Kimberly K. Egan, Esquire
DLA Piper US LLP
Washington, DC
The press, blogosphere, the United States Congress, and the plaintiffs’ bar have all taken aim at FDA’s August 22, 2008, final rule codifying its “longstanding view” of when it is appropriate for a sponsor to use the “Changes Being Effected” (CBE) rules to update a label without prior agency approval. The final rule explains that sponsors may unilaterally add or update a warning, precaution, or contraindication in only one circumstance: when “newly acquired information” provides “sufficient evidence of a causal relationship” between a drug, device, or biologic with an adverse reaction.
In a letter to FDA, Representative Henry Waxman (D-CA) asserted that this evidence-based test is a new “higher standard” for prescription drug warnings, “apparently designed to bolster the argument by companies defending against lawsuits that the regulations precluded them from adding contraindications, warnings, precautions, and adverse reactions in the absence of FDA approval.” He claimed that the new rule would “drastically limit the situations in which a manufacturer is permitted to add or strengthen a contraindication, warning, precaution, or adverse reaction without waiting for FDA to approve such a change.” Jan. 23, 2008, Waxman Letter to FDA. The American Association of Trial Lawyers asked FDA to withdraw it.
The New Rule Reaffirms FDA’s Evidence-Based Labeling Requirements
Waxman’s concerns are misplaced. The final rule does not, in fact, change anything about FDA’s implementation of the CBE rules. Nor does it modify a sponsor’s underlying labeling obligations. It simply codifies FDA’s “existing labeling standards and policies,” policies that have been in effect since at least 1982, when FDA noted that the CBE procedure was appropriate only for “newly discovered risks” and “new information about the safe use of a drug product.” 47 Fed. Reg. 46622, 46623, 46635 (FDA Oct. 19, 1982). The underlying labeling obligations remain, specifically, that warnings and precautions are limited to “clinically significant adverse reactions . . . [and] other potential safety hazards” for which there is “reasonable evidence of a causal association with a drug; a causal relationship need not have been definitively established.” 21 C.F.R. 201.57(c)(4)(ii)(6).
Requiring drug labels to be based on scientific evidence is a reasonable regulatory objective. The alternative—allowing sponsors to include speculative, unreliable, erroneous, or misleading safety information in labels—defeats the purpose of a label, which is to ensure that doctors and patients are adequately informed about a medicine’s potential risks.
Speculative Labeling Is Dangerous
Speculative labeling is not just unscientific, it can be dangerous. Speculative and unsubstantiated safety information can prevent patients from receiving necessary medical care. Labeling that is overly alarming or misleading might lead prescribers and patients to avoid a medicine that would have been beneficial and possibly less risky than alternatives. In 1979, FDA said that including theoretical concerns in drug warnings “would cause that very important section of the labeling to lose its significance.” 44 Fed. Reg. 37434, 37453 (June 29, 1979). In its 2006 labeling guidance, FDA explained that “[e]xhaustive lists of every reported adverse event, including those that are infrequent and minor, commonly observed in the absence of drug therapy or not plausibly related to the drug therapy should be avoided . . . Such lists are not informative and tend to obscure the more clinically meaningful information.”
As a result, prudent and responsible sponsors typically consult with FDA before adding risk information, regardless of whether the information qualifies for a CBE submission. This is because even with a CBE submission, FDA may later reject the new label language and could deem it false or misleading, subjecting the sponsor to enforcement action. Sykes v. Glaxo-SmithKline, 2007 WL 957337, at *18 (E.D. Pa. Mar. 28, 2007) (“manufacturers typically consult with FDA before [changing labeling] to avoid . . . changes with which the agency ultimately might disagree”).
If It’s Not New, Why Did We Need It?
So why did FDA issue the new rule? To fill a gap in the language of the regulations that had been repeatedly misused by plaintiffs’ lawyers in lawsuits claiming that pharmaceutical companies failed to warn adequately of a medicine’s risks. Plaintiffs’ lawyers argue that sponsors can and should update their labels whenever any data – old or new—shows a numeric imbalance in any adverse event over either placebo or active comparators. Plaintiffs’ lawyers argue that sponsors must do so even if FDA has already reviewed the information and declined to impose labeling. Plaintiffs’ lawyers argue that sponsors should do so even if there is no biologically plausible explanation for the events, even if the events are more likely to reflect the underlying condition for which a patient is being treated, and even if the rate of events is indistinguishable from the rate in the general population.
FDA’s final rule explains that none of these arguments is correct. Numeric imbalances in adverse events are not sufficient to warrant a warning, precaution, or contraindication—whether preapproved by FDA or not—because standing alone, a numeric imbalance is not reasonable evidence of an association between a medicine and an adverse event. A numeric imbalance is reasonable evidence of a numeric imbalance—no more, no less.
FDA’s final rule also gives us a regulatory definition of “newly acquired information.” It is information that has either never been submitted to FDA or that is of a “different type or greater severity or frequency than previously included in submissions to FDA.” It includes “data derived from new clinical studies, reports of adverse events, and new analyses of previously submitted data,” but it also includes new analyses of previously submitted information, including meta-analyses or new interpretations of existing clinical trial data. In its responses to public comments, FDA said clearly that if a sponsor
“submits adverse event data information to FDA, and then later conducts a new analysis of data showing risks of a different type or of a greater severity or frequency than did reports previously submitted to FDA, the sponsor meets the requirements for ‘newly acquired information.’” Because the new rule explicitly covers new analyses of old data, it arguably broadens the scope of data subject to the CBE rules, not limits it as Waxman fears.
Even before FDA issued the final rule, courts around the country had interpreted the CBE rules to apply to new information only. See, e.g., Weiss v. Fujisawa Pharm. Co., 464 F. Supp. 2d 666, 675 (E.D. Ky. 2006) (under §314.07 “a drug manufacturer may warn patients and healthcare providers should they discover new evidence of a particular risk following the approval of the original label); Perry v. Novartis Pharm. Corp., 456 F. Supp. 2d 678, 685-86 (E.D. Pa. 2006) (“This particular regulation was promulgated precisely to allow drug-makers to quickly strengthen label warnings when evidence of new side effects are discovered”); Zikis v. Pfizer Inc., 2005 WL 3019409, at *3 (N.D.Ill. Nov. 8, 2005) (§314.70 “allows for an amendment to a label without extended delay when a drug manufacturer learns of new dangerous side effects”).
Finally, the rule reiterates that drug labeling is the province of FDA and that its labeling rules pre-empt contrary state authority. In a bipartisan letter on pre-emption, five former FDA chief counsels explained that “[i]f every state judge and jury could fashion their own labeling requirements for drugs and medical devices, there would be regulatory chaos for these two industries that are so vital to public health . . . .” This closely watched issue is currently before the Supreme Court in Wyeth v. Levine, a pre-emption case in which plaintiffs argue that Wyeth should have warned about certain potential side effects even without FDA approval.