The Sisyphean task called human clinical trial
I honestly hate sitting on panel discussions. I like offering solutions to problems through my keynotes, but panel discussions rarely provide a chance for that. Still, I accept invitations and sit on these panels with the thought that you never know when you’ll get a challenging question. At an event organized by the pharmaceutical industry, I sat with representatives from major pharma companies. I had given a talk about how disruptive innovations would transform the pharma industry entirely. To my surprise there were patients in the audience.
One patient directed his question about human testing to another member of the panel. I was glad because I covered the issue in my talk and wanted to hear what pharma representatives had to say about it. Current pharma models involve a lengthy and expensive process of clinical trials whose purpose is to assess the safety and efficacy of potential new drugs.
New drugs are approved through human clinical trials. These are rigorous, starting in animal trials and gradually moving to patients. It typically costs billions of dollars and takes many years to complete, sometimes more than a decade. Patients in trials are exposed to side effects that cannot be predicted or expected. If the trial is successful, it may or may not receive approval of the respective regulatory agency, e.g. the US Food and Drugs Administration (FDA).
If a pharmaceutical company jumps through all the hoops and wins approval, they can sell their new product for a limited time under patent protection. If it does not win approval all their investment goes down the drain.
How can we change the process?
There are two lines of thoughts concerning possible changes to the system. There are many who would like to improve the existing system, and there are those who see the future of clinical trials in something completely different, rather revolutionary, the so-called in silico trials.
Technology to assist clinical trials
One way to help clinical trials is through various online services which make it possible for more and more patients to participate in the process of drug creation. TrialReach tries to bridge the gap between patients and researchers who are developing new drugs. If more patients have a chance to participate in trials, they might become more engaged with potential treatments or even be able to access new treatments before they become FDA approved and freely available. TrialX similarly matches clinical trials to patients according to their gender, age, location, and medical condition. The number of such services is growing to accommodate an increasing demand from patients.
An unexpected hindrance to the success of clinical trials is linked to transportation. Although in the United States, patients who decide to participate in clinical trials often get taxi vouchers and bus passes to enable them to show up at trial sites. However, more often than not, it is still not enough for patients to actually show up. So ridesharing companies such as Uber or Lyft decided to help. Uber recently launched its Circulation initiative to get patients to clinical trial sites, using a system that allows trial organizers to coordinate rides themselves in a software backend so they know exactly where the patient is. The patient interacts with the system via text message. And Lyft launched a similar initiative.
In the United States, also the government supports actively the improvement of clinical trials. Investigators at the University of Utah, School of Medicine’s Data Coordinating Center have been awarded in 2016 a seven-year, $25 million grant from the National Center for Advancing Translational Sciences (NCATS) of the National Institutes of Health, to form one of three Trial Innovation Centers (TIC) to improve clinical research. Their goal is three-fold: (1) to eliminate and help actors get through the obstacles of bureaucracy in initiating multicenter trials and to comply with national policies; (2) to support the development and implementation of practical study protocols within feasible budgets and (3) to support for randomized drug trials, including industry-sponsored trials.
The rapid development of artificial intelligence might also aid clinical trials – with AI they might require significantly less time and they might be brought closer to the medical institutions and patients themselves. For example the company, Atomwise uses supercomputers that root out therapies from a database of molecular structures. Last year, Atomwise launched a virtual search for safe, existing medicines that could be redesigned to treat the Ebola virus. They found two drugs predicted by the company’s AI technology which may significantly reduce Ebola infectivity. This analysis, which typically would have taken months or years, was completed in less than one day. Imagine how efficient drug creation would become if such clinical trials could be run at the “ground zero” level of healthcare, namely in pharmacies.
Are in silico trials the future?
And what if we take a radical turn? What if it is time to use disruptive innovations to change how clinical trials are performed? Imagine the following: neither animals, nor humans are the subject of the lengthy and costly drug creation process, but their characteristics are so perfectly simulated that the clinical trial can be carried out in less time, with less money and still amazing results. This method is called an in silico trial.
An in silico clinical trial is an individualized computer simulation used in the development or regulatory evaluation of a medicinal product, device, or intervention. While completely simulated clinical trials are not feasible with current technology and understanding of biology, its development would be expected to have major benefits over current in vivo clinical trials, and research on it is being pursued.
Imagine if we could test thousands of new potential drugs on billions of virtual patient models in minutes? What would it take to achieve such a capability? At the very least, the virtual patients must almost perfectly mimic the physiology of the target patients, with all of the variation that actual patients show. The model should encompass circulatory, neural, endocrine, and metabolic systems, and each of these must demonstrate valid mechanism–based responses to physiological and pharmacological stimuli. Probably cognitive computers would be needed to deal with the gargantuan amount of resulting data.
HumMod is one of the most advanced simulations in this respect. It provides a top–down model of human physiology from whole organs to individual molecules. It features more than 1,500 equations and 6,500 variables such as body fluids, circulation, electrolytes, hormones, metabolism, and skin temperature. HumMod aims to simulate how human physiology works, and claims to be the most sophisticated mathematical model of human physiology ever created. HumMod has been in development for decades and it is still far from completion. It may take decades to get there.
Might Organs-on-Chips offer help with clinical trials?
Maybe supplementary technologies are needed. The Organs-on-Chips technology is able to use stem cells to mimic organs of the body with a series of devices. Many experts believe that this technology could revolutionize clinical trials and replace animal testing completely. Organs-on-Chips are engineered to mimic how the lung or the heart works at the cellular level. They are translucent, and so can provide a window into the inner workings of a particular organ. The Wyss Institute plans to build ten different organs–on–chips and connect them together. Doing this may mimic whole–body physiology better, and thus better assess responses to new drug candidates.
However, we should note that although the experiments are promising, these are still far from a real and total-body simulation of human physiology. Even if organs could be mimicked, connecting the models to each other is more complicated than we would think.
Food for thought
After the members of the panel had tried to break the question of human testing into smaller pieces, I jumped in and said I hoped that technology will soon allow us to test drugs not on patients but in silica. You might assume that developing the supercomputers for this or simulating the wonderfully complex human body on a chip are the biggest challenges to bring this about. But I think the biggest obstacle will be the resistance of pharma companies and authorities that do not like to change a very old process.
Also, we need to ensure absolute safety for patients. We need to make sure any in silico finding is suitable to be used in practice. But I am confident I will tell my kids in a decade or so that when I became a medical doctor in 2009, we used to test drugs on patients. It was a barbaric era of medicine until artificial intelligence, supercomputers and organ-on-chip technologies came along.