In Silico Clinical Trial Replicates Results Of Traditional Trials

Researchers at the University of Leeds in the UK have conducted a virtual ‘clinical trial’, using parameters from a real patient population and in silico modeling, to investigate the use of a flow diverter device in brain aneurysms. The virtual trial successfully predicted the results of real trials in humans, suggesting that the approach could be used to supplement and expedite clinical trials in the future.

Clinical trials are a massive investment of time, money, and effort. When the results are not positive, it is hugely disappointing for everyone involved. Techniques that can help to speed up the proces and make it less expensive would be hugely welcome by those that organize and conduct clinical trials. Virtual trials may be one way to extensively test medical devices before they are tried in humans. This could potentially help to minimize the risk that experimental treatments won’t work, while reducing the number of human participants in clinical trials and even the number of animals required for preclinical testing.

​

“In-silico trials offer an opportunity to do virtual experiments that could explain concepts that are difficult to study in conventional clinical trials,” said Alex Frangi, a researcher involved in the study. “The current approach to improve our understanding of new medical devices is slow, as conventional trials can easily take five to eight years, from their design to completion. In-silico trials could reduce this period to less than six months in some circumstances, making knowledge and therapeutic technologies safer and more promptly available to clinicians and patients.”

This latest project is a proof-of-concept study of virtual trials in the treatment of aneurysms using a flow diverter, which is a mesh tube that is placed into a blood vessel using a catheter to reduce blood flow into an aneurysm. To perform the virtual trial, the researchers created a sample of 82 virtual patients, akin to Sims in the popular computer game, using patient data from clinical databases. The ‘Sims’ closely resembled real patients from previous flow diverter trials.

The researchers then used in silico modeling to predict the results of flow diverter implantation in the virtual patient group. The virtual trial predicted that 82.9% of the patients in the sample would experience successful flow diverter treatment. This was strikingly close to the results of three real-world flow diverter trials (86.8%, 74.8% and 76.8% success rate), suggesting that the virtual trial was largely accurate in predicting treatment success.

“The results demonstrate the huge potential of in-silico trials,” said Frangi. “We have shown that the approach can replicate the findings of traditional clinical trials – and they do that in a fraction of the time it normally takes, and at a fraction of the cost.”

Source