Stephanie Fox-Rawlings, PhD, National Center for Heath Research; December 12, 2018
I am Dr. Stephanie Fox-Rawlings and I’m glad to speak today on behalf of the National Center for Health Research. Our center conducts research and scrutinizes data to provide objective health information to patients, health professionals, and policy makers. We do not accept funding from drug and medical device companies, so I have no conflicts of interest.
The FDA says that the agency is willing to accept more uncertainty in the risk-benefit profile for a Breakthrough Device for patients who lack effective treatment options. However, even Breakthrough Devices are supposed to demonstrate a “reasonable assurance of safety and effectiveness.” The evidence in the clinical trials presented today do not demonstrate this.
In both clinical trials, patients are either treated with the new device or nothing. Since there was no sham or other special treatment to compare the device to, research tells us there was likely to be a substantial placebo effect. The effectiveness endpoints can all be influenced by patients’ confidence in the treatment and their desire to get well. Thus, the lack of blinding for patients and the lack of an alternative treatment for controls would be expected to greatly influence these results.
Despite those expectations, there are very small differences in these trials between patients receiving treatment and those who had no treatment. That small difference could be due to a placebo effect since the patients know they are getting a breakthrough treatment. Even with the advantage of the placebo effect, the two clinical trials do not provide reasonable assurance of effectiveness. The pivotal trial showed that the device was not better than nothing until post hoc analyses suggested a subgroup might benefit.
Post hoc analyses can provide exploratory data that is useful for designing future research, but are not statistically appropriate as the basis for FDA approval.
The confirmatory trial is presented as the main evidence for effectiveness. But here too the differences between implanted patients and those with no treatment are small enough that their clinical meaningfulness is unclear. The unusual clinical trial design and statistics further complicates their interpretation.
In addition, a disproportionate number of patients in the studies are white males. Congress has passed legislation on FDA’s need to demand diversity in clinical trials, and heart disease trials were a priority. Data primarily on white men may not be generalizable to the millions of women and people of color with chronic heart failure.
Over all, the clinical trials leave too many unanswered questions about whether the Optimizer actually works.
So what about safety? All surgeries pose risks. The device itself has additional risks. Those risks would be worth taking if there was a good chance of a benefit, but instead the benefit is questionable.
When patients are desperate, some people believe that it makes sense to approve a treatment as long as it doesn’t have major risks. But that means patients would be paying for a surgical treatment that might not work. It is more ethical to ask a small number of patients to participate in clinical trials where they know that the treatment is experimental. That is a better option than having larger numbers of patients pay to be experimental test subjects and may not even realize it.
In addition, we know that sponsors are much more likely to complete pre-market studies in a timely manner, while post-market studies are often delayed or poorly implemented. As a result, many post-market studies do not provide useful data about safety or effectiveness.
In conclusion, the Optimizer is intended to treat patients with chronic heart failure and who have few other options. I commend the sponsors for trying to find a novel way to treat these patients. But the real question here is whether the Optimizer actually helps patients, and if so, which patients? The evidence does not clearly demonstrate that it helps patients and doesn’t provide scientifically valid information about which patients are more likely to benefit. Another well designed and executed study is needed to demonstrate a reasonable assurance of effectiveness and safety.
Please consider that this device is the first of its kind. That means that a low standard for approval will set the standard for future devices. But a higher standard will ensure patients will benefit.
We all want to help these patients, but new treatments need to provide a real benefit. The questions that we need to answer are does this device help patients and should patients continue to have the option of receiving an experimental treatment as part of a clinical trial, or should they pay for an experimental treatment that has been described as a breakthrough?
The Circulatory System Devices Panel voted in favor of approving the device- that the data demonstrated that it was effective 11-2, that it was safe 12-1, and that the benefits outweigh the risks 12-0 (1 abstain).