Pitt Logo LinkContact Us

An Adaptive Two-Stage Dose-Response Design Method for Establishing Proof of Concept in Drug Development

Tanaka, Yoko (2011) An Adaptive Two-Stage Dose-Response Design Method for Establishing Proof of Concept in Drug Development. Doctoral Dissertation, University of Pittsburgh.

[img] PDF - Primary Text
Restricted to University of Pittsburgh users only until 31 January 2016.

Download (1808Kb) | Request a copy

    Abstract

    In clinical drug development, searching for the true dose-response curve is ethically and logistically challenging. Establishing evidence of dose-response or Proof of Concept (PoC) is the first step for both determining the best dose-response model and optimizing a treatment dose correctly for clinical use. To overcome these challenges, we employ an adaptive two-stage design where both adding and dropping treatment arms is possible between stages. In the first part of this dissertation, we develop a method extending the Multiple Comparison Procedures and Modeling (MCP-Mod) approach into this adaptive two-stage design. Our goal is to establish "global" PoC across the stages. Between stages, we propose using an Adding and/or Dropping Treatment Adaptation Rule (ADTAR). In the ADTAR method, dose specifications in the second stage depend on the first stage's results. Treating the unobserved doses and imbalanced aggregate sample sizes in the second stage as missing data, we derive weights and adjust the test statistics in the second stage. Specifically, we assume that the missing data mechanism caused by ADTAR is missing at random. At the end of the second stage, we perform the global PoC test combining the test results from both stages. To preserve the family-wise error rate, we use a Conditional Error Function. Using simulation studies, we evaluated our design method and compared it to a conventional (one-stage) study design and different fixed two-stage designs. Our method showed overall robust high power for detecting the global PoC across three forms of true dose-response curves. In the second part of this dissertation, we find constraints for choosing doses in the original and extended MCP-Mod methods. Specifically, we establish lower bounds of the number and levels of doses for each method using simulation studies. Our proposed method is a viable tool in searching for a dose-response relationship. In accordance with ICH guidelines, our method helps to provide optimal doses of drugs for treating or preventing different diseases. Since drugs are widely used in human populations, such methods have a great Public Health impact in appropriately treating or preventing many types of diseases.


    Share

    Citation/Export:
    Social Networking:

    Details

    Item Type: University of Pittsburgh ETD
    Creators/Authors:
    CreatorsEmailORCID
    Tanaka, Yokoyoko.tnk@gmail.com
    ETD Committee:
    ETD Committee TypeCommittee MemberEmailORCID
    Committee ChairAnderson, Stewartandersons@nsabp.pitt.edu
    Committee MemberWahed, AbdusWahedA@edc.pitt.edu
    Committee MemberSampson, Allan Rasampson@pitt.edu
    Committee MemberBertolet, Marniebertoletm@edc.pitt.edu
    Title: An Adaptive Two-Stage Dose-Response Design Method for Establishing Proof of Concept in Drug Development
    Status: Unpublished
    Abstract: In clinical drug development, searching for the true dose-response curve is ethically and logistically challenging. Establishing evidence of dose-response or Proof of Concept (PoC) is the first step for both determining the best dose-response model and optimizing a treatment dose correctly for clinical use. To overcome these challenges, we employ an adaptive two-stage design where both adding and dropping treatment arms is possible between stages. In the first part of this dissertation, we develop a method extending the Multiple Comparison Procedures and Modeling (MCP-Mod) approach into this adaptive two-stage design. Our goal is to establish "global" PoC across the stages. Between stages, we propose using an Adding and/or Dropping Treatment Adaptation Rule (ADTAR). In the ADTAR method, dose specifications in the second stage depend on the first stage's results. Treating the unobserved doses and imbalanced aggregate sample sizes in the second stage as missing data, we derive weights and adjust the test statistics in the second stage. Specifically, we assume that the missing data mechanism caused by ADTAR is missing at random. At the end of the second stage, we perform the global PoC test combining the test results from both stages. To preserve the family-wise error rate, we use a Conditional Error Function. Using simulation studies, we evaluated our design method and compared it to a conventional (one-stage) study design and different fixed two-stage designs. Our method showed overall robust high power for detecting the global PoC across three forms of true dose-response curves. In the second part of this dissertation, we find constraints for choosing doses in the original and extended MCP-Mod methods. Specifically, we establish lower bounds of the number and levels of doses for each method using simulation studies. Our proposed method is a viable tool in searching for a dose-response relationship. In accordance with ICH guidelines, our method helps to provide optimal doses of drugs for treating or preventing different diseases. Since drugs are widely used in human populations, such methods have a great Public Health impact in appropriately treating or preventing many types of diseases.
    Date: 31 January 2011
    Date Type: Completion
    Defense Date: 17 September 2010
    Approval Date: 31 January 2011
    Submission Date: 14 October 2010
    Access Restriction: 5 year -- Restrict access to University of Pittsburgh for a period of 5 years.
    Patent pending: No
    Institution: University of Pittsburgh
    Thesis Type: Doctoral Dissertation
    Refereed: Yes
    Degree: PhD - Doctor of Philosophy
    URN: etd-10142010-124612
    Uncontrolled Keywords: Clinical Drug Development; Dose-Response Models
    Schools and Programs: Graduate School of Public Health > Biostatistics
    Date Deposited: 10 Nov 2011 15:03
    Last Modified: 07 May 2012 14:36
    Other ID: http://etd.library.pitt.edu/ETD/available/etd-10142010-124612/, etd-10142010-124612

    Actions (login required)

    View Item

    Document Downloads