Innovative Study Design
To optimize speed and improve the likelihood of completing successful, valuable research, sponsors are increasingly focused on two specific kinds of design: synthetic data and platform trials.
1. Synthetic Data
While randomized clinical trials remain the gold standard for evaluation of experimental therapeutics, they present challenges for both patients and sponsors. Reliance on placebo control arms can create challenges in both recruitment and retention. Some patients may be uninterested in taking a chance of receiving a placebo or have already rejected the standard of care. In the trial itself, patients may drop out if they suspect they aren’t receiving any level of treatment.
Synthetic data uses information previously collected in place of a control arm. That data may come from historical control data, real world data, or the generation of a companion data set from other sources.
That data creates an external control arm that becomes the comparator, meaning that every patient in the active study will receive the experimental treatment. While external control arms were certainly used before COVID-19, they’ve been thrown in the spotlight as both a moral imperative and a source of speed in 2022.
The ethical advantages of external control arms range from removing selection biases and easing concerns around placebos to ensuring strong real-world comparators within carefully controlled trials. The speed factor is undeniable—and likely why this approach has been brought into that spotlight now—because sponsors need only enroll half the usual number of patients. That gets trials completed more quickly and approved therapeutics to people faster.
Both the US Food and Drug Administration (FDA) and European Medicines Agency (EMA) have been active in working with sponsors using synthetic data. In the years between 2000 and 2019, the FDA approved 45 treatments with an external control arm.2 Twenty European countries expanded the label for a non-small cell lung cancer treatment based on a similar design. NICE has continued to evaluate the use of external data in its submissions and found increasing use over the last two years.
2. Platform Trials
Platform trials may be the great signal of the industry’s commitment to appropriate acceleration. They’ve challenged a series of steps that previously worked sequentially to work simultaneously. Or, at least with important overlap.
Working from one master protocol, platform trials allow multiple treatments to either enter or exit the trial over the course of the study. That gives researchers the ability to adapt based on the response from different cohorts. They can quickly drop treatments that are not performing and declare one or more approaches superior. The interim data may even lead them to add assessments or treatments that might ultimately best reveal impact for a particular population. Moreover, patients may be matched more effectively with the appropriate trial protocol through streamlined screening measures, reducing the burden on both patients and sites.
Platform trials are not new. But they proved their value during the pandemic—not just in the US, but in Europe. A large platform trial in the UK called RECOVERY was the first to prove that dexamethasone significantly reduces deaths among hospitalized COVID-19 patients, and the same platform further showed that the malaria treatment hydroxychloroquine is of little use.3 These results had an immense impact on clinical responses to COVID-19 around the world.
In 2022 and beyond, you’ll see this approach increasingly used in drugs expected to have multiple indications, letting investigators learn about multiple uses at once. They’re also expected to play critical roles for investigational treatments that only have a very limited period of time to recruit, due to resources, patient need or the competitive environment. In those cases, biopharma leaders are increasingly turning to platform trials where the traditional models won’t work.
As more sponsors adopt platform approaches, look for new conversations about data speed, interim evaluation considerations and statistical rules to prompt action.
Biopharma’s approach to integrating real world data into overall development has moved from hesitancy to reliance. Synthetic arms (discussed above) are one use of real world data, but within the new clinical model, real world data is also building engines of discovery that fuel continuous improvement. Although there are many, three particular "engines" stand out as we enter 2022:
1. Right-now understanding
Randomized controlled clinical trials are a snapshot in time. They’re how particular people responded to a particular drug given a particular pathology. Investigators are using real world evidence to respond when one of those variables changes. A clear example is COVID-19. The vaccine clinical trials were built around the original variant of the disease. As the coronavirus spread and mutated, new variants emerged that simply didn’t exist in the trials. Real world data let innovators build models for how effective the vaccines were likely to be against those variants. Similar models are being used to understand differential impact in subpopulations, the impact of multiple treatments, and the role of underlying health.
2. Uncovering rare opportunity
Real world data (RWD) has become a crucial engine for uncovering new evidence in rare disease. It starts with understanding the disease progression and overall impact on the patient over time. RWD charting that natural disease progression fuels important discovery and research design. But, increasingly, real world is standing in for both initial research and ongoing study over time. In some of the more than 7,000 rare diseases, there simply are not enough patients to support a typical trial. More importantly, many times the treatments are so essential to a patient population with significant unmet need that they simply cannot wait for traditional research.
3. Pro-patient approvals
In 2021, the FDA’s controversial approval of Biogen’s Aduhelm™ was welcomed by Alzheimer’s advocates and organizations across the US. During consideration of the drug’s surrogate endpoint and FDA precedent, those same advocates accused the agency of not being “pro patient.” For them, not giving them access to the drug meant not giving them access to hope. Even with the approval many questions remain about this and other drugs that were reviewed with surrogate endpoints instead of clinical ones. Real world data becomes a critical tool for the biopharma company and the agency to continue to build a picture of that drug over time, one focused on real world efficacy, safety and value. That use case appears even more often in rare diseases where complex, specialized rare gene and cell therapies that can alter DNA or RNA lead regulators to require five, 10 or even more years of data in the post-approval setting. Satisfying regulators, however, is just one step in bringing medicines to patients. Especially in rare diseases, manufacturers must also consider how payers will look at their evidence dossiers at a time of great change in the market-access environment. Thanks to a burst of research productivity, patients and doctors for the first time can choose among multiple therapies for rare conditions such as Duchenne muscular dystrophy and spinal muscular atrophy. In response, recent survey data5 show that payers will apply tools such as step therapies and formulary tiers to help manage costs. The emergence of treatment competition in rare conditions will change how payers look at a drug’s value proposition and continue to affect benefit design.
Manufacturers must anticipate these changes and upgrade their go-to-market procedures and philosophy, which may include accelerating the timelines on confirmatory trials so payers can be confident their budgets are spent on treatments that work.
