Who Is Now Responsible for Discovering and Warning About Adverse Effects of Generic Drugs?
Aaron S. Kesselheim, MD, JD, MPH; Michael D. Green, JD; Jerry Avorn, MD
When a patient in the United States fills a prescription, the likelihood that a generic drug will be dispensed is overwhelming—and increasing. Generic drugs account for more than 80% of prescriptions, but only 20% of total drug spending, because of their lower cost. The move to generics is estimated to have saved more than a trillion dollars in the last decade.1
Although generic drugs are bioequivalent to their brand-name counterparts, important legal distinctions between the 2 categories have resulted from 3 Supreme Court decisions. The first, Wyeth v Levine, arose after a woman developed gangrene in her forearm caused by intra-arterial injection of Wyeth’s antiemetic Phenergan (promethazine); the patient was awarded $6.7 million in damages based on her claim that the drug’s label did not sufficiently warn about its risks. In 2009, the Court upheld the award, noting that brand-name manufacturers are primarily responsible for providing accurate safety warnings for the drugs they produce.2 The decision makes sense because premarket testing does not reveal the full range of a drug’s adverse effects, postmarket surveillance by the US Food and Drug Administration (FDA) is insufficient, and state tort liability incentivizes manufacturers to continue monitoring the safety of their products.
A second decision, PLIVA v Mensing (2011), was based on similar facts. A woman developed tardive dyskinesia after receiving PLIVA’s generic version of the gastrointestinal motility agent metoclopromide (Reglan). The patient also claimed that the manufacturer provided inadequate warning about this potential adverse effect. But the Supreme Court ruled against her, because she received a generic product. Justice Thomas held that the Food, Drug, and Cosmetic Act requires generic manufacturers to provide the same warnings provided with the brand-name version, so generic producers could not change their labels to reflect new safety knowledge.3 This superseding federal legal requirement preempted the patient’s lawsuit.
On June 24, 2013, the Supreme Court ruled on a third case, Mutual Pharmaceutical Co v Bartlett,4 in which a New Hampshire woman was prescribed the brand-name nonsteroidal anti-inflammatory drug Clinoril for shoulder pain. The patient’s pharmacy filled the prescription with sulindac, a generic version of the drug manufactured by Mutual Pharmaceutical Company. The patient developed Stevens-Johnson syndrome, a rare hypersensitivity reaction, which progressed to toxic epidermal necrolysis, resulting in loss of nearly two-thirds of the outer layer of her skin and permanent near-blindness.
In suing the drug’s manufacturer, the patient’s lawyers relied on a slightly different argument than was used in PLIVA v Mensing: they argued that sulindac had a “design defect” that made it unreasonably dangerous. Such claims usually relate to durable goods designed with excessive dangers, as would occur if an automobile manufacturer produced a car without airbags or with a weak frame. But legal scholars have long known that drugs are different,5 because the same molecular components that produce benefits also lead to adverse effects. Therefore, drugs generally cannot be redesigned in a safer way.
Some states, including New Hampshire, nonetheless permit consideration of design defect claims for drugs. After trial, a jury agreed that sulindac’s risks outweighed its benefits, awarding Bartlett more than $21 million. The First Circuit upheld that verdict, reasoning that although a generic manufacturer cannot change its active ingredient or warning label without FDA approval, it can cease selling a drug such as sulindac. But writing for a 5-4 Supreme Court majority, Justice Alito overturned the lower court’s decision.4 He agreed that the law prevented the generic manufacturer from adding a warning not included in the label for brand-name Clinoril. But he did not accept the stop-selling logic, reasoning that an “actor seeking to satisfy both his federal- and state-law obligations” should not be “required to cease acting altogether in order to avoid liability.”4
Because PLIVA v Mensing prevented Bartlett from making a failure-to-warn claim, she had no recourse for her injuries. Justice Alito noted that it might be unfair for patients to have different rights based on the fortuity of whether their pharmacist substituted a generic version of a brand-name drug, but he considered this to be an inevitable result of the FDA decision to approve sulindac initially despite its known risks and Congress’ decision “to regulate the manufacture and sale of generic drugs in a way that reduces their cost to patients but leaves generic drug manufacturers incapable of modifying either the drugs’ compositions or their warnings.”4
Manufacturers of brand-name and generic drugs now face different legal responsibilities for warning about the risks of what are the same drugs. Brand-name manufacturers must closely monitor the safety of their products after approval and update the adverse effects sections of their labels as necessary or else be subject to pay substantial damages based on liability to injured patients. By contrast, generic manufacturers do not face such liability. Thus, once a brand-name company’s exclusivity ends and its market share declines, or it stops production altogether, active pharmacovigilance is likely to end, reducing the chance of discovering rare or delayed adverse effects later in a drug’s market life that could vitally inform prescribing practice or change a product’s risk-benefit assessment.6
In the week following the Bartlett decision, the FDA announced its intention to propose a rule that would allow generic manufacturers to add new safety information to labels beyond the brand-name products’ warnings.7 But if this rule is implemented, who would undertake the studies to detect unanticipated risks of drugs that no longer have patent protection? Individual manufacturers of generics? The FDA? An outside group? What would be the source of funding to perform such ongoing postmarket surveillance activities? If generic drug labels were untethered from those of branded products, would information about adverse effects have to be uniform for all generic versions of the same medication, as the outline for the FDA’s proposed rule seems to require?
These questions bear directly on the rapidly evolving state of pharmacoepidemiology in the United States. Acting on a 2007 mandate from Congress in the wake of the Vioxx debacle, the FDA built a nationwide postmarket safety surveillance program that now has the capacity to systematically monitor adverse events in more than 100 million US residents. The FDA continues to refer to the system, initially named the Sentinel System, as the Mini-Sentinel Pilot project, and after years of development it has still not generated a large number of appropriately adjusted analyses of drug risks. With no specifically appropriated funds to continue this work, and lack of clarity over who will be responsible for its front-line implementation after current contracts expire in the coming year, this program may not have adequate capacity to address the additional need for the pharmacovigilance of generic products that the Supreme Court cases have highlighted and that the FDA’s intended rulemaking would make possible.
As generic drugs come to make up an ever-larger share of medication use, the likelihood will increase that knowledge about a drug’s risks will remain incomplete or poorly characterized at the time a generic version is introduced. This trio of Supreme Court decisions, followed by the FDA’s recent stated rulemaking intentions, sets the stage for creation of an independent organization with pharmacoepidemiologic expertise to assess the adverse effects of generic drugs in concert with the agency. Whatever the future of Sentinel, this research effort could leverage the rapid proliferation of public and private data sets that make enormous populations available for such systematic adverse-effect surveillance. The results of these analyses could be provided to the FDA, which would determine whether the common generic drug label needed to be adjusted or whether accumulating data about the relative risks of potentially fatal adverse events for drugs like sulindac meant that it remained as safe as others in its class, required additional labeling, or should be taken off the market.
Funding for such work could come from the newly established generic drug user fees8 or from additional small levies on each dispensed generic drug. The latter funding stream could also be used to compensate patients injured by unwarned safety problems caused by generic drugs; a similar approach exists for vaccine-related damages and appears to be working reasonably well.9 This would be a more sensible solution than leaving such determinations to local juries in design defect cases or putting the duty on a single small generic manufacturer.
The Supreme Court’s recent decisions have created a paradox of unequal recourse for patients harmed by serious adverse events caused by branded vs generic medications, even though these drugs are bioequivalent and freely substituted under state law. A rule along the lines that the FDA has signaled offers a potential solution. Realizing that promise will require a disciplined and adequately supported agenda of safety surveillance for generic products, probably through a new partnership between government and academia. Besides allowing for greater equity for patients, this approach could also go a long way toward guiding proper use of the increasingly expanding generic armamentarium so as to increase its benefits and contain its risks more effectively.
