LONDON – Tracking the pandemic in all its manifestations – from symptoms and spread, to viral genomics and repurposing drugs – has massively increased appreciation of the importance of real world data, with significant implications for the future of drug discovery and clinical development, the use of patient data and health care as a whole.
Digital technologies adopted to cope with the crisis have unleashed “revolutionary behavior,” said Steve Bates, chief executive of the U.K. Bioindustry Association. “COVID-19 has meant this area of work has come on at a rate seemingly impossible a few years ago, where the public, and business, have been forced to adapt to this new normal.
“The phrase I keep returning to when describing the effect of COVID-19 is sometimes nothing happens for decades, but sometimes decades happen in a weekend,” Bates told attendees of the (virtual) UK Bioscience Forum.
The impact has been felt across the sector. A year ago, COVID-19 was unknown; there was no idea about the symptoms, how to diagnose the infection or how to treat it.
Gaps remain, but said Angela McFarlane, senior principal at health data specialist Iqvia, “I doubt there has ever been a time when real world data has been more important.” It has been essential in identifying and quantifying symptoms, predicting which patients are at the highest risk, in developing therapies and vaccines, and in real world pharmacovigilance, McFarlane said.
With 150-plus COVID-19 vaccines being rushed through development in one year rather than 10, real world data will be “a cornerstone for success” in running trials and ensuring public trust in these products, said McFarlane.
At the same time, rapid programs to advance COVID-19 therapies can light the way, demonstrating how recruitment and delivery of clinical trials can be underpinned by real world data.
The most notable exemplar is the U.K. Recovery trial, which has recruited 13,755 patients in 176 hospitals, and already has provided statistically significant evidence that dexamethasone reduces mortality, while hydroxychloroquine and the HIV combination treatment lopinavir/ritonavir do not.
The Recovery trial is poised to deliver data in two further potential therapies, with 1,000 patients now randomized to Roche Holding AG’s rheumatoid arthritis drug, Actemra (tocilizumab), vs. control, and 900 patients randomized to convalescent plasma vs. control. Earlier this month Recovery began recruiting patients to test Regeneron Pharmaceuticals Inc.’s antibody cocktail, REGN-COV2.
At the heart of running such a large study in the thick of a pandemic is a smart trial design that rather than requiring hard-pressed clinicians to record reams of data, is tapping into existing records. These include details of pre-existing conditions and medication from primary care; information on admissions, discharges, treatments and deaths from hospital records; and results from the national COVID-19 testing system.
Regulators, too, have been forced by the COVID-19 crisis to speed up the application of real world evidence in their processes, said Jackie Mulryne, partner at the London law firm, Arnold & Porter. As one example, a report on an EMA workshop on COVID-19 and real world data recently concluded evidence generated by observational research is “fundamental” to understanding the safety and efficacy of drugs.
“It’s clear regulators are increasing the focus on real world evidence,” Mulryne said. But while COVID-19 is a spur, there is some way still to go, with use of observational data mostly limited to orphan drugs. “Real world evidence has been used more for early approval for novel products [which is] then supplemented by post approval data,” said Mulryne.
More data needed
A recent analysis of 280 drugs approved from 2009 onward showed real world evidence was used in assessments of only 17 of those. “All the products are orphan, that is, regulators use real world evidence for rare diseases where randomized trials are not appropriate,” Mulryne noted.
COVID-19 may be focusing minds, but regulators are “playing catch-up” and several limitations must be addressed for the impetus that has come from the pandemic to translate through to systematic use of real world data in regulatory decision making.
Those include the fact that real world data currently are not generated or collected for use in regulatory filings; the absence of a legal framework to underpin the use of this type of data in the context of drug approvals; wide variations in the quality of data collection and methodologies for analysing it; difficulties in merging disparate sources; and the fact that the weight given to any piece of evidence depends on how it is to be used.
As things stand, there is no one size fits all. Companies have to clearly identify the questions to be answered in advance of the collection of data, and to justify its use in proving their case. “It comes down to knowing your data,” Mulryne said. “You have to validate it, so regulators will trust it.”
If the regulatory push remains modest, there is increasing pull from health technology assessment bodies, which have always ranked evidence of how a drug works for a real life patient population in a particular health care setting above data generated in clinical trials.
“We find problems with the evidence base – that is traditional evidence from randomized controlled trials. Hence why real world evidence is so important,” said Adrian Jonas, associate director for data analytics at the National Institute for Health and Care Excellence (NICE).
NICE guidance is intended to give access to the best drugs whilst providing the best value for the National Health Service, two objectives that often are in opposition.
“We gather evidence for final decision-making [on access] following approval,” Jonas said. “But we could do with more data.”
That would enable NICE to develop and update guidance more rapidly, get answers to questions that cannot be found through existing approaches, and improve monitoring of the uptake and impact of its guidance.
There are opposing forces at work in moves to make greater use of real world evidence, with the opportunity to improve products, the greater availability of data, new data collection methods, and initiatives to encourage national implementation, fighting against methodological problems, lack of resources, privacy and issues around reproducibility.
NICE has set out an interim framework for assessing real world evidence in the context of COVID-19 therapies, diagnostics and vaccines. Next year, it will set out a framework for the broader embrace of real world evidence in health technology assessment, which aims to capitalize on the potential of all the new digital tools that are available.
“NICE wants to encourage innovation; we want to get drugs to patients. But the evidence has to be there,” Jonas said.