This is the 22nd in a series of interviews between CFAH President and Founder Jessie Gruman and patient and consumer group leaders about their experiences with and attitudes toward comparative effectiveness research (CER).
Gruman: Tell me about yourself and your organization.
Marty Tenenbaum: I started Cancer Commons in response to my own situation: A decade ago I was diagnosed with metastatic melanoma, a disease that had no effective therapies and a dire prognosis. I initially consulted several local oncologists, all of whom had different recommendations. I then undertook a nationwide search for clinical trials, several of which appeared promising. But no one could tell me which treatment or trial was likely to work best for me. I gathered all the information I could and then guessed, betting my life on a vaccine trial that ultimately failed to demonstrate efficacy in a large multi-center clinical trial. Still, some participants in that trial responded well, and I was fortunate to be one of them.
Why did the vaccine work for me and not for others?
This was the impetus to establish Cancer Commons. Thanks to the genomics revolution and related technologies, such as computational and systems biology, we now have the means to answer such questions, and, more importantly, to predict which treatments are likely to work best for an individual cancer patient.
To help patients get the best possible outcomes, we plan to collect data on treatments and outcomes from as many patients as possible (starting with lung cancer and melanoma) and correlate it with those patients’ biomarkers. In effect, we’ll be doing CER one patient at a time, rather than through clinical trials. Of course, we’ll also take clinical trial information into account, but since every cancer is unique at the molecular level, the population statistics from trials are of limited value for patients at the tails of the bell curve. The clinical results from these “N of 1” experiments will be used to continually improve our database of biomarkers and treatments, where feasible, participating scientists will also try to validate whether recommended therapies had the desired effect through post treatment biopsies and molecular analysis.
Gruman: Where do you see CER fitting in the effort to improve the effectiveness of health care?
Marty Tenenbaum: In cancer, there have been several well-publicized instances in which a small fraction of patients with a given tumor type (sometimes less than 1%) respond to an available drug or investigational agent. Off label use of drugs that might work for one patient are not paid for by insurance. In CER for cancer, the data comes from large trials. We believe it’s more effective – and increasingly possible – to get really granular data for the individual patient. Only then can it be determined what is best for him or her.
Given that every cancer is unique and there are hundreds of new targeted therapies in the pipeline, which will likely be used in cocktails, there is no way to systematically test all combinations of drugs on all patients. Traditional trials are slow, expensive and often fail with most patients. There has to be a better way to match patients and drugs. We think our approach is that better way.
Gruman: Tell me how your organization views the relationship between CER and patient-centered outcomes research (PCOR).
Marty Tenenbaum: What we are doing constitutes evidence-based PCOR done in an extremely deep way. For this to work, lots of patients have to be engaged as full partners in the trenches with the researchers who will work with them. Our approach appeals to patients’ own interests to have the best scientists in the world looking at the data from their tumor. We also promise people that if someone finds something with their data that is useful to them, they will be the first to know. If there’s a trial that’s put together with their data, they’ll be first in line for that study. Such a close partnership requires trust and openness. We and others have consistently found that patients with advanced disease are less concerned about privacy than they are about staying alive.
Gruman: Can you give me an example of how your constituents have been affected by CER?
Marty Tenenbaum: As I mentioned, my life was apparently saved by a vaccine that failed its trial but appeared to help some patients. Unfortunately, we’ll never know because the manufacturer destroyed the remaining stock to mitigate liability concerns.
You can read about similar cases in the news all the time: a drug trial fails but a few of the patients seem to have benefited; or an approved targeted drug like Gleevec and Herceptin, for example, benefit a few patients in other cancers that happen to have the right mutation; or a drug like Xalcori, approved for a subset of lung cancer patients with an ALK transformation is found to benefit an even smaller number of lung patients with an ROS1 mutation.
Gruman: Is the technology available today prepared to do this kind of computational biology?
Marty Tenenbaum: Many researchers would say not yet, maybe in 5 years. I say, “Tell that to a patient.”
The reality is that doctors routinely use molecular testing to guide cancer treatment. The only way we can accelerate learning is to capture the methods, the reasoning and the outcomes. We are now recruiting researchers, patients and clinicians to participate in rapid learning communities, and we’re creating a rapid publishing platform to immediately disseminate what we find.
Gruman: You identify Cancer Commons as a patient-centric organization.
Marty Tenenbaum: We’re dedicated to helping patients achieve the best possible outcomes by pushing the boundaries of science with compassion and “hair on fire” urgency. Patients help us break through structural barriers that impede cancer research by, for example, donating their data for collaborative analysis, and participating in small, proof of concept studies of highly promising drugs. Not all patients will want to participate in such studies, which might involve sequential biopsies after each treatment. Our contribution is to tell them about options. Their contribution is to tell us how it worked.
Although we do have the ability to recruit patients directly, to date we’ve mainly partnered with patient advocacy groups, who encourage their members to participate in PCOR through Cancer Commons. Major partners include the Melanoma Research Foundation and the Bonnie J. Addario Lung Cancer Foundation.
Gruman: Some professionals believe that patients are opposed to CER. Do you think this is accurate? If so, what do you think is behind this view?
Marty Tenenbaum: Patients aren’t opposed to CER. They don’t like to participate in clinical trials, sometimes because they’re ill-informed, often because they don’t like the idea of being a human guinea pig. Unfortunately, many professionals believe that only randomized controlled trials provide reliable evidence for CER.
We now realize that cancer is hundreds, if not thousands, of unique molecular diseases. With hundreds of new targeted therapies and diagnostics in development and the likelihood that combination therapies will be required in most cases, there simply isn’t enough time, money, patients or specimens to sustain the testing of mono-therapies on large groups of patients, who likely have many different sub-diseases.
We need a smarter approach, one that exploits genomics to generate much more information from every patient, particularly information about which drugs are likely to work best in specific patients. Such an approach enables drugs to be tested with far fewer patients. It lets us leverage what we learn at the molecular level across many cancers and rapidly develop new off label uses for targeted therapies that are approved. Most importantly, it provides each patient with the optimal treatment for them, removing a major psychological – and in our opinion, ethical issue – that discourages patients from enrolling in traditional trials where the primary objective is not to cure them, but to validate a drug or diagnostic.
Gruman: What are your fears and hopes for CER?
Marty Tenenbaum: My biggest hope? I don’t expect that technology can cure cancer, but I believe that we can significantly improve outcomes for a large proportion of today’s cancer patients by matching them with the right drugs and learning as much as we can from their responses.
There is a large disparity of information across the medical world. If you consult 6 doctors, you’ll likely get 6 opinions about how to treat your cancer. And 5-year survivals may vary as much as 50%. This is inexcusable. The problem is solvable, however. The data from the treatment of many patients can be aggregated in order to inform patients how best to respond to their individual cancer.
My biggest fear is fighting the bureaucratic barriers to this approach. But many patients are willing to share their data and participate in proof of concept studies in exchange for a better chance for a cure. We’re hoping that an army of a million patients can move these big institutions.
More CER Interviews by Jessie Gruman