Recent FDA Guidance on PCCPs 

To help manufacturers navigate this new option, the FDA has already released three guidance documents: 

 

1. AI/ML-Enabled Devices: On April 3, 2023, the FDA issued draft guidance titled “Marketing Submission Recommendations for a Predetermined Change Control Plan for Artificial Intelligence/Machine Learning (AI/ML)-Enabled Device Software Functions.” While the comment period for this guidance has ended, it’s important to remember that FDA guidance—whether draft or final—serves as recommendations rather than strict legal requirements. We often advise that if you plan to deviate from these recommendations, having a strong scientific or technical justification is crucial. 

1. Antimicrobial Susceptibility Test (AST) Devices: On September 29, 2023, the FDA published final guidance on updating breakpoints in AST system devices, titled “Antimicrobial Susceptibility Test (AST) System Devices – Updating Breakpoints in Device Labeling.” Unlike the AI/ML draft guidance, this guidance was not open for public comment and was directly released in its final form. 

1. General PCCP Guidance: On August 22, 2024, the FDA released draft guidance titled “Predetermined Change Control Plans for Medical Devices.” This guidance is open for public comment until November 22, 2024, and manufacturers should take advantage of the opportunity to provide feedback. 

 

What You Need to Know About PCCP Applications

All PCCPs must be submitted and cleared as part of your pre-market submission before any changes can be made. However, the application of PCCPs varies depending on the type of device. For example, an approved PCCP for updating breakpoints on AST devices can be applied to future or legacy devices under the same classification and product code, provided they share the same intended use and technological characteristics. For most other devices, PCCPs are device-specific and cannot be applied to another device, even if the technology is similar. 

We should emphasize that once a PCCP is approved, it cannot be modified, unless through another submission. PCCPs must also be implemented through the manufacturer’s Quality Management System (QMS).

 

What Modifications Can a PCCP Cover?

Here are a few examples of common modifications that might be covered under a PCCP:

• Adding new primers or probes to detect additional alleles for a gene already targeted by the device.
• Implementing a new nucleic acid extraction method in a molecular assay.
• Incorporating different makes or models that input data or signals to a device’s algorithm.
• Extending the shelf-life of a device using a different, but well-established method.
• Adding a new sample type, such as lithium heparin plasma, to a device already approved for serum.
• Introducing a new potassium ion-selective electrode in a device that measures potassium in serum.

 

What Modifications Are Not Suitable for a PCCP?

On the flip side, certain changes do not fit within the scope of a PCCP, including:

• Adding a new collection device or sample type (e.g., switching from saliva to buccal swab).
• Introducing new primers/probes for a gene or locus not previously targeted.
• Removing any limitations from the device’s use.

 

International Collaboration on PCCP Development

The FDA has collaborated with Health Canada and the U.K.’s Medicines and Healthcare products Regulatory Agency (MHRA) to establish guiding principles for developing PCCPs, particularly for Machine Learning-Enabled Medical Devices (ML-DSF). These principles highlight the importance of focusing on risk assessment, transparency, and evidence-based decision-making. They advise that PCCPs should be focused and bounded. They also emphasize considering the Total Product Lifecycle (TPLC) when developing PCCPs.

 

PCCPs – New Valuable Tool for AI/Machine Learning Devices

As the use of machine learning in medical devices continues to grow, PCCPs are becoming an increasingly valuable tool for manufacturers. At MDC Associates, we pride ourselves on staying at the forefront of these regulatory advancements, ensuring that our clients are well-prepared to implement PCCPs, whether for breakpoint updates or AI/ML-enabled devices.