Are Some Cancer Drugs Costing Too Much for Too Little?

December 12, 2016

Newer cancer drugs are being approved on the basis of shorter studies and on outcomes that might not predict living longer or feeling better.

Newer cancer drugs are being approved on the basis of shorter studies and on outcomes that might not predict living longer or feeling better, according to a new article published in JAMA Internal Medicine.

Investigators looked at 18 cancer drugs approved by the US Food and Drug Administration (FDA) between 2008 and 2012. All of these drugs were not shown in postmarket studies to significantly prolong overall survival (OS). 

The researchers found that the estimated annual costs of the 18 drugs ranged from $20,237 (rituximab) to $169,836 (cabozantinib-S-malate). They found that 13 drugs had annual costs that exceeded $100, 000 and the most expensive drug (cabozantinib) did not improve OS and worsened quality of life (QOL) compared with placebo.

“I was very surprised by the findings. Keep in mind, that these studies were conducted and funded by the companies whose products were being evaluated, not by outsiders. I was even more surprised when I talked to FDA officials and heard from officials from some of the companies involved. They were all very defensive, saying essentially well even though the data so far show that the drug has no benefit for overall survival or quality of life, we think that it really does have a benefit," Diana Zuckerman, PhD, who is the President at the National Center for Health Research, Washington, DC, told OncoTherapy Network.

Zuckerman and colleagues searched PubMed for clinical trial data and analyzed all peer-reviewed findings and FDA review summaries for QOL. They also calculated the annual costs for these. “They keep analyzing the data in different ways until they finally find something that looks more promising. For example, maybe a small genetic subset of patients do a little better, but the label won't be that specific, it will say the drug is approved for all patients with lung cancer, or colorectal cancer, or whatever,” said Zuckerman.

The PubMed search identified 466 articles; however, only 31 met the study inclusion criteria of a clinical trial with a comparison group using a validated QOL survey instrument. They also analyzed FDA review summaries, but publicly available summaries were only available on 15 of the 18 drugs. The QOL studies compared seven drugs with placebo or observation groups and found that two drugs demonstrated worse effects on QOL. The remaining five drugs showed no statistical difference or had mixed results.

The researchers question whether drugs that are costing this much money should retain FDA approval and remain on the market when postmarket studies show they no clinically meaningful benefit compared with placebo or observation.

“The most important message for oncologists is to check out the data on new cancer drugs for themselves whenever possible. Cancer patients care about living longer and feeling better, and that is more important than whether the tumor shrinks for a while,” Zuckerman. “We encourage oncologists to look at results for overall survival not progression-free survival.”