Big data, blockchain technology and artificial intelligence have given industry the chance to streamline drug trials, say experts, who agree with recent FDA calls for more efficient clinical research models.
Dr Scott Gottlieb’s brief, energetic tenure as US FDA Commissioner is likely to shape agency oversight of the pharmaceutical industry for years to come, particularly in relation to clinical development.
In April, shortly before stepping down, he suggested industry’s complicated and costly approach to trials stymied drug research.
Gottlieb blamed “legacy business models that discourage collaboration and data sharing”. He also warned that unless industry finds ways to streamline trials, therapeutic opportunities made possible by advances in genomics and precision medicine would be missed.
“Without a more agile clinical research enterprise capable of testing more therapies or combinations of therapies against an expanding array of targets more efficiently and at lower total cost, important therapeutic opportunities may be delayed or discarded because we can’t afford to run trials needed to validate them.”
R&D inefficiency
The criticism is not without foundation, says Patrick Pilch, managing director and national leader of the BDO Center for Healthcare Excellence & Innovation.
“I would agree with Dr Gottlieb – as the health and life sciences sectors continue to converge and the ecosystem overall shifts towards value-based models that reward positive patient outcomes, inefficiencies in clinical trials can’t be ignored.”
The argument is supported by the fact that increased R&D spending has not resulted in more approvals. According to a PhrMA survey, industry invested $71.4bn in R&D in 2017, up from $29.8bn in 2001. While the number of approvals grew over the period, the increase was nowhere near proportional.
Pilch told us the disparity “makes the need to improve clinical trials’ speed and efficiency more urgent than ever”, adding that “if we want to continue to incentivise spending on finding treatments for rare diseases, clinical trial speed and efficiency need to be addressed”.
Challenging trials
Gen Li, founder and president of data-driven drug development contractor Phesi, acknowledged that industry’s approach to clinical research could be improved by making better use of data and technology.
“Over the years, our industry has been generating and has accumulated an unbelievable amount of data. The opportunities to pool, understand and utilise these dynamic and ever increasing data are tremendous, as Dr Gottlieb suggested.
“Specifically, to his point, our industry has collected and accumulated a large amount of data from placebo and comparator arms. It is absolutely possible to utilise this ‘useless’ data to reduce the sample size and potentially eliminate the placebo and comparator arms. The implications are huge: reduced costs, shortened cycle times and better and more ethical patient care,” Li said.
The caveat is that the sources of this data need to be taken into consideration. Li continued: “Without objective and quantitative understanding of the steps leading to the generation of data in the placebo and comparator arm, it will be risky to accurately and reliably utilise the data. With analytics tools,however, this is now possible and the opportunities this presents are significant.”
Blockchain
Better data use will only help industry streamline clinical development if maintenance and exchange technologies are effective, according to Pistoia Alliance President Steve Arlington.
“While data has huge potential to greatly increase trial efficiency, the challenge is in making it interoperable and shareable across the industry – so it can be used to inform how trials are designed and executed, to improve the accuracy of conclusions drawn from results and accelerate the approvals process.
“Today, organisations are harvesting information in many different formats without any standardisation and with few protocols for sharing. Greater use and adoption of data standards will be essential if the industry is to overcome barriers to efficiency in the clinical trial process.”
Blockchain is one possible trial data management solution. The technology creates a digital record of products or information as it is exchanged. Each transaction is a ‘block’ and these are chained together to create a record that all parties involved can access, but not edit.
Arlington said blockchain could be used to build trust levels and assure regulators that data from a trial is ‘true’.
“Each block in a chain can be linked to show that methodology has been followed, keeping the trial transparent, showing the data’s provenance and providing an audit trail.
“This would enable regulators to see the entire clinical package with all documents traceable back through the blockchain to give confidence in the trial’s history and outcomes – and could significantly decrease the time and money necessary to develop a drug.”
To implement a clinical trial blockchain, all parties would need to agree on privacy protocols and data formats, as well as decide who holds which access keys and this, Arlington said, would take considerable negotiation.
“To embrace innovations like blockchain, as Gottlieb asks, all relevant stakeholders need to be ‘in the same room’. This would require technology vendors and device manufacturers, pharma and biotech companies, and global regulators to all work on the same problem. No single organisation will be able to go it alone, and collaboration will be critical in ensuring the industry can realise
the potential of technology in making trials more efficient, more productive and safer.”
Trial training
Patrick Pilch from BDO also sees collaboration and the establishment of common standards as key to streamlining trial data management.
Patrick Pilch
“Increasing standardisation and improving user- friendliness of data collection technology is one of the greatest opportunities to improve the speed and efficiency of clinical trials.”
He explained that many clinical trials are outsourced by pharma companies to physician practices, which may be simultaneously conducting multiple trials for different companies using differing data collection technologies.
“Reducing the time spent learning and operating multiple different data collection technologies would reduce the administrative burden placed on investigators at clinical trial sites and speed-up the overall process. Technologies like artificial intelligence and robotics may be able to help as well,” Pilch said.
However, while automating some administrative tasks may speed-up research, sponsors will need to ensure their systems are robust. Pilch continued: “While digital transformation offers much promise, technology is not infallible. Organisations that are keen to automate their administrative work need to rigorously vet and test the technologies they plan to incorporate.
“Additionally, they’ll need to have secure systems in place to monitor technology for any errors and protect against cyberattacks.”
Patients are a virtue
Recruitment is another area in which trials could be streamlined. According to research by the Tufts Center for the Study of Drug Development (TUFTS), most clinical studies do not meet their patient recruitment goals.
Gen Li from Phesi commented that “challenges in patient recruitment are only the symptom of underlying clinical trial dysfunction”, citing the inability to find successful investigators and the lack of effective methods of finding patients as the major hurdles.
A more systemic approach could help streamline recruitment according to Li, who continued: “A seamless synthesis of the most efficient and validated approach to strategy, design, execution and monitoring is preferable.”
Patrick Pilch also commented that industry needs to get better at planning and executing recruitment strategies.
“Investment in patient enrolment and fulfilling compliance requirements are likely to remain staples of clinical drug trials – and for good reason. Compliance requirements are in place to ensure patient safety when a drug ultimately comes to market.
“Performing due diligence when determining patient eligibility for a trial is important for the safety of patients participating in the trials and those that will ultimately receive the treatment after commercialisation.”
Here again data and technology can help. Pilch continued: “There are several innovative approaches that we’re monitoring right now. For instance, there’s a health tech start-up focused on connecting experimenters with qualified patients to improve clinical trial efficiency.”
The organisation – Elligo Health Research – provides a platform to connect physicians with qualified patients. The aim is to reduce the administrative burden associated with scheduling and collecting data from trials.
Pilch said: “This [Elligo] platform, and ones like it, are examples of innovative approaches that are working to tackle the most significant barriers to efficiency in clinical trials.”
Research QbD
Pamela Tenaerts, executive director of the Clinical Trials Transformation Initiative (CTTI), also sees opportunities for patient recruitment to be streamlined. She told us sponsors should apply QbD manufacturing principles in a clinical setting.
Pamela Tenaerts
“The benefits of thinking differently with QbD as an approach to protocol design can be significant – essentially, you should focus resources on errors that matter to decision-making primary endpoints and patient safety, which can lead to better data, faster enrolment and reduced patient burdens.”
She continued that survey results from CTTI’s recruitment work indicate that identifying patients who meet eligibility criteria is the primary barrier to meeting recruitment goals.
“Participant recruitment must be considered at the initial outset of trial planning. Evidence-based trial feasibility, site selection and communication strategies must also be taken into account,” Tenaerts added.
Industry’s approach to trial patient safety – specifically the work carried out by Institutional Review Boards – could also be streamlined, according to Tenaerts, who continued: “For nearly a decade, CTTI has championed the adoption of single IRBs for multicentre clinical trials.” She added that an sIRB for such studies can improve quality and efficacy.
“We have developed a number of recommendations and other resources for sIRB implementation. Today, CTTI is building on this work by collaborating with an NIH workgroup to develop a comprehensive plan for assessing the NIH’s sIRB policy, and by developing resources to assist researchers and institutions in implementing the sIRB model.”