BIO recently released the largest study of clinical drug development success rates to date, which reported that the average overall likelihood of approval (LOA) by FDA from Phase 1 for all developmental candidates was only 9.6 percent – and of the 14 major disease areas, oncology had the lowest LOA at 5.1 percent. In addition to that, phase 2 clinical programs continue to experience very low success rates, with only 30.7 percent of candidates advancing to phase 3.

A failed clinical trial has immense repercussions for a pharmaceutical company. To illustrate, when it was reported earlier this year that FDA failed to approve Sarepta Therapeutics' drug eteplirsen, in Phase 2b of a clinical trial for Duchenne muscular dystrophy (DMD), the company’s stock crashed by 50 percent. When companies experience such a catastrophe, they often have to reduce workforce, shutter research sites, consolidate business units, and narrow their therapeutic focus areas. It can have a huge impact not only on the company, but on employees, shareholders, and ultimately patients.

The cost of failed clinical trials is compounded by the fact that there are fewer blockbuster drugs coming to market. Ultimately, each successful drug that goes to market must cover not only its own R&D costs, but the costs for all the unsuccessful drugs as well – and with drugs now focusing on smaller patient populations, each drug has a smaller market in which to recoup those R&D costs. Reducing the rate of trial failure is thus critical.

Why the high rate of failure?

There are of course many factors that contribute to a failed clinical trial, but the key issue is that, despite the hype around precision medicine, pharma as an industry is not yet serious about biomarker discovery and implementation.

The BIO study mentioned above noted that rare disease programs and programs that utilized selection biomarkers had higher success rates at each phase of development vs. the overall dataset. The importance of biomarkers for patient selection is clear. Over 10 years ago FDA even issued guidance that all drugs should be accompanied by a companion diagnostic (CDx). The industry has not demonstrated that it’s serious, however, about applying a thorough understanding of biology and the immune system to develop robust and meaningful biomarkers. The problem is that, while the industry believes that leveraging biomarkers and companion diagnostics is critical to precision medicine, developing them is challenging and expensive, and therefore has not been a priority.

For example, if we look at oncology and immuno-oncology specifically, detecting sustainable, predictive biomarkers and developing robust CDx requires the correlation of genomics and complex tissue data, understanding the spatial relationship between immune cells within the tumor environment, and the development of highly sensitive, precise, quantifiable and reproducible assays. This is a challenging process and requires a significant time and cost investment; a single assay can cost millions of dollars. While developing a through drug discovery and development program is the more thoughtful and effective approach in the long-term, that money can be – and often is – used toward more short-term goals such as recruiting additional patients into a clinical trial in order to meet FDA requirements, or on a marketing campaign to promote an approved drug.

Certainly there are some companies that are putting a true effort behind biomarker-based research, particularly those that are focused on immuno-oncology, which requires an in-depth understanding of key markers and the immune system’s response to cancer. Trials in this area are not insusceptible to failure, however. Recently the industry received a great surprise when Bristol-Myers Squibb’s Opdivo failed to meet its endpoint in lung cancer, costing the company $20 billion in market value. However, some say this could be because the company enrolled too broad of a pool – patients whose tumors had relatively low levels of PD-L1, which Opdivo tries to block, since there is data that suggests patients with higher levels of PD-L1 are more likely to respond.

Put your money where your mouth is

There is a saying in Germany that translates to, “It’s just a movement of the lips, it’s not a real commitment.” If the industry is serious about reducing clinical trial failures, it needs to stop paying lip service to its commitment to precision medicine, and be serious about its claims 10 years ago about selecting the right patient for the right drug, with the right dose, at the right time. This means applying a dedicated focus to understanding the immune system and using biomarkers in the drug discovery and development process.

Biomarkers are the key to fewer failures. Biomarker testing can help identify efficacy or lack thereof, enabling pharmaceutical companies to identify promising drug candidates and halt unsuccessful drugs earlier in the process. Diagnostics facilitate better patient stratification and identify patients that will most likely benefit from the drug, enabling pharmaceutical companies to select the correct patients for a clinical trial – and not enroll patients who are unlikely to respond and contribute to poor trial data.

There are now more advanced tools and technologies than ever before to enable pharmaceutical companies to better understand the immune system and identify biomarkers. Big Data can also play a role in reducing the rate of trial failures. Mass amounts of data and knowledge have been collected in isolation over the years, but we are finally now in a position to standardize and correlate that data. With new storage and machine learning technologies, algorithms and analysis tools, we can start leveraging previously disparate data to develop new strategies, ideas and findings.

Clinical trial failure – particularly in later stages – can be disastrous to a pharmaceutical company. For the last 50 years, the industry has dedicated the bulk of its resources toward the traditional model of developing, commercializing and marketing a drug. Going forward, if we’re going to actually identify the right patients for the right treatments and reduce the rate of trial failures, more money needs to be put into data mining, diagnostics and biomarker testing efforts rather than conventional drug development efforts.

About the Author

Ralf Huss is Chief Medical Officer of Definiens and has more than 20 years of training and experience in histopathology and cancer research. Prior to joining Definiens, he also served as Global Head of Histopathology and Tissue Biomarkers at Roche Diagnostics.