William A. Hyman, Professor Emeritus of Biomedical Engineering, Texas A&M University, & Adjunct Professor of Biomedical Engineering, The Cooper Union08.16.16
A recurring question in the regulation of medical devices, including orthopedic implants and instruments, is whether or not certain changes in the design, materials, or manufacturing of an already marketed device require additional premarket filings with the U.S. Food and Drug Administration (FDA) and a corresponding clearance (for Class II devices) or approval (for Class III devices). This question is addressed for Class II devices in a new FDA Draft Guidance Document (DG) entitled “Deciding When to Submit a 510(k) for a Change to an Existing Device”.1 This guidance, when finalized, would replace the 1997 guidance of the same name.2 In 2011, there was a different proposed replacement draft that was so popular that Congress passed legislation ordering it to be withdrawn and removed from the FDA’s website. However, copies of that draft are still around for the curious.
A 510(k), also known as premarket notification, is the regulatory path for Class II devices while Class III devices follow the Pre Market Approval (PMA) pathway. Class II devices are “cleared” for marketing while only Class III devices are “approved.” The difference between cleared and approved is not, I would argue, of interest only to regulatory geeks since it reflects what evidence was required for the device to be marketable. Notably, Class II devices do not typically require a formal clinical trial while Class III devices typically do. This difference should also be important to end-users. There are both Class II and Class III orthopedic devices even for the same general application. For example, in FDA nomenclature, a “hip joint metal/composite semi-constrained cemented prosthesis” is Class II while a “hip joint metal constrained cemented or uncemented prosthesis” is Class III.
The change principle addressed in the new DG regarding when a new regulatory filing is required for a Class II device is relatively clear. Changes that could significantly affect the safety or effectiveness of the device require a new 510(k). As with many rules, the principle is easier to define than the application of that principle. Here this is reflected in in the phrase “could significantly effect.” The DG seeks to clarify the FDA’s thinking on the considerations that go into “could significantly effect.” Even if this draft emerges as a final guidance document, it will in theory not be mandatory, since it will be guidance and not a regulation.
For material changes, the DG outlines the primary considerations for the “significantly” test. The first question is “Is this a change in material type, material formulation, chemical composition, or the material’s processing?” where change is explained to mean any change. This includes a change in supplier or manufacturer material processing or finishing steps. The second question is “Will the changed material directly or indirectly contact body tissues or fluids?” If the answer is yes, this obviously raises biocompatibility concerns and the possibility of needing new testing. However, the DG does allow for a conclusion that new testing is not required based on detailed knowledge of the chemistry and prior use of the new material in its final finished state in another device from the same manufacturer, provided that other device has the same or a more demanding risk profile. This provision might preclude a strategy of bringing a low risk device to market and then using its materials to segue into higher risk applications. It is also noted that the use of the new material by a different manufacturer is not definitive unless the exact formulation and processing of the device, in ways that may affect the safety and effectiveness of the final finished product, can be verified. This may be an example of the FDA becoming more demanding about complete characterization of a material beyond, for example, its generic name. The DG also “recommends” that a manufacturer consider whether the material change could affect the performance of the device by affecting properties such as strength or hardness, or whether the new material could be affected by any labeled cleaning, disinfection, or sterilization process for the device. These questions may overlap with other question area in the DG such as labeling; technology, engineering, and performance changes; and general risk assessment.
What is not changed is that if a decision is made that a new 510(k) is not required this should be documented in the device file. This has sometimes been called the “no 510(k) rationale.” The DG includes guidance on the content of such documentation including the basic requirement that it should be prepared in a way that an FDA investigator or other third party can understand what the change is and the rationale underlying the manufacturer’s conclusion that a new 510(k) is not required. It should be noted here that third parties could include those involved in litigation, who may be seeking to test the appropriateness of the avoidance of further attention by the FDA. The DG states that the documentation should include detailed and “robust” analysis beyond just “yes or no,” or otherwise perfunctory answers to the questions. An outline for doing this is provided. Regardless of the level of documentation, no new 510(k) decision needs to be communicated to the FDA and multiple changes over time can accumulate without new FDA premarket scrutiny. This also means that a device being sold can end up differing to varying degrees from the device that was cleared in the original 510(k). In the past, the FDA has suggested that there should be periodic reporting to the FDA of device changes that were implemented without a 510(k), but this is not addressed in this DG.
If the 1997 Guidance is replaced, one curiosity that will disappear is its Appendix entitled “Draft Controlled Vocabulary from the FDA Biomaterials Compendium.” This lists generic names for a variety of materials that have seen use in implants. However, the Biomaterials Compendium never came into existence; thus, this draft is to a phantom document. Nonetheless, I have seen at least one manufacturer cite this appendix as having some meaning with respect to the suitability of a generic material for implantation. Furthermore, the FDA is in general moving away from the value of generic material names in favor of greater specificity.
It is not unreasonable to expect that implanted devices, such as orthopedic implants, will evolve over time, including the materials from which they are made. Such change can possibly be for the better and/or be made to accommodate other factors, such as when a material that is being used becomes unavailable. When device changes do occur, manufacturers face the basic regulatory question of whether or not the changed device requires and/or should have new FDA review. This regulatory question of course has at least in parallel the scientific and engineering questions of what the effect of the changes will be on device performance. Assuming that regulation is effective, significant changes should be evaluated by the FDA in the fulfillment of its mission of protecting the public. On the other hand, the FDA and manufacturers would be overly burdened by a requirement for new submissions for routine changes that don’t meet the important but vague significance threshold. This DG, if it ever becomes a full guidance rather than a draft, will add some nominally, if voluntary, structure to the decision making process, but it will still be up to the manufacturer to perform a comprehensive and meaningful analysis before deciding what the effects of the change might be and whether a new 510(k) is required. The user and patient communities are largely dependent on the manufacturer’s skill and integrity in making these decisions.
References
1FDA, Deciding When to Submit a 510(k) for a Change to an Existing Device, Draft Guidance for Industry and Food and Drug Administration Staff, August 8, 2016, http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/UCM514771.pdf?source=govdelivery&utm_medium=email&utm_source=govdelivery
2FDA, Deciding When to Submit a 510(k) for a Change to an Existing Device, 510(k) Memorandum #K97-1, January 10, 1997, http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm080235.htm.
A 510(k), also known as premarket notification, is the regulatory path for Class II devices while Class III devices follow the Pre Market Approval (PMA) pathway. Class II devices are “cleared” for marketing while only Class III devices are “approved.” The difference between cleared and approved is not, I would argue, of interest only to regulatory geeks since it reflects what evidence was required for the device to be marketable. Notably, Class II devices do not typically require a formal clinical trial while Class III devices typically do. This difference should also be important to end-users. There are both Class II and Class III orthopedic devices even for the same general application. For example, in FDA nomenclature, a “hip joint metal/composite semi-constrained cemented prosthesis” is Class II while a “hip joint metal constrained cemented or uncemented prosthesis” is Class III.
The change principle addressed in the new DG regarding when a new regulatory filing is required for a Class II device is relatively clear. Changes that could significantly affect the safety or effectiveness of the device require a new 510(k). As with many rules, the principle is easier to define than the application of that principle. Here this is reflected in in the phrase “could significantly effect.” The DG seeks to clarify the FDA’s thinking on the considerations that go into “could significantly effect.” Even if this draft emerges as a final guidance document, it will in theory not be mandatory, since it will be guidance and not a regulation.
For material changes, the DG outlines the primary considerations for the “significantly” test. The first question is “Is this a change in material type, material formulation, chemical composition, or the material’s processing?” where change is explained to mean any change. This includes a change in supplier or manufacturer material processing or finishing steps. The second question is “Will the changed material directly or indirectly contact body tissues or fluids?” If the answer is yes, this obviously raises biocompatibility concerns and the possibility of needing new testing. However, the DG does allow for a conclusion that new testing is not required based on detailed knowledge of the chemistry and prior use of the new material in its final finished state in another device from the same manufacturer, provided that other device has the same or a more demanding risk profile. This provision might preclude a strategy of bringing a low risk device to market and then using its materials to segue into higher risk applications. It is also noted that the use of the new material by a different manufacturer is not definitive unless the exact formulation and processing of the device, in ways that may affect the safety and effectiveness of the final finished product, can be verified. This may be an example of the FDA becoming more demanding about complete characterization of a material beyond, for example, its generic name. The DG also “recommends” that a manufacturer consider whether the material change could affect the performance of the device by affecting properties such as strength or hardness, or whether the new material could be affected by any labeled cleaning, disinfection, or sterilization process for the device. These questions may overlap with other question area in the DG such as labeling; technology, engineering, and performance changes; and general risk assessment.
What is not changed is that if a decision is made that a new 510(k) is not required this should be documented in the device file. This has sometimes been called the “no 510(k) rationale.” The DG includes guidance on the content of such documentation including the basic requirement that it should be prepared in a way that an FDA investigator or other third party can understand what the change is and the rationale underlying the manufacturer’s conclusion that a new 510(k) is not required. It should be noted here that third parties could include those involved in litigation, who may be seeking to test the appropriateness of the avoidance of further attention by the FDA. The DG states that the documentation should include detailed and “robust” analysis beyond just “yes or no,” or otherwise perfunctory answers to the questions. An outline for doing this is provided. Regardless of the level of documentation, no new 510(k) decision needs to be communicated to the FDA and multiple changes over time can accumulate without new FDA premarket scrutiny. This also means that a device being sold can end up differing to varying degrees from the device that was cleared in the original 510(k). In the past, the FDA has suggested that there should be periodic reporting to the FDA of device changes that were implemented without a 510(k), but this is not addressed in this DG.
If the 1997 Guidance is replaced, one curiosity that will disappear is its Appendix entitled “Draft Controlled Vocabulary from the FDA Biomaterials Compendium.” This lists generic names for a variety of materials that have seen use in implants. However, the Biomaterials Compendium never came into existence; thus, this draft is to a phantom document. Nonetheless, I have seen at least one manufacturer cite this appendix as having some meaning with respect to the suitability of a generic material for implantation. Furthermore, the FDA is in general moving away from the value of generic material names in favor of greater specificity.
It is not unreasonable to expect that implanted devices, such as orthopedic implants, will evolve over time, including the materials from which they are made. Such change can possibly be for the better and/or be made to accommodate other factors, such as when a material that is being used becomes unavailable. When device changes do occur, manufacturers face the basic regulatory question of whether or not the changed device requires and/or should have new FDA review. This regulatory question of course has at least in parallel the scientific and engineering questions of what the effect of the changes will be on device performance. Assuming that regulation is effective, significant changes should be evaluated by the FDA in the fulfillment of its mission of protecting the public. On the other hand, the FDA and manufacturers would be overly burdened by a requirement for new submissions for routine changes that don’t meet the important but vague significance threshold. This DG, if it ever becomes a full guidance rather than a draft, will add some nominally, if voluntary, structure to the decision making process, but it will still be up to the manufacturer to perform a comprehensive and meaningful analysis before deciding what the effects of the change might be and whether a new 510(k) is required. The user and patient communities are largely dependent on the manufacturer’s skill and integrity in making these decisions.
References
1FDA, Deciding When to Submit a 510(k) for a Change to an Existing Device, Draft Guidance for Industry and Food and Drug Administration Staff, August 8, 2016, http://www.fda.gov/downloads/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/UCM514771.pdf?source=govdelivery&utm_medium=email&utm_source=govdelivery
2FDA, Deciding When to Submit a 510(k) for a Change to an Existing Device, 510(k) Memorandum #K97-1, January 10, 1997, http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm080235.htm.