Real-World Observational Data Factor into FDA's Decisions

FDA in February started celebrating the 10th anniversary of the project that enables the agency to actively identify and analyze risks in products on the U.S. market and lessen reliance on spontaneous reports.

The Sentinel Initiative, proclaimed Commissioner Scott Gottlieb during a February 7 public workshop in Bethesda, Maryland, “today provides a truly 21st-century surveillance system for the agency.”

FDA launched the Sentinel Initiative in 2008, describing it as a project to detect signals of potential adverse events unanticipated in products’ clinical trials or during regulatory review. Authorization for the project, as well as a $25-million annual appropriation for the first 5 years, came from the Food and Drug Administration Amendments Act of 2007.

“Better, faster, cheaper” is how some of the speakers at the daylong workshop described the Sentinel System, which came online 2 years ago after a pilot program known as Mini-Sentinel.

“Sentinel has become one of the largest and most successful users of big data in healthcare—a much more efficient system that can generate much more high-quality evidence in a shorter period of time,” Gottlieb said. “And though it continues to develop, we already realized many of the benefits that come from the ability to use real-world observational data as a tool to monitor medical product safety and identify further concerns and be able to evaluate them.”

As of February 7, information had been published online regarding several Sentinel analyses whose results, according to the project’s coordinating center at the Harvard Pilgrim Health Care Institute, “contributed in some material way to inform an important regulatory discussion or action.”

The results of one such analysis figured into FDA’s December 2017 announcement that the labeling for all gadolinium-based contrast agents will identify pregnant women as being among the patient groups possibly at higher risk for retention of the heavy metal.

From querying the Sentinel Distributed Database, FDA determined that the project’s U.S. data partners had electronic records on magnetic resonance imaging procedures conducted during 3,726,555 pregnancies from 2006 through 2015. A gadolinium-based contrast agent was administered during 8,842 of these pregnancies, usually in the first trimester, FDA reported at a September 2017 advisory committee meeting. That level of usage—1 in 421 pregnancies—was 8 times the rate reported in a Canadian retrospective study that found a significant risk for stillbirth or neonatal death after exposure to a gadolinium-based contrast agent.

“It’s amazing how large numbers get small fast,” remarked Richard Platt, principal investigator for the Sentinel Coordinating Center, on the general challenge of examining outcomes that happen infrequently.

Another Sentinel analysis was cited as support for FDA’s decision, in response to a 2015 citizen petition, not to withdraw the agency’s approval of applications to market ketoconazole 200-mg oral tablets.

The marketing of oral medications containing ketoconazole was suspended in the European Union in 2013, after regulators concluded that the risk of liver injury exceeded the benefits from treatment.

But ketoconazole, FDA stated, is the only agency-approved antifungal that is formulated for oral use and indicated for the treatment of coccidiomycosis. The antifungal is also indicated for the treatment of blastomycosis and histoplasmosis.

“FDA’s updated analyses support the [agency’s] conclusion that the risks and benefits of ketoconazole remain favorable for treatment of patients with these systemic fungal infections who do not have other treatment options,” Janet Woodcock, director of the FDA Center for Drug Evaluation and Research (CDER), wrote to the petitioners in September 2017.

After querying the Sentinel Distributed Database, which had electronic records associated with about 160 million unique person identifiers, FDA determined that 13 or fewer patients per year during the period 2009–15 were prescribed ketoconazole to treat coccidiomycosis; 10 or fewer patients per year were prescribed ketoconazole to treat blastomycosis or histoplasmosis.

FDA also determined from its 2016 query that usage of ketoconazole had decreased substantially since 2013, once the labeling was revised to warn about severe liver injuries and adrenal gland problems and instructed clinicians not to prescribe the drug to treat fungal nail infections, ringworm, or candidal infections.

All of the electronic data are held by Sentinel Initiative partners, not FDA. As of February, according to the project’s website (www.sentinelinitiative.org), 18 data partners such as Aetna and Hospital Corporation of America can respond to queries.

The database started integrating Medicare data in January, said Michael Nguyen, CDER’s lead for Sentinel activities.

[This news story appears in the March 15, 2018, issue of AJHP.]