Using Virtual Trials to Screen for Potential Glioblastoma Therapies
The existing drug development system “has failed miserably” for people with glioblastoma brain tumors, according to Al Musella, DPM, President of the Musella Foundation for Brain Tumor Research & Information, Inc. In partnership with Cancer Commons, his organization’s Brain Tumor Virtual Trial aims to speed discovery of promising new glioblastoma treatments.
Curious Dr. George: How might your Virtual Trial approach speed the process toward either validating or invalidating potential therapies for glioblastoma?
Al Musella, DPM: I have watched how the current drug development system works since 1992, and think we could do much better. There are many problems but the top ones, as they relate to malignant brain tumors, are:
- The current system is set up to try to find one magic bullet that will successfully treat the disease. This concept has worked reasonably well for some diseases, but for brain tumors it has failed miserably. I feel that the ultimate cure for brain tumors will be a combination approach, and under the current system, it is very difficult to get the individual components of that ultimate cocktail approved by the U.S. Food and Drug administration (FDA).
- There are not enough patients for the number of trials needed. About 10% of adult glioblastoma patients—about 2,000 per year—enter clinical trials. There are currently 318 glioblastoma trials open for enrollment, which means only six patients are available per trial per year. And some trials require large numbers; it takes years to accrue enough patients, and many trials never do.
- Phase 3 trials are too long and inflexible. In the last 25 years, only one phase 3 glioblastoma study showed a statistically significant improvement in survival, adding only 3 months to median survival. As patients are treated, we learn how to use the treatments better, but the rigid structure of a phase 3 trial doesn’t allow for modifications midstream. Rather than stop a trial, discard the results and start over, researchers complete a trial anyway, knowing that there may be a better way to use the drug.
I propose an alternative approach:
- All patients must be watched in a regulatory-grade registry so we learn from every patient—not just the 10% who may not even represent the typical patient.
- A formal randomized trial is perfect to prove that a combination works, but we are not yet at the point where we have a combination worthy of a full-scale phase 3 trial. So, if a patient desires entry into a formal trial, I propose they be given access to the registry to pick the most promising trials.
- We need a system in which experimental treatments get fast conditional approvals when they are shown to be relatively safe, with early evidence that they have the intended effect—even if that effect doesn’t result by itself in prolonged life. This allows us to use them in combinations and figure out how best to use them.
- Paying for medical drugs is a problem. The only way this can work is if patients do not have to pay high costs for their treatments, and if drug companies get paid for the treatments. This will allow access to the drug. The Virtual Trial system should keep drug costs down, as it will take the vast majority of time and money out of the drug development process and break the monopolies currently preventing many new drugs from being developed. In fact, I have proposed a bill (see below) that basically requires Medicare and Medicaid to pay for them and encourages private insurance to pay. We are working out the exact wording and are considering a “pay for performance” model, or a compromise in which costs for non-standard treatments cannot exceed what the standard of care would have cost.
- Doctors, or teams of researchers, can then think up the best combinations for each individual patient, and test them in small, fast, inexpensive (or free) virtual trials. They can see the ongoing results in all patients and quickly determine if a conditionally approved drug is worthy of continued usage—perhaps tweaking how it is given or the combinations used, or if it should be dropped. Once the ultimate cocktail is found, it can then be tested in a formal phase 3 trial.
I have been working on the components of this plan. We already have “right-to-try” laws passed in the U.S., but these did not work out as intended, mainly because of cost issues and fear that the FDA would hold usage of this pathway against a drug company. So, I am now working on a much-improved version of early access called the “Promising Pathway Act,” which was recently introduced into Congress. It fixed most of the problems with right-to-try laws and the FDA’s Accelerated Approval Program. It provides for conditional approval of treatments and requires all patients who use any conditionally approved drug to participate in a registry.
My organization, in partnership with Cancer Commons and its for-profit spinoff xCures, has set up a patient-navigation program that should be the model for how we approach all serious diseases. Our team of neuro-oncologists, PhD researchers, and nurse navigators—aided by an artificial intelligence engine—can look at our rapidly growing patient registry (we now have over 1,250 patients) and evaluate medical research, the rationales presented in tumor board meetings, and patients’ medical histories in order to come up with a list of treatment plan options that may be best for any given patient. The patient and their own doctor can select a treatment from our list, or they can elect to try whatever other treatments they want. Whatever they choose, our team then follows up on every patient to see the outcome of their chosen treatment. This way we learn from every patient.
We try to help patients get access to their chosen treatments, but the most promising options are usually impossible to access because the components are not yet approved, and either they are unavailable under expanded access or the drug companies won’t allow us to combine the experimental drugs.
Nonetheless, any doctor or researcher who is using a new combination of FDA-approved drugs for a patient can submit that treatment plan option to our system to be run as a virtual trial, allowing the doctor or researcher to quickly test the theory in their patient and use our system as a screening tool. They can try many combinations quickly, and if a particular combination shows a lot of promise, it can then be promoted to a traditional phase 3 trial to validate the findings.
Any treatment plan options validated in this way would be made up of FDA-approved treatments, so any doctor could prescribe them, allowing every brain tumor patient to have access—not just those perfect patients that fit into existing clinical trials.
In summary, our Virtual Trials could be used as a screening tool to figure out the best combinations, and how best to use treatments, which then can be tested in formal trials. Conversely, this approach could be used to quickly eliminate bad combinations. Having all results available in a central registry would help doctors from around the country avoid trying the same ineffective combinations.
Dr. Musella can be reached at firstname.lastname@example.org.