Clinical trials are expensive – they can cost tens and thousands of dollars per patient.
If a trial has hiccups, which many do, the process has to start all over again, often lasting many years, turning it into a massive financial sinkhole.
Perhaps the answer is to radically redesign clinical trials to run as computer simulations of patient groups, instead of recruiting people for a real-life clinical trial. It has the potential to save both time and money, encourage innovation, and even save lives.
More or less akin to the way virtual populations are built in The Sims, in-silico clinical trials could revolutionise how clinical trials are conducted. They can reduce the time and costs of getting new medical devices and medicines developed, while minimising human and animal harm in the process, especially if the simulations are run on green HPC.
Putting in-silico trials to the test.
Medical trials at present are carried out in one of two options, either in-vitro, which is on cells in a petri dish, or in-vivo, which is in live subjects, either animal or human. Both methods have their pros and cons – in-vitro is safe but very far removed from a real-life situation, while in-vivo is more realistic but carries higher risks.
Published in the journal Nature Communications, a team of scientists pioneered the most comprehensive trial for the third option – in-silico trials.
Led by Professor Alejandro Frangi from the University of Leeds, the team investigated whether an in-silico trial could replicate the results of three real-life clinical trials that assessed the effectiveness of a medical device called an intracranial flow diverter. The device, like mini stems, goes into blood vessels in the brain and prevents the walls of the vessels from bulging and accumulating blood, which leads to brain aneurysms. Unsuccessful treatments can cause vessels to burst, leading to hemorrhagic stroke.
How are in-silico trials carried out?
In their study, the researchers created a virtual population using real patient data drawn from clinical databases, making sure that the virtual patients closely resembled the patients used in the real-life clinical trials in terms of age, sex and aneurysm characteristics.
Using cloud-based computing, they then constructed a computational model that analysed how the implanted device would affect blood flow in each of the virtual patients. Different physiological conditions for each patient, such as normal and high blood pressure, can be uniquely identified. They can even perform analysis on patient subgroups with different types of aneurysms.
Results from the in-silico trial fit very neatly with the three trials conducted in the real world, making this particular test a huge advance as it shows that the virtual method could replicate traditional clinical trial findings.
Will virtual clinical trials ever become a reality?
Powerful computer simulations have existed for a long time, so what’s preventing clinical trials from jumping on board the digital train?
Well, to make in-silico clinical trials a reality, what we need are highly accurate models of the comparatively small-scale things that we are studying. And that’s the problem. Accurate models of the entire human body – simulating how one part of the body affects another – don’t exist yet. Blood flow in the brain can be modelled precisely, but that isn’t necessarily replicated in other parts of the human body.
On top of that, evidence arrived from in-silico trials is also rarely accepted by regulators to approve medical treatments. At this time, scientists are still working on validating the method before stand-alone in-silico trials can become a reality.
Certainly, real patients will not disappear from the clinical trial landscape any time soon. The real goal of implementing virtual clinical trials is to streamline how many clinical trials on real patients need to be conducted. This could ultimately allow medical practitioners to cut costs and save valuable time, while also making clinical trials more representative of the population.
As the evolution of high-performance computing makes previously impossible science possible, expect more in-silico clinical trials to mature in the coming years.
