Prozac Treatment of Major Depression: Discontinuation Study

This study enrolled 627 subjects with Major Depressive illness at New York State Psychiatric Institute and Massachusetts General Hospital. Subjects were treated with fluoxetine 10-60mg over a 12 week period. The "responder" group was defined by those no longer meeting criteria for Major Depression at week 12, along with CGI ratings of "much improved" or "very much improved" as determined by an independent evaluator. At week 12 "non-responders" were withdrawn from the study and received open label treatment; responders were randomized in double-blind fashion to either fluoxetine continuation (20-80mg daily) at response dose or placebo switch for up to 24 weeks. The responder group was stratified by "specific or true" drug response (late onset and persistent once attained) and "nonspecific or placebo" response (early onset or nonpersistent) patterns. Subjects were evaluated at one week and two week intervals at different phases of continuation treatment, and depression relapse was determined by agreement between study psychiatrist and independent evaluator CGI and Ham-D ratings, as well as administration of the MDD section of the Mood Disorders Module of the Structured Clinical Interview for DSM-IV Disorders at those visits. A subset of study participants also provided DNA samples to determine whether there are any DNA markers of response type. Data were analyzed to test the following hypotheses: that during continuation fluoxetine treatment improved patients with a "true drug" acute response pattern randomized to placebo had a poorer outcome than those maintained on active drug; that during continuation fluoxetine treatment improved patients with a "placebo" acute response pattern randomized to placebo had no worse an outcome than those maintained on drug; that during continuation fluoxetine treatment patients with a "true drug" acute response pattern randomized to continue on fluoxetine were more likely to maintain their benefit than those with a "placebo" pattern.