2:52 PM
 | 
Oct 05, 2017
 |  BC Innovations  |  Strategy

Fixing FDA’s FAERS

FAERS is an underexploited but undercurated resource for de-risking tox

Last week FDA announced a new searchable format for its adverse events reporting database, making it easier for drug developers to use the tool to de-risk drug candidates. But according to an August publication from Novartis AG and UCSF, the database needs some scrubbing before it can live up to its promise.

The FDA Adverse Event Reporting System (FAERS) was created to collect information on adverse events that arise outside of the carefully controlled settings of clinical trials after a drug has made it onto the market, including low frequency AEs and those caused by combinations of therapies not tested in the clinic.

FAERS data became publicly available in 2013 and the database is emerging as a key resource for post-marketing pharmacovigilance. It contains roughly nine million AE reports across a vast array of licensed drugs, with about a million new reports added annually.

But the database also represents a potential treasure trove for drug developers aiming to look for clues to human toxicities their preclinical drug candidates might produce.

Using FAERS, any company can compare the chemical structure and targets of its drug candidates to those of marketed therapies, offering a powerful predictive tool to complement standard in vitro and animal toxicology studies. Tapping the data should help companies select the safest targets and compounds, and begin developing risk mitigation strategies early in the drug development process before a compound enters the clinic.

However, Laszlo Urban, global head of preclinical safety profiling at Novartis Institutes for Biomedical Research (NIBR), told BioCentury FAERS contains reporting errors and structural flaws that have prevented its wide adoption by industry.

“First you had to uncover the reliable data from this rubble of confounding data. Our paper centered on that massive and still uncompleted task.”

Brian Shoichet, UCSF

“We used FAERS data, but not to a large extent because we recognized there are biases and confounding factors in FAERS. In fact, some of the side effects reported are actually disease symptoms,” said Urban.

Last week FDA made searching the data more user-friendly. Previously, companies had to download FAERS data and create their own programs to analyze it or work with a third party company that had already built such a program. Now, companies can search FAERS by drug name or query reports from a specific time frame to assess the impact of specific events on side effect reporting.

But Urban said the changes don’t address the structural problems with FAERS. In eLife in August, he and Brian Shoichet at the University of California San Francisco characterized the most pressing issues with the database and created workarounds to address some of them. Shoichet is a professor of pharmacy at UCSF and an academic advisor for NIBR.

According to Schoichet, most...

Read the full 2284 word article

User Sign in

Trial Subscription

Get a 4-week free trial subscription to BioCentury Innovations

Article Purchase

$85 USD
More Info >