View Full Version : Lithium vs Depakote in child bipolar

08-01-05, 08:27 AM
Lithium Vs. Depakote for Long-term Maintenance treatment in Pediatric Bipolar Disorder

This study purports to be the first to look at maintenance treatments in pediatric bipolar disorder in a double-blind, randomized and controlled fashion. It is difficult to do clinical trials in children for many reasons. First, it is costly to do clinical trials and once drugs are approved, they can be used "off-label" (meaning without FDA approval) in children as long as there isn't evidence that they are unsafe in children. Also, it can be difficult to do experiments on children because of parental concerns and concerns over possible long-term effects that are of less consequence potentially in adults.

However, this study looked at two medications that were commonly used in pediatric bipolar disorder, Lithium and Diavalproex sodium (Depakote, Depakene). The authors selected potential subjects who were diagnosed with bipolar disorder at an early age. They were then given both Lithium and Divalproex until such time that their symptoms waned (avg of 10 weeks.) After that, they were randomly assigned to either receive Divalproex or Lithium for up to the next 76 weeks. Nearly 300 patients were initially screened for the research and based on inclusion/exclusion criteria, 30 patients entered the second phase in each group. They were monitored for side effects by a doctor who was not involved in the research and could remove them from the study if there were side effects that became problematic.

Overall, it was shown that there was little to no difference statistically between the groups. Nearly 75% of patients reported side effects with subjects having to withdraw from the study for alopecia (hair loss) in both groups (1 in each group.) Other main side effects were headache and abdominal pain in the divalproex group and enuresis (bed-wetting) and vomiting in the Lithium group. One patient from the Lithium group dropped out of the study for enuresis and one dropped from the divalproex group for a low platelet count and another dropped out for a change in thyroid hormone tests. When the authors looked at the length of time to relapse, there was no difference between the groups. Based on the statistics, they would have needed nearly 100x more patients in the study in order to be able to show a difference. This means that the difference was so small as to not have matter clinically.

One methodological advantage to this study is that it allowed patients who were also diagnosed and treated for other psychiatric conditions to be in the study. For example, if a patient had ADHD and was on a stimulant for that disorder, they could be part of this study. That helps to replicate what real life is like instead of only taking patients that have no complicating factors which is different from what a "typical" patient will present like. This study also had a long followup duration which helped to improve the quality of the data.

Limitations of the study are that it was a small cohort of patients. Only 30/group is not a lot and based on how close the results were to each other, it would have taken over 3000 patients to notice a difference between the groups. Also, they mention that it would have been helpful to have had a combo lithium/divalproex group to further compare if the combo therapy is better than either individually. The authors also mention that they would like to have a placebo group to compare against, but also that is difficult ethically as it would mean that some subjects would not receive medication despite our feeling that it is helpful.

Overall, given that lithium is significantly less expensive than divalproex, it makes sense that it would be a logical first choice if all other things are equal. However, individual circumstances may dictate that divalaproex is better to start with. Also, if one doesn't work, it is reasonable to try the other drug before going to other medications. Not all drugs work for any particular individual so it is always helpful to work closely with a psychiatrist to determine the best medication regimen for an individual patient, but these data indicate that for a population of people on average, lithium and divalproex are equally effective.

The Stanley Medical Research Institute primarily supported this study. It was also supported in part by an NIMH Developing Centers for Interventions and Services Research Grant (P 20 MH-66054). Nursing and pharmacy activities were supported in part by NICHD Pediatric Pharmacology Research Unit (PPRU) contract HD 31323-05. Medications were provided in part by Abbott Laboratories.