As per other blogging occasions when I've talked about the antiepileptic drug (AED) valproate and its use specifically during pregnancy, I caution that no medical or clinical advice is given or intended on this blog. Anyone in any doubt or seeking an opinion on this other or any other AED should consult their prescribing physician. Don't mess with epilepsy. I repeat: don't mess with epilepsy.
The findings however reported by Areti Angeliki Veroniki and colleagues  (open-access) synthesising the available peer-reviewed evidence to compare "the safety of antiepileptic drugs (AEDs) on neurodevelopment of infants/children exposed in utero or during breast feeding" provides an all too familiar picture specifically in relation to valproate use during pregnancy. Namely: "Valproate alone or combined with another AED is associated with the greatest odds of adverse neurodevelopmental outcomes compared with control."
Utilising something called a Bayesian random-effects network meta-analysis (NMA), researchers set about synthesising and ranking various AEDs previously studied in terms of their safety when it comes to *association* with various childhood developmental diagnoses including autism and things like cognitive developmental delay. Based on the results from 29 studies reporting on AED use and 'neurological outcomes' including some 5000 patients, authors made some important observations.
Not to be alarmist but: "results suggest that AEDs generally pose a risk for infants and children exposed in utero or during breast feeding." The authors caution that their results whilst methodologically strong are not without limitations; not least that the studies in this area are observational only and therefore have "inherent biases because of confounding and shortcomings of these studies." That being said, the tide of such observational research cannot be readily ignored. If there is one or more confounding variables that is/are somehow being hidden behind something like valproate use during pregnancy and its potential risks of offspring developmental outcomes, they seem to be very well hidden indeed.
"Valproate was significantly associated with more children experiencing autism/dyspraxia, language, cognitive and psychomotor developmental delays versus children who were not exposed to AEDs." Valproate came 'top of the pops' when it came to those neurological outcomes. Other AEDs also showed associations with specific outcomes - "oxcarbazepine and lamotrigine were associated with increased occurrence of autism" - but the data pointed to valproate as potentially showing greatest risk to offspring. I should point out that there is a mouse model of autism that relies on offspring valproate exposure which, whilst subject to shortcomings (see here), kinda suggested that there may be an important association. Said animal research also reveals some potentially important avenues for perhaps mitigating any deleterious effects of valproate use on offspring too (see here) with the requirement for more experimental study.
"Future studies should assess the genetic contribution from the biological father, maternal seizures during pregnancy, exposure through breast feeding only, types of epilepsy and maternal family history." Of course the authors are right to talk about other factors that are probably important to various childhood developmental outcomes. But unlike a related area talking about medicines taken during pregnancy potentially affecting risk of something like autism and how one might tease apart medicine use from underlying health issue(s) (see here for example), I don't think it too likely that there is going to be a large bank of pregnant mums out there who present with epilepsy or other seizure disorder that is not going to be already managed via AEDs. Indeed, it would be rather unethical as well as unwise to leave epilepsy untreated under any conditions.
To close, where we're up to when it comes to government (UK) advice on AEDS such as valproate and pregnancy use. Again, if it doubt, talk to your prescribing physician.
 Vernoniki AA. et al. Comparative safety of antiepileptic drugs for neurological development in children exposed during pregnancy and breast feeding: a systematic review and network meta-analysis. BMJ Open. 2017; 7: e017248.