PHARMACORESISTANCE IN EPILEPSY - HOW SHOULD IT BE DEFINED

Although terms such as 'treatment-resistant', 'pharmacoresistant' and 'refractory' epilepsy are ubiquitous in the epileptological literature , only seldom has an attempt been made to define them clearly. To labe l these terms as intuitive is simply avoiding the problem, and ignores the fact that 'intuitions' can vary widely. In fact, when definitions have been proposed, they have been as disparate and numerous as the s cientists using them. For example, there is no agreement on how many d rugs should be tried before a patient can be considered as having 'pha rmacoresistant' epilepsy, and there is uncertainty as to whether a def inition should include specific requirements about dosages, minimum se izure frequency and minimum duration of the disease. While it is ackno wledged that definitions may vary depending on the issue being address ed, proliferation of too many definitions or, even worse, avoidance of definitions altogether may hamper the interpretation and comparabilit y of studies assessing the effects of anticonvulsants. The problem is especially acute when selecting patients with pharmacoresistant epilep sy for trials of new anticonvulsants: heterogeneity in the characteris tics of these patients due to a loose definition of pharmacoresistance is the main explanation for the wide differences in outcome between s tudies using the same drug and comparable designs. The selection of pa tients for such studies could be improved greatly by introducing a gra ded system of drug resistance. According to a recent proposal, grade I resistance may include epilepsy that is unresponsive to a maximally t olerated dosage of 1 primary drug; grape II may include epilepsy that is unresponsive to 2 primary drugs used sequentially (IIA) or in combi nation (IIB); and grade III may include epilepsy that is unresponsive to 3 or more primary drugs used sequentially (IIIA) or in combination (IIIB). The main advantage of these categories is that they allow for a prediction of the probability of response to a subsequently added dr ug. A case can also be made for differentiating true pharmacoresistanc e from pseudoresistance, a condition defined as the persistence of sei zures because of poor compliance, inappropriate treatment, inappropria te assessment of response, or inadequate dosage or administration regi men.