Early Response to Aripiprazole May Predict Treatment Success in Adolescents With Schizophrenia

Deborah Brauser

November 12, 2009

November 12, 2009 (Honolulu, Hawaii) — Early clinical response to the antipsychotic aripiprazole predicts future treatment success in adolescents with schizophrenia, according to post hoc analysis from a large trial presented here at the American Academy of Child & Adolescent Psychiatry (AACAP) 56th Annual Meeting.

In fact, an early response by aripiprazole-treated patients at week 3 demonstrated 59% sensitivity, 99% specificity, 86% positive predictive value (PPV), and 68% negative predictive value (NPV) in predicting future response.

For remission at week 6, the early response at week 3 showed 72% sensitivity, 76% specificity, 83% PPV, and 62% NPV.

"This is the first confirmation that, as in adults, the majority of response occurs early in adolescents with schizophrenia," lead investigator Chrisoph Correll, MD, medical director and associate professor of psychiatry at the Zucker Hillside Hospital/Albert Einstein College of Medicine in Glen Oaks, New York, told Medscape Psychiatry.

"In addition, patients who did respond early are likely to maintain that status at 6 weeks," added Dr. Correll. "In other words, a nonresponse at 3 weeks predicts pretty reliably a nonresponse at 6 weeks."

Few Randomized Controlled Trials in Adolescents With Schizophrenia

Approximately 39% of males diagnosed as having schizophrenia present with symptoms of psychosis before the age of 20 years. However, there have been relatively few double-blind, placebo-controlled trials performed in adolescents.

"The fact that past studies have shown meaningful improvements in adults treated with aripiprazole led us to want to evaluate its effects in adolescents," explained Dr. Correll.

He reported that there have been recent challenges to the commonly held belief that a delay exists between the start of drug therapy and improvement. "If response and remission in adolescents with schizophrenia could be predicted early in the course of treatment with antipsychotics, this could save both time and resources," he said.

In the original 6-week study, 302 adolescents (13 – 17 years old, 57% male, 60% white) were enrolled at 101 sites in the United States, Europe, South America, Asia, and South Africa. Adolescents were randomized to receive either a 10-mg (n = 100) or 30-mg (n = 102) dose of aripiprazole or matching placebo (n = 100).

In this analysis, the main outcome measure was the mean change in the Positive and Negative Syndrome Scale (PANSS) total score in 258 patients (mean age, 15.5 years) from baseline to the study's 6-week endpoint. Early response was defined as a 20% or more reduction from baseline in PANSS score at weeks 1, 2, and 3.

Remission was defined as a score of 3 or less (mild or less) on 8 PANSS subscore items at week 6, including delusions, conceptual disorganization, hallucinatory behavior, unusual thought content, mannerisms/posturing, blunted affect, social withdrawal, and lack of spontaneity/flow of conversation.

"We basically wanted to find out: can we predict a response or nonresponse at the end?" said Dr. Correll.

Response Achieved in First 2 to 3 Weeks

Results showed that the combined aripiprazole responders achieved nearly 50% of the PANSS total score reduction by week 2 and up to 75% by week 3.

In weeks 2 and 3, the patients treated with aripiprazole, 10 mg, achieved 51% and 68%, respectively, of the overall reduction in PANSS total score from baseline to week 6. Those treated with the 30-mg dose achieved 49% and 75% at the same time points, respectively.

In other words, said Dr. Correll, most response was achieved in the first 2 to 3 weeks of the trial. However, he noted, because aripiprazole has a 3-day half-life, the 2-week timeline was not the best predictor of response, although the 3-week time point was.

"I think the number-1 takeaway for clinicians is this: if you don't see at least a minimal response at 3 weeks with aripiprazole, and most likely at 2 weeks, you should think about switching [medication] or doing something else in order to improve response."

Dr. Correll reported that his team is now looking at other response predictors.

"There has been a suggestion that early extrapyramidal side effects can maybe predict nonresponse, which would be a biological marker that there isn't enough dopamine in some areas of the brain, while there's too much in the limbic system. So we're now looking at other pathways that could be related to nonresponse."

Too Early for Prime Time?

Asked by Medscape Psychiatry to comment on the study findings, Neal D. Ryan, MD, AACAP program committee chair and professor of psychiatry at the University of Pittsburgh Medical Center in Pennsylvania, said that given the extensive analysis there is a good chance that the essential finding that relatively early response is a reasonably good predictor of ultimate response is correct.

"However," Dr. Ryan added, "it is also very likely that the particular analysis that gives the very best predictions is an overestimate of the real effect, and we can’t know how much of an overestimate it may be unless this is subsequently examined on a separate data set. Therefore, we may need more data before this nice work can directly be used to improve our treatment algorithms."

This study was funded by Otsuka Pharmaceutical Co, Ltd. Dr. Correll reports he has been a consultant to or received honoraria from AstraZeneca, Bristol-Myers Squibb, Cephalon, Eli Lilly, Medicure, Janssen, Otsuka, Pfizer, Shering-Plough, Supernus, and Vanda. He has also served on the speaker's bureau of AstraZeneca, Bristol-Myers Squibb, Otsuka, and Pfizer.

American Academy of Child & Adolescent Psychiatry (AACAP) 56th Annual Meeting: Abstract 2.47. Presented October 28, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.