Challenges in clinical drug development

Some challenges facing drug development

How do we avoid rofecoxib-type cases?

Avoiding post-marketing withdrawal

  • Rofecoxib demonstrates that some adverse effects are difficult to identify in clinical drug development, but it is not clear we have better alternatives

  • We want safe and effective medicines in a timely and cost-efficient manner

    • Tightening regulatory requirements costs more, takes longer, and may not work

    • Can we tolerate the risks of not tightening regulation?

One success story in this area has been the requirement to demonstrate cardiovascular safety in newly developed diabetes drugs. This guidance was put in place in 2008 and in the intervening period additional data regarding cardiovascular safety is available for several drugs.

Key aspects of the guidance US Food and Drug Administration (2018), 8.

  • Establishment of an independent cardiovascular endpoints committee to adjudicate cardiovascular events
  • Phase 2 and Phase 3 trials should include patients at a higher risk for cardiovascular events in order to obtain sufficient endpoints to allow a meaningful estimate of risk
  • Prior to marketing, the upper bound of the 2-sided 95% confidence interval should be less than 1.8 with a reassuring point estimate
  • A postmarketing trial may be necessary to show that the upper bound of the 2-sided confidence interval is less than 1.3

In the eight drugs considered between the 2008 guidance and the 2018 report, none lead to increased risk of major adjudicated cardiovascular events and a number of studies suggested reduced risk for specific drugs.

This is especially clear for SGLT-2 inhibitors, e.g. empagloglozin. The EMPA-REG OUTCOMES trial suggested a reduction in major adjudicated cardiocasvular events in participants taking empagliflozin (Zinman et al. 2015).

Efficiency of drug development

People tend to use the term “pipeline problem” in different ways, sometimes to suggest that there is a lack of new drugs (in general), sometimes to refer a lack of new drugs in a specific area, and sometimes to reduced efficiency in the drug development process.

The pipeline problem

  • Less or stable new drug and biologic applications

  • Increase in costs of development

  • Drug development not targeted to public health needs

  • Inefficiencies in evaluation of drug safety and efficacy

Two clinical areas that are often offered as examples of the pipeline problem are the development of analgesics and antibiotics (Williams 2010; Boucher et al. 2009).

Informative drug development

Learning and Confirming

  • Learning and confirming are two distinct activities within drug development (Sheiner 1997)

  • There is a focus on learning in early development and a focus on confirmation late—but learn-confirm cycle occurs throughout

  • Learning and confirming imply different goals, study designs and analysis

Learning and Confirming throughout clinical drug development

  • Phase I the focus is on pharmacokinetics, metabolism, pharmacology: learning

  • Phase IIa the focus is on proof of concept (drug works as proposed in patients with an adequate safety margin): confirming

  • Phase IIb the focus is on selecting an appropriate dose and gaining information that with aid design of Phase III studies: learning

  • Phase III the focus is on confirming clinical efficacy: confirming

There is much overlap of the phases of drug development and often both learning and confirming are occurring. There is, however, a shifting focus throughout the clinical drug development process that (roughly) follows the path described.

Sheiner (1997) made the argument that a key source for inefficiency in drug development was the insufficient use of the right kind of methodological tools in the learning phases of drug development.

This argument has been accepted and informative drug development—including the use of adaptive trial designs, Bayesian analysis and pharmacokinetic-pharmacodynamic models—is increasingly adopted as the standard approach to clinical drug development.

There are a lot of different terms associated with this focus on better using the information available during drug development. While many of these terms are distinct, there are aspects in which they overlap: informative drug development, pharmacometrics, model-based drug development, model-informed drug discovery and development, quantitative pharmacology, MODSIM, …, and a range of others.

While strict “Traditional drug development” no longer exists, it is useful to make the comparison between “traditional” approaches and and informative, or model-based, approaches to drug development

Traditional Drug Develop. Informative Drug Develop.
Statistical approach Frequentist Bayesian/Frequentist
Use of data Data from trials considered independently; trials run in parallel Sequential trials; prior evidence explicitly incorporated
Design Focus on randomized controlled trials Greater flexibility
Analytical approach Hypothesis testing Use PKPD information to model what we know about the drug; use these models to measure and predict outcomes and to optimise design

Further reading

  • Marshall et al. (2019) provides an overview of the current implementation of model-informed drug discover and development
  • Trivedi, Lee, and Meibohm (2013) discusses the application of pharmacometrics to the clinical development of anti-infectives
  • Orloff et al. (2009) discusses adaptive trial designs

References

Boucher, Helen W, George H Talbot, John S Bradley, John E Edwards, David Gilbert, Louis B Rice, Michael Scheld, Brad Spellberg, and John Bartlett. 2009. “Bad bugs, no drugs: no ESKAPE! An update from the Infectious Diseases Society of America.” Clinical Infectious Diseases : An Official Publication of the Infectious Diseases Society of America 48 (1): 1–12. https://doi.org/10.1086/595011.

Marshall, Scott, Rajanikanth Madabushi, Efthymios Manolis, Kevin Krudys, Alexander Staab, Kevin Dykstra, and Sandra A. G. Visser. 2019. “Model-Informed Drug Discovery and Development: Current Industry Good Practice and Regulatory Expectations and Future Perspectives.” CPT: Pharmacometrics and Systems Pharmacology 8 (2): 87–96. https://doi.org/10.1002/psp4.12372.

Orloff, John, Frank Douglas, Jose Pinheiro, Susan Levinson, Michael Branson, Pravin Chaturvedi, Ene Ette, et al. 2009. “The future of drug development: advancing clinical trial design.” Nature Reviews Drug Discovery 8 (12): 949–57. https://doi.org/10.1038/nrd3025.

Sheiner, L B. 1997. “Learning versus confirming in clinical drug development.” Clinical Pharmacology {&} Therapeutics 61 (3): 275–91. https://doi.org/10.1016/S0009-9236(97)90160-0.

Trivedi, Ashit, Richard E. Lee, and Bernd Meibohm. 2013. “Applications of pharmacometrics in the clinical development and pharmacotherapy of anti-infectives.” Expert Review of Clinical Pharmacology 6 (2): 159–70. https://doi.org/10.1586/ecp.13.6.

US Food and Drug Administration. 2018. “FDA Background Document Endocrinologic and Metabolic Drugs Advisory Committee Meeting.” https://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM623913.pdf.

Williams, M. 2010. “Productivity Shortfalls in Drug Discovery: Contributions from the Preclinical Sciences?” Journal of Pharmacology and Experimental Therapeutics 336 (1): 3–8. https://doi.org/10.1124/jpet.110.171751.

Zinman, Bernard, Christoph Wanner, John M. Lachin, David Fitchett, Erich Bluhmki, Stefan Hantel, Michaela Mattheus, et al. 2015. “Empagliflozin, Cardiovascular Outcomes, and Mortality in Type 2 Diabetes.” New England Journal of Medicine 373 (22): 2117–28. https://doi.org/10.1056/NEJMoa1504720.

Previous