Treating refractory epilepsy in adults

Most adult patients with refractory epilepsy have partial (focal) seizures with or without secondary generalisation. During the 1970s and early 1980s studies showed that in 70-80% of adults with newly diagnosed epilepsy, seizures were controlled successfully by carefully monitored monotherapy with any of the four standard antiepileptic drugs—phenobarbital, phenytoin, carbamazepine, or sodium valproate—all of which seemed to have similar efficacy in partial epilepsy in later comparative trials of monotherapy.^1-4 Furthermore, adding a second drug for patients with continuing seizures on optimum monotherapy led to modest benefit in no more than one third, a deterioration in seizure control or unacceptable toxicity in about a quarter, and no change in the rest.²,⁵

These studies led to important questions. Should patients unresponsive to the optimum use of the first drug be switched to alternative monotherapy or treated with polytherapy? If so, which drug or drug combination is appropriate?

Twenty years later these questions remain un-answered. Meanwhile 10 new drugs have been licensed and marketed in the United Kingdom as adjunctive therapy in adults for resistant mainly partial epilepsies: clobazam, vigabatrin, lamotrigine, gabapentin, topiramate, tiagabine, levetiracetam, oxcarbazepine, pregabalin, and zonisamide.

The only pragmatic controlled clinical trial of adjunctive therapy in partial epilepsy that was unresponsive to a single drug showed that the probability of remaining free of seizures over the next year was 16% for patients on adjunctive therapy and 14% for those switched to alternative monotherapy.⁶ The authors emphasised that the trial was statistically underpowered and that they had had difficulty in recruitment because of financial competition from commercial sponsors targeting similar patients for new drug trials. In a prospective observational study of 422 newly diagnosed patients, 47% became seizure free on the first drug and only an additional 14% on alternative monotherapy with a second or third drug, whereas only 3% were seizure free on a combination of two drugs—all of which implies a need to consider surgery in appropriately selected patients earlier.⁷ Interestingly, freedom from seizures was almost identical, whether patients were treated first with a standard drug or a new drug.

The recent guidelines from the National Institute for Health and Clinical Excellence (NICE) on newer drugs for epilepsy in adults is therefore vague on the management of refractory epilepsy because of a lack of data comparing the new drugs with each other or with standard drugs, either as monotherapy or combination therapy.⁸ NICE recommends that combination therapy should be considered only when attempts at monotherapy have not resulted in freedom from seizures but gives no guidance on the number of attempts at monotherapy or which combination to try. I estimate that the bewildered general neurologist or physician can choose from up to 13 options for monotherapy and 91 options for combination therapy. This is clearly an unsatisfactory situation for patients and physicians alike.

Doctors tend to opt for their own favourite combinations, influenced by marketing pressures or based on speculative concepts of different mechanisms of drug action or synergy—concepts that originally led to the subsequently discredited combined capsules of phenobarbital and phenytoin.⁹ Switching to alternative monotherapy is more time consuming and requires careful clinical monitoring. It is easier to add a second drug, and if there is some clinical improvement, as occurs in up to one third of patients, it is tempting but possibly erroneous to assume that improvement is due to the combination rather than to the second drug. Such thinking perpetuates the phenomenon and scale of polytherapy.² Furthermore, combination therapy increases the risk of side effects, including teratogenicity,¹⁰ especially if the drugs are similar (for example, carbamazepine and oxcarbazepine) or if they interact (for example, lamotrigine and valproate or carbamazepine).

In treating epilepsy in childhood, similar problems arise. Compared with adults, however, children have a much higher incidence of idiopathic generalised epilepsy syndromes, for which some standard or new drugs, such as carbamazepine or vigabatrin, may be inappropriate. In addition, fewer of the new drugs have been licensed for use in children, and comparative data on different drug treatments are even scarcer.¹¹

The pharmaceutical industry finances 90% of all clinical trials in the UK.¹² But industry has no interest in supporting large scale, long term pragmatic trials that might provide evidence to reduce much unnecessary polytherapy and therefore restrict the market for many of the newer drugs. Academics specialising in epilepsy will therefore have to clarify, through research, how much real progress has been made in managing resistant epilepsy since the era of phenobarbital and phenytoin.