October 26, 2009

Researchers at CRI have been conducting CNS development studies for 30 years.  Over this period of time, we have been active participants in the changes in CNS development landscapes which have resulted in the current challenges. 

The approach to drug development with the SSRIs 30 years ago was to immediately go from phase I to large multi-site, multi-arm, phase II studies in patients suffering from either their first episode or recurrent episodes of major depression.  Yet, even in this era, only one out of two of these studies was positive. 

However, SSRIs changed the landscape of the treatment of patients with clinical depression such that now 10% of Americans over the age of 6 are on a SSRI or related antidepressant and those numbers are from epidemiological studies done before SSRIs became generic.  This increased use of antidepressants is in part the result of a major marketing effort by many major pharmaceutical companies including direct-to-consumer advertising. 

Coincidental with this change, the difficulty of recruiting patients into clinical depression trials and the placebo response rate has increased.  Both of these developments have in turn contributed to an increase in cost and time to do antidepressant trials and increased failure rate due at least in part to higher placebo response rates.

Several approaches have been taken to deal with these problems.  One has been to stay with the traditional approach but to increase either the number of subjects enrolled in these trials and/or the number of trials done to achieve two positive studies.  Another has been to use centralized rating and even centralized diagnosis.  Both of these approaches have added expense and increased the time to study completion and thus have limited the number of studies that can be done to explore the efficacy of the drug in different conditions.  Nevertheless, there is no compelling evidence to support the contention that centralized raters reduce high placebo response rates. 

In contrast is a third approach which is to randomize subjects after a prospective unsuccessful trial of an existing antidepressant.  This approach was successfully used in three aripiprazole augmentation studies:  Each of the three studies yielded statistically significant difference (p < 0.001) between aripiprazole versus placebo augmentation.  Another study using this approach involved a single site and tested the antidepressant effects of the intravenous infusion of a novel NR2B NMDA antagonist versus placebo infusion in patients who had not benefitted from a 6 week trial of paroxetine. 

A significant challenge with the latter study was the fact that the investigational treatment required admission overnight to a clinical research unit (CRU) to receive an intravenous infusion of the investigational antidepressant.  The reason that poses a challenge is that both the CRU admission and the intravenous infusion could reasonably be expected to increase placebo response rates.    Nevertheless, this study with a small number of subjects (n = 30) demonstrated a statistically significant difference between the active and placebo infusion.

If the investigational drug has a truly novel mechanism of action, the issue of how to best proceed becomes even more daunting because there may be greater uncertainty about what dose to use and even how to determine the best dose to use. 
 
For this reason, adaptive trial approaches have been developed.  Such studies use an even larger number of doses than the 2-3 doses used in a traditional fixed dose study.  These studies use computer algorithms to preferentially assign patients going forward in the trial to the dose or doses that have the greatest likelihood of separating from placebo based on the data gathered by that point in the study.  Blind is maintained by having the dose assignment done by the computer algorithm analyzing the data and assigning prospective patients to specific dose options.

We at CRI have considerable experience with novel designs, the use of special populations, and adaptive designs to advance the early development of novel drugs in a cost and time effective way.  If you are interested in these approaches, we would be helpful to send you reprints of studies that we have been able to publish or discuss these approaches with you.