Bipolar Disorder

September 15, 2003

September 1, 2003 from ObGynNews By Lee S. Cohen, M.D.

As the use of anticonvulsants to treat bipolar illness has grown over the past decade, so has the number of women successfully treated with these medications who have questions about whether they should discontinue these drugs before they attempt to conceive, or what to do if they are already pregnant.

The anticonvulsants that have been most widely used for bipolar illness are sodium valproate and carbamazepine, and more recently, gabapentin (Neurontin), lamotrigine (Lamictal), oxcarbazepine (Trileptal), and tiagabine (Gabitril). Until recently, there have been few reproductive safety data available on the newer anticonvulsants.

Many women and their physicians are caught in a particularly vexing bind because two of the mainstays of bipolar therapy, lithium and sodium valproate, are known teratogens, though the teratogenicity of these two compounds is
particularly different. The risk associated with first-trimester exposure ranges from a relatively modest 0.05% risk of
Ebstein’s anomaly with lithium to an approximately 8% risk of cardiovascular malformations and neural tube defects
with sodium valproate. The latter is based on recent findings from the Antiepileptic Drug Registry at Massachusetts General Hospital (Am. J. Obstet. Gynecol. 187[6 pt. 2]:s137, 2002).

But the data that are accumulating on lamotrigine, approved in June for maintenance treatment of bipolar disorder, provide some welcome news for reproductive-aged women with bipolar disorder. An interim report on cases collected by the lamotrigine pregnancy registry maintained by the manufacturer, GlaxoSmithKline, since September 1992 indicates that the drug does not appear to be teratogenic. The report does note, however, that the sample size is not large enough to make definitive conclusions.

As of March, the pregnancy registry had collected information on more than 500 first-trimester exposures in women
treated with the drug for bipolar illness and for epilepsy, which did not demonstrate an increase in major birth defects associated with first-trimester exposure, supporting earlier reports.

The risk of teratogenicity was significantly increased with first-trimester exposure to the combination of lamotrigine and sodium valproate (more commonly used for epilepsy), but not with lamotrigine monotherapy: Among the 302 pregnancies exposed to monotherapy in the first trimester, there were 9 (3%) major birth defects, compared with 7 (10.4%) major birth defects among the 67 cases of firsttrimester exposure to both drugs. There were 5 (3.5%) major birth defects among 148 cases of first-trimester exposure to polytherapy that did not include sodium valproate.

The clinical implications of these long-awaited data on lamotrigine are relatively clear and present an opportunity to navigate the tricky course of maintaining euthymia across pregnancy and minimizing exposure to drugs that might be harmful to the fetus.

For example, sodium valproate can be deferred for a medicine such as lamotrigine in some patients, particularly those who do not respond to or who have not tolerated lithium. Although lamotrigine has not demonstrated efficacy for the treatment of acute mania, the anticonvulsant can be combined with medicines that are helpful in treating this phase of bipolar disorder. Such adjunctive medicines include high-potency typical antipsychotics like haloperidol or trifluoperazine.

Unfortunately, the reproductive safety data available for the newer atypical antipsychotic olanzapine (Zyprexa)–efficacious for both acute mania and for prophylaxis against recurrent mania–are exceedingly sparse. Clinicians are left with the task of trying to minimize exposure to medicines we know very little about, such as olanzapine, and to medicines we know a lot about that appear to be particularly harmful to the fetus, such as sodium valproate.

Lamotrigine is the only one of the newer anticonvulsants for which there are enough exposed cases to allow for
some reliable quantification of teratogenic risk. Manufacturers of the other anticonvulsants have not established independent registries. The Antiepileptic Drug Registry at Massachusetts General Hospital is collecting data on a spectrum of newer anticonvulsants, but to date the numbers are too small for any conclusions, except on lamotrigine.

One caveat with respect to use of lamotrigine lies in the very small but quantifiable risk of Stevens-Johnson syndrome associated with lamotrigine therapy. To reduce risk, the manufacturer recommends titrating patients gingerly, by no more than 25 mg weekly.

DR. LEE COHEN is a psychiatrist and director of the perinatal psychiatry program at Massachusetts General Hospital, Boston. Related information is available on the program’s web site at www.womensmentalhealth.org.