|
Activity Number:
|
100
- Statistical Innovations for Drug Approval and Reimbursement for Rare Disease
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 3, 2020 : 1:00 PM to 2:50 PM
|
|
Sponsor:
|
International Indian Statistical Association
|
|
Abstract #309264
|
|
|
Title:
|
On Analysis of Single Arm Trial with Natural History Controls
|
|
Author(s):
|
Qing Liu*
|
|
Companies:
|
Quantitative and Regulatory Medical Science, LLC
|
|
Keywords:
|
Gene therapy;
Rare disease;
Natural history controls;
Real-world evidence;
Virtual matched controls;
Tipping point analysis
|
|
Abstract:
|
The 2018 FDA draft guidance on gene therapy for rare diseases states that a single arm trial with natural history controls can be used as the basis for assessing comparative effectiveness of a new gene therapy with an existing control. To adjust for many confounding risk factors, we propose a virtual matched control method such that the same set of natural history control data is used to provide virtual matching for each patient of the single arm trial. This 1-to-many virtual matching results in matched differences that are statistical dependent and cannot be analyzed with existing one sample test procedure. We resolve this issue via an exact conditional intra-patient (ECIP) test of patients receiving the treatment where the null distribution of the test statistic is evaluated according to the actual control distribution. In contrast, a traditional one-sample test assumes a symmetric distribution of the error terms about zero under the null hypothesis. By shifting the control distribution, we can identify the tipping point at which the ECIP test loses statistical significance. The tipping point analysis allows assessment of the robustness of results of the single arm trial.
|
Authors who are presenting talks have a * after their name.
