Drug safety analyses should factor product benefits into the equation and not just focus narrowly on reducing risks, FDA Acting Chief Scientist Jesse Goodman said during an address at the FDA/Food and Drug Law Institute annual conference on April 22.

"How do we define safety?" Goodman asked. "I think there are two variables. Safety is really about minimizing risk and maximizing benefits." Goodman's slides note that the benefit side of the equation is the one that is discussed "the least."

Goodman's remarks echo some of the ideas expressed by former FDA Deputy Commissioner Scott Gottlieb, now at the American Enterprise Institute. Gottlieb has posited that the data mining of health claims databases may help illuminate product benefits, not just product risks, that were not observed during initial registration studies (¹ Also see "Comparative Research Needs New Approach To Answer Real-Life Questions" - Pink Sheet, 23 February, 2009., p. 3).

But the view that safety considerations should include benefits does not mean that nebulous risks can be ignored, Goodman indicated.

Industry long has been concerned about FDA's program of "preliminary communication" about drug safety information while the agency continues to investigate whether an observed problem actually is linked to a product. Firms believe such alerts may needlessly discourage patients from taking a beneficial medication (² Also see "FDA List Of Potential Safety Signals Has 3 Pfizer Drugs, Echoes Drug Watch Site" - Pink Sheet, 8 September, 2008., p. 3).

"With the drug safety issues that have come up, there are often data that [are] preliminary, that we don't know how to interpret, so frequently our communications deal with the challenges of uncertainty," Goodman said. "But I believe that ... in this area, transparency may be challenging in the short run, communication may be challenging, but I think even in the face of uncertainty, transparency is the right approach, and the 'see no evil' approach is not really an option."

"We need to get much better at communication," Goodman acknowledged. "But the better we are at communication, the more trust there is and the better decisions people will make."

The "revolution in informatics offers tremendous opportunities" to improve safety, Goodman noted. "Molecular science can be used to design products, predict toxicity, link clinical outcomes to the genome, deal with production."

Personalized medicine vs. comparative effectiveness

The informatics revolution also may help resolve another hot-button issue: how the government's new push for comparative effectiveness will impact the growing science of personalized medicine.

"It's really been cast as either/or," Goodman said, in which "there's going to be personalized medicine now and that's it, or people looking at randomized trials or comparative effectiveness are ignoring individual data."

"We have to reach for a synthesis here, sort of a new epidemiologic paradigm," he said. "Years ago, doctors just said, well, my patient got better when I did whatever, and therefore I'll do that to the next 10 people - sometimes right, sometimes wrong, but not always effective so we moved to controlled trials."

"So now we have the ability within carefully managed studies and within health care practice to look at individual and subgroup responses to therapies, and I think not just to identify the best therapy for individuals and not just the best for populations, but the best therapies for groups within these populations," he said.

"They may not even be single drugs, they may be interventions, they may be public health interventions," he noted.

At present, "most people are focused on targeting effectiveness or preventing harm," he said, such as efforts to identify who should receive warfarin. "Ultimately, where I think we are going to get to is more shades of relative benefit which encompasses safety and effectiveness."

"We can move from everybody should or shouldn't get this treatment and some people are going to benefit and some people aren't - and some people are going to get hurt and some people aren't - to knowing that maybe for 90 out of 100 people, we are really giving the best therapy there is out of the whole menu."

-M. Nielsen Hobbs (n.hobbs@elsevier.com)

Drug Safety Should Focus On “Maximizing Benefit,” Not Just “Minimizing Risk” – FDA’s Goodman

More from Archive

More from Pink Sheet