By: Laurie Reed
Last week I had the pleasure of attending MassMEDIC’s Regulatory Roundup Conference, a one-day event held in Boston’s Seaport district. This was my first time attending this particular conference, and I was happy to learn more about some of the broader aspects of the medical device regulatory landscape beyond human factors.
Here are my top takeaways from a few of the presentations I attended:
“Accelerate Regulatory Approvals & Time to Market with AI-Powered Clinical Data and Virtual Twins” by John McCarthy of Dassault Systèmes
Did you know it’s possible to supplement clinical trial data with synthetic (or ‘in silico’) data generated through AI-Powered predictive models? I didn’t, but how cool! Apparently in November 2023, FDA released a guidance document titled “Assessing the Credibility of Computational Modeling and Simulation in Medical Device Submissions” which provides direction on using such predictive models in premarket submissions. The folks at Dassault Systèmes have been doing some groundbreaking work in this area and shared some of the practical uses of AI-powered data and virtual twins.
In layman’s terms, a virtual twin is a digital 3D model of some part of the human anatomy. In the examples that Dassault shared, they had created virtual twins of hundreds of human hearts with mitral valve disease, representing the spectrum of disease states and anatomical differences that might be found amongst the target population. Using AI-power algorithms, they were able to digitally test out the use of a new medical product to treat mitral valve disease within the virtual twin models and determine what the clinical outcomes would have been if real human patients mirroring those models had received the treatment.
In this way, they were able to generate a large set of synthetic data to augment their client’s data, which had been collected from actual humans in their clinical trial. This has valuable implications for healthcare. Not only does this technology ensure the suite of potential disease states are fully represented, but it also eases the burden on enrolling patients into clinical trials and could help decrease overall time to market. Of course, V&V must be performed on the synthetic datasets and the credibility of the computational models must be assessed and provided in the submission.
“Driving Value with Adoption of AI in Healthcare QARA (Quality Assurance and Regulatory Affairs) Systems” by Michael King of IQVIA Technologies & Chris Hart of Foley Hoag LLP
This discussion between Michael (who is a product developer) and Chris (who is an attorney) was very insightful for getting up to speed on the current state of AI usage in the medical device space. I learned that the U.S. has been lagging behind Europe in terms of regulating the use of AI in the development of medical devices. In the U.S., our laws are more fragmented, slower to develop, and have been focused on very targeted or specific areas. However, Europe has released the first-ever and most comprehensive legal AI regulation to date (called the “EU AI Act”).
FDA has yet to draft something similar to this in the U.S., although some individual states are starting to create their own AI regulations. It was hypothesized that an overarching FDA AI regulation won’t happen anytime soon due to our impending government administration change. Furthermore, President Joe Biden had issued an executive order in October 2023 on the topic of AI and safety, but it is predicted that it will be scrapped soon after the new government administration takes office.
“FDA Keynote & Moderated Q&A” with Dr. Michelle Tarver, Director of CDRH
I had the privilege of (remotely) meeting Dr. Tarver, the newly appointed Director of FDA’s CDRH division. Dr. Tarver was unable to attend the conference in person but provided a fantastic overview of her background and the division’s strategic initiatives.
Here are some interesting points that I took away from her presentation (as well as the Q&A period):
Dr. Tarver holds a PhD and is both an ophthalmologist and epidemiologist. She is incredibly passionate about involving patients in the risk-benefit analysis of new medical devices and was the driver behind FDA’s Guidance entitled “Incorporating Voluntary Patient Preference Information over the Total Product Life Cycle” (which is still open for comments until 12/5/24).
FDA reorganized the Office of Regulatory Affairs in Oct 2024; the CDRH is now the umbrella for this body and they renamed it to OII – Office of Inspections and Investigations.
Dr. Tarver seeks to continue to increase the number of new, novel, and innovative medical technologies that are submitted to the FDA first (or in parallel to other markets). CDRH specifically plans to prioritize technologies that address health disparities and inequities in the U.S., as well as devices that facilitate access to healthcare in rural areas (including in homes).
Another of CDRH’s important strategic initiatives revolves around re-designing the premarket program to improve core business processes. Their goal is to achieve a higher volume of single-cycle reviews. To assist with this, they are working on streamlining the review process, making it more predictive, minimizing submission idle times, and improving their internal SOPs.
CDRH also has AI on their minds and is working on publishing a white paper on the topic. They are even exploring how they might use AI within the Agency to assist with premarket reviews, noting that there is a vast amount of medical data that is unused in the regulatory space.
Dr. Tarver discussed a number of other goals and recent achievements, including: the TAP pilot program; release of new draft guidance on remote clinical trials and real-world evidence; creation of the NEST department; and initiation of the ‘Home as a Healthcare Hub’ initiative, all of which you can find out more about by visiting the FDA’s website.
“FDA’s New Quality Management System Regulation (QMSR)- What’s New and How to Get Ready for the Feb 2026 Implementation Deadline” by Aaron Snyder & Alexis Compton of Allotex
The FDA’s new QMSR was no surprise as it has been in development for some time, but it was helpful to hear an overview of the key changes as a result of this new regulation. FDA’s 21 CFR 820 has been in place for well over 20 years, so the industry is excited for this update. The main goal of the new QMSR is to achieve better harmonization with ISO13485, with a secondary goal of increasing the role of risk management and utilizing more contemporary risk terminology.
Toward this end, the QMSR is better aligned with the ISO13485 standard and includes a pointer to it. Where the old 21 CFR and the current ISO standard differed, this new QMSR generally adopted the ISO approach. However, there are a few exceptions; the FDA added the following 3 clauses:
820.10 requirements for a QMS
820.35 control of records
820.34 device labeling and packaging controls
These changes are viewed as wholly positive by both the Agency and industry since everyone will now be speaking the same language and relying mostly on ISO terminology. For companies who are already ISO13485 certified, achieving compliance with the new QMSR should be fairly stress-free.
“Insights Into Development of IEC60601-1 4th Edition” by Yaqing Liu & Todd Konieczny of Intertek Boxborough
Yaqing and Todd both serve as integral members of the committees responsible for revising IEC60601-1, so they offered useful insights into the current state of the revised standard and some of the changes that will be coming down the pike.
The committees are currently actively working on the new draft of the standard. The goal is to have a final standard published in 2029. Following that, a grace period is expected and companies should be in full compliance with the new standard by the year 2032.
I learned that much of the general ISO60601-1 content will remain, but a lot more detail will be added to requirements, and of course, new requirements will be added. Even with the deletion of some items that are no longer relevant (such as requirements for cathode ray tubes or gases that are no longer in use), the new standard is expected to be very lengthy.
Some of the expected changes include:
Bringing in requirements for new technology like AI and new test methods.
Adding in requirements for non-human subjects (i.e. animals).
Bringing in requirements for system sub-assemblies
Removing cross-references as much as possible to prevent users from having to navigate all over the document to find specific content.
Transforming requirements into single sentences that relate to one discrete idea so it will be easier to show compliance.
Folding all usability-related items together into their own section.
“Laboratory Developed Tests as In Vitro Diagnostics: Regulation, Litigation & Legislation” by Greg Levine & Beth Weinman of Ropes & Gray
This presentation was educational and fascinating, as Greg & Beth presented the history of the FDA’s approach to laboratory developed tests (LDTs), and the recent controversy around the ruling to regulate them as In Vitro Diagnostic (IVD) devices.
As background, LDTs tests are designed, manufactured, and used within a laboratory for clinical purposes. One example of an LDT is a custom blood test developed by a single academic institution’s laboratory for diagnosis of a rare virus. Tests like these are often not commercialized if there is no good business case for return on investment. For decades, FDA chose not to regulate these LDTs. However, as technology has advanced, the tests have become more complex, and FDA has identified risks associated with their safe use (think Theranos). Thus, the Agency changed its mind and decided to exert its authority over them.
In 2023, the Agency drafted “The LDT Rule”, a regulation which initiated over 6,000 contentious comments from industry. Despite this, it was published and took effect in July 2024. Immediately following this, two lawsuits were filed against the FDA in Texas to ask the courts to invalidate The LDT Rule.
The plaintiffs put forth a host of arguments against the rule, including that:
Regulating LDTs is outside of FDA’s authority
LDTs don’t qualify as true medical devices
LDTs are already regulated under CLIA
Regulating LDTs as IVDs puts up new barriers to rare disease testing
On the other hand, FDA poses that:
Regulating LDTs has always been within their authority, but they had chosen not to because they weren’t as risky before
LDTs do indeed meet the definition of a medical device
CLIA isn’t equipped to effectively assess the validation and safety of these tests
In parallel to all of this controversy, in June 2024, a landmark Supreme Court ruling referred to as the Chevron case was overturned in a new case called Loper Bright vs. Raimondo. This ruling invalidated the idea that the courts should defer to government agencies (such as FDA) when there is ambiguity about how to interpret a regulation outside of their main expertise area. Instead, the Loper & Bright ruling took the stance that deference to government agencies like FDA is a violation of the ‘separation of power’ principle. Because of this recent ruling and the tendencies of the impending new government administration, there is wide speculation that ‘The LDT Rule’ will not be upheld and that other Agency regulations may also be next on the chopping block.
This should be an interesting space to stay tuned to!