Why Do Clinical Trials Fail?
November 7, 2018
Large quantities of time, money, and intellectual capital are expended to plan and execute clinical trials. Yet, despite all of these investments, clinical trials don’t always succeed. But why?
Clinical Trial Success Rates
As a first step, let’s get a sense of how successful drug developers are with clinical trials by taking a look at some estimates of clinical trial success rates. One of the most recent assessments was performed by researchers at the MIT Sloan School of Management led by Charles E. and Susan T. Harris Professor Andrew Lo. Their database of clinical trial information and drug approval data spanned over 15 years (2000 – 2015) and included over 400,000 separate data points representing over 21,000 compounds. Data sources included ClinicalTrials.gov, press releases, and analyst reports, along with a number of other primary and secondary sources. Their findings were published in Biostatistics. Based upon their analysis, the overall probability of success (POS) for industry-sponsored drugs entering Phase 1 trials to obtain FDA approval was 13.8%. This number was slightly higher than other recent analyses of similar data sets (Thomas 2016; Hay 2014). Not surprisingly, the highest POS was for drugs moving from Phase 1 to Phase 2 (66.4%), while the lowest POS was for drugs moving from Phase 2 to Phase 3 (58.3%). Interestingly, the POS to get from Phase 3 to approval was only 59.0%. This means that around 2 out of every 5 drugs that reach Phase 3 “confirmatory” trials still fail to win approval for the indication being investigated. Even when looking only at lead indications, still about 30% of drugs in Phase 3 fail to reach approval. When viewed by therapeutic area, the authors observed overall POS values that ranged from 3.4% for oncology to 33.4% for vaccines. Oncology, which is a notoriously difficult area of drug development, was the only therapeutic area with a POS <15%. When oncology was excluded from the assessment, the overall POS increased from 13.8% to 20.9%.
When Good Trials Go Bad
Now that we’ve got a better understanding of success rates across phases and indications, let’s dig a little deeper to uncover why it is that so many clinical trials fail. For any failed trial, there can be many potential culprits for the failure. Sometimes these failures are unavoidable, but most of the time they arise due to poor planning or a misunderstanding of key biological and/or drug development principles. During early phases of clinical development, the cause of a clinical trial failure could be as simple as human subjects not responding to the drug the way that animal models do. This is a key concern for a number of therapeutic areas, but especially for fields like immuno-oncology where truly translatable animal models are often lacking. Clinical trial failures may also arise when the body of knowledge about a disease or condition is sparse (e.g., orphan diseases). Orphan drugs are those used to treat rare diseases (defined by the FDA as affecting <200,000 people in the US) and are often more difficult to develop, in part because they tend to be understudied. In agreement with this, Lo and colleagues observed a lower success rate for orphan drug trials compared to trials for non-orphan indications (POS of 6.2% for orphan drug trials vs. 13.8% overall). Clinical trial failures can also result from an insufficient understanding of how the investigational product interacts with the body (i.e., pharmacokinetics and pharmacodynamics). These interactions can vary (sometimes widely) between animals and humans, between healthy subjects and patients, and between one demographic group and another. Beyond these biological considerations, a number of other factors can also derail a clinical trial. These may include:
- Inadequate Study Design
- Improper Dose Selection
- Non-optimal Assessment Schedules
- Inappropriate Efficacy Metrics/Markers
- Issues with how the data are analyzed.
Even placebo response has been implicated as a contributor to clinical trial failures, especially in Phase 3.
How to Minimize the Chance of a Failed Clinical Trial
Many of the issues that result in trial failures can arise at any phase of clinical development. As such, it is critical that steps be taken early to prevent these failures. It’s important to consult with experts who have a lot of experience navigating the sometimes-choppy waters of clinical drug development. Allucent's scientists and senior consultants understand how to design and execute successful clinical trials and get drugs approved. The following tips will help minimize the chance of clinical trial failure.
Ensure Appropriate Clinical Trial Design
It’s important to make sure your study design is appropriate and optimized from the very beginning. There are lots of options for trial designs out there and many places for drug developers to go wrong. Here are a few questions you may want to ask when designing your clinical trial:
- Do you need a parallel study or a crossover?
- What about an adaptive design?
- Is your study population appropriately sized?
- How about those inclusion and exclusion criteria?
- Have you determined the optimal dosage(s) and pharmacokinetic (PK) sampling schedule?
- What about food effects, drug interactions, renal impairment, and hepatic impairment?
In the end, you need to ensure that you’ve got an efficient design that is fit for your purpose, defensible with regulatory authorities like the FDA, and that it is capable of providing the information you seek.
Utilize Modeling and Simulation to Inform Your Drug Development Decisions
Another important consideration is the use of modeling and simulation techniques to inform clinical trial and overall drug development decisions. Modeling and simulation approaches are useful for tasks ranging from predicting appropriate and defensible doses to understanding the relationship between drug concentration and desirable/undesirable pharmacodynamic (PD) responses. Modeling and simulation can also characterize the PK/PD variability of your drug and assist in understanding the factors that contribute to that variability. When used at strategic points in the development program, model based drug development can save you significant time and money and get your product to market faster.
Select Appropriate Efficacy Endpoints
Selecting the appropriate efficacy endpoint(s) can mean the difference between an efficient and informative trial and one that is neither. The most relevant endpoints are those that give a direct readout of clinical benefit. These are also the ones that regulatory authorities are most interested in understanding. However, there are times when alternative endpoints can be the most efficient means to a speedy approval. As part of an ongoing effort to provide safe and effective therapies to the public as efficiently as possible, many drug developers and regulators have been exploring the use of biomarkers for patient selection and even as efficacy readouts for approval. This usually happens with post-marketing commitments to examine direct benefit (for example, the overall survival in the case of oncology drugs). Biomarkers are measurements like blood pressure, cholesterol, viral load, and tumor size that reflect the activity of a disease process and generally quantitatively correlate with disease progression. Biomarkers that are believed to be reflective of clinical benefit are known as surrogate markers or surrogate endpoints. As an aid to drug developers and the public, the FDA provides a list of surrogate endpoints that have been used as the basis for drug and biologic approvals. While the use of biomarkers has been subject to ongoing debate and their ability to accurately predict clinical benefit has been variable, they can still be viable options, especially in therapeutic areas like oncology. Regardless of whether you choose more traditional endpoints or opt for alternatives like biomarkers, the decision should be made with careful consideration of all potential risks and benefits.
Clinical trials involve significant investments of time and money. Even so, sometimes clinical trials fail. Understanding why clinical trials fail and how failure can be prevented is essential to getting your drug to market faster. Allucent’s team has extensive experience helping our clients get their drugs approved. Whether you need a top-quality protocol written, a second opinion on your PK sampling schedule, expert advice on modeling and simulation, or a comprehensive clinical development plan, our team stands ready to help you meet your goals. Check out our Services page for a full list of our capabilities.
- Wong C.H., Siah K.W., and Lo A.W. (2018). Estimation of clinical trial success rates and related parameters. Biostatistics. https://doi.org/10.1093/biostatistics/kxx069
- Thomas D. W., Burns J., Audette J., Carrol A., Dow-Hygelund C., and Hay M. (2016). Clinical Development Success Rates 2006–2015. San Diego: Biomedtracker.
- Hay M., Thomas D. W., Craighead J. L., Economides C., and Rosenthal J. (2014). Clinical development success rates for investigational drugs. Nature Biotechnology 32, 40–51. https://doi.org/10.1038/nbt.2786