Brooklyn, N.Y. – A risk calculator for bipolar disorder that has reached late stages of development might merit a trial to test whether treating prodromal symptoms delays or prevents the disease from developing in young patients, an expert said at a pediatric psychopharmacology update held by the American Academy of Child and Adolescent Psychiatry.
Dr. Boris Birmaher
Clinicians often are confronted with children who have prodromal symptoms of bipolar disorder, but only about half of those children eventually convert to full disease expression. As a result, treatment is not routine practice, said Boris Birmaher, MD, Endowed Chair, Early Onset Bipolar Disease, at the University of Pittsburgh.
A risk calculator that more effectively identifies those at highest risk of converting might be a tool that could allow early intervention to be tested. Such a calculator has been in development for some years, and the most recent research suggests that it is nearing a degree of accuracy that is clinically meaningful (J Am Acad Child Adolesc Psychiatry. 2018;57:755-63).
Accuracy might improve further with the identification and incorporation of more predictive variables. High-risk features for conversion include a parent with bipolar disorder, particularly one with childhood onset, and specific clinical features, such as prominent episodes of mania.
“Once replicated, the risk calculator will be instrumental to predict personalized risk to develop bipolar disease,” Dr. Birmaher said. He compared it to risk calculators now in use in other fields of medicine, such as cancer and cardiovascular disease, which generate information used by patients and their physicians for treatment decisions.
In the most recent study with the risk calculator, which has been the topic of several previous publications, 140 children with a diagnosis of bipolar disorder not otherwise specified (BP-NOS) were assessed every 7 months for a median of 11.5 years. BP-NOS is the diagnosis of a prodromal syndrome that includes mood lability and other features of bipolar disorder but not at levels reaching DSM-5 diagnostic criteria. The primary outcome of the study was conversion from BP-NOS to bipolar I or II, which are DSM-5 categories.
At the end of follow-up, 53.6% of the population had converted to bipolar I or II, which is consistent with previous risk estimates in this population. In specific patients, the correlation between predicted and observed conversions was “excellent.” Furthermore, the risk calculator was reported to have provided “good” discrimination between converters and nonconverters.
The risk calculator already is accessible online (http://www.cabsresearch.pitt.edu/bpriskcalculator/). Dr. Birmaher invited clinicians to visit and “play around” with its features, but he also issued a warning. “Be careful because we need to further validate this, which we are doing now, to see if it’s truly accurate or not. If it is, it would be a very good tool for us,” he said.
There are many potential clinical applications of the risk calculator, but Dr. Birmaher emphasized its possible value in selecting at-risk patients for therapy studies. Although several therapy trials already have been conducted in high-risk children on the basis of clinical presentation alone, such as a double-blind trial in BP-NOS patients that associated aripiprazole with a reduction in mania (J Child Adolesc Psychopharmacol. 2017;27:864-74), Dr. Birmaher sees the risk calculator as potentially more precise in identifying candidates.