By Turna Ray
Under a collaboration with personalized medicine-focused bioinformatics company MolecularHealth, the US Food and Drug Administration will evaluate software that will allow the agency to analyze and predict drug safety issues caused by a variety of factors, including pharmacogenomic interactions.
After evaluating the Molecular Analysis of Side Effects, or MASE, software the FDA will likely use a version of the system to gauge safety issues with marketed drugs. "The immediate plans are to use [MASE] in the post-market setting," an agency spokesperson told PGx Reporter.
MASE is a software-as-a-service offering that MolecularHealth is marketing to a variety of end users, including drug developers and payors. As reported by PGx Reporter sister publication BioInform, the company launched MASE last month (BI 1/20/2012).
MASE allows users to assess drug safety by analyzing data on therapeutic mechanisms of action, treatment-treatment interactions, as well as associations between molecular markers, diseases, and drugs in published literature. The software also includes visualization and analytical features that enable users to explore drug safety issues from a statistical or molecular perspective.
The capabilities of MASE are aligned with the FDA's expressed intent to detect and analyze drug safety issues at a mechanistic level. "The FDA Predictive Safety Team has been meeting with Molecular Health … for over a year," the agency spokesperson said, noting that the project builds on work FDA has been doing internally "for years" to mine its Adverse Event Reporting System — a computerized system to monitor drug-related adverse events — to understand mechanisms of drug toxicity.
Adverse reactions are often not detected for several years after a drug has been approved. In recent years, however, as genetic testing firms and bioinformatics companies have begun to amass large databases of molecular and phenotype data on individuals treated with a variety of drugs, the FDA has expressed a desire to use this information to detect treatment-related toxicities faster and with more precision.
The commercially available version of MASE combines patients' drug-related clinical and molecular data with information from a publicly available version of FDA's AERS. MolecularHealth has said that MASE users can use the system to analyze mechanism-based safety data for all drugs currently on the market, as well as identify potential adverse events for drugs under development.
According to the FDA spokesperson, MolecularHealth has created a new version of MASE that the company will present next month to multiple offices in the Center for Drug Evaluation and Research.
"Specific projects for collaboration will be discussed at next month's meeting," the agency spokesperson said. "The Predictive Safety Team has interest in hepatotoxicity and cardiotoxicity. Other areas of interest will likely include some focus areas related to the Sentinel Initiative."
In 2008, the FDA launched the Sentinel Initiative, under which the agency is implementing a proactive, electronic system for tracking adverse events associated with drugs, biologics, and medical devices that the agency has approved for marketing. Under this effort, the agency hopes that with a more refined mechanistic understanding of drug safety it can pick up on potential toxicity signals associated with investigational drugs and ink risk-mitigation strategies with sponsors to avoid adverse events when the drug goes on the market.
In 2010, the FDA's Advisory Committee for Pharmaceutical Science and Clinical Pharmacology held a meeting to discuss plans to build a predictive safety system that integrates pharmcogenomics, chemical structure data, and systems biology approaches to predict drug-induced adverse events before they happen (PGx Reporter 3/24/2010).
Then, last year the FDA held several drug safety workshops in which the agency discussed its intent to develop a mechanism-based drug safety assessment and prediction program. MolecularHealth's collaboration with the FDA around the MASE system evolved during that time.
MASE is exactly the type of technology that FDA has described, according to Jeffrey Marrazzo, MolecularHealth's chief business officer. "It supports not only the ability to look at adverse event information from a statistical basis, but it allows you to link understanding of how drugs work in the human system, thereby allowing you to identify, either proactively or even afterwards, safety signals with a drug," Marrazzo told PGx Reporter.
In its meeting with the FDA, MolecularHealth discussed how it could launch MASE commercially but also garner input from the agency. Marrazzo said the FDA will likely tweak the system for its own purposes.
"MASE will continue to use only the publically available AERS data at this time," the FDA spokesperson explained. "Internally, we may perform our own data-mining analyses using the in-house AERS data as a quality-control assessment."
According to the agency, its collaboration with MolecularHealth is one of several public-private partnerships focused on improving its adverse event predictive capabilities through better understanding of mechanistic-based drug safety issues. "The goal of the Predictive Safety Team is to bring together expertise from within the FDA, academia, and pharmaceutical sponsors," the agency spokesperson said. "The FDA has other Research Collaboration Agreements in place and is looking to create new RCAs."
According to the spokesperson, the agency is interested in inking additional drug safety collaborations around ontologies, cheminformatics, bioinformatics, and systems biology.
Have topics you'd like to see covered in Pharmacogenomics Reporter? Contact the editor at tray [at] genomeweb [.] com.