Generic substitution of antiepileptic drugs: a systematic review of prospective and retrospective studies.

OBJECTIVE To systematically review the literature on generic antiepileptic drugs (AEDs), evaluate the efficacy and safety of generic AED substitution, and perform pharmacokinetic (PK) analysis using the American Academy of Neurology (AAN) scheme to classify evidence. DATA SOURCES PubMed and Cumulative Index to Nursing and Allied Health Literature searches from January 1, 1980, to October 15, 2010, were performed using the search terms anticonvulsant, antiepileptic drug, carbamazepine, divalproex, ethosuximide, gabapentin, lamotrigine, levetiracetam, oxcarbazepine, pheno-barbital, phenytoin, primidone, topiramate, valproate, valproic acid, and zonisamide; bioavailability, bioequivalence, bioequivalency, bioequivalent, and substitution; and generic. STUDY SELECTION AND DATA EXTRACTION Retrospective and prospective controlled studies of generic substitution of AEDs were included in the review. Non-English-language articles and uncontrolled clinical studies were excluded. Published articles were categorized using the AAN criteria for systematic reviews. DATA SYNTHESIS We identified 156 articles. Of these, 20 met our inclusion criteria; 7 were retrospective studies, 6 were prospective studies in patients with epilepsy, and 7 were prospective studies in healthy subjects. All articles were rated Class I to Class III, using AAN criteria. The retrospective studies were categorized as Class III and showed a significant relationship between generic substitution and increased use of health care resources because of seizures or AED toxicity. Prospective studies were categorized as Class I, II, and III. Prospective studies in patients showed no differences between brand and generic drugs in PK parameters of bioequivalence. Three prospective studies in healthy subjects reported significant differences in maximum drug concentrations. Comparison of brand and generic drugs revealed no significant difference in seizure frequency; however, some prospective studies showed significant differences in PK parameters, primarily those not used for bioequivalence determinations. CONCLUSIONS There is inconsistency between retrospective and prospective studies of generic AED substitution. The highest levels of evidence indicate that there should not be a problem with generic substitution, although some patients are more prone to problems with the generic products. Some evidence suggests that switches between multiple generic AED products in certain individuals may be problematic.