Ellen Dillavou (WakeMed Heart Center, Raleigh, USA) asks whether the high bar for evidence being sought to bring new de novo devices to market is stifling innovation in the field of vascular access.
In the last decade we have seen remarkable innovation in the dialysis access space—percutaneous fistulae, external supports, drug-eluting technology, immediate access grafts and novel catheters for treating thrombosis and stenosis. However, getting these tools into the hands of practitioners is a challenging and unpredictable process.
At the last VEITH meeting (19–23 November, New York, USA) I had the privilege of debating these challenges with Robert Lee, formerly of the US Food and Drug Association (FDA) and head of the committee which was tasked with approving these novel technologies. As a clinician focused on access, the following are some of my thoughts on this topic.
While randomised controlled trials (RCTs) are the gold standard for clinical evaluation, their necessity for de novo devices, particularly in dialysis access, warrants reconsideration. While rigorous evaluation is essential, the “least burdensome” principle, enshrined in the FDA Modernization Act of 1997 (FDAMA), mandates that the FDA utilises the least burdensome means of demonstrating that the probable benefits outweigh the probable risks, while protecting patient safety, and the level of evidence required to tip the scales of approval is commensurate with the level of potential risk a device may pose. In the context of novel dialysis access devices requiring de novo classification, the risk level is often low to moderate as device-related mortality and/or unresolvable morbidity are rarely probable. In dialysis access, RCTs can often represent an overly burdensome approach, potentially hindering innovation and delaying patient access to new technologies.
The dialysis access space presents unique challenges for RCT design. First, patients requiring dialysis often have complex comorbidities, limited treatment options, and a high risk of complications which can complicate both enrolment and interpretation of results.
The recent Vascular Therapies RCT on the juxta-anastomotic delivery of sirolimus to support maturation—ACCESS—is a prime example of complexity in the patient population leading to a negative result. The treatment group had an over representation of complex patients that skewed the endpoint in favour of the control. Even though a subgroup analysis demonstrated a potential clinical effect in the most challenging patient population, the failure of the overall endpoint resulted in the FDA requiring another RCT before a de novo approval could be granted. Luckily, Vascular Therapies was able to pull together the funding to run another study—ACCESS 2—and hopefully we will be able to offer our patients access to this therapy in a few years. However, the lack of any safety risk reported in the study combined with the probable benefit in complex patients begs the question why de novo status was not granted for at least a limited indication for use.
The ethical consideration of randomisation also cannot be ignored. In many cases, there may be a strong clinical rationale to believe a new device offers advantages over existing options. Randomising patients to a potentially inferior treatment, even temporarily, raises ethical concerns. Additionally, providing a treatment to a patient that may not align with best practices can be equally concerning. The recent WAVE trial of the Wrapsody endoprosthesis (Merit Medical) highlights this issue. Multiple RCTs have demonstrated that primary use of stent-grafts is superior to primary angioplasty of a stenosed access. However, current clinical guidelines dictate we give angioplasty a chance first because nothing is left behind that could compromise venous real estate. In the Wrapsody study design, as guided by the FDA, primary stenting was performed on patients who may have responded well to angioplasty. Conversely, angioplasty was performed on patients with multiple failed angioplasties who were indicated for a stent-graft. Neither addresses the problem that Wrapsody was designed to solve, which is the edge stenosis and in-growth failure mechanisms of legacy stent-grafts, and this randomisation strategy potentially harmed patients in the quest for patient safety.
Finally, the FDA should give credence to physician judgement. Unlike pharma, we can directly observe how the device is performing and assess benefit on a case-by-case basis. If given sufficient data to understand the risks, we can determine what is appropriate for our patients. The endovascular arteriovenous fistula (endoAVF) cases are excellent examples of single-arm studies that allowed for the dialysis access community to make sound clinical judgements. Generally, we decided that endoAVF has value in specific circumstances for a subset of patients. However, this was only realised because the FDA allowed access to the device in a real-world setting at an early timepoint.
The recently approved and reimbursed VasQ device (Laminate Medical) is undergoing that same process. The single-arm study demonstrated promise and comparators of fistulas created by the same surgeons in the study suggested superiority of VasQ access creations. While surgeons now have a chance to assess the value of the device in the real world, a post-market RCT is being conducted which will add to the level of evidence.
Unfortunately, the FDA defers to lone medical reviewers that may have their own biases. To be clear, this is not a criticism of these reviewers, but the process that requires them to make complex decisions on an island. No one celebrates the reviewer when a good device is approved, but everyone will let them know when an approved device leads to excessive patient harm. This dynamic leads to a conservative approach that seeks comfort in an RCT rather than clinical judgement. For this reason, the reaction to the lack of endoAVF adoption may have swung the pendulum to default to RCTs for several new devices on the horizon, regardless of the safety profile.
Whilst having the highest level of data is enticing, the cost of RCTs and risk of design issues may stifle investment into innovation for dialysis access. We have seen a recent surge of investment since the successful FDA approval, Centers for Medicare & Medicaid Services (CMS) reimbursement and acquisition of the endoAVF devices. Devices like those currently in clinical studies like Venostent (VenoStent), Velocity (Venova Medical), and EchoMark (Sonavex) have great potential to improve the lives of dialysis patients. However, if the barrier to enter the market is increased, we may see the renaissance end. Trials require substantial resources, including funding for patient recruitment, data collection, analysis, and regulatory submissions. For smaller companies, which are often the source of innovation in the medical device space, this can be prohibitive. This can create a barrier to entry, preventing potentially beneficial devices from reaching the market. It’s like trying to tip the scales when one side is already weighed down by immense costs.
So, what are the alternatives? The least burdensome principle calls for exploring other scientifically valid methods. In the dialysis access space, this could include:
- Well-designed single-arm studies: These studies can be valuable for evaluating the safety and effectiveness of a new device, particularly when historical controls or registries are available for comparison.
- Real-world data: Real-world data can complement data from clinical trials and provide a more comprehensive understanding of the device’s safety and effectiveness.
- Patient registries: These organised collections of data on patients with specific conditions can be used to track the long-term outcomes of devices and identify safety signals.
These alternative approaches, when rigorously implemented and analysed, can provide substantial evidence of safety and effectiveness while reducing the burden on manufacturers, aligning with the intent of the least burdensome principle.
While RCTs remain a valuable tool for evaluating medical devices, they should not be considered the default or only method for demonstrating safety and effectiveness for FDA approval, especially in the challenging landscape of dialysis access. For de novo devices in this space, requiring RCTs can often be overly burdensome, hindering innovation and delaying patient access to potentially life-saving technologies. By embracing alternative approaches, when scientifically justified, the FDA can better adhere to the least burdensome principle, tipping the scales towards a more balanced and patient-centered approach to medical device regulation.
Ellen Dillavou is the division chief of Vascular Surgery at WakeMed Hospitals (Raleigh, USA). She reports that she is a speaker for WL Gore and 3M/KCI, a consultant/speaker for Merit Medical and a scientific advisor for Venostent and WL Gore.
