SUMMARYObjective: To investigate the relationship between antiepileptic drug (AED) polytherapy in pregnant women and the risk of fetal malformations as prescribing practice changed, with valproate being used less often and at lower doses. Specifically, the risks associated with two of the most common AEDs included in polytherapy over recent years, levetiracetam and topiramate, were examined. Methods: An observational cohort study in which malformation rates were analyzed in 1,461 pregnancies exposed to AED monotherapy, and in 484 exposed to antiepileptic drug combinations, from the Australian Register of Antiepileptic Drugs in Pregnancy over a 15-year period (1999-2014). Results: Fetal malformation rates had fallen over time in monotherapy pregnancies, but increased in polytherapy pregnancies, despite decreasing use and lower dosages of valproate. The rise in polytherapy malformation rates began around 2005 when levetiracetam and topiramate use began to increase. Excluding pregnancies involving valproate exposure, malformation rates were higher in the remaining polytherapy pregnancies as compared with the monotherapy ones (6.90% vs. 3.64%; odds ratio [OR] 1.96, 95% confidence interval [CI] 1.14-3.39). Malformation rates were similar in polytherapy pregnancies whether or not levetiracetam was included (7.14% vs. 8.38%), but were higher in polytherapy pregnancies involving topiramate (14.94% vs. 6.55%: OR 2.507, 95% CI 1.23-5.10). Logistic regression showed that topiramate in polytherapy had a positive dose relationship with teratogenicity risk (p = 0.025). Significance: The malformation risk associated with AED polytherapy depends on the specific drugs involved. Topiramate, when used as part of AED polytherapy that did not include valproate, was associated with a dose-related increased risk of fetal malformations. KEY WORDS: AED polytherapy, Fetal malformation, Levetiracetam, Topiramate, Valproate.After the association between antiepileptic drug (AED) exposure during pregnancy and the occurrence of fetal malformations was recognized, several investigations reported that the risk of such malformations was greater when combinations of antiepileptic drugs were employed in pregnant women than when only a single antiepileptic drug was involved.1-5 As a consequence, the belief was expressed that prescribing AED combinations in pregnant women should be avoided as far as possible. Then, in 2010 and 2011, three articles appeared in the literature, including one from our group, each demonstrating that the critical factor that determined the increased malformation risk associated with AED polytherapy was not the use of two drugs simultaneously per se, but the inclusion of valproate-by this time known to be a dose-dependent teratogen-in the combinations. [6][7][8] Because prescribing practice in pregnant women with epilepsy has changed over the last decade and valproate is now being employed less often and at lower