According to a new study in JAMA, there is no evidence that the majority of the U.S. Food and Drug Administration (FDA) approved cancer drugs over the last five years work. The FDA used substitute measures, laboratory measures rather than clinical evidence, to approve two out of three cancer drugs, without any indication that the drugs either improve health or prolong life.  Indeed, in follow-up studies, 86 percent of drugs approved through substitute measures showed no clinical benefits.

In the case of bevacizumab, a breast cancer drug treatment, the FDA allowed the drug to go to market based on substitute measures that showed “progression-free survival” or PFS. But, a later study showed substantial toxicity and no improvement in life expectancy. And, the FDA ended up removing authorization to market it.

In 2009, the U.S. Government Accountability Office took the FDA to task for failing to do postmarketing studies on all drugs that are approved using substitute measures. The GAO reported that the FDA had plans to improve oversight, but it was not prepared to say whether those initiatives would be effective. According to the GAO, there are no specific conditions that require the FDA to speed up the withdrawal of a drug from the market if a drug company does not conduct a follow-up study in a timely manner or if a drug is shown not to be beneficial in a follow-up study.

The drug companies prefer getting drug approvals using substitute measures that do not focus on efficacy or health outcomes. Substitute measures allow drug companies to bring drugs to market faster and at less cost. When the drug companies use clinical measures that look at health outcomes, they must spend more time and money on their trials.

The problem is that substitute measures may show good performance when in fact health outcomes are not any better.  Health complications may also not be evident through the substitute measures. So, drugs approved using substitute measures may actually hurt patients.

The price of a new cancer drug can easily be $10,000 a month, regardless of either the drug’s worth or its cost of development.  The cost of cancer drugs can literally bankrupt a person with cancer.  In 2014, the least expensive new cancer drug approved cost more than $120,000 a year.

Today, the drug industry helps cover the cost of the FDA’s work. And, we are left to wonder whether the FDA’s independence has been compromised at the expense of patient safety. At the same time, the public may wrongly assume that FDA approval means the drugs are safe.