Public Health Watch: FDA Uses ‘Extrapolation’ in 1 in 4 Drug Approvals

Extrapolation is used often in scientific research to, as a paper published in 2004 puts it, “project, extend, or expand (known data or experiences) into an area not known or experienced so as to arrive at…knowledge of the unknown area by inferences.”

In drug approvals, however, the US Food and Drug Administration (FDA) often uses extrapolation to apply pivotal clinical trial findings “to a broader population of patients than was represented in those trials,” according to the researchers behind an analysis published on April 19 by JAMA Network Open. This is often necessary, the researchers wrote, because the pivotal trials used to gain FDA approval “are necessarily limited in the populations they cover.”

Most recently, the agency initially used extrapolation in its approval of aducanumab in 2021 for all patients with Alzheimer’s disease, “even though it was studied only in those with mild disease,” they noted.

If this sounds a bit scary—or even “antiscience”—perhaps it should.

“I think one message [from out study] is to scrutinize the trials on which approval was based and not necessarily assume that because something is approved by the FDA that it’s not important to examine the details of the approval,” study coauthor Aaron S. Kesselheim, MD, JD, MPH, a professor of medicine at Harvard Medical School and a faculty member in the Division of Pharmacoepidemiology and Pharmacoeconomics Brigham and Women’s Hospital, both in Boston, told Contagion.

The analysis of 105 novel drugs approved by the FDA between 2015 and 2017 found that 23 involved extrapolations of the trial population characteristics in the approved indication. In most cases, these included extrapolation of the trial findings to patients with greater disease severity (n=14 drugs) or with different disease subtypes (n=6). The remaining 3 cases concerned concomitant medication use.

Of relevance to Public Health Watch readers, at least 1 antibiotic—ceftazidime with avibactam—was among the 23. The antibiotic combination is indicated for adults with complicated urinary tract infections caused by Escherichia coli, Klebsiella pneumoniae, Citrobacter koseri, Enterobacter aerogenes, Enterobacter cloacae, Citrobacter freundii, Proteus species, and Pseudomonas aeruginosa. However, several pathogens included in the indication—Pseudomonas aeruginosa, Citrobacter freundii, Citrobacter koseri, Proteus mirabilis, and Enterobacter cloacae—made up less than 5% of those in the experimental arm of the study, according to Dr. Kesselheim and his colleagues.

“A drug that works in a milder form of a condition may not work the same way in a more severe form,” Kesselheim said. “This is why it’s extremely important to be aware when such extrapolation occurs and to make sure that we follow up on drug use after FDA approval to ensure that the actual benefits and risks are consistent and update the labeling if not.”