refractory epilepsy who did not have learning disability (17). Sachdeo et al. (8) found a 20.6% median percentage reduction in monthly seizure rate compared with baseline during a double-blind trial of adjunctive TPM in Lennox–Gastaut syndrome. This may be a reflection of greater seizure severity in these younger patients. As in other studies, we found no relation between seizure control and extent of learning disability (18).

Side effects in this study were consistent with those reported in the regulatory trials with TPM (19). In four patients, however, reversible worsening of preexisting hemiparesis developed, a problem we have previously reported (20). This resolved in all patients with dosage reduction, but led to discontinuation of TPM in one individual. The mechanism responsible for this clinical problem is unclear. TPM treatment was discontinued in a further eight patients, in six of whom, adverse effects developed at low dosage. Worsening of seizures occurred in the remaining two patients, a phenomenon occasionally seen with the introduction of most AEDs (21). For those patients who remained on TPM treatment, mean caregiver scores for behaviour, alertness, and sleep improved while appetite remained unchanged. This is a likely consequence of improvement in seizure control or a reduction in AED burden rather than being a direct effect of TPM itself.

There was wide variability in TPM dosing and plasma concentrations in all the outcome groups. Many patients responding well to the drug did so at daily TPM doses of ≤200 mg. This supports the suggestion of possible complementary modes of action with TPM in some of these combinations (22). No AED combination worked particularly well, although TPM with LTG seemed to be promising, in keeping with our previous observation (23). Concomitant AEDs were discontinued in 10 patients with a range of severity of learning disability. In the nine patients in whom TPM was discontinued, poor tolerability was obvious within 8 weeks of starting treatment.


TPM was effective as add-on therapy for a range of difficult-to-control seizures in patients with learning disability. The drug was generally well tolerated, demonstrating no overall deleterious effect on appetite, behaviour, alertness, and sleep in those patients who continued to take it. Seizure freedom is a realistic goal in this patient population.