In 1991 the National Institute on Aging began funding the Alzheimer's Disease Cooperative Study (ADCS), a consortium of 29 University based sites located throughout the United States. The coordinating center is located at the University of California in San Diego. The lessons learned and the process developed in establishing the study provide an excellent parallel and case study for developing similar studies focused on Multiple Sclerosis.
The ADCS operates under a specific two-fold mandate: 1.) to conduct clinical trials in Alzheimer's Disease using compounds that would not be tested by industry and 2.) develop novel instruments for use specifically in Alzheimer's clinical trials. Typically, the multi-centered studies have used compounds that do not require industry sponsorship. Usually, this is because they lack patent protection, either because they are available over the counter, or they are already off-patent and are thus unlikely to be developed by industry for use in Alzheimer's disease. Examples of this include Vitamin E, prednisone and estrogen (for women).
More recently, the ADCS entered into partnerships with industry. An example is a study which compared the impact on disease progression in Alzheimer's patients treated with a cholinesterase inhibitor that is made by Pfizer, Vitamin E, or a placebo
Several instruments developed or improved upon through the ADCS are now being used in industry-sponsored trials. Typically Alzheimer's clinical trails include several domains, including cognition, behavior, and function.
The ADCS has also conducted several studies aimed at symptom management, including agitation and sleep disorders. While these studies are not aimed specifically at targeting the progression of AD, they do evaluate important behavioral symptoms that impact the quality of life for both the patient and the caregiver.
The structure of the ADCS is overseen by the principal investigator, Dr. Leon Thal. In addition, each of our 29 member sites has a site principal investigator and, together, this group comprises the steering committee. The steering committee reviews the status of ongoing studies, provides input into the development of new protocols, and votes to carry forward any new protocol. No study can go forward without a majority vote from the steering committee. Since 1991 more than 10 clinical trials have either been completed or are underway.
A number of subcommittees carry on the work of the ADCS: The industry liaison committee; the internal and external ethics committees; the instrument development committee; the publications committee; and the data safety monitoring committee for each protocol. Working groups work with the project director to develop and refine each protocol.
The database for the ADCS is located centrally at UCSD. All of the data is entered and cleaned there. Any of the site PIs may submit an analysis request. Typically, the project director with a few key investigators conducts the primary analysis of any project. Secondary and tertiary analyses are open to any member site PI or their colleagues.
The careful planning and timely initiation of a clinical trial requires a great deal of planning. Some of the major milestones in conducting a clinical trial include protocol development, identification and training of personnel, developing an implementation timeline, managing the ongoing study, and conducting an appropriate closeout of the protocol.
There are a number of interesting questions and hypotheses that can relate to any single study. It's essential to identify the pivotal question to be addressed in the study to form a primary hypothesis. The number of secondary hypotheses should be limited. The study design is dependent on the primary hypothesis. It also reflects a trade-off between what might be ideal in theory, and what is practical in terms of resources and enrollment.
The next step is to identify the appropriate outcome measures to assess the primary and secondary hypotheses. The number of outcome measures should be small to avoid being overly burdened to both the patient and the site staff. Usually, we evaluate which key domains are studied in the protocol (cognition or behavior or function) and select from our existing instruments to identify the ones that are most appropriate.
Power calculations are done through a variety of sample sizes, and they are used to develop the most powerful protocol possible given the available resources. The protocol should also include a prespecified statistical plan, particularly for the primary hypothesis. The protocol needs to specify whether or not an interim analysis will be conducted. If it is conducted, there will be an associated loss of power that needs to be addressed in the statistical plan. Also "stopping rules'' need to be specified in advance. These may be based on safety issues, efficacy, or lack of efficacy, but they do need to be written into the protocol. There are a number of other logistical issues including, figuring out how many sites you need to adequately enroll for your target patient population. It's important to realistically evaluate the centers involved and determine how many patients they truly can enroll.
The inclusion and exclusion criteria section, along with the schedule of events, are the portions of the protocol most often referenced, once a study is underway. The inclusion and exclusion criteria need to be worded very carefully and clearly and should avoid using any vague language.
Early in protocol development, it is important to evaluate the coordinating center's staff. The coordinating center's staff typically consists of a project director, administrative personnel, the clinical monitors, the database manager, and statistical staff. Together, this core group implements the clinical trial.
A carefully planned timeline is essential for a smooth implementation. Timelines are developed at least a year before the anticipated start-up date for any study. At least 12 months prior to start, it's important to have a draft protocol that's being worked on by the working group.
Also, it's important to identify key personnel at the coordinating center early on. This will allow a window of opportunity to hire and train any newly needed staff. At least 9 months prior to starting the protocol, the site and the protocol PIs (who may be different from the steering committee site PIs), should be identified. Every effort should be made to finalize the protocol and the budget at this point, so then contract negotiations can begin with the participating sites.
Closeout of the study also requires careful planning. The goal is to minimize the amount of time between the time of the last patient completing the protocol, and the data being prepared for analysis. The data cleanup is a two-phase process. The first is to collect any data that is remaining onsite, and get it into the coordinating center as quickly as possible. The second aspect is that once it's in house, get the data clean and complete, and ready for analysis. While the data cleanup is going on, the monitors should be going to the sites and counting on reconciling any remaining study medication and arranging for destruction. The data analysis should follow the plan that was prespecified in protocol. At this point in the analysis, it should really focus on the primary hypothesis, and one or two of the secondary hypotheses. But any subanalyses should be left for later. The preparation of the primary paper addresses issues such as: who among the participating sites will work with the project director to prepare the paper, and who will be listed as coauthors. This is also the best time to obtain feedback from sites on improving future procedures. They are the ones who have the most experience in following the procedures manual and guidelines for the study, and they are invaluable resources.
Go to TOP.
|
|
|