The FDA’s system for regulating prescription drug use after initial marketing approval is ineffectively protecting patient safety, depriving patients of best available information on the full therapeutic potential of approved drugs and draining billions of dollars from the pharmaceutical industry in penalties and False Claims Act payments for providing “unapproved” scientific information about their drugs to physicians who are free to prescribe those drugs for any purpose they think appropriate. That system, like much of historic government regulatory regimes, is built on faulty scientific premises and desperately needs reform to reflect the reality of how medical knowledge evolves and to take advantage of internet information gathering resources unimaginable at the time the system was conceived. Such reform is both possible and practical.
In 1962, Congress erected the modern system of drug management on the dual foundation of safety and efficacy. Before a new drug may be made available for sale to the public, its sponsor must persuade the FDA by in-vitro and clinical testing that the drug’s therapeutic benefit in treating specified diseases or conditions outweighs the patient safety risks it may create. Those risks, while identifiable in the aggregate, generally cannot be traced to an ascertainable human characteristic and thus must be treated as falling randomly within the entire treated population. The FDA must also approve a drug label, explain the characteristics of the drug, its best method of use and the treatments for which it has been approved. The approved label must also identify patients who should not be given the drug and warn physicians of safety risks that other patients may encounter. The approval process is extended and expensive and approval review, largely funded by user fees, is the principal focus of FDA’s drug regulation resources.
Once a drug has been approved for sale, FDA’s regulatory involvement in assessing its safety and efficacy greatly diminishes. FDA regulations require a drug’s sponsor to collect and review adverse event reports generated by prescribers and patients and reported to the sponsor, as well as relevant literature, and to modify its label warnings to reflect a newly discovered reasonable association between the drug and a serious health hazard. FDA relies on private tort litigation by injured patients to enforce those duties. But FDA itself does not actively monitor accumulating safety information. Despite incipient efforts to reach out to hospitals and other operational databases through the Sentinel program, FDA has no comprehensive or disciplined information arising from the ongoing use of an approved drug. Moreover, FDA lacks the resources to review the information it does collect programmatically and focuses its resources on a drug safety issue only when stimulated by external pressure.
The tacit premise of FDA’s post-approval safety monitoring and of the 1962 Act is that the comprehensive review required for approval largely suffices for ongoing drug management. Experience has demonstrated, however, that use of a drug by larger numbers of patients for longer periods reveals characteristics not discerned in pre-approval testing. Thus, expediting and enhancing the accumulation and analysis of post-approval safety-related experience data should be an FDA priority. Identifying safety issues and preventing injury is far better than depending on after-the-fact tort litigation and case specific evaluation of alleged drug hazards to protect the public health.
FDA’s post-approval regime is even weaker on the efficacy side. Clinical experimentation with approved drugs, unlimited by legal restraints, can involve variations in dosing, use in populations not covered by FDA-approved labeling (pediatric), or treatment of diseases or conditions not included in FDA-approved labeling (“off-label” use). As the results of these uses are disseminated by word of mouth, journal articles, CME seminars and medical texts, “off-label” use expands, not infrequently becoming best medical practice and being recognized for coverage by government programs and insurance formularics. Nevertheless, the product’s sponsor is severely restricted in providing “off-label” use information, despite being best positioned to assess the value of “off-label” use taking into account clinical trials data unavailable to others and feedback from sales calls on physicians. FDA considers off-label promotion to be a law violation and the government holds sponsors accountable for False Claims Act treble damages if “off-label” promotion “induces” physicians to prescribe off-label to patients whose prescriptions are covered by federally-funded programs. The line between permissibly providing scientific information and engaging in forbidden “off-label” promotion is fuzzy at best and FDA has refused to remedy that lack of clarity by regulation or definitive guidance. The result is a clearly sub-optimal information flow about the full potential of approved drugs and a foreclosure of health benefit to those who could be treated by those drugs.
FDA regulations permit a sponsor to expand approved uses by filing a supplemental New Drug Application (“sNDA"). As the name implies, sNDAs require the same kind of expensive and time consuming clinical trials as initial approval creating a disincentive to bringing uses that affect only small populations on-label. Even when an sNDA is contemplated, the statutory incentive for obtaining sNDA approval for a new use is a three year exclusivity period where no other manufacturer can claim that use. Where that three year period overlaps the initial exclusivity period arising from approval of the underlying drug or a patent-based exclusivity period, it has limited economic benefit unless active promotion of the new use leads to sales sufficiently exceeding “off-label” prospects to warrant investing in the sNDA. And where a drug is subject to generic competition for its original approved use, the three year sNDA exclusively restricts only generic labeling and does not preclude generic distribution or preclude generic substitution under state pharmacy law making the economic benefit even more questionable. Not surprisingly, the number of sNDAs approved for significant new use is small before generic entry and effectively zero thereafter. Patients are thus denied both important information from drug sponsors about unapproved but potentially valuable uses and the considered judgment of FDA about the efficacy of those uses.
Extracting patients from this current lose-lose post-approval quagmire should be an industry and FDA priority. The way forward to better collecting and disseminating both safety and efficacy information post-approval lies in effective use of internet information policy as, for example, demonstrated by Wikipedia. Wikipedia, in fact, already includes substantial volumes of drug information voluntarily submitted and edited although not comprehensively collected or reliably evaluated. But a more rigorous web-based approach along the following lines is clearly achievable and well worth considering.
FDA could require each approved manufacturer of a drug, pioneer or generic, to pay a proportionate share of establishing, maintaining and monitoring an official drug-specific public website. Manufacturers would be required to submit, but not limited to submitting, all information now reported to FDA including adverse event reports and relevant scientific literature. Manufacturers also could submit any additional data or analyses relating both to approved and “off-label” uses they deemed relevant and reliable.
Site monitors would be nominated by participating manufacturers subject to FDA confirmation of their independence and scientific qualification. Once confirmed, a site monitor would be responsible for: (1) developing a format(s) for submissions by manufacturers, physicians and others; (2) reviewing submitted information for completeness and compliance with formatting instructions; (3) organizing submissions to permit physicians to access them by relevant categories (e.g., adverse event reports, clinical efficacy observations, publications); and (4) periodically posting an independent evaluation of the significance of the submitted data for safety and efficacy purposes. The monitor’s periodic evaluations also would be submitted to FDA together with recommendations, if appropriate, for labeling changes. In addition, the monitor would be responsible for providing FDA prompt notice of serious safety hazards.
Active independent and continuous monitoring of clinical experience and academic literature by qualified administrators would permit FDA to have the full benefit of that experience and analysis without increasing its work force or its budgetary expenditures. FDA would no longer have to vest principal responsibility for updating approved labeling in manufacturers whose incentives may not be perceived to align with those of FDA or rely on the outcomes of less than rigorous tort actions. With the assistance of scientifically-qualified monitors, FDA would have the ability and responsibility to modify labeled safety information on a timely basis and binding all authorized manufacturers to conform. FDA also would be in a position to provide authoritative guidance to physicians about “off-label” uses that could benefit patients willing to take a drug prior to formal FDA approval of its use for their disease or condition.
As required by the Federal Food Drug and Cosmetics Act, new uses including those recommended by site administrators, would be incorporated into FDA-approved labeling only upon approval of an sNDA. In the interim, FDA could use monitor evaluations and the data underlying them to determine what further clinical testing, if any, should be required for sNDA purposes easing the financial burden of sNDA preparation. In addition, FDA could permit certain monitor-recommended use modifications which FDA determines to have adequate scientific support to be incorporated into approved labeling in a “qualified” use category. “Qualified” use designation would inform doctors of potential benefit to their patients and specific risks arising from qualified uses while making clear that FDA’s full approval requirements remain to be satisfied. Any promotion of “qualified” uses would be required to include disclosure that final FDA approval for the use had not yet been obtained.
FDA claims broad regulatory authority to condition new drug approvals on continuing monitoring and disclosure requirements and could invoke that same authority to require both pioneer and generic drug manufacturers to maintain and monitor drug specific websites as part of the approval process. FDA regulations also establish procedures and requirements for amending approved drug labels and could be equally adapted to the proposed monitoring regime. FDA already has statutory authority to add new safety information to drug labels at its own initiative and broad discretion to establish the content of drug labels which could extend to the “qualified” use concept. With welfare in mind, there should be no legal impediment to promptly implementing an enhanced “patient first” post-approval safety and efficacy monitoring program.
It is time to acknowledge that FDA “off-label” enforcement actions, false claims penalties and vague in terrorem guidance to manufacturers cannot and should not suppress “off-label” use and that a continuing dynamic approach to drug evaluation would optimize patient outcomes and help restore the FDA approved label as the definitive source of up-to-date drug information. The information/communications revolution has created the possibility of taking full advantage of clinical experience and scientific expertise in post-approval drug management and the public health demands that FDA promptly take full advantage of it.
About the author
Bert W. Rein is a founding partner of Wiley Rein LLP.