Pushing the accelerator

Though our understanding of human biology and medicine has increased tremendously over the past decade, the pace at which drugs are brought to market has remained remarkably slow, and the process very expensive. It still takes an average of 12 years for a drug to go from lab bench to bedside, and while costs vary greatly the Association of the British Pharmaceutical Industry has estimated the average to be £2bn.

There are good reasons for this pragmatism. A great deal of time must be dedicated to background research, not to mention rigorous clinical trials and due diligence throughout the drug testing phases. However, the process is still rife with inefficiencies. It takes between three and five months for a pharmaceutical company to set up a trial, and considering that some drugs require as many as 70 trials before receiving regulatory approval, these months can eventually add up to years spent on preparation instead of progress.

What is the limiting factor?

While software can’t speed up the research to develop the molecule or biologic, it can have a significant impact on speeding up the process. Most of the process inefficiencies in today’s drug development environment stem from longstanding operational silos between internal stakeholders. With each department working in relative isolation and using only their own data sets to inform their decisions, pharmaceutical companies have yet to make use of all the information at their disposal in a cohesive, advantageous way.

Progress has been made of course. Over the last 10 years the industry has started to move from clunky, manual processes to eClinical technologies such as Electronic Data Capture (EDC), Clinical Trial Management Systems (CTMS) and Clinical Data Management Systems (CDMS). These have improved the way data is used at individual points in the drug development chain and sped up distinct pockets of analysis, but they are still point solutions that work independently of each other and are prone to the same redundancy and data entry errors as manual processes.

A catalyst for faster, safer trials

eClinical technology has only fulfilled part of the promise of an all-digital ecosystem, but the emergence of cloud-based eClinical software is paving the way for significant improvements. Cloud-based systems are ideally suited to unifying disparate systems, and can allow pharmaceutical companies to link every element of their drug development cycle to each other and to a single central database. As a result, they help to eliminate duplicate processes, allow all teams to work off a single and complete set of data that is only entered once, and most importantly speed up trials so drugs can be tested more quickly.

Improved visibility into data will drive faster and better decision-making across the organisation. For instance, teams will be able to more quickly prepare submissions for biostatistical analyses and share their learnings. Quintiles, the world’s largest contract research organisation, has differentiated itself by providing its customers with a real-time view of clinical trial data so they monitor their progress and quickly adjust their approach when required.

There are also efficiencies to be gained from a regulatory perspective. With a transparent view of where data is stored and how it flows between teams, companies can access compliance reports and respond to health authority requests more quickly. By some estimates, an integrated eClinical platform has the potential to cut set-up time for clinical studies by 50-80%.

Discovering new patterns in the data

In addition to working more quickly, a more universal approach to data will allow researchers to uncover hidden relationships in their data sets that might provoke new, potentially life-saving discoveries.

Take the case of a pharmaceutical company that worked with PwC to find out why a promising cancer drug kept failing in some phase III trials and to pinpoint suitable patient groups for future testing. An analysis of clinical and biomarker data from phase II and phase III trials allowed the company to divide patients based on specific gene expressions and mutation signatures that demonstrated drastically different survival rates, and ultimately attributed the drug’s failures to a genetic imbalance in a portion of patients afflicted with a rapidly progressing disease. This discovery led to the identification of several valuable biomarkers and helped the organisation determine which patient groups to target in future trials and which to exclude.

The potential of machine learning

The ability to collect, analyse and process all of a company’s data in a centralised way will also form the foundation for success for new innovations like machine learning and artificial intelligence (AI), which allow researchers to augment their analyses even further with new algorithmic techniques. While we are still in the early days of AI, the technology will increasingly help research teams to identify the best candidates for particular drug trials.

Machine learning will help deliver operational cost savings as well. Intelligent algorithms will enable pharmaceutical companies to predict potential supply and demand for new drugs based on a range of historical and market data, or to automatically reallocate manufacturing capacity to avoid product shortages.

Machine learning and AI also lend themselves to more careful candidate selection for clinical trials, which will result in safer testing. Historical data collected over thousands of trials will reveal a range of flags indicating a potential safety risk for patients. As this data set grows the level of insight it reveals will increase, which means the risk of choosing at-risk candidates will drop. Pharmaceutical companies are under enormous pressure to develop drugs faster while still ensuring patient safety, and advances like this will be crucial to helping them achieve this.

New technology requires new culture

The ultimate aim of clinical trials is to provide the public with life-saving or disease-curing drugs as quickly and safely as possible. This is not a goal that can be realised by technology alone, and it will take a cultural shift across the pharmaceutical development industry if companies are to make the most of eClinical platforms.

Researchers have been conditioned for years to work in isolation, both by the organisational structure in which they operate and the limited technologies at their disposal. This is why the shift from manual processes to fully digital ones, and the subsequent shift from on-premise solutions to cloud platforms, will not happen overnight. Even today, pharmaceutical companies are still adjusting their way of working as they move to digital data management systems and begin to experiment with cloud-based capabilities.

The step-change promised by fully cloud-based eClinical platforms will demand a significant change in approach and more collaborative research methods, but these are both advantageous in their own right. With the right people, processes and roadmap in place pharmaceutical companies may finally be in a position to speed up and improve drug development by significant margins.