Robert Chu

	We cannot build a world where every patient will access, let alone afford the personalized treatments we are being promised with current drug discovery costs. As medicine turns its attention to rare disease, the gold standard for evaluating new drugs, randomized control trials (RCTs), can no longer be the only viable option. Shrinking patient populations make it impractical and costly to assign similar patients into two groups, one with the new treatment, the other with a placebo. To scale precision medicine, we must shift to agile, near real-time analysis of novel data sources, including patient-generated healthcare data (PGHD), outcome reports, and Internet of Things (IoT) information. As life science companies adopt remote monitoring at scale, we will move to value-based pharma, where drugs are priced on real-world effectiveness. We believe this must happen sooner than later.

	Indeed, 95% of rare disorders still have no FDA-approved therapies. This will come as no relief to the 350 million rare disease patients worldwide, with gene mutation affecting less than five in 10,000 people. For patients, joining clinical trials is often a last resort, when they have exhausted every other option. In rare disease studies, recruiting patients can seem like finding a needle in a haystack. Participation stands at a dismal 5%. Faced with a similarly distressing situation, AIDs activists in the late 1980s demanded better access to experimental drugs. This resulted in the introduction of “compassionate use” programs. Life sciences companies and the FDA were led to broaden clinical trial inclusion.

	We must strike a new trade-off between research focus and compassion in rare diseases. The challenge is formidable in the absence of a widespread epidemic. There are many rare disease patients, but each requires an almost unique treatment. Nevertheless, we must invent a pathway towards broader inclusion in novel therapeutics. This should be the roadmap for 21st Century care.

Expanding drug monitoring beyond clinical trials

	To prove the efficacy of new drugs, RCTs have remained the gold standard since their systematic introduction in the 1930s. While efficient at proving outcomes, studies typically suffer from multiple bottlenecks. Life sciences companies hire contract research organizations (CRO) to select clinical sites (hospitals, clinics, etc.). They recruit investigators to manage the studies, who, in turn, enroll patients. There is no single source of data to target eligible patients. Registries imperfectly put together by advocacy groups or academics stand in disparate silos. Researchers struggle to recruit enough patients. Fear of receiving a placebo leads many patients to opt out. Data collection remains primarily paper-based. Patients have no access to their data. The information collected in one trial is not standardized, hardly accessible for further studies, let alone to serve as a control group for subsequent research.

	Finally, patient cohorts dissolve after trials, making drug sponsors blind to what happens in the real world. As the current opioid crisis has illustrated, FDA approval is no safeguard against doctors over-prescribing drugs to patients who may abuse them. Study periods or population sizes are insufficient to assess how long treatment works (e.g., waning immunity of vaccines) or identify all serious adverse events. This needs to change.

Fast-tracking drug evaluation in the digital age 

	Without a paradigm shift, personalized medicine will remain an unaffordable dream, available to only a privileged few in elite hospitals. The industry needs a faster, cheaper process to bring new therapies to more patients. Finding eligible patients quicker and collecting more data over longer periods is precisely what the digital health revolution is about. Drug discovery can no longer remain blind to the new wealth of health data. Patients are now monitoring themselves at home with continuous real-time data collected from wearables, sensors, and smartphones. In the last 10 years, medical records have almost all been digitized. The data that has emerged at the periphery of life sciences must be leveraged.

	But drug companies have, at best, only started digitizing their current drug validation process. What actually needs to be done, is to rethink the validation process itself. We should move away from the sequential process of submitting historical paper-based evidence to regulators, with damaging time delays. This means shifting to a process where data can be streamed in real time from the real world. Smart registries could ultimately become in silico control arms for future clinical trials.

Making real-world data ‘regulatory grade’ 

	In an effort to step up this continuous monitoring of drugs, the U.S.’s 21st Century Cures Act, signed into law in 2016, has given FDA a mandate to use real-world data (RWD) to bring new innovations to “patients who need them faster and more efficiently.'' The law builds on FDA's ongoing work to incorporate the perspectives of patients into the development of drugs and devices in the agency’s decision-making process.

	Using RWD can address issues of cohort studies. For instance, electronic medical records (EMRs) can help keep track of participants in the long run, analyze changes in medication, and allow long-term monitoring. It can elevate the reliability of trials by informing on the actual demographics of patients eligible for a drug. For rare diseases, smart registries offer the potential to fast-track research, provided data is correctly standardized to reflect clinical features, treatment, and outcomes.

	In that sense, RWD raises challenges of its own. We must make data reliable and homogeneous. Data is typically collected in a multiple settings and subject to bias. Since this data is a byproduct of treating patients, it has not been organized for specific research focus. Quality can be low. Data quality management (DQM) is key. Large samples are necessary to eliminate bias and select relevant samples. In addition, EMR data is often not enough. Tailor-made apps may be necessary to track actual drug adherence.

	To move faster, we need to recognize that the market opportunity is huge. Before its sale to Roche in 2018 for $2 billion, Flatiron Health had offered a roadmap into make EMR records “regulatory grade,” through a combination of technology and manual data classification. Roche was able to rely on this to convince regulators from 20 countries to expand the label for Alecensa (alectinib) in non-small cell lung cancer. Flatiron’s oncology director, Amy Abernathy, is now FDA’s deputy commissioner, perhaps a sign of things to come.

	Building “synthetic control arms” for clinical trials has revolutionary implications. It could help trials become faster and cheaper. Trials could prove more appealing to patients, who would all receive active treatments in the study. The transformation for life sciences could be more radical still, as all pharmaceuticals become, in effect, remote monitoring companies. It could even bring about a shift in the business model. Instead of charging for pills, organizations would begin charging for services and actual patient outcomes. Imagine a world where drugs are assessed and priced on their actual effectiveness. Why would anybody pay for an expensive drug that patients refuse to ingest? Welcome to value-based pharma. Or it could lead to a Netflix-type subscriptions for holistic patient management. Either way, the digital disruption will come to life sciences.

	Because patients cannot wait, and often cannot pay, we believe in accelerating this transition. This is why we have started standardizing RWD for evidence submission. With participation from FDA members, we have formed a consortium that brings together a group of determined digital health, bioinformatics, and life sciences experts to make it happen. Our hope is that the life sciences industry will support this approach not just for short-term financial gains due to faster drug approvals, but because they will stand to win with patients by inventing this new value-based world with them.

, MD, CIO of Beth Israel Lahey, Professor at Harvard Medical School; 

, CEO of Embleema; Alexis Normand, Head of the Embleema Consortium

Articles

The case for evaluating drugs with real-world digital health data.

Robert Chu

Articles

What’s Next for Drug Validation in the Age of Precision Medicine?