“Clinical trials are the most expensive part of developing a drug. And, it’s very hard to do a clinical trial testing the new drug’s interaction with every other drug that might be out there.”William F Feehery, CEO of Certara
Traditional clinical trials are equivalent to billions of dollars and years of hard work with no guarantee for the new drug to be approved by regulatory bodies, not to speak about the dangers of testing medication on animals or humans. According to CB Insights, on average, it costs $2.6B to research and develop a successful drug and takes 10+ years to come to market. It’s estimated that in-vivo testing (testing on animals and humans) accounts for more than 75% of the total cost, with recruitment alone being one of the most significant barriers to drug development — only 6% of clinical trials are ultimately completed on time.
Then there’s the issue of clinical effectiveness. According to the US Food and Drug Administration (FDA), medication ineffectiveness ranges from 38 to 75% for various illnesses ranging from depression to osteoporosis. The primary cause is each individual’s genetic makeup. It is so diverse and their interaction so unique that medicines designed for the “ideal patient” may not be appropriate for the “actual patient.”
Amid a global health crisis, the challenges only multiply: Covid-19 interrupted an estimated 80% of non-Covid-related clinical trials. One way to modernize the drug testing process is by applying technologies to the traditional framework, such as online platforms to seek out participants. An alternative method is to build an entirely new setting. That new setting leverages the multiplier effect of several exponential technologies to virtualize the clinical trial process.
If you’ve been following this blog, you know that I’ve written on several of the critical topics before in my “Straight Talk” series of posts. Here are the major ones:
- Some Straight Talk on Remote Patient Monitoring in Health Care
- Some Straight Talk on XR in Health Care
- Some Straight Talk on Digital Health Sensors in Health Care
- Some Straight Talk on the Internet of Medical Things (IoMT) in Health Care
- Some Straight Talk on How Technology Will Impact Clinical Trials
- Some Straight Talk on Edge Computing in Health Care
For this post, I want to focus on two additional areas recently getting attention: digital twins and in-silico trials. These are generally lumped together. But for this review, I’ve chosen to keep them separate.
As part of the move to personalized medicine, researchers are interested in developing digital twins that could integrate known human physiology and immunology with an individual patient’s clinical data in real-time, then produce predictions of what would happen during various medical events. A digital twin is a virtual representation of a single person where every known medicine for that person’s illness can be tested. This will allow the best treatment to be determined. It can even monitor the virtual “person” and notify you if a medical condition develops as a side-effect enabling preventive actions. As a result, the digital twin has numerous applications across multiple therapeutic areas in healthcare.
It’s been reported that 66% of healthcare executives expect increasing investment in digital twins over the next three years. This is because digital twins improve healthcare organization performance, discover areas for improvements, provide customization and personalization of medicine and diagnosis, and enable the development of new medications and devices.
If we return to the list of exponential technologies above, digital twins use all of them to create a complete picture of an individual’s vitals, medical state, response to drugs, therapy, and the surrounding environment. Companies are creating digital twins to specifically look at chronic diseases like diabetes, where a chronic diabetes patient’s lifestyle, daily food habits, and blood sugar data are analyzed. The model notifies the patient about prescriptions, dietary habit modifications, medical consultations, and so on.
Another excellent example of the advancement of the field is the Oncosimulator project. The In Silico Oncology Group is developing an in silico experimental platform, as well as an advanced medical decision support tool called Oncosimulator, in collaboration with several research centers in Europe and Japan to optimize cancer treatment. The oncosimulator is an integrated software system simulating in vivo tumor response to therapeutics within a clinical trial environment. It aims to support clinical decision-making for individual patients.
Several companies have created digital twin representations of human organs. For instance, Hewlett Packard Enterprise collaborated with Ecole Polytechnique Fédérale de Lausannes (EPFL) on the Blue Brain Project, using its supercomputer to develop digital models of the brain for scientific purposes. Siemens Healthineers offers a Digital Twin model, and Philips offers their own virtual heart. Dassault Systèmes launched its Living Heart Project in 2014 to crowdsource a virtual twin of the human heart. The project has evolved as an open-source collaboration among medical researchers, surgeons, medical device manufacturers, and drug companies. Meanwhile, the company’s Living Brain project is guiding epilepsy treatment and tracking the progression of neurodegenerative diseases. The company has organized similar efforts for lungs, knees, eyes, and other systems.
Although digital twins have a promising future in health care, the full impact of the technology will be determined by its capacity to integrate knowledge into accurate medical advice at scale. Better data, new interactions between patients and providers, and a regulatory framework to confirm these promises will be required to support this transformation.
In-silico trials simulate the effects of a new treatment using virtual populations to supplement or even partially replace in vivo testing. Researchers can use modeling and simulation to predict trial outcomes before advancing to real-world clinical trials and ultimately design studies that are more likely to succeed. Virtual populations can diversify the biological variability of traditional trials and enable the exploration of irregular phenotypes that would be difficult to recruit for. Control or placebo arms in trials can be simulated so that real-life patients who need treatment are guaranteed to receive it. This helps encourage potential subjects to enroll in the first place. Finally, in silico methods can lead to more exploratory research outcomes that might not be feasible with conventional trials. For instance, a recent in-silico trial looked at the same virtual population twice to see how the presence or absence of a secondary risk factor affected treatment.
The technology is mainly early-stage, but it has recently seen increasing adoption from medical device and pharma players. Using statistical models of disease progression, researchers can better simulate clinical outcomes for a given cohort of patients, down to the level of how specific traits impact treatment. This could result in a hyper-personalized approach to assessing a patient’s fit for a given intervention.
In-silico technology, though, is not without its drawbacks. For one, using computer-generated patient populations relies on real-life, historical data for modeling, making it tricky to test for unexpected or novel side effects to treatments. Instead, in-silico trials might be more suited to test a treatment’s efficacy (i.e., to validate “expected” results). Soon, in-silico testing will primarily be used to augment or optimize traditional in-vivo testing rather than replace it altogether. Watch the regulators too. The U.S. Food and Drug Administration also picked up on the potential of in-silico trials, and it’s actively supporting the development of virtual models – for the testing of new medical devices. The FDA and the EUA in Europe are creating frameworks outlining best practices for collecting and analyzing data like digital evidence. And the FDA is already planning for a future in which more than half of all clinical trial data will come from computer simulations.
Big tech has a vital role to play here. In 2019, Verily, the health and life sciences company under Google parent company Alphabet, announced it was moving into the clinical trials space. Last month, Amazon Web Services announced a collaboration with Thread Research intended to decrease clinical trial costs while improving research access and data quality. And the Apple Watch has been used in a variety of studies, recent and current, both to investigate the efficacy of treatments (in areas where the watch’s efficacy has already been satisfactorily demonstrated), as well to further investigate the watch in other use cases. Watch for new partnerships with Google, Amazon, Apple, and others to gather and collect data to support in silico clinical trials.
After reviewing the current literature, I would sum up the benefits and drawbacks of virtual clinical trials like this:
- Larger number of trial subjects
- Decreased costs
- No consequences for either animals or humans
- Better patient engagement
- Can lead to more exploratory research outcomes
- Not compatible with all types of clinical trials
- Access issues
- Difficult to test for unexpected side effects
- Credibility factor
- EHR interoperability challenges
While completely simulated clinical trials are not feasible with current technology and understanding of biology, their development would be expected to have significant benefits over current in-vivo clinical trials. Under the right conditions, they could rapidly supplant traditional in-person approaches and dramatically enhance the scale, data collection, geographic range, cost-effectiveness, and speed of clinical trials. Certainly, decentralized clinical trials are here to stay. As we’ve seen in other areas of health care, the last 24 months have crystalized the potential of the virtualized research model, driving a rate of deployment and progress that might otherwise have taken 5-10 years to materialize.