Addressing High Failure Rates in Drug Development

FDA has published an excellent report on 22 instances in which phase 2 and phase 3 trials had divergent results. Appendix A of the report provides justifications for essential elements of modern clinical trials, including a control group, randomization and blinding. In addition, Appendix A highlights risks of techniques for improving efficiency in early development, including adaptive designs, enrichment designs and use of historical controls. The report seems to weigh against any relaxation of current development methodology. The FDA report on divergent phase 2 and phase 3 results is a timely reminder that early approval of novel drugs will not always benefit patients and other stakeholders.

On the other hand, the biopharmaceutical industry, CROs like Health Decisions and patient advocacy groups, especially in indications with significant unmet medical needs, all tend to believe that current development methodology is sometimes more restrictive than necessary. Many in the industry favor a shift away from traditional phase 3 testing toward allowing earlier market access for novel drugs, with increased postmarket surveillance to reduce the risk of dosing the public when knowledge about safety and efficacy is limited. A Health Decisions white paper, Phase 2c: Expedited Development for All, advances one of many proposals to reduce the burden of premarket testing while adding postmarketing safeguards.

The Leading Cause of High Failure Rates in Drug Development

Amidst the debate about trial methodology and development efficiency, Derek Lowe at In the Pipeline provides an eloquent cautionary statement:

It’s hard, sometimes, for people who’ve worked in other industries to appreciate this. Drug development is a unique combination of very high regulatory burden and very high failure rates, so it’s tempting to say that the regulations cause the failures. But that isn’t true. Biology causes the high failure rates – specifically, our lack of understanding of biology…

This stark reminder that there are bigger challenges in drug development than regulations and methodology is timely as we consider proposals for providing earlier public access to investigational drugs. To be sure, it is important to mitigate the huge delays in rewarding industry investment in drug development as this reduces investment and the pace of innovation in therapeutics for all health conditions. Furthermore, denying patients timely access to safe and effective new medicines can cause substantial harm, especially to patients with significant unmet medical needs. In the extreme case, patients could die while awaiting approval of improved therapeutics. On the other hand, there is risk of great potential harm in allowing large patient populations timely access to new medicines when evidence of safety and efficacy is limited. Nobody wants a news cycle dominated by stories about serious adverse events associated with a new drug. With the scenario of early distribution of a novel drug to a very large population, a hypothetical nightmare case might involve the premature deaths of numerous patients undergoing treatment for a non-life-threatening health condition.

Blockbuster Drug, Blockbuster Ethical Dilemma

Even the largest phase 3 trials may not be large enough to detect serious safety issues that affect a very small percentage of a very large patient population. (The history of novel drugs for type 2 diabetes, with global prevalence of 422 million cases in 2014, is illustrative.) While today’s oral contraceptives are remarkably safe, this is another class of drug with a very large patient population. In 2012, 9.72 million women in the United States were taking oral contraceptives and globally the number of women using oral contraceptives was approximately 104 million. Rapid adoption of any truly novel oral contraceptive might involve millions of women. With such a vast patient population, even a minuscule rate of any specific type of adverse event could affect thousands of women worldwide. Postmarket surveillance seems the only practical approach to early detection of relatively rare safety issues in very large populations. However, three questions remain unresolved about how to address such cases:

  1. In collecting data on safety and efficacy, what is the optimal point for the transition from premarket testing to postmarket surveillance?
  2. What level of postmarket surveillance will be adequate to detect safety issues?
  3. Is there a reasonable way to impose step-wise limits on early distribution to prevent patient exposure from far outpacing the accumulation of knowledge about potential safety issues?

The Baseline for Evaluating Policy Proposals

It would be foolish to underestimate the challenge facing the industry, academics, patient advocates and regulatory agencies in defining a policy that optimizes public benefit from novel drugs when knowledge about them is on balance favorable but also limited. Like drug development itself, defining such a policy is a fundamentally hard problem with high risk of failure and potentially substantial consequences for human health.

FDA currently provides access to four expedited programs that can accelerate market access: fast track designation, breakthrough therapy designation, accelerated approval and priority review. The FDA determines access to these programs on a case-by-case basis. When the fundamental problem is our lack of understanding of biology, any proposal for an across-the-board reduction in premarket testing of novel drugs requires careful consideration. In the interest of public safety, the burden of proof on proposals to accelerate market access for novel drugs rests on any proposal that would allow early access without a case-by-case review.