Proposal: A Virtual Trial in Medulloblastoma
‘End the Insanity’
Marty Tenenbaum, Al Musella, Rider McDowell
Introduction {adapted from the GBM virtual trial whitepaper}:
For most patients with recurrent medulloblastoma, the prognosis remains grim: average survival with currently approved treatments is less than 2 years. These patients can’t afford to wait a decade or more for new drugs to go through the regulatory process. The good news is that the oncology drug development pipeline for brain cancer is full of promising targeted- and immuno-therapies, some of which have already been approved for treatment of other cancers. These drugs are, in principle, available today off-label, in trials, and through compassionate use programs. However, no one knows the optimal way to use these ‘available’ therapies, because they are seldom tested head to head or in combinations. When used off label, nobody is tracking the results, dooming patients to repeatedly using bad combinations and not quickly identifying the best combinations. Testing multi-drug regimens in clinical trials is problematic because of the explosive combinatorics.
In the absence of definitive clinical studies, the fastest way to improve outcomes is by aggregating the insights, experiences and intuitions of our best clinicians, and continuously validating and refining them based on real-world outcomes data. Every day, patients who have exhausted the standard of care are treated with off-label drugs and rational cocktails. Unfortunately, these individualized (‘N-of-1’) experiments are uncoordinated, and their results seldom reported, so little is learned. If we can capture these results and rapidly share them with community oncologists on the front lines, we’re confident many lives can be saved – or at least meaningfully extended.
We propose two bold and synergistic initiatives that address these issues: Virtual trials and Conditional Approval. Virtual trials capture treatment and outcome data at the point of care, and use that information to inform future treatment decisions, thereby transforming the everyday practice of oncology into a global adaptive search for better treatments and cures. Conditional Approval facilitates drug access by creating a new regulatory pathway whereby any physician can treat an advanced cancer patient with promising investigational drugs that have demonstrated safety and a biological effect in a small group of patients. Patients receiving such therapies would be required to participate in a Virtual Trial, where their treatments, outcomes and side-effects could be captured and used to rapidly improve the outcomes of future patients.
The proposal:
- Cancer Commons creates a virtual trial for MB, in partnership with:
- PNOC, a consortium of 15 leading pediatric cancer centers
- A new consortium of biotechs, pharmas, and academic scientists with promising investigational treatments for MB
- The FDA
- PNOC serves as a virtual tumor board proposing / prioritizing promising treatment combinations for testing in the Virtual Trial.
- The companies agree to make their treatments available to patients in the Virtual Trial through several existing regulatory pathways
- An open clinical trial, if one exists
- Compassionate use / expanded access
- Accelerated approval, where the FDA grants conditional approval to make treatments available to patients in the Virtual Trial, so that docs can experiment with them in rational combinations.
- Companies to consider: in discussion.
- Many of these companies have promising treatments that are stuck in limbo. All have demonstrated safety and efficacy for some patients, but not enough to gain normal approval. Meanwhile, the trials have fully accrued or the companies have run out of money – and no one can get the treatments. In a disease where so many die, we must end this insanity, by getting these treatments into the toolkits of physicians so they can experiment with them in cocktails – much like was accomplished with HIV drugs.
- Partner with the FDA to get accelerated approval for the drugs stuck on limbo, provided the patients participate in the Virtual Trial.
- With accelerated approval, they become available, and physicians can experiment with combinations. And companies can get reimbursed, easing their financial squeeze.
- Here’s how the Virtual Trial works in practice:
- PNOC functions as a virtual tumor board, generating and prioritizing treatment hypotheses for each patient in the virtual trial.
- If there’s a clear consensus on what to do, go for it.
- If there’s a split decision, run the experiment.
- Coordinated teams of PNOC experts and AI algorithms function like an ‘Air Traffic Control System’ assigning patients to treatments with the dual aims of optimizing individual outcomes and maximizing collective learning.
- We will use the latest technology for real time monitoring of treatment response, so that if a treatment isn’t working for a given patient, that patient can get more shots on goal.
Safety
- Safety committee passes expert judgement on each new combo.including recommending pre-clinical testing.
- The combo is dose escalated. (needs to be done at somepoint, do it on patients) Informed consent in dire conditions.
- Cocktail can be less toxic cause the drugs are chosen to be synergistic and in complementary pathways, so they don’t have to be at maximum tolerated dose. And adverse events go way down as you back off.
Benefits
- Virtual Trials can search the combinatorial space of treatment combinations far more efficiently than traditional trials because:
- No start up costs – the platform is always on, monitoring all patients, on all treatments all the time. Physicians are similarly engaged. ‘Always on’. New treatment hypotheses can be injected at any time.
- Fewer patients required for trials: – far fewer on average than with individual drug trials, because all hypotheses are tested in parallel, and patients are assigned to treatments dynamically based on the latest clinical results. Successful treatments are rapidly replicated and failures rapidly quashed.
- Unlike traditional cancer research, virtual trials have the potential to make an immediate difference for today’s patients, because we don’t need to invent new treatments; we just need to optimize the use of currently available treatments — off label and in multi-drug regimens; and do prospective experiments to determine which treatments work in which patients.
- Together, we can collectively search for optimal treatment regimens far more efficiently than we can as individuals.
Sustainable business models
- Pediatric drug vouchers
- Create investment vehicles that can provide funding to biotechs out of money to make their drugs available to the Virtual Trial.