New study shows FDA's loose cardiac device regulation led to significant recalls
Current device research gets a bump to the front of the regulatory line and suffers from poor methodology
Few fields have had such a great advancement from the biotech side as cardiology over the past half-century. PCI and catheterization-based techniques revolutionized the space in the 1970s and 80s. As devices for minimally-invasive interventions continued to improve, TAVR and other procedures began dominating and changed the conversation surrounding who needed open-heart surgery. Each of these breakthroughs showed very real improvements on patient outcomes. The latest bunch that have joined the list include leadless pacemakers, subcutaneous ICDs, and wearable event monitors. This does not even include the large strides made in cardiac imaging modalities to better understand heart disease.
However, I came across a recent Annals of Internal Medicine article which detailed the poor state of our current regulatory apparatus surrounding cardiovascular medical devices. This is incredibly unfortunate, as these devices were designed to be employed in the most emergent and complex of situations and should merit even stricter approval requirements than drugs.
In the study, published by a group out of UCSF led by internal medicine intern Dr. Claudia See, 137 Class I FDA recalls from 2013 - 2022 that affected 157 cardiovascular devices were analyzed. The results elucidate several key themes surrounding the research (or lack thereof) behind devices that underwent recall along with regulatory gaps in the FDA’s current approval process. Let’s go through them one by one.
Lack of pre-market testing - Of the recalls, only a paltry 19.1% of devices had undergone any pre-market clinical testing. Why is this the case? According to our current regulatory process, devices that are considered moderate-risk (Class II) or lower are not required to undergo premarket clinical testing if they can demonstrate that their technology and indications for use are similar to an approved device that is already on the market. In other words, if you reengineer a current device for the same purpose, you essentially get to cut to the front of the regulatory line without conducting any new study, either against the currently-used device or against alternative standard of care practices. This is not the same as in drug approval, where new medications of the same class or mechanism of action are subject to the same clinical trial process as the original. Say you want to invent a new beta-blocker - you’re going to have to conduct nearly the same amount of study as all the beta-blockers that have been approved before.
Poor pre-market testing methodology - Of the studies that did undergo pre-market testing, almost 80% utilized surrogate endpoints and not patient outcomes. A number of others utilized non-blinded and nonrandomized trial strategies among small sample sizes with no control and poor follow-up. Perhaps the most egregious example of this is the VIRTUS trial, which led to the approval of the Vici venous stent for symptomatic iliofemoral venous insufficiency in 2019. This was a single-arm study with only 170 patients where the primary outcome was simply venous patency at 12 months. Little attention was given to more clinically meaningful endpoints such as improvements in symptoms and quality of life for affected patients. Vici went on to be recalled in 2021 due to reports of stent migration. Another example is the Valiant EVO trial, which led to the approval of the Valiant thoracic endograft for descending aortic aneurysms in 2018. Once again, this was a non-randomized study of 87 patients that were followed-up for only 30 days and included no comparison to current standard of care. Valiant was later recalled due to reports of endoleaks and stent fractures.
Lack of post-approval surveillance - Under our current structure, only high-risk (Class III) devices are required to undergo any post-approval study. In the article, many of these devices had not even met this requirement. Dr. See finds that 14% of studied devices were required to have post-approval investigation, yet almost two-thirds of these had not completed this requirement or reported delays in publishing post-approval research by the time of recall. From what I can glean, there are surprisingly no penalties (fines, pulling from market etc) for reporting delays. It is almost certain that in this timeframe, additional units were sold and utilized in patients. Remember, these are high-risk devices. In my opinion, post-approval testing should be required for moderate and high-risk devices (Class II and III) and delays in meeting these requirements should be met with severe penalty - including immediate cessation of use. This is the only way to disincentivize “gaming the system” with devices that do not work.
An additional word on the first point - some might argue that creating a stricter approval process for “equivalent” devices increases the ability of monopolies to form in the market. While this is certainly true, it does not seem very fair for market competitors to essentially “piggyback” the research of already-approved products claiming equivalency in order to reach market status faster. After reading this study, monopolization of the market also became less of a concern for me - I would rather have a monopoly on a device that works than competition in a space between many devices that do not work or provide no additional benefit. If anything, the FDA should require studies on “equivalent” novel devices to be compared to their already-used counterparts as the control arm to determine any new benefit before approval.
As someone studying to join a profession that appeared to be progressing at light-speed, the article’s findings disappointed me greatly. I always envisioned devices as the bleeding edge of medical research, reflecting our era’s advanced engineering capabilities. I had hoped that in some cases, devices could grow to replace pharmaceuticals or at least work faster than them, avoiding long-term costs and effects associated with medication. But it seems this is little more than a lofty ideal. Recall is a serious thing, and should be very rare concerning approved devices that patients pay big money for and depend on for their lives. However, it appears that the medical device industry is rapidly becoming the new Big Pharma - pushing devices onto the market with weak research and taking advantage of loose regulatory systems to turn a profit with little true regard for patient outcomes.
But it’s not just their fault. The FDA is purported to be an institution that we trust as the public to evaluate the science and bring products to market with our best interests in mind. Cardiovascular disease has been, for a while now, the leading cause of death in the United States. It will continue to be so as obesity and chronic disease rates rise. Cardiovascular devices in particular have the potential to be used in large segments of our population, but these devices alone account for nearly a third of all Class I recalls. Taking another look at the numbers, each recalled device had been utilized in a median of 7,649 cases, with the IQR ranging from 953 to 28,446 cases. Doing a little math here, that means that the studied recalls likely affected over one million patients, if not more. I shudder to think of the total number of people affected by all recalled devices over a longer time period.
This FDA approval strategy not only has little regard for patient outcomes, but also stifles new innovation. Due to the ease of bringing “equivalent” devices to market, manufacturers are not incentivized to innovate and bring new products with new mechanisms to the table. It costs much less to reengineer and approve a current device than it would to bring a completely novel device to the market from scratch. Our regulators must keep in their mind that manufacturers are incentivized to bring products to market as fast as possible for the potential to earn millions.
Now do I think that the FDA is intentionally trying to harm anyone or ensure that device manufacturers turn a profit? Not necessarily. I think in most cases, people see a shiny new device that theoretically makes sense accompanied by study published in a fancy journal with a p-value < 0.05. The excitement of innovation is certainly something that the FDA, and the government as a whole, want to deliver to the public. And I get that. However, it should not come at the cost of safety or stringency in our approval processes. We have to also remember our past - over most of modern medical history, there have been thousands, if not tens of thousands of treatments that ended up being disproven or being recalled. A high degree of skepticism is needed for any and all new treatments, even if the results on paper seem superb.