Margaret Dunkle, George Washington University, 2011
I am Margaret Dunkle, Senior Research Scientist at the Department of Health Policy of The George Washington University. I also direct the Early Identification and Intervention Collaborative for Los Angeles County. Among other awards, I have received the American Academy of Pediatrics’ Dale Richmond Award for outstanding achievement in the field of child development; and Vice President Al Gore’s “reinventing government” Hammer Award.
I have testified nine times before Congress on issues affecting women, children, and families and have authored more than 100 publications on these subjects.
I am here today solely because of my interest in ensuring that medical products for women be proven to be safe and effective. I am not representing any organization. I am not being paid or otherwise compensated.
The 2006 FDA approval of silicone gel breast implants was controversial because there were so many unanswered questions about the long-term safety of implants, including when they might break and leak inside a woman’s body, and what the consequences of such leakage might be.
“As conditions of approval,” the FDA required that the two implant companies (Allergan and Mentor) each study at least 40,000 women. The FDA also required that these studies be at least 10 years long since some serious adverse outcomes might not become obvious, or even be measurable, in just a few years. And it takes both time and a large sample to identify subpopulations of women who might be at greater risk for problems, such as autoimmune reactions, or reproductive or connective tissue concerns.
It is now five years later, when the studies should be close to half done. But the data for both the Allergan and Mentor studies are not what they need to be for a study of this importance.
Why do I say this? Because of the extraordinarily high “loss” of study participants, as the following table shows.
In just three years, Mentor (now a subsidiary of Johnson & Johnson) has somehow managed to “lose” four out of five of both the silicone-implant patients and the augmentation patients enrolled in its 10-year study. Their paltry overall enrollment rate, with only 21% to 29% of patients still in the study, makes their data virtually useless in terms of assessing either safety or effectiveness.
As you can see, Allergan’s rates are better, but still likely too low to meet the study’s goals. Simply put, when you lose half or more of your sample, you cannot assume that the data for the few left in the study also represent those who were “lost.”
The Mentor rates, especially, would be laughable, were it not for the seriousness of the issues. Let me put it this way. If I can easily find my antediluvian high school classmates with a few “advanced search” clicks on Google, how can it be that these well-resourced medical-device companies cannot keep track of much younger patients that they have already enrolled and consented into their clinical trials? If the will were there, there are many sound methodological ways to ensure high rates of ongoing participation in these studies.
So, what to do about this bad situation?
The FDA’s job is to protect public health by ensuring that such things as medical devices implanted into women are both safe and effective. My five recommendations are aimed at helping the Agency do just that job.
First, rescind approval for the Mentor implants. They did not uphold their part of the bargain for the post-approval study. If they can’t competently do this research to ensure their product’s safety, they should not be putting their product into the bodies of more than a hundred thousand women every year.
Second, do not restore approval of the Mentor implants until they have at least two years of credible data with 80% or more retention of clinical-trial recipients. They could accomplish this by re-enrolling or “finding” the thousands of women they have “lost to follow up.” To ensure credibility, Mentor would need to pay an independent entity to check the validity of their data.
Third, extend the duration of both the Mentor and Allergan studies to 15 years. The FDA’s own “Update” report indicates that 10 years may not be long enough to identify long-term problems from cancer to rarely diagnosed autoimmune diseases, or to detect populations especially susceptible to adverse outcomes.
Fourth, require ongoing transparency and disclosure of the data from these studies by making public meetings such as this, and the June 2011 “Update” report, the rule, rather than the exception.
Fifth, implement a broad FDA policy to put companies on notice that there will be consequences of not doing or incompetently doing FDA-required post-approval research. Consequences could include withdrawing a product from the market, or substantial financial penalties for doing a bad job of the research, and then using that “penalty money” to fund research controlled by AHQR or independent researchers.
In conclusion, with almost 400,000 women getting breast implants each year for augmentation or reconstruction these studies need to be done right. The current situation also provides a window for the FDA to improve the incentive system so that manufacturers do not again ignore post-approval research requirements to the detriment of the public’s health.
Thank you for this opportunity to comment. And I ask that my full written statement be inserted into the meeting record.