Wyeth vs. Levine is an important legal case being tried before the US Supreme Court. You may have read about the lawsuit in the New York Times, NEJM, JAMA, the Wall Street Journal, or my own blog. It revolves around the tragic story of a woman (Ms. Levine) who experienced an extremely rare side effect (severe tissue damage resulting in the amputation of her right arm) because a drug was administered improperly (into an artery rather than a vein). Ms. Levine is arguing that her injury could have been avoided if the drug label had stronger warning language, and the Vermont Supreme Court ruled in her favor, awarding her $7 million. The court ruled that a jury in the state of Vermont had the right to hold Wyeth accountable for a different labeling standard than the one approved by the FDA.
The plot thickens, however, in that Wyeth’s FDA-approved label very clearly discourages injection of their drug into or near an artery, and it also describes the potential consequence (including gangrene) of such an action. The FDA approved Wyeth’s label in full knowledge of the potential risks and benefits of the drug. In fact, Wyeth asked to strengthen the language of the label before Ms. Levine was injured, and the FDA declined to make the change because label changes are based on new information about a drug’s frequency or severity of risks. Wyeth had nothing new to disclose.
Why didn’t the FDA allow Wyeth to strengthen the warning language? Imagine if you were responsible for the drug labels of 10,000 different drugs, and you had to review and approve every proposed change? You’d probably want to limit the changes to substantive requests (rather than matters of semantics and word preferences) or else you’d have ten times the work with no real benefit to consumers. Imagine all the synonym swaps that would be requested: “Dear FDA, could you please change the word “rather” to “quite” on page 7 of drug label 986.01?” It would drive you nuts.
So to help drug companies understand which sort of changes should be submitted, the FDA requested that label changes be related to new information only (meaning: they’d be happy to change the label to reflect a change in risk based on new data suggesting a different frequency or severity of potential injury – otherwise leave the language alone).
The State of Vermont essentially ruled that local juries had the right to re-assess the risk/benefit analysis described on drug labels, and hold pharmaceutical companies to the jury standard, regardless of what the FDA had decided previously about appropriate labeling. Wyeth has appealed this to the US Supreme Court.
I have written quite extensively about what’s at stake in the Wyeth vs. Levine case at Science Based Medicine. Feel free to add your 2 cents – should juries set drug labeling standards, or should the FDA’s approval process preempt litigation in cases related to known risks (without new information)?