In response to repeated drug busts, the drug industry came together in a way that flew in the face of the usual institutionalized secrecy surrounding development data. In 2010, a group of major pharmaceutical companies commenced sharing data from failed clinical trials to try to figure out what went wrong in the studies and what could be done to improve drug development.
The Critical Path Institute (C-Path), founded in 2005 in Tucson, Arizona, has facilitated this cooperation and data sharing. Through its Coalition Against Major Diseases (CAMD) program, C-Path intends to accelerate the development of therapies for neurodegenerative diseases by advancing drug development tools for regulatory approval.
"I think the greatest accomplishment of C-Path is what we have learned from the over 6,000-patient database in 22 clinical trials," Raymond L. Woosley, M.D., Ph.D., president emeritus of C-Path, told Genetic Engineering & Biotechnology News (GEN). "At the end of the year (2011) we looked at the aggregate data and found you couldn't even tell the average age of a study patient because the investigators had recorded data in different ways. Neither were the cognition scales and endpoints in the clinical trials consistently recorded."
C-Path manages an industrial consortia of companies willing to share precompetitive knowledge and work in support of projects that are identified as high priority by the FDA and are in the interest of public health. Since 2005, over $20 million (M) in grants and $10M in contributions have been received to support its work. To serve as a neutral and trusted third party for collaborators, C-Path says that it does not accept money from organizations that develop products regulated by the FDA or that would create a real or perceived conflict of interest.
Created jointly by C-Path and the Engelberg Center at the Brookings Institution, CAMD's members include six nonprofit groups representing patients' interests, 15 pharmaceutical companies, FDA, EMA, the National Institute on Aging (NIA), the National Institute of Neurological Disorders and Stroke (NINDS), and representatives from academia.
Broadly stated, the CAMD's initial work involves the sharing of clinical trial information by its corporate members to generate the core information needed to build new disease models that represent populations. These models incorporate data from concurrent research as well as biomarker information from other research efforts. CAMD has the active participation of the FDA and its commitment to review data for possible qualification. The coalition plans to seek qualification for these models and biomarkers and place them into the public record to enhance research and drug development.
In 2010, seven of CAMD's member organizations agreed to share their data from 11 recent AD clinical research studies. Data was standardized, pooled and made available to qualified researchers around the world. They invested significant in-kind resources to remap the retrospective data to the new format that is now part of the Clinical Data Interchange Standards Consortium (CDISC) standard.
Those organizations included Abbott Laboratories, Alzheimer's Disease Cooperative Study, AstraZeneca Pharmaceuticals, GlaxoSmithKline, Johnson & Johnson, Pfizer and Sanofi-Aventis. C-Path also worked with Ephibian to build a secure online data repository. Data from additional drug makers and the NIH will be added in the future, according to CAMD.
Increasingly, CAMD's approach integrates a "disease model" and a "drug model," which together are used to simulate and design clinical trials with a greater chance of success. During the process, participants remapped their data to the new CDISC standard and combined them into a 6,000-patient database now being actively used by CAMD to test initial scientific results before formal submission to the FDA. To generate new data for this number of patients would cost millions of dollars and take more than six years in new clinical trials, the institute says.
"One of the things we learned was that getting the data was just the first step," Dr. Woosley noted. "For example, no drug could have ever been found effective in these studies because many of the patients never progressed. Looking at trial data, it was all over the place.
"Once you got 6,000 patients in a usable database, at the first cut last December, we were able to produce a quantitative disease model that would allow us to track progression in three different patient subsets: nonprogressors, slow progressors, and rapid progressors. So for future studies you can design a study with adequate power and duration and identify an effective drug.
"To me," Dr. Woosley added, "it demonstrated that the industry was willing to put in the time to get the work done to understand the data well enough to merge it. Setting data standards was our biggest accomplishment."
CAMD has made some considerable progress thus far. It is evaluating Parkinson's disease (PD) clinical trial data from NINDS, GE Healthcare, and the Institute for Neurodegenerative Diseases under the new CDISC Parkinson's disease data standard. This data will be used to further develop and enrich a Parkinson's disease model first developed by the FDA.
The coalition also aims to create a similar pooled database for tuberculosis, and the data will be available to all the participating drug makers and outside researchers with a valid scientific question. Beyond data sharing and cooperating in the development of clinical data standards, CAMD is also helping FDA define the qualification process for drug development tools.
As CAMD moves further toward its goal of developing and qualifying the most modern tools and methods for AD and PD studies, drug developers may have more confidence that shared resources will enhance their ability to design meaningful clinical trials for neurodegenerative disorders.
Patricia F. Dimond
Genetic Engineering & Biotechnology News