Site-Independent Confirmation of Subject Selection for an Alzheimer’s Disease Trial.

Lyketsos, C.,  Targum, S.,  Drake, K.,  Pendergrass, J.C.,  Munro, C., Smith, G.,  Lozano, A. “Site-Independent Confirmation of Subject Selection for an Alzheimer’s Disease Trial.” Poster session presented at: Alzheimer’s Association International Conference, July 11th – July 17th 2014, Copenhagen, Denmark.


Several studies have failed to detect a significant signal in clinical trials for Alzheimer’s disease (AD) due, in part, to inappropriate participant selection related to diagnosis or sufficient evidence of progressive symptom severity.  A “second” independent opinion regarding participant selection may improve selection of control groups that are likely to be progressive.   We created an enrollment review committee (ERC) to provide site-independent historical, medical, and diagnostic verification and progressive symptom severity confirmation prior to randomization in an ongoing clinical trial. Methods: ADvance ( identifier NCT01608061) is a double-blind, placebo-controlled evaluation of deep brain stimulation targeting fornix (DBS-F) in patients with mild probable AD being conducted at 7 academic sites within the United States and Canada.  Entry criteria include an ADAS-cog-11 score between 12-24 (inclusive).  As part of this study, we introduced an ERC that was responsible for review and approval of all participants proposed by the trial sites.  The ERC is composed of 3 clinicians who are not affiliated with the trial site proposing the participant: two AD clinicians (neurologists and/or psychiatrists) and a neurosurgeon.  Challenged cases generate adjudication between the site investigator and a member of the ERC. Results: To date, 40 of 80 potentially eligible subjects (50%) failed to meet eligibility criteria at screen (or baseline).  26 of these eligibility failures related to ADAS-cog-11 scores that were either too low (<12) or too high (>24).  Other sources of ineligibility included diagnostic, psychosocial, or surgical issues that would affect study participation or preclude accurate assessments. Conclusions: In this survey of patients proposed as eligible for a mild probable AD study, symptom severity (either too high or too low) as measured by the ADAS-cog-11 was the most frequent reason for eligibility failure.  The use of a site-independent review strategy for subject validation prior to randomization may improve the rigor of the assessment and facilitate better signal detection, in part by assuring the progressive nature of cognitive and functional change in the year prior to enrollment.