The more detailed response is that these results should be a wake-up call to the: a) key opinion leaders in the scientific community, b) sponsors of clinical trials, c) FDA, d) NIH and e) legislators/policy formulators that the old and familiar ways of therapy development are not producing successful results in our efforts to develop treatments for Alzheimer’s disease.

To avoid future disappointments or failures we need a radical recalibration in the current thinking and paradigms of therapy development. Specifically, many of our assumptions and firmly held ideas require reevaluation.  The list of these key topics include: a) conceptual models or theories about the disease, b) promising therapeutic targets, c) pathogenesis of the disease, d) the design of clinical trials, e) regulatory process and requirements for therapy approval, f) the costs and financing of prolonged clinical trials, g) laws governing patent, intellectual property and exclusivity and h) barriers to public-private partnerships in therapy development.

This process of rethinking is being pursued by the Campaign to Prevent Alzheimer’s Disease by 2020 (PAD2020, www.pad2020.org).  Initially, this effort was started by a series of annual think-tank meetings organized by the Leon Thal Symposia (2007-2009).  Today, the PAD2020 Campaign is offering the scientific community and all other stakeholders an opportunity to participate in a comprehensive strategic planning process.  The penultimate goal is the development of a national plan of action on how to address the multitude of complex problems and challenges confronting therapy development.

            As a ‘Case Study’, the very disappointing results of the Dimebon trial may not be so much due to ‘bad or ineffective drug’ being tested but a reflection on the failure in the: a) study design, b) criteria for subjects selection, c) study duration, d) regulatory constraints/barriers and e) cost of trial or financial consideration.  The Dimebon trial is an excellent lesson illustrating the very difficult situation facing the scientific community and pharmaceutical companies particularly to develop therapies for Alzheimer’s disease.

In the end, the combination of insensitive measures, questions about the drug’s mode of action and a fundamentally different subject population defeated this trial. In the Russian study, the subjects were drug naive and consequently may have been more fertile for intervention.  Presently most Alzheimer’s disease clinical trials in the US are conducted either as an ‘add-on’ to an ongoing AChEI treatment regiment, which likely exert some effect, or with populations that have failed other therapies; thus hardly drug naïve subjects.  Therefore, if the compound is only modestly efficacious it will not rise above these obstacles.  There are also challenges in running a trial in multiple countries with multiple cultures and languages.  These are not trivial issues and pose barriers to a successful trial in AD, but ultimately should not be insurmountable.

Our field, and especially pharmaceutical R&D companies that run these massive and expensive trials, tends to latch onto primary outcome measures (in this specific case, the ADAS-Cog) merely because they appeared to work once or twice before, and because the FDA approved an NDA (e.g., Aricept) based on data from that test.  The fact that there have been a few successes with that measure in pivotal trials does not, in any way, take away the fact that the measure has very poor psychometric properties, it has had repeated problems in multiple validation trials – especially at the mild end of the disease spectrum, it is resistant to ANY meaningful change over a 12 month period in mild AD (because of the poor psychometric characteristics), and we can frankly do MUCH better.  But, there is a steadfast insistence on calling a test a “gold standard” merely because it worked once or twice before.  We all have to get it through our heads that there is no “gold standard”, and the utility of individual outcome measures for a given study will always depend on specifics of the population (severity of disease), practical study limitations, etc…

Alzheimer’s disease, along with other neurodegenerative disorders, starts many decades before the first clinical signs or symptoms appear. Current practices in clinical trial design [based on regulatory requirements] typically require patients to have some clinically recognized/accepted indications such as Mild Cognitive Impairment or mild forms of the disease.

While this traditional approach with currently accepted outcomes and measures makes it easy to recruit subject and fulfills a regulatory requirement of demonstrating efficacy, it works against compounds that may putatively act to restore or maintain nerve cell functioning.  Thus, compounds such as Dimebon have no chance of showing efficacy in these late stages of neurodegeneration (Mild to Moderate AD) when cells and functional systems have deteriorated beyond repair.  This situation is true for a number of compounds or trials in Alzheimer’s disease.  The solution or a true test for the efficacy of such compound is to start the trial during asymptomatic stages of the disease, but then we run into a number of other problems.

So, in the case of Dimebon, too much was riding on the hope that there would be movement on a scale that is tough to move, in mild-to-moderate disease, in the space of a year. In addition, well, the drug just may not work.  In this case, it may be a mixture between poor actual drug effectiveness and poor sensitivity to change of the measures used.  Back to the drawing board!

The PAD2020 Campaign aims to find a way out of this Catch-22 situation. [visit www.pad2020.org]."