Discussion: The role, position, and function of the FDA—The past, present, and future

The health of the nation and the world has been profoundly positively influenced by the pursuit and implementation of evidence-based approaches for the prevention and treatment of diseases. The US Food and Drug Administration (FDA), along with other regulatory authorities throughout the world, have made critically important contributions to this pursuit. The FDA recognizes their mission is to be “responsible for protecting public health by ensuring the safety, efficacy, and security of human and veterinary drugs, biological products, and medical devices; and by ensuring the safety of our nation’s food supply, cosmetics, and products that emit radiation... FDA is responsible for advancing the public health by helping to speed innovations that make medical products more effective, safer, and more affordable and by helping the public get the accurate, science-based information they need to use medical products and foods to maintain and improve their health.”

The FDA has effectively addressed their mission by providing consistently strong leadership and oversight that has been of integral importance to protecting public health. The FDA has been highly influential in the attainment of substantial evidence of efficacy and safety of approved drugs, biologics and devices by ensuring achievement of proper standards for design, conduct and analysis of registrational trials. Among these significant favorable influences have been:

∼ Encouraging stepwise clinical development;

∼ Having proper controls in registrational trials, often best achieved through use of randomization;

∼ Properly addressing multiplicity through pre-specification of primary and secondary outcome measures and primary methods of analysis;

∼ Recognizing reliability and interpretability of efficacy is enhanced by assessing effects on direct measures of how a subject “feels, functions or survives,” and recognizing the likelihood of being misled when relying on replacement endpoints such as biomarkers that have simply been shown to be correlated with direct measures of tangible benefit and often do not capture risks adequately;

∼ Recognizing the value of blinding when outcome measures are subjective;

∼ Guiding proper designs of non-inferiority trials;

∼ Ensuring clinical trials are designed and conducted in a manner to assure the safety and ethical protection of study participants;

∼ Enhancing the quality of trial conduct by encouraging timely enrollment, best real-world achievable adherence to experimental and control regimens, pro-active efforts to maximize retention, and maintaining confidentiality of efficacy data in ongoing trials;

∼ Recognizing that exploratory analyses usually should be viewed as generating hypotheses; and

∼ Recognizing the need for confirmatory trials, especially in settings where positive predictive probabilities for efficacy are not strong (such as when obtaining marginally positive results in a setting where the pre-trial likelihood of meaningful benefit was rather low), or when post hoc analyses would properly be viewed as being hypothesis generating, or when important safety signals have been identified.

The FDA’s role is of integral importance in addressing conflicts of interest in medical research. There is a pervasive interest in obtaining positive results that would enable increased options for caregivers to offer their patients, and would provide enhanced reputations for researchers and financial benefits to sponsors. While these are valid interests, they can conflict with the greater interest in ensuring the integrity and reliability of clinical research. There is an inherent high level of multiplicity in clinical research, arising from multiple outcome measures, methods of analysis, subgroups of patients and analyses over time. Without proper oversight including that consistently provided by FDA, such multiplicity provides strong risks for misleading conclusions, especially when data are explored by those looking for positive results (Fleming, 2010). This was discussed in a presentation in May 2016 by commentator John Oliver in “Last Week Tonight,” (John Oliver’s commentary, 2016) who recognized such risk for bias “because of scientists feeling pressured to come up with eye-catching positive results ... to get those results, there are all sorts of ways that, consciously or not, you can tweak your study, you can alter how long it lasts, or make your random sample too small to be reliable, or engage in something that scientists call P-Hacking ... it basically means collecting lots of variables and then playing with your data until you find something that counts as statistically significant but is probably meaningless.” These issues are of particular concern in the pursuit of personalized medicine due to the inherently high magnitude of multiplicity in identification of subgroups of patients who benefit from new therapies, the need for accurate clinical laboratory tests to appropriately guide personalized medicine strategies, and the still evolving regulatory oversight process for laboratory tests. Negative consequences of poorly understood or weakly applied regulatory oversight processes for laboratory developed tests have been vividly demonstrated. There are numerous examples of clinical tests promoted for personalized medicine that were made clinically available prematurely (FDA “20 examples”, 2015) or that were incorporated into clinical trials where they were used to guide treatment decisions but later found to be based on flawed evidence or even fraudulent data (IOM Omics Report, 2012). Failure to insist on good clinical and laboratory practices, apply rigorous standards for the design, conduct, and analysis of biomedical research, and implement safeguards to address conflicts of interest poses threats to the integrity of biomedical research and exposes patients to potential harms.

The FDA’s ability to address its responsibility “for advancing the public health by helping to speed innovations” is enhanced by its privileged position of having broad access to results of clinical research. The efficiency of the scientific process is increased if we are enlightened by the lessons learned from past experiences, whether they be successes or failures. However, for pharmaceutical and biotech companies that collectively lead a substantial component of the clinical research effort to develop and evaluate drugs, biologics, and devices, clear competitive advantages are achieved by protecting their confidential and proprietary information and trade secrets. This provides strong motivation for these companies to reveal as little as possible about the successes and failures in their scientific process. Fortunately, regulatory authorities have broad access to sponsor’s data through their oversight responsibilities. These broad experiences enable FDA to be more enlightened when they guide clinical trial development and to make more informed judgments about whether evidence is sufficient for marketing approval and about how to proceed when it is not.

Reducing FDA’s regulatory authority, for example by substantially lowering the standards for strength of evidence required for marketing approval, and in turn increasing the reliance on post-marketing observational studies to provide enlightenment about efficacy and safety, likely would be treacherous for the public. Interventions often have meaningful positive or negative effects on the rate of key efficacy or safety outcomes. When the effects do not alter these rates by at least 5- to 10-fold, there is considerable risk that they would not be detected through evidence from the marketplace that lacks formal controls. For illustration, hundreds of thousands of patients received encainide and flecainide for arrhythmia suppression after myocardial infarction, until the randomized Cardiac Arrhythmia Suppression Trial established such use actually tripled the death rate; hormonal replacement treatment was widely used by menopausal women, until the randomized Women’s Health Initiative trial established such use increased the risks of heart attacks and strokes; the COX-2 inhibitors rofecoxib and valdecoxib were widely used for pain relief in osteoarthritis and rheumatoid arthritis patients until randomized clinical trials collectively involving fifty thousand patients revealed these agents increased the risks of MACE, (i.e. cardiovascular death, stroke, or myocardial infarction); and beta-carotene was widely used until the randomized Beta-Carotene and Retinol Efficacy Trial revealed such use induced increased incidence and risk of death from lung cancer. Numerous other examples reveal peril that would result from a regulatory strategy heavily dependent on evidence from the marketplace.

We will continue to hear arguments about the merits of weakening regulatory standards. While doing so could reduce costs of drug development prior to regulatory approval, such an approach could induce far greater costs to the public and even to drug developers.

For the public, there would be important opportunity costs, occurring when patients chase some novel intervention having evidence that is more hype than substance: these patients may be giving up other options that would have provided benefit. Furthermore, allowing widespread expenditures on interventions for which there is only scant evidence of true benefit and safety would further accelerate the rate of increase in costs of health care and, given that resources available for health care are not unlimited, would adversely impact the level of resources available for safe and effective interventions.

For drug developers, there could be increased risks from litigation due to safety events either that are real but were not discovered by the less rigorous regulatory process, or that are only alleged to be related to a product but where its sponsor does not have the scientific evidence to justify its safety.

Weakening FDA’s regulatory standards in order to reduce the burden of the drug development process could fail to achieve the intended outcome since reliable scientific evidence about the efficacy and safety of drugs, biologics and devices, usually best provided by controlled clinical trials that are properly designed, conducted and analyzed, is of critical importance not to only to the FDA, but also to other regulatory authorities throughout the world and surely to patients, payers and providers who face the complex challenges of choosing among available health care options.

It might be argued that stronger rather than weaker regulatory standards are needed. Evidence is provided by Liz Szabo’s article, appearing in USA Today on February 9, 2017 (Szabo, 2017), entitled “Treating Cancer: Hope vs Hype-Dozens of New Cancer Drugs Do Little to Improve Survival, Frustrating Patients.” Concerns are reported “that the Food and Drug Administration is approving cancer drugs without proof that they cure patients or help them live longer,” and that more cancer drugs are being approved simply based on effects on the biomarker-based endpoint, “progression free survival.” Vinay Prasad of Oregon Health and Sciences University provided evidence that the link between effects on “progression free survival” and effects on overall survival often is weak. FDA’s Richard Pazdur conveyed the proper intention, “FDA wants to give patients a chance to benefit as soon as possible,” at times relying on long term post marketing studies to provide the evidence needed to reliably establish efficacy and safety. Researcher Diana Zuckerman of the National Center for Health Research reported that many of these post marketing studies have not provided clear evidence, with Prasad recognizing it is uncommon for post marketing studies to be completed establishing survival advantages when drugs are approved without such evidence. Otis Brawley, chief medical officer at the American Cancer Society, expressed concerns that FDA is lowering its standards, and that “Studies suggest that both patients and doctors tend to overestimate drugs’ benefit, but underestimate their risks and side effects.” This access has come at significant financial cost. Referencing the Center for Health Policy and Outcomes at Memorial Sloan Kettering Cancer Center, the article reports that for a single patient “Cancer drugs approved last year cost an average of 171,000 a year.”

Is the FDA in need of a major change in the way it regulates? While the FDA has been very effective in addressing its mission, it will be important to pursue thoughtful and creative approaches to adapt to the changing landscape of medical product development and enhance the ability of the FDA to protect public health in the future. The following should be among such approaches:

Enhance the FDA’s ability to address their mission by ensuring they have necessary resources and by reducing the influence of political or special interests that conflict with their primary goals of protecting the public.

Encourage ongoing consideration of creative approaches to the scientific process and to regulatory oversight that would improve efficiency while maintaining or increasing reliability, and that would effectively address new types of challenges that may emerge. Some potential approaches include:

∼ Increase support for pragmatic trials, conducted in real-world settings with streamlined procedures for screening and enrollment, with endpoints (such as mortality in oncology trials) that are highly relevant to subjects and that can be readily assessed in a straightforward manner, with reduced burdens to subjects and researchers and with reduced risk of missing data that would compromise the ability to obtain unbiased assessments of treatment benefits and risks.

∼ Reduce the cost of clinical trials by reducing the collection of information that is not of integral importance to the benefit-to-risk assessment and, in particular, reduce line-by-line on site monitoring of data collection forms (efforts related to those of the previous FDA Commissioner, Robert Califf (Eisenstein and others, 2008)).

∼ Pursue creative approaches to achieve timely and reliable evaluation of personalized medicine interventions, recognizing the importance of identifying outcomes having sufficient sensitivity that intervention effects can be addressed in smaller sample sizes, recognizing reliability is greatly enhanced by having randomized controls and using direct “feels, functions, survives” outcome measures, and recognizing the importance of increasing the number of participating clinical sites and expanding their geographic reach and accessibility.

∼ Continue use of regulatory approaches for providing early public access, such as accelerated approval, while improving the ability to achieve timely completion of validation trials that reliability address effects on “feels, functions, or survives” outcome measures, and improving the ability to withdraw marketing approval when validation trials do not provide evidence in a timely manner that reliably establishes substantial evidence of efficacy and safety.

∼ More strongly encourage that data from completed randomized clinical trials be shared in a timely manner to enhance transparency and enlightenment, as advocated by the Institute of Medicine.

∼ Increase the collaboration between regulatory authorities worldwide; this is particularly important when addressing emerging public health emergencies.

If we continue to maintain regulations in areas other than health care, such as authorizations required to drive a motorized vehicle including having a driver’s license, abiding by laws regarding use of seat belts and helmets, and carrying insurance, then surely there is need for proper regulatory oversight of the healthcare process. Public health is enhanced by evidence based medicine, where the interests of the public are protected by the oversight of the research process provided by regulatory authorities such as FDA. The goal is not simply to ensure availability of interventions for the treatment and prevention of diseases to enable the public to have “choices,” but rather to ensure there are adequate insights about benefits and risks of these interventions to enable “informed choices.” Attainment of the best possible outcomes for patients will require careful balancing of benefits and risks for the full spectrum of medical products (Califf, 2017). A strong FDA has an integral role in achieving that goal.