Phase II/III Seamless Design for Promising Investigational Drug

• Activity Number: 458
• Type: Contributed
• Date/Time: Wednesday, August 6, 2014 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Lei Pang*+
• Companies: Merck
• Keywords: seamless design ; combination test ; phase 2/3 ; dose response

Abstract

When efficacy is seen in early clinical data, the development time frame could be potentially reduced by a seamless phase 2/3 clinical trial. Current methods for seamless phase 2/3 trial design address the multiplicity issue from treatment selection by the means of combination test and p-value adjustment to control the family wise error rate. However, for compounds with strong proven biological effect, often we observe a monotone dose-response relationship in phase 2 trials. Recent examples are the PCSK-9 antibodies in reducing blood LDL level and BACE-1 inhibitors in reduction of CSF and brain amyloid protein level. In this research, a modified p-value adjustment method for the combination test is proposed, to avoid some of the efficiency penalty while still maintain adequate type-1 error control. The proposed method relies on the pre-specification of the order of the treatment group test statistics in the phase 3 go criterion. This method will reduce the overall sample size, while still maintain type 1 error control. The author shows that, for drugs with superior efficacy profile, the loss in power is minimal and can be controlled by study design. Simulation study was conducted