Elena Scotti/FUSION

Just over five years ago, Gardiner Harris wrote a profile of Dr. Richard Pazdur for the New York Times. Pazdur was—and still is—the man in charge of approving or disapproving new cancer drugs at the FDA: unless and until he approves it, you’re not allowed to prescribe it. And the NYT didn’t go easy on him: the headline was “Where Cancer Progress Is Rare, One Man Says No.”


Since then, three things have happened. One is that Pazdur’s wife died of ovarian cancer; another is that the FDA is getting faster at approving new drugs. And the third is that Harris has revisited his subject, for a new profile. This time, the headline reads “FDA Regulator, Widowed by Cancer, Helps Speed Drug Approval.”

So, what’s at stake here, what has changed, and did it really take Pazdur’s wife dying of cancer for things to start getting better at the FDA?


The big picture is that modern medicine has done an astonishingly good job of finding cures for formerly-fatal diseases, sometimes even eradicating them outright. But cancer, of course, is the exception. Which means that as people die of fewer and fewer other diseases, cancer is going to end up taking more and more of us—until such time as the fabled Cure For Cancer arrives.

For the time being, however, the hundreds of different pharmaceutical companies submitting cancer drugs to the FDA generally aren’t thinking that big. Most of them are putting forward drugs that may or may not help some relatively small subset of people with some very specific cancer. And it’s the FDA’s job to approve those drugs, if (and only if) they really do work.

The problem is, not only is it incredibly hard to determine whether or not a particular drug works; it’s hard even to come up with a useful and workable understanding of what it means for a drug to work.


There are two related issues here. The first is the idea of proof, in a world dominated by statistical extrapolations. Consider the classic double-blind clinical trial, where one group of patients takes the drug in question, and another group takes a placebo. How big do the differences between the two groups have to be, in order to prove that the drug works, especially considering that placebos are getting more and more effective, these days? And, secondly, what exactly are you measuring?

The criteria that allow a cancer drug to be approved by the FDA are significantly looser than almost any other drug. For one thing, cancer drugs don’t need to save lives, or even prolong them: because people with cancer, especially early-stage cancer, can live for many years before dying of the disease, it’s incredibly difficult to prove to any scientist’s satisfaction that an early-stage drug prolongs life. What’s more, drugs don’t even necessarily need to prolong life to be effective. Many people with late-stage cancer, for instance, would happily take a drug that was statistically expected to cut their life expectancy in half, if it had just a small chance of curing them of cancer entirely.


It’s easy to understand the stance that when someone is dying of a fatal disease, it’s silly to worry about whether any given drug is “safe.” What’s important is just to have a glimmer of hope; indeed, it can seem downright inhumane to deny a drug to a dying person—a drug they and their doctors might desperately want to try—just because some trials had shown adverse side effects in certain patients, or otherwise had failed to pass muster with the FDA. That’s why angry people often call Pazdur a murderer: let’s say that, thanks to the FDA, 1,200 dying cancer patients were denied a potentially life-saving drug. Even if that drug ended up saving only 1% of those patients, that would still mean the difference between life and death for a dozen people, not to mention the huge difference in outcome for their friends and their families.

The problem is that it’s not as simple as that: what makes sense on an individual level can be bad policy on a societal level. For one thing, of course, there’s the money issue: FDA approval can mean a windfall of hundreds of millions or even billions of dollars for pharmaceutical companies, many of which have only one drug in development. As Harris explains,

Because patients are desperate and insurers are often forced by law to pay, prices are soaring, making cancer medicines among the best ways for drug makers to make money.


Viewed through this lens, the FDA is the only agency preventing unscrupulous pharma companies extracting billions of dollars from the weakest and most vulnerable of victims. Your drug might do no good at all—it might, in fact, do measurable harm—and yet unless the FDA stops you, you can more or less name your price and sell the same drug, over and over again, to a never-ending stream of dying people. That’s great for drug companies, but it’s dreadful for any society, especially one trying to get a grip on healthcare expenditures that are spiralling out of control.

A very large part of the FDA’s job is to ensure not only that drugs are safe, but also that companies don’t sell dream-infused snake oil. Capitalism can be gamed: just go into business selling hugely expensive lottery tickets, and tell everybody that for every million people with cancer who buy a ticket, one or two will be cured. You’ll get a lot of takers, especially if insurance companies are forced to buy the ticket whenever a doctor writes a prescription. Only the FDA can prevent such evil business models from taking root and even flourishing.

Pazdur did very well in his job as last line of defense against companies trying to introduce drugs that would be extremely profitable even if they didn’t work. And he probably improved many patients’ lives, too, although few of them knew it. Cancer drugs are nasty things, and tend to have gruesome side effects. By not taking such drugs, cancer patients can end up with a significantly higher quality of life, and can even end up with a much more pleasant death. These things matter, a lot. The U.S. healthcare system is not generally set up to maximize your standard of living until you die; it’s much more likely to throw everything it can at you in an attempt to delay the inevitable. And again, Pazdur’s team at the FDA is one of the very few groups in the country capable of working for the good of the many, rather than for the interests of the few.



So, what happened over the past few years, beyond Pazdur getting an unpleasant first-hand experience of what life is like for terminal cancer patients and their families? The biggest difference between then and now seems to be that when the FDA does approve a drug, it does so more quickly. That’s got to be a good thing: if the FDA is going to approve a drug anyway, there’s no point in dragging its feet. And the agency really can move fast: in one case, the FDA received a drug’s trial results weeks before the executives at its manufacturer were able to see them. The result was that the drug was approved three months ahead of schedule.

The problems, of course, arise if the FDA cuts corners in its haste, or starts accepting drug-company pseudoscience. Pazdur claims that the increase in drug approvals is largely a function of increased funding, on the one hand, and better biotechnology, on the other: new genomic technology can demonstrate the effect a drug has much more easily than enormous blinded trials. Critics say that Pazdur has gone soft, and/or has become captured by the pharmaceutical companies.

Still, it’s worth taking Pazdur at his word when he says that he has “evolved from regulator to regulator-advocate.” And it’s obvious that his move has certainly been great for biotech companies. But, sadly, there’s no easy way of telling whether it has been good for cancer patients writ large. When you identify too much with the desires of individual cancer sufferers, you can end up causing unnecessary pain for patients as a whole. Sometimes, regulators really do know better than patients themselves what’s good for them. Which means that when regulators cease to be able to effectively stand up to the companies they regulate, all of us lose.