Special designations and the US Food and Drug Administration’s “dual mandate”

Drug regulation has, in the past 50 years, undergone a sea change. With an eye toward faster access to new drugs so that more patients can benefit from them, Congress has mandated (and the US Food and Drug Administration [FDA] implemented) initiatives like the orphan, breakthrough, and accelerated approval programs. Whether these programs have helped society achieve important outcomes—reductions in development time and/or incremental public health benefit—without eroding clinical practice’s evidence standards is a matter of vigorous debate. This debate is particularly fierce in oncology, which has become the most active area of drug development in part because of these programs’ incentive structures. Nowhere is this debate more germane than the United States, because of its outsized influence on the cancer drug market and pharmaceutical pricing that has been largely unchecked by payers, resulting in cancer-related spending expected to exceed $250 billion per year by 2030 (1).

With such large sums at stake, debate about the logic and effectiveness of special designation programs is oftentimes emotional. However, in this issue of the Journal, Michaeli and colleagues (2) add objectivity and dispassionate descriptive analysis into the fray. Their work is a comprehensive description of oncology drugs receiving approval by the FDA between 2000 and 2022 and the special designations granted to each along the way. They assess the associations between special designations and a host of societally relevant outcomes, including development time, clinical and biological innovativeness, pivotal clinical trial’s design and primary endpoint, and cost-effectiveness. The descriptive analysis, one of the largest of its kind, contains several interesting observations. First, four-fifths of approved oncology drugs received priority review designation prior to their approval, raising questions ranging from the practical (eg, How “prioritized” can review be if 80% of drugs receive it?) to the epistemological (eg, Does applying an abbreviated review process to so many drugs erode confidence in those that achieve full regulatory approval?). Second, to their credit, Michaeli and colleagues (2) attempt to assess the incremental clinical benefits (as determined by pivotal clinical trial point estimates), population burden of disease, and the prices of approved drugs (based on Medicare average sales price), finding that drugs with more designations tended to treat fewer patients but confer greater efficacy at somewhat similar costs as compared with drugs with fewer designations. Ironically, these same drugs were less likely to have been tested in proper randomized, controlled trials, casting doubt on the true nature of the benefit. Finally, the authors raise the issue of eroding evidence standards in oncology, citing the nearly two-thirds of approvals based on nonrandomized data and the relative absence of active comparators from clinical trials involving drugs with multiple special designations.

The work is not without its shortcomings, as the authors note (2). Determinations of innovativeness are inherently subjective and, although they may correlate with special designations, they may not correlate with public health outcomes. Because their analysis focuses only on those drugs that succeed in achieving full regulatory approval, Michaeli and colleagues (2) lack information about not only those drugs that failed to achieve full regulatory approval but also those clinical trials involving approved drugs that were not submitted to the FDA. Given that designations are likely correlated with one another, whether it is truly the number of special designations or their combinations that explain differences in societal outcomes remains unknown. Finally, and perhaps most critically for the public health, to what extent are drugs that receive special designations achieving the goals of the special designations and meeting the needs of patients?

Whether a given FDA designation serves the public health is a thorny question for multiple reasons. First, a policy’s success should be defined a priori. Second, the approach should eschew by-drug analysis. Third, care must be taken because the introduction of the policy itself might be a poor instrumental variable. To wit, accelerated approval may make sense as a policy if certain outcomes were both met and attributable to the policy itself. It may be the case that 9 of 10 drugs receiving accelerated approval fail to demonstrate an overall survival or quality-adjusted life-year benefit against a good standard of care control in a confirmatory phase III trial. Yet, accelerated approval may be a dominant strategy if the 1 drug that does confer overall survival benefit saves enough lives or quality-adjusted life-years over the standard of care during the time period in which it had accelerated but not full regulatory approval, through a large survival benefit in a small population or even a marginal survival benefit in a large population. An alternative framework for evaluation of the policy emerges: construct counterfactuals from imputed estimates of what the standard of care would have been expected to achieve in the population that received the novel drug during the accelerated approval period and compare that expectation with what was achieved by the policy. Such studies are needed.

The notion of meeting unmet medical needs raises thorny questions for drug developers and regulators. Do we genuinely meet an unmet medical need when we provide a 2-month overall survival benefit? What about when we provide an active drug that shrinks tumors but does not improve overall survival? Although meeting an unmet medical need does not appear to have an objective standard from a regulatory perspective, many of us would look at drugs approved by these pathways and dispute that they meet an unmet need. For example, avelumab in Merkel cell carcinoma is cited as clinically innovative, as it was approved by the accelerated approval pathway in 2017 (3). At the time that it was approved in this indication, though, another drug with identical mechanisms of action demonstrated activity in Merkel cell carcinoma 1 year prior to avelumab’s accelerated approval (4). The unfortunate truth is that there is a strong business case for seeking approval on the basis of unmet medical need. Accelerated approval can be cited as supporting evidence for anticompetitive method-of-use patents that lead to greater cost, despite the substantial differences between the FDA’s “unmet need” and the US Patent and Trademark Office’s “long-felt need” standards (5).

As Michaeli and colleagues (2) correctly state, the FDA will not be a comparative effectiveness agency for the time being. Instead, in the United States, this will likely fall to government payers. Facing long-term fiscal challenges, government payers may be forced to become the arbiters of reasonableness by raising evidence standards for their coverage determinations (eg, Centers for Medicare and Medicaid Services and the aducanumab saga) or through negotiating the price of drugs (eg, the Inflation Reduction Act allowing Medicare negotiation on drug prices). Given the political economy of drug development and the political pressure it faces to utilize the special designations, the question then becomes what a well-intentioned FDA can do. First, it can approach regulatory decisions through a risk management lens by asking what magnitude of harm would be expected were a mistake made. Two corollaries to this are that evidence standards should scale with the size of an indication, because getting the decision wrong would have substantial clinical (and financial) downsides, and safety through dosage optimization should be pursued, so that if a drug is not in truth efficacious then at least its risk of toxicity is reduced (6). Second, for special designations like accelerated approval to be all things to all people, postmarketing requirements that seek to protect future potential recipients need to be rigorously enforced and drugs that fail to meet or complete postmarketing requirements should promptly be removed from the market (7). Third is to make clear the need for well-designed and rigorously conducted studies by asking scientifically challenging questions of the Oncologic Drugs Advisory Committee. Regulators may be able to create better policy by asking stakeholders for strictly scientific judgments rather than preferred policies. For example, the FDA recently convened an Oncologic Drugs Advisory Committee meeting to determine whether a registrational phase III trial that had already been published in a high-impact journal was, in fact, an interpretable study. Most members caveated their final vote with a comment on the question’s narrow framing (8). Whereas the scientific question can be answered only in scientific terms, an approval question can be interpreted vis-á-vis patients, such that a vote against approval is clouded by the perception that doing so would take away from individuals a potentially efficacious option—something oncologists would be loath to do.

Ultimately, the FDA is a public health authority tasked with using science to ensure that the drugs entering the US market are safe and effective. The presence of special designations introduces to the FDA something resembling the “dual mandate” of the Federal Reserve. Whereas the Federal Reserve’s mandates are full employment and price stability, the FDA’s mandates increasingly appear to be to ensure the safety and effectiveness of drugs entering the US market by rigorously applying the scientific method and enhancing access to those drugs by strategically cutting the scientific method short. How to balance the 2 mandates depends on whether it is in the public’s interest to trade evidentiary certainty for future potential recipients of the drug for more expedient access to a drug (with less certain benefit) for the patients who would receive the drug sooner. Stuck between a rock and a hard place, it can be useful to go back to first principles. In the case of the FDA, its establishment was animated by the need for scientific rigor, so if forced to preference one mandate over another, scientific rigor should rule the day.

Data availability

No new data were generated or analyzed for this editorial.

Author contributions

Garth W. Strohbehn, MD, MPhil (Conceptualization; Writing—original draft; Writing—review & editing) and Mark J. Ratain, MD (Conceptualization; Writing—review & editing).

Funding

No funding was received in support of this editorial.

Conflicts of interest

GWS is an employee of the US Department of Veterans Affairs; the views expressed do not reflect those of the US federal government. GWS has received consulting fees from VIVIO Health and EBSCO Information Systems and is a co-inventor of a filed method-of-use patent in dose optimization. MJR has received funding from Arnold Ventures. The authors are co-inventors of a filed method-of-use patent (held by the University of Chicago) in dose optimization of tocilizumab for viral infections and directors of the Optimal Cancer Care Alliance, an Ann Arbor, MI, USA-based 501(c)3 nonprofit organization.