Contributed Commentary by Jae Chung, Oracle Health Sciences

January 14, 2022 | The traditional drug development and clinical study design process have always been, and to an extent remain, sequential and siloed. Rigid research traditions with a clear separation of responsibilities have resulted in a clinical development ‘one-size-fits-all’ process. The result: studies show that at any given time about one-third of all clinical trials are behind schedule. Perhaps the most disturbing fact is that cycle times associated with starting clinical trials have not changed in more than two decades.

There is no question the pharmaceutical industry needs to change. Trials need to be accelerated so treatments can be brought to market sooner. To accomplish this, the industry must be more flexible and agile in how it conducts clinical trials.

The pandemic has forced the industry to find and adopt new ways of working, designing trials, and managing the data that they yield. According to a recent study by Oracle and Informa, 63% of organizations have adopted some kind of protocol redesign as they move to more of a decentralized trial model. And recent and upcoming guidelines from the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) guidelines have also encouraged the modernization of clinical trials, specifically the adoption of risk management practices, use of new technology, improved quality by design (QbD), and more critical thinking.

A New Approach

Technology will continue to play a large role in how we conduct trials, but new ways of working will have to be implemented to truly drive change. For example, during the COVID-19 vaccine trials, several new approaches emerged that helped speed the trials and bring the vaccines to market in record time. One such approach is fast-tracking, or parallelization.

Using parallelization, activities normally performed in sequence take place simultaneously, at least for a portion of their duration, without affecting the performance of each other. The result is that multiple tasks can be completed in a shorter span than if they were sequential.

For example, after a site has been identified as an ideal candidate for a clinical trial and the Confidential Disclosure Agreement (CDA) has been signed, the feasibility survey and other activities must be completed before the site can be formally selected for the study, assuming no red flags are raised. This process takes on average 7-8 weeks according to research conducted by the Tufts Center for the Study of Drug Development (Tufts CSDD).

Parallelization allows for this step in the selection process to be overlapped with the activation process if it is assessed that the likelihood of the site being rejected is low. In this scenario, the Study Package can be sent to the site as soon as the CDA is signed, reducing the average study start-up schedule by nearly six weeks. Project managers typically look at using this technique first as it does not incur additional costs. Because it involves conducting activities in overlapping, rather than consecutive, intervals, there is increased risk with this approach. However, the potential rewards—lower costs and reduced study timeline—create a situation where the risk level is acceptable.

Pfizer’s COVID-19 vaccine trials employed the parallelization approach and the company successfully brought a highly effective and safe vaccine to market in less than one year by treating the whole clinical development process (Phase I, II, III, and to some extent IV) as one protocol, managed by one team. Instead of waiting for certain steps to be completed before they moved on to the next area of the trial, activities were completed concurrently, enabling them to move much more quickly. Moreover, because one team managed the entire process, they were able to make decisions swiftly and decisively.

Thanks to parallelization, the team was able to start doing patient recruitment before all the initial start-up paperwork was finalized. This reduced their timeline significantly because by the time all the initial paperwork was done they were well into the recruitment phase.

Predictive Analytics Help Speed Clinical Trials

Pfizer’s success with using the parallelization approach for its COVID-19 vaccine trials raises the bar for how other pharmaceutical companies can replicate it. The answer lies in the technology that enables it.

In reality, the inability to see where optimizations in the process can be made and when to apply new techniques like parallelization is a big challenge for clinical trial managers due to the normally siloed structure of trials and multiple teams engaged in very different aspects. But today technologies like predictive analytics can help clinical trial managers spot opportunities to implement parallelization techniques. As we’ve seen with the Pfizer example, this helps keep trial costs down and improves accuracy and efficiency.

Predictive analytics can guide clinical operations teams in milestone planning with in-application planning assistance and use of leading indicators (e.g., sites, start month) for associated milestone estimations. This type of proactive planning assistance can alert teams to unforeseen issues that allow for decisions to be made before studies incur risk due to missed timelines. It can also offer guidance as to when the parallelization of various processes should begin while minimizing risk exposure.

Machine learning tools and automation free clinical project managers from tedious, repetitive activities so they can focus on strategic activities, drive optimal proactive planning in study execution, and aid in-depth internal reviews of organizational processes, resource allocations, study costs, and quality assessments.

There is no one-size-fits-all approach or solution to clinical trials. But forward-thinking organizations that have support from key decision-makers who support a culture of innovation will break through these entrenched barriers and ultimately continue to reduce runaway timelines and costs. Companies like Pfizer that embrace a unified eClinical platform with new strategic approaches like parallelization have already made it possible to move the needle significantly and bring therapies to market in record time. The ability to break down organizational silos, develop more robust and extensible predictive capabilities lead to further reductions in cycle times, greater site engagement in studies, better adherence to study budgets, and audit readiness. It’s a critical change at a critical time.

Jae Chung is VP of Strategy and Product Management for Oracle Health Sciences. He focuses on delivering advanced solutions for getting clinical studies started in the cloud. Prior to joining Oracle he founded goBalto, which was acquired by Oracle Oct 2018. He also co-founded Celltrion, a leading biopharmaceutical group. Jae received an MBA from NYU Stern School of Business. He can be reached at jae.chung@oracle.com.