In late December, the Food and Drug Administration (FDA) revoked approval of the cancer drug Avastin for metastatic breast cancer. The decision set off a firestorm of reaction: the right condemned the denial of a potential life-saving drug for breast cancer patients, while the left cheered the withdrawal of an expensive drug that seemed to offer little hope of improvement to patients based on clinical studies. Both reactions missed the larger point: rapidly advancing diagnostic technologies should make wholesale drug withdrawals a thing of the past. Subgroups of patients who benefit from specific cancer treatments would still have access to them, while the rest can seek other alternatives. But to get there, the FDA, the National Institutes of Health (NIH), and private companies must aggressively work together.
The tragic complexity facing cancer patients today is this: more and more, it’s becoming clear that there is no one-size-fits-all treatment for people suffering from the very same diagnosis. Avastin was a case in point: some patients responded well to the drug, while others had no response or even life-threatening side effects. No one – at least yet – seems to know why. Withdrawing the drug’s approval for breast cancer treatment, though, is an archaic approach to a new reality. The key is to find ways to distinguish responders from non-responders, and get the drug as rapidly as possible to the patients who are most likely to benefit.
Today, the FDA basically supervises this process from the top down, with drug companies conducting the trials. There have been some notable successes in recent years. Data shows that Iressa, for lung cancer, is most effective in Asian non-smokers. Herceptin, a breast cancer drug, is targeted at the 25-30% of breast cancer patients whose disease over-expresses the HER-2/neu protein. Two new drugs for metastatic melanoma are known to be most effective in patients whose hair turns stark white after 6-12 months of treatment. In that subset of patients, the cancer has either been stabilized or shrunk to undetectable levels – an incredible advance over existing therapies.
The problem is that the FDA’s top-down approval process is no longer adequate to the complexity of the diagnostic landscape. What we need, instead, is a bottom-up model of experimentation and approval.
New diagnostic tools could speed and improve the use of tailored cancer therapies by creating a rapid feedback system that would allow physicians to detect cancers at earlier stages, pinpoint sub-groups of patients, and determine quickly if they’re responding to treatment. Rather than waiting for the FDA to approve new ways to use drugs after laborious and time-consuming trials, physicians could quickly tailor drugs to patients as they go, learning along the way. This approach could advance personalized medicines for cancer patients from a dream into a reality – and not in ten years, or even five, but in two or three.
The Many Faces Of Cancer
Cancer is among the most complex human diseases. What we call breast cancer is actually not one disease, but at least four. And the same seems to be true for many other forms of cancer. In fact, cancer is a genetically unstable disease, giving it the power to metastasize, evade the body’s immune defenses, and develop resistance to chemotherapy. This is why cancer drugs are often used in “cocktails” to try and knock out the disease before it can adapt.
Blood tests to detect circulating tumor cells (CTCs) are one of the most promising, though still emergent, technologies in the cancer diagnostic field. These tests work in the same basic way as panning for gold in a river. Doctors take a small tube of blood from a patient during a routine office visit and scan it for CTCs. They use technologies that sift out cancerous cells by one of several different processes that rely on cancer-cell surface-protein expression, tumor-cell size, or other characteristics.
In principle, the tests may give doctors a quick and inexpensive way to screen for cancer that can replace expensive or invasive tests like mammograms or colonoscopies; to tailor cancer treatments and adjust them based on how many or what type of CTCs are found in a patient’s bloodstream; or make physicians more comfortable adopting a “watch and wait” approach for elderly patients in cases where the underlying cancer may grow so slowly that it will never become life-threatening.
Consider a disease like pancreatic cancer, a rapidly fatal affliction with few early symptoms. Validated CTC tests could identify pancreatic tumors when they are still operable or identify tumor types that would be likely to respond to treatment. Both would be astonishing advances for patients.
Unfortunately, there are a host of unanswered questions in the CTC field. Different companies use different technologies to sift out CTCs, and there is no common agreement on what the different “scores” mean, or what CTCs indicate about the underlying cancer. Until these issues are sorted out, FDA regulators will be loath to approve new drugs developed using those technologies, or allow them to be marketed directly to consumers and doctors. And even though the FDA created its Critical Path Initiative in 2004 to address basic questions like these, the agency’s budget for advancing regulatory science initiatives remains paltry – about $20 million, far less than the budget for a single large cancer clinical trial. In 2007, Congress created the Reagan Udall Foundation to help advance innovative drug-development technologies – but then repeatedly withheld funding out of the wrongheaded belief that anything that speeds up drug development is a giveaway to the drug industry and would imperil patient safety.
The Way Forward: The Key Role Of The NIH
So how can we move forward to enable a more nimble and effective diagnostic, treatment, and approval process? Here is where the National Institutes of Health must play a leading role. With a budget of over $30 billion (compared to the FDA’s budget of about $4 billion), the NIH must step forward and offer itself as a “neutral ground” of pure objective research (whereas the FDA is often caught up in controversies and accusations that it’s too cozy with drug companies). The NIH should collaborate with drug and biotech companies to establish core standards for evaluating and validating new CTC diagnostics. Once these standards are established, the FDA can then enforce them.
By working closely with industry and academic experts to set basic, common standards for CTC tests on a cancer-by-cancer basis, the NIH could empower a wave of cancer diagnostic innovation. Improved diagnostics would also further the NIH’s mission by helping translate discoveries in basic cancer biology into real-world treatments faster than is currently possible.
Cancer treatment is poised for the same burst of innovation that revolutionized the treatment of HIV/AIDS—where the development of blood tests that accurately measured the amount of the HIV virus circulating in patients’ blood led to rapid drug development in the field. Today, AIDS is a serious but manageable chronic illness with a robust pipeline of powerful drugs whose effectiveness is monitored using a relatively simple blood test. We have a chance to get to a similar place with many cancers, but only if the institutions governing drug development in the U.S. catch up with the science that has already helped so many patients stricken with this dread disease.