October 28, 2022
I’m Dr. Diana Zuckerman, president of the National Center for Health Research. We scrutinize the safety and effectiveness of medical products, and we don’t accept funding from companies that make those products.
My expertise is based on post-doc training in epidemiology and public health, my previous policy positions at Congressional Committees with oversight over FDA, my previous position at the US Department of Health and Human Services, and as a faculty member and researcher at Harvard and Yale.
This is a terrible disease and the personal stories are very important. What is the evidence that this treatment actually works?
The sponsor has provided data from 2 studies:
- A single center, single arm trial (Study 03-133) included 94 pediatric patients ages 0.9 to 13 years with CNS/LM relapsed neuroblastoma who received intracerebroventricular infusions. Three doses were used: 25 mCi, 33.5 mCi, or 50 mCi, based on age. The Primary Endpoint was overall survival and the results showed 45% alive at 3 year.
- The second study is not completed yet. The Interim report of this multi-center single arm study (Study 101) shows 20 patients with imaging evidence of CNS/LM disease. The Sponsor reports 7 responders, but 3 of these are not confirmed. That matters because transient responses are not necessarily meaningful, and therefore can’t be considered evidence. The Primary Endpoint is 3-year overall survival, but those data are not yet available.
Safety
It is important to note that 19% of patients permanently discontinued using the drug due to an adverse reaction in Study 03-133 and 28% of patients in Study 101. Most were due to myelosuppression. However, 3% were due to chemical meningitis (3 cases in 03-133 and 1 case in Study 101), and there was one case of fatal intracranial hemorrhage
Shortcomings of the Studies
Only one study was completed and it wasn’t randomized or blinded or controlled. The external control group was very small and they differed from the patients receiving the treatment, because the latter had more intensive prior treatment. There were significant population differences as well as treatment differences between the treatment group and the external controls. In addition, overall survival for this condition has improved since the control data sample was collected.
Therefore, we agree with FDA’s conclusion that differences in overall survival can’t be “reliably attributed” to the drug. We also agree that “the application does not include reliable response rate data to provide supportive evidence.”
In Study 101, no patient demonstrated a response that can be unequivocally attributed to the drug. And there was no overall response rate calculated for Study 03-133 and only very limited overall response rate data in Study 101.
FDA has been the Gold Standard? Do these data meet that standard?
The FDA Summary states: The comparator is too dissimilar to the treatment group. There is no reliable information on tumor response rate. Therefore, the submitted study cannot be considered an adequate and well-controlled trial necessary to establish effectiveness. That’s the standard that the law requires for FDA approval.
On the other hand, there is an unmet need. These children need treatment. Should FDA be flexible even though the data are clearly inadequate? Why isn’t this an accelerated approval application instead of a regular application? Are the data even good enough for accelerated approval?
Are FDA Decisions Perpetuating Poor Studies?
Why didn’t the sponsor conduct a randomized, double blind controlled trial? This isn’t a rhetorical question. What’s the incentive for any company to conduct a well-designed study if a poorly conducted study with questionable findings results in approval? Those controversial decisions set a dangerous precedent.
Unfortunately, when FDA approves a drug based on inadequate data, all companies lose the incentive to conduct well-designed studies. Not just the company whose product is approved, but all other companies as well.
Patients Deserve Better
We do patients and their families no favors by approving treatments that aren’t proven to work.
FDA’s Expanded access program is the way to give patients access to experimental drugs, because it provides free treatments in a well monitored situation where the patients and families know that they are taking a risk using an experimental drug. Why should patients pay to take an experimental drug, thinking it is proven safe because it is FDA approved?
I can’t believe I have to say this: Without an appropriate control group, it is not possible to provide evidence that patients and doctors need to make informed decisions. Even a small study with a small well matched control group is better than nothing.
Unfortunately, the preponderance of evidence doesn’t support approval for this drug.
The Advisory Committee agreed with our assessment and voted 16-0 that the studies did not prove the drug improved overall survival. The FDA subsequently rejected the application.