Oxcarbazepine is one of the most overrated drugs in psychiatry. It is used frequently, even though it has little to no scientific basis for being used at all. It’s an anticonvulsant, but psychiatrists generally don't treat epilepsy. It’s use for psychiatric conditions is questionable at best.
Why would carbamazepine “work” for mood illnesses, but not oxcarbazepine?
For the same reason that chlorpromazine “works”, but imipramine doesn’t. Literally, the difference between carbamazepine and oxcarbazepine is one carbonyl bond (C=O). Similarly, the exocyclic difference between chlorpromazine (Thorazine, an antipsychotic) and imipramine (a tricyclic antidepressant) is one chlorine bond. One is an antipsychotic that works for mania, but doesn't appear to work for depression (chlorpromazine); the other is an antidepressant that causes mania and doesn't work for schizophrenia (imipramine).
One carbonyl bond can make a huge difference.
This doesn't meant that similarity of chemical structure is irrelevant; sometimes it can imply similar clinical effects, whether for efficacy or for side effects. Forinstance, chlorpromazine and imipramine share many similar side effects (anticholinergic effects of dry mouth and constipation, and antihistamine effects of sedation and weight gain). But their clinical effects are quite different.
Similarly oxcarbazepine and carbamazepine share some side effects (like hyponatremia), but have important differences (e.g., no drug interactions with oxcarbazepine). So you can’t assume their clinical effects are similar.
It would be like saying that since the same company makes Volkswagen and BMW, the two cars must be more or less the same.
One bond can make a big difference.
In short, you shouldn't assume clinical efficacy based on chemical similarity. As emphasized in the inaugural January 2015 PL issue, clinical research, not biological speculation, is needed for clinical claims.
What does the clinical research show about the efficacy of oxcarbazepine in psychiatric conditions? Most of the research is in mood illness. In acute mania randomized trials, some studies indicate similar efficacy to haloperidol, but without placebo controls, we can’t infer efficacy since mania resolves by itself spontaneously within weeks to months. There are no randomized trials in acute bipolar depression. Observational reports, including by the PL editor, suggest some benefit, but they can’t be assumed to be correct without randomized data, because of confounding bias. Two maintenance RCTs found no benefit over placebo (A Vasudev et al, Cochrane Database Syst Rev. 2008;(1):CD005171).
More evidence that oxcarbazepine is ineffective overall involves studies of its active metabolite, licarbazepine, which was studied as a possible new agent by Novartis (oxcarbazepine is now generic, and no profits exist with it). Licarbazepine was found to be equivalent to placebo in multiple mania RCTs. These negative data were never published.
This inefficacy was proven again by another pharmaceutical company, Sunovion, which conducted two randomized clinical trials of an enantiomer of licarbazepine (eslicarbazepine), which found again that the agent was equivalent to placebo for acute mania. These data have been published recently (H Grunze et al, J Affect Disord. 2015;174:70-82).
In sum, oxcarbazepine and/or its active metabolite have been proven ineffective in acute mania and in maintenance prophylaxis of bipolar illness. It’s never been proven effective in any randomized trial of an acute depressive episode.
Hence oxcarbazepine and/or its active metabolite is proven ineffective for mood illness.
Like most anticonvulsants, oxcarbazepine raises the seizure threshold via sodium channel blockade. Any biological effects that are relevant to mood are unknown.
The main reason mental health clinicians like this agent is because it doesn't have drug interactions, unlike carbamazepine, and thus it can be combined with other agents, such as dopamine blockers or monoamine agonists (antidepressants). It also has less risk of rash or leukopenia. It has more sedation than carbamazepine, however, and it has one potentially serious medical risk that can lead to seizures, namely a 2% risk of hyponatremia. Like carbamazepine, there’s no weight gain.
PL cautions clinicians in the use of this agent in persons with eating disorders, who often wish to take medications without weight gain. Frequently, such persons will drink water excessively, as a way of maintaining weight loss. This over drinking of water, when combined with oxcarbazepine, can lead to dangerous hyponatremia. If sodium levels fall below 120, seizures can occur.