Yesterday, the Oncologic Drugs Advisory Committee issued its recommendation to the FDA on the use of Avastin for the treatment of breast cancer: no. Based on clinical trial results, ODAC members voted unanimously that the drug is not safe and effective for the treatment of patients with this disease. I’ve written about this Avastin saga before, but this landmark vote is worthy of a post of its own.
There are many issues threaded throughout the FDA’s initial approval of Avastin and its subsequent rejection (that initial approval was, essentially, temporary, pending further trial results, which have proven to be unpersuasive) (also, the FDA has not yet officially rescinded its approval, but this ODAC vote basically solidifies that move) (and remember, the FDA regulates how pharmaceutical companies can market their drugs, not how doctors use them).
One of the most gripping issues is that of patient versus population. The ODAC hearings were stocked with patients who had benefited from the drug — or at least, a combination regimen that included Avastin plus chemotherapy. Knowing that there are individual breast cancer patients who might benefit from the drug makes it hard to say no to its approval. But approvals are based on studies, not individuals. Data drawn from a large population are what scientific evaluations are based on. That might seem like a cold, hard fact against the tears of a patient given precious extra months to live. But reliance on the statistics is founded in the scientific method: that the data give the facts, and that the larger the population, the more accurate the data. An individual’s response could be a fluke; hundreds or thousands of people show what the case actually is.
[That being said, there is growing interest in so-called outliers, patients whose responses lie at either extreme of the average: an extremely good outcome or an extremely bad outcome raises questions about genetic variations or other factors that are more easily measurable because of the extreme response. A topic for a later post.]
So, when it comes to patients versus populations, which way is right? Should the FDA reject a drug’s approval based on a study even though there are still patients who might respond? Or should a drug be approved on the basis that it might help some people, and that is reason enough to make it available? The FDA is charged with the job of the former, because only large studies can confirm a drug’s safety and efficacy. Still, it’s a good question.
In the end, it might not matter. Avastin is still approved for other types of cancer, and so doctors are free to use the drug “off label.” Insurers might not agree to cover the drug for breast cancer patients because of the FDA’s decision. However, many insurers follow the guidelines issued by the National Comprehensive Cancer Network (NCCN) and other such authorities. If the NCCN keeps Avastin in its list of recommended treatments for breast cancer, then some insurers might cover the $80K+ price tag. (Though of course there are some who will accuse insurance companies of somehow swaying the FDA’s decision so that they have a reason to not cover the treatment.)
Of course, none of these understandings will stop people from slinging accusations about death panels, which is a shame because all that noise misses the opportunity to have a real conversation about some vital, core issues, questions, and dilemmas.