The healthcare industry is slow to innovate. One cure can wait for decades of clinical trials in the time it takes to design 20 distinct iPhone models. The only way healthcare providers can deliver the promise of personalized medicine is with simulation.
“Leading companies have made great innovations to save lives, but many are never realized as treatments because it’s too expensive to provide the required evidence to get them approved by regulatory authorities,” says Thierry Marchal, global industry director for healthcare at ANSYS.
It’s a frightening thought — the only thing between a patient and a known cure sitting in someone’s lab is years — sometimes decades — of testing.
You can barely blame the pharmaceutical companies. If they come out with a new drug that cures millions but seriously endangers one due to an allergic reaction what would the headlines say?
Some could also be tempted to blame the Food and Drug Administration (FDA) for approving a treatment that fails only once. So, the FDA might tighten the regulatory process and require the testing of another 1,000 people — just to be safe. With this increased cost many companies may think twice and opt to scrap promising treatments.
But if it takes this long to validate treatments, how is personalized medicine possible?
A key to personalized medicine is to validate treatments based on computer models of patients by partially replacing clinical trials with in silico clinical trials. In silico clinical trials are run on computers using a large cohort of virtual patients. The researchers can run thousands of simulations in the computer to ensure that a new treatment is safe and beneficial for the target population.
Simulations Create Clinical Trials of Personalized Medicine
“The biggest problems with lab testing and clinical trials is that they take a long time and create models that are not totally accurate,” says Marchal. “You can test 50,000 people but you will never be able to capture the variability of the 7.4 billion people on this planet.”
Simulation, on the other hand, is much more affordable, faster and versatile.
You can run a decade-long traditional clinical trial on 1,000 people — hoping the data applies to everyone.
Or, you can perform computed tomography (CT) scans of you, your children, your parents, your neighbors and your aunt Betty. You then create a computer model for everyone in your network and run tens of thousands of simulations on this large digital population to ensure that none of them would suffer from the treatment if they needed it in the real life.
That is the beauty of a properly verified and validated (V&V) in silico clinical trial. It could run a simulated clinical trial on thousands of virtual patient in weeks without risking any human or animal.
These simulations will be able to show how some treatments would improve your aunt Betty’s condition, worsen her condition, trigger any side effects, maintain her health for the future or lead to dramatic consequences.
The regulatory authorities and the government won’t go after you if a hundred digital Betty clones suffer to save the life of a real-world Betty. They won’t be as kind if the real Betty’s condition worsened dramatically due to ignorance about your medical product’s unfortunate interaction with her specific body.
Simulations Can Verify Treatments for Whole Populations
Simulations can go far beyond traditional clinical trials. What if it could ensure that a treatment will be safe for all 7.4 billion people on this planet and anyone who may live in the next 2,000 years? This is theoretically possible.
Let’s say you know the historic world record for the largest and smallest human heart. You can use this data to determine the variations of key parameters, or material properties, of the organ.
You can use this data in a parameter study to verify a treatment for every heart that lands between those world records.
At this point, you can effectively verify that the procedure would work for everyone without a clinical trial.
“In the future, I think we may completely replace clinical trials for numerous cases,” says Marchal. “This is already discussed in medical conferences now that the industry and regulatory bodies are gaining confidence in simulation. We are not there yet, but the seed has been planted. Today’s in silico clinical trials are already saving a lot of money for the industry — they also help manufacturers release treatments years faster.”
Since clinical testing could cost up to $40,000 a person, even partially replacing clinical trials with simulation could save tens of millions of dollars. It’s beginning to happen already. Marchal explains that companies have reduced their clinical trials by hundreds, even thousands, of people based on simulation results.
Regulatory bodies are also accepting these computer models as a form of validation — assuming the V&V of the models and simulations. In fact, that might be the biggest hurdle to the dream of personalized medicine — validating every simulation and model to a level acceptable to the regulatory authorities.
Clearly, as Marchal points out, the FDA and other regulatory organizations are starting to embrace simulations. So, he might be right. Perhaps it’s only a matter of time and commitment until most clinical trials are a thing of the past.