Patient-Specific Devices Regulation and Questions

Patient-specific devices are nearing standard of care in several areas and many hospitals are eager to take the near automated workflows being sold for general purposes by 3D Printing companies in-house to produce these perfectly matched parts more efficiently for their patients. Does this make the hospital institutions medical device manufacturers, or does the regulatory burden fall on the 3D Printing companies who are providing the workflow and equipment? There are still more questions than answers at this point, but let’s take a look at where we are and how we got here.

In 2019 the FDA CDRH Additive Manufacturing Working Group set out a framework to guide the regulation of point-of-care 3D printed devices. It includes five categories inclusive of personalized and even serialized products, grouped by risk, and associated methods of regulation. A discussion paper, published by FDA in December 2021, formalized and opened for public comment these categories and considerations for regulations that could surround them. Reading the comments formally posted on the FDA’s document docket system, shows the divide that still remains between traditional manufacturers and point of care facilities, questions, and ideas on level of regulation that should exist. 

With just the discussion paper to guide industry, no clear path exists for point of care facilities or medical device manufacturers. 

Manufacturers, who have been pursuing market clearance for patient specific devices since the mid 2000s and have become adept at finding the appropriate regulatory pathways for each indication, appear to be hesitant to produce the turn-key-solutions for complicated products that have, for many years, held a high bar of entry.  This paired with the responsibility of managing customers’ (point of care facilities) use of the systems, which currently have many opportunities for off-label use, have potentially led to the slow pickup of this type of system.

Point of care facilities are struggling too, living somewhere between understanding there will be enforcement discretion, arguably in practice now, or that they will need to have quality systems and registrations fully compliant to 21 CFR 820 if they want to do any printing of personalized devices in-house. Formal guidance will need to be issued by FDA to answer these questions and push the industry forward in the desired direction.

The one area where turn-key solutions exist is in the image processing to diagnostic anatomic model area, and this area has seen tremendous growth in not only standard processing of DICOM images, but through machine learning and automated segmentation systems in the last 3 years, but even the regulation of these systems has not been done under formal guidance.  The LLZ pathway was introduced in a joint meeting between FDA and the RSNA SIG and really only exists as word of mouth information, and even that has begun to change.

In 2018 the first product was cleared under this new framework under product code LLZ. It was Materialise’s Mimics InPrint, a segmentation software, and included a true turn-key system for printing of diagnostic quality anatomic models. This allows a customer to purchase the software system capable of reading and segmenting (editing out only anatomic features desired for the application) medical imaging data, with clear, tested, instruction on which anatomies can be printed, which 3D printers and materials to use, and how to print, post process, and quality check the physical replicas printed. This turn-key method aligned perfectly into the forthcoming framework for classifying higher risk 3D Printed devices.  Several other companies followed with clearances in this space in the last 3 years.

However, seeming to align with the timing of this Rule, and as segmentation systems continued to gain complexity, growing deeper dependance into machine learning allowing for more push button segmentation / less manual processes, another product code, QIH has began to grow.  QIH is a product code under the same regulation, 21 CFR 892.2050, as LLZ, but is defined as: Automated Radiological Image Processing Software. Products cleared under the QIH product code can be found here.  This product codes appears to be where image segmentation software, including those capable of outputting diagnostic images suitable for printing or further processing, are being regulated and will be going forward instead of the LLZ code.

This evolving regulation and product code journey is not uncommon, but guidance usually follows to help industry navigate the current thinking on regulatory paths for devices. The 3D printing and personalized device industry is overdue for some clear guidance on the categorization of devices and how they are regulated, but the most telling information currently available is to watch new products come to market and trace the regulatory path they navigated.