LSHC Horizons Brochure 2024 - Flipbook - Page 15
Hogan Lovells | 2024 Life Sciences and Health Care Horizons | Digital Health and AI
15
Current regulatory landscape of Artificial Intelligence
and Machine Learning-enabled devices
AI and Machine Learning (AI/ML) based
technologies are transforming health care at a
rapid pace. Software incorporating these cutting
edge technologies has become a critical part of
an increasing number of medical devices. The
surge of Software as a Medical Device (SaMD) in
recent years has enabled the speedy adaptation
of AI/ML algorithms for a wide range of
applications, such as earlier disease detection
and diagnosis, development of personalized
diagnostics and therapeutics, among many others.
A powerful feature of AI/ML-based technology
is its ability to continuously learn from real-word
use and then utilize that information to improve
device performance. The ability to adapt over
time presents a unique problem to FDA that
requires both short and long term visions that
challenge the conventional idea of a static device
around which existing regulations and policies
are established. This sentiment is echoed in a
recent interview with FDA Commissioner Robert
Califf, who discussed the exciting potential of
such algorithms benefitting public health but
also the need for oversight to prevent unintended
harm.1 While the regulatory framework for AI/
ML-enabled devices is evolving, one thing is
certain: the traditional paradigm of medical
device regulation needs to be reconsidered,
especially for adaptive and generative algorithms.
To date, only locked algorithms - an algorithm
that provides the same result each time the same
input is applied and does not change with use have been cleared or approved by FDA.
Kelliann H. Payne
Partner
Philadelphia
FDA recognizes, however, its need to adapt
so as not to stifle innovation. Accordingly,
in April 2023, FDA issued a draft guidance
document that sets forth a proposed
regulatory framework for modifications to
locked algorithms following FDA clearance or
approval through the use of a Predetermined
Change Control Plan (PCCP). According
to FDA, “a PCCP, as part of a marketing
submission, is intended to provide a means to
implement modifications” to AI/ML-enabled
devices “that generally would otherwise require
additional marketing submissions prior to
implementation”.2
We are continuously assessing the evolution
of this regulatory framework to provide the
latest insight on FDA’s approach towards this
rapidly burgeoning technology. As more device
manufacturers innovate with AI/ML, there
is both uncertainty and opportunity. Device
manufacturers would be wise to engage with
FDA early, as the agency has made it clear
that it intends to lean on industry experts,
professional organizations, other regulatory
bodies, and also real-world data from its own
database on how to best approach regulating
AI/ML-enabled devices.
1 Dan McKay, FDA ‘Can’t Do This Alone,’ Wants Help Vetting AI In
Healthcare, LAW360 (Jan. 10, 2024, 4:48 PM EST), https://www.
law360.com/articles/1782163/print?section=lifesciences.
2 Draft Guidance for Industry and FDA Staff, Apr. 3, 2023, Marketing
Submission Recommendations for a Predetermined Change Control
Plan for Artificial Intelligence/Machine Learning (AI/ML)-Enabled
Device Software Functions, https://www.fda.gov/media/166704/
download.
Eriko Yoshimaru
Senior Director of
Regulatory Affairs
Washington, D.C.
Jason Russell
Senior Associate
Philadelphia
Read the Life
Sciences and
Health Chapter
of our AI Hub
