More than 500,000 U.S. patients have received metal-on-metal hip prostheses, most of which were implanted between 2003 and 2010. These prostheses entered the market through the 510(k) pathway at the Food and Drug Administration (FDA), whereby manufacturers need only demonstrate substantial equivalence to a device already on the market to gain approval. Unfortunately, there is now compelling evidence that these implants fail at a higher rate than hip prostheses made of other materials1; indeed, one type of metal-on-metal hip has a failure rate of nearly 50% at 6 years.2 Moreover, a number of unresolved questions related to these devices remain, including what the relationship is between serum metal ion levels and the occurrence of local and systemic adverse events, what relationship exists between serum metal levels and the need for revision surgery, how much product-level variation there is in revision rates and adverse events, and which patient- and clinician-level variables are associated with those rates and events.
On May 6, 2011, in response to these public health concerns, the FDA ordered manufacturers of metal-on-metal hip implants to conduct postmarket surveillance studies on their products. (The FDA can require such studies under the Food, Drug, and Cosmetic Act [FDCA] if failure would be reasonably likely to have serious adverse health consequences, if the device is expected to have considerable use in pediatric populations, if it’s intended to be implanted for more than 1 year, or if it’s intended to be a life-sustaining or life-supporting device used outside a device user facility.) The FDA directed the implant manufacturers to examine both adverse events and patients’ pre- and postimplantation levels of chromium and cobalt.3 The agency recommended a cross-sectional study design to capture data on patients from the time of initial implantation to 8 years later. It also recommended that manufacturers conduct a failure analysis to evaluate the devices that had been explanted from patients who had participated in clinical studies and all other reasonably available explants.
Unfortunately, data from these studies, which are funded by the manufacturers, will not be available to protect the public health anytime soon. As of June 18, 2012, the FDA and manufacturers had reached agreements on study protocols for less than one quarter of the devices, most of the study plans had not been finalized, and it was unclear whether any studies had begun. Moreover, methodologic issues will limit the usefulness of the information that emerges from these studies. These problems illustrate some of the challenges associated with postmarketing surveillance of medical devices in the United States — and are some of the fraught issues for the FDA to consider when it convenes the Orthopaedic and Rehabilitation Devices Panel of the Medical Devices Advisory Committee to discuss metal-on-metal hip implants on June 27 and 28, 2012.
According to the FDA’s website, which is updated monthly, just 24 of the 104 metal-on-metal hip products (23%) for which manufacturers face an active order to complete a study were categorized as having a “Study Pending,” which indicates that the FDA has approved the study plan. The studies for the remaining 80 products were listed as having either a “Plan Pending” or a “Plan Overdue.” (We exclude products with a study status of “Other” or “Terminated,” which indicate that there is not an active study for these products — because they were never marketed, for example.)
Several factors may contribute to delays in study initiation: the development of a scientifically sound study protocol is time-consuming and resource-intensive; manufacturers lack incentives to conduct studies that may reveal adverse information about their products; and delays at the FDA may slow down the process of finalizing study protocols. Strategies for reducing delays must address at least one of these factors.
Unfortunately, even when the studies proceed, limitations may constrain the amount of useful information that emerges. One significant shortcoming is that each manufacturer is permitted to conduct its own independent study on its product or products. The resulting lack of harmonization among studies will lead to challenges in pooling the data and making cross-product comparisons. For example, variations in definitions of outcomes, collection of patient- and provider-level data, and patient follow-up could undermine attempts to understand product-level differences. Similarly, companies may measure chromium and cobalt levels differently, using varied assays, laboratories, and protocols and introducing uncertainty into attempts to pool results. Although identical protocols would not necessarily be appropriate for all products, this lack of harmonization limits the public health benefit of the studies.
The FDCA grants the FDA the authority to ensure that “the [study] plan will result in the collection of useful data that can reveal unforeseen adverse events or other information necessary to protect the public health”; in certain instances, a requirement of study harmonization may fall within that authority. Given that many protocols for metal-on-metal–hip studies are not yet finalized, the FDA may still have an opportunity to maximize the amount of information that will be provided.
Another weakness is that current law prevents the FDA from requiring studies such as these to last more than 3 years for most devices. Given that hip implants are anticipated to last for 15 years, a 3-year study will neither capture all adverse events nor obtain the 8 years’ worth of data that the FDA requested for a device that only recently entered the marketplace.
To improve this type of study, we recommend a few key approaches. First, the FDA and manufacturers must collaborate to ensure that protocols are finalized and studies initiated as quickly as possible. If the manufacturers are causing the delay, the FDA should use its available enforcement tools — warning letters, fines, and removal from the marketplace — to demonstrate its commitment to postmarketing surveillance and provide incentives for manufacturers to complete the studies. In addition, given recent increases in the number of these studies ordered by the FDA, the agency must be sure it has sufficient numbers of staff members and the expertise to ensure that the studies use appropriate methods and are conducted promptly.
Second, the study methods should be harmonized to the extent possible. The FDA should explore ways of improving the coordination of studies conducted by multiple manufacturers to ensure that data can be pooled and cross-product comparisons can be made where appropriate. Coordination could be improved, in part, through creative engagement with external stakeholders. Possible models include the FDA advisory committees and efforts like the FDA-initiated International Consortium of Orthopedic Registries.
Third, the infrastructure needs improvement. Among other factors, the lack of comprehensive national medical-device registries, the absence of unique device identifiers, the inability of data systems to “talk” to one another easily, and inadequate funding hamper efforts to conduct postmarketing surveillance studies in the United States. Indeed, the first concerns regarding high revision rates for metal-on-metal hip implants arose out of the Australian joint registry in 2007, and additional data emerged from the National Joint Registry of England and Wales in 2010. These studies could be launched more quickly if there were an established U.S.-based infrastructure to support them. Infrastructure investments would benefit all stakeholders: manufacturers could conduct less costly postmarketing trials and use the feedback for iterative device improvements; patients and clinicians would gain information on the risk–benefit profiles of devices; and the FDA would have more confidence that adverse health effects would be quickly identified.
While the FDA and other stakeholders struggle with these systemic issues, the problems with metal-on-metal hip implants will continue to occupy the agency, clinicians, manufacturers, and thousands of affected patients for the foreseeable future. The upcoming advisory committee meeting would be an opportune time for the FDA to address the slow start to these studies — and to signal that substantial penalties may be assessed against any manufacturer that is responsible for delays in finalizing protocols.