I'll be writing more about ASCO revelations in subsequent posts, but when it comes to cancer drug discovery we're living in an incredibly exciting time. Researchers are gaining an increasing understanding of the underlying "circuitry" that drives cancer cell growth and are coming up with many new targets to attack - particularly with monoclonal antibodies that bind to cancer cells and (for the most part) avoid collateral damage to other tissues and organs associated with traditional cancer chemotherapy and radiation treatments.
There's definitely a push-pull mechanism at work driving advances in both cancer research and clinical care. At the research end, decades of federal funding (through the NIH and NCI) have greatly improved our understanding of the basic biology of cancer, creating a "target rich environment" for biopharmaceutical developing innovative new medicines.
At the "pull" end, the FDA has been (generally) good at using regulatory tools like accelerated approval and Fast Track to get promising new cancer treatments to patients. New targeted biologics treatments for cancer also command premium pricing in the market, and face little or no generic competition (although with the FDA developing a new biosimilars pathway, that is likely to change over the next several years). And even where good generic cancer drugs are available, they are often used in combination therapy with newer medicines, which also helps to blunt pricing pressures for new drugs. (The situation is somewhat different in Europe, where "health technology assessment" and drug price negotiations can slow the uptake of innovative medicines. If the U.S. adopted similar strategies, incentives for cancer drug innovation would certainly weaken.)
Still it's not surprising to read that (hat tip: Pharmalot) cancer leads the way in biopharma deal making, with almost a quarter of all new deals in 2011, and that (according to PhRMA) there are almost 1,000 new compounds in development to treat cancer in the U.S.
That's all good news.
But if we're ever going to successfully transition from adding a few months or years to cancer patients' lives to effective long term control of complex cancers (akin to how we now treat HIV) we're going to have to re-think how we develop, test, and approve new cancer drugs.
Cancer is a disease that is, literally, evolution in real time; a networked disease that will, given enough time, mutate to develop one or more pathways around even our most powerful targeted drugs. For instance, cancers can develop additional growth promoting mutations that blunt the impact of targeted drugs; even when one particular signaling pathway is blocked, a second, or even a third tumor promoting pathway can come online; or the disease can metastasize to parts of the body (like the brain) where many drugs cannot reach.
So, to be successful, cancer combination therapies ideally should:
Begin as early as possible. Early detection is key to controlling and curing cancer. Better diagnostics (perhaps through blood tests for circulating tumor cells or other assays) would help identify cancers early in their progression and guide treatment decisions to improve outcomes.
Attack multiple growth pathways simultaneously, as we do with HIV therapies, to limit or even halt the development of drug resistance. Some new cancer treatments may, in fact, turn out to be useless on their own - but will also be key blockers of second or third line signaling pathways. So we'll have to think not so much of shrinking tumors as we will of shutting down the molecular hubs that drive cancer cell growth through personalized drug cocktails.
Ramp up the immune system's ability to fight off or control cancer once the growth of the cancer has been halted. This will likely involve re-programming the immune system to recognize and attack cancer cells, defeating cancer's ability to evade the immune system.
In other words, we want to catch the cancer when it's young and vulnerable, pummel it with a stream of targeted drugs, and then unleash the immune system to keep it under long term control. Under that scenario, it might be possible to shift from more aggressive treatments to maintenance therapy with occasional blood tests to ensure the cancer remains in check.
But how do we get from here to there? It won't be easy.
Developing combination therapies of investigational products faces many hurdles - from IP to uncertainty over what standards the FDA will use to approve novel products that are tested in combination. Still the FDA has created guidance (2010) for companies testing investigational combination therapies, and important new trials (like I-SPY 2, and BATTLE 1 & 2) have provided proof of concept that such trials can be executed successfully.
FDA has also recently announced a new endpoint for accelerated approval for new drugs to treat breast cancer that could streamline trials even further. Neutral third parties - like the Foundation for the NIH - can also play a pivotal role in accelerating pre-competitive research, exploring what combinations work best. There's certainly more to be done in this space, but initial results are promising, and the FDA and industry can build on successful experiences in other therapeutic areas, like testing combination HIV therapies.
When it comes to cancer clinical trials, the challenges are thornier. Only about 5 percent of all cancer patients enroll in cancer clinical trials. Even assuming that many of the hundreds of compounds in testing today turn out to be ineffective, it will be impossible to test them all in combination therapies given today's slow, bureaucratic, and hidebound clinical trials system.
We also have a long way to go in improving biospecimen collection (tumor tissue, etc) that can not only improve our understanding of cancer's molecular networks, but help guide patients into the right clinical trials. Clinical trials will have to be as flexible as the cancers they're fighting, and right now they're not. Given the tailored nature of many new therapies, it should be possible (at least in theory) to run very fast trials in small numbers of patients whose cancers have the specific combination of mutations we're trying to attack, but first we have to make sure we can find those patients (and their doctors), and get them enrolled quickly.
Finally, not every patient lives near a Memorial Sloan Kettering or M.D. Anderson Cancer Center. Instead, most cancer patients will be treated by community oncologists, who don't have access to the latest and best decision support tools, cancer genomics screens, and most up to date information on cancer clinical trials. This has to change. There's no reason, in the heyday of cloud computing, that cancer patients across the country shouldn't benefit from state of the art cancer therapy protocols and cutting edge clinical trials. And the more patients we can enroll in rapid "virtual" clinical trials and collect information on outcomes, the more we can cycle that information back into the discovery and development pipeline to accelerate the discovery of new targets and improve the use of both new and old drug combinations.
The challenges are serious, but far from insurmountable, although it will require rapid coordination on several fronts simultaneously to make the kind of progress that we know is possible in the war on cancer.
How to do that? One approach would be to create a competitive framework for cancer innovation networks based around leading medical centers and:
Creating a virtual network of cancer clinical trials to rapidly increase enrollment through targeted IT initiatives and reduced paperwork. Clinical trial enrollment should be based on a cancer's molecular signature, and exclusion criteria should be sharply reduced so many more patients are eligible for clinical trials.
Shift the FDA's role from approving new drugs to validating the biomarkers, algorithms, and surrogate and clinical endpoints (in collaboration with NIH and NCI, academic medical centers, patient's groups and industry) used to identify and prioritize new molecular targets and drug combinations. From conception to marketing, Pfizer's crizotinib still took about six years to reach patients. That's historically fast, but still too slow. New combination therapies should be tested as early as possible, when they have the best chance of actually controlling the cancer. Thankfully, the FDA has already signaled that it's willing to embrace new approval pathways for breakthrough products, for instance breast cancer drugs that produce complete pathologic response (pCR) when used in the treatment of early stage triple-negative breast cancer. This may provide a paradigm for using novel therapies in high risk cancer populations, when the prognosis is poor or the risk of drug resistance is high. Confirmatory trials are still required, but the initial trials could be completed in a matter of months in a minimal number of patients if the drugs are successful.
Reward companies (large and small) that bring new therapies to market with companion diagnostics. Much as we do now to encourage pediatric drug testing, companies that develop (or cooperate in developing) companion molecular diagnostics should be granted an additional six months of patent protection. This would also be a boon for the diagnostics industry, which would benefit from an infusion of industry cash.
There's a lot more we can do to advance cancer care (and plenty of other problems to overcome), but focusing on rapidly developing and validating combination therapies would accelerate the entire endeavor. We can do this by enhancing the use of virtual trials based on molecular diagnostics, shifting FDA approval decisions based on targeting known hubs of cancer promoting pathways and surrogate/clinical endpoints, and expanding patent protection for drugs developed with companion diagnostics.
This isn't a radically new approach; the core can already be found in ASCO's excellent blueprint document, released in November 2011 and legislation like the MODDERN Cures bill. The only thing that is lacking today is a sense of urgency and leadership from Washington.
We don't have to remake the entire U.S. cancer care system overnight. But we should be able to ramp up a prototype system that brings together all of the stakeholders and generates proof of concept. Assuming that it works, the rest of the system will rush to embrace what works. Or risk getting left behind as reimbursement and research funding gravitates to the true industry leaders.
The technology is moving so fast today we finally have the ability to beat cancer at its own game. Evolution, after all is comparatively slow. The key growth promoting pathways of cancer evolved over millions of years - and they probably number in the scores, rather than the hundreds.
Our researchers, our clinical trials, and our companies can be retooled to become much more nimble than even the most ferocious cancer.
That is, if we let them.