CDRH Wraps Up Guidance for Nitinol Devices

The FDA’s Center for Devices and Radiological Health (CDRH) recently posted a pair of guidances, one each for devices containing nitinol and for biocompatibility testing for devices that come into contact with the skin. Perhaps the more critical of these is the final guidance for nitinol devices, given that the agency had previously stated its concerns about the use of several materials in devices, including nitinol.

Several Key Tweaks to Nitinol Final Guidance

The final guidance for nitinol devices adds a few terms to the glossary not found in the draft, including definitions for R-phase (the intermediate phase between austenite and martensite states), and the start and finish temperatures for the R-phase. There was some fine editing of the content for the final version, including a change in the background section, which uses “thermomechanical” consistently in the final whereas the draft used the terms “mechanical” and “thermomechanical” more or less interchangeably. Both versions list the device types in which nitinol is used, but the final broadens the description of “cardiovascular stents” to “stents.”

One key difference between the two documents is found in the section for transformation temperatures under the technical considerations section. The draft had stipulated that the manufacturer had to provide austenite finish (Af) temperatures for a device for which the pseudoelastic properties of nitinol are key to device performance. The final allows the manufacturer to report either Af temperatures or one of two alternative data sources, the R-phase finish (Rf) temperature or body temperature mechanical test data.

This modification appears to have been in response to comments provided by W.L. Gore & Associates, Inc., of Flagstaff, Ariz., which had endorsed the option of Rf temperature or “body temperature tensile data” as possible substitutes for Af temperature data.

The section for experimental testing considerations is expanded somewhat in the final, starting with a discussion of mechanical testing. The draft states little more than that such testing be conducted under strain or displacement control unless the manufacturer can justify the use of an alternative. The final explicitly offers the alternative of pressure control with displacement monitoring for radial pulsatile fatigue of a vascular stent. The FDA would accept this approach when conducted under the provisions of ASTM F2477.

The section for computational stress/strain analysis is similarly edited for several key points, such as the description of material model parameters. The draft states that these parameters can be obtained from ASTM F2516, which can be used at the relevant clinical conditions, such as for temperature and preconditioning. However, the final guidance drops the mention of preconditioning and replaces it with surface processing steps. The FDA added the requirement that computational analysis be conducted at a temperature that is representative of the clinical use environment, such as 37°C for implantable devices.

Also under computational stress/strain analysis, the section dealing with fatigue analysis is significantly modified. The draft states that nitinol fatigue life is strain-dependent, and thus traditional approaches, such as Goodman/Soderberg diagrams, are not applicable. The draft had recommended that the sponsor calculate fatigue safety factors by means of a constant life curve defined by mean and alternating strains for a “relevant number of cycles.”

The final guidance reiterates much of the content of the draft, but adds a notation that fatigue life can be affected by pre-strain, such as crimping of a stent onto a delivery catheter. Pre-strain can positively or negatively affect fatigue life, and thus the final recommends the manufacturer discuss the effect of pre-strain on the device in the premarket filing. The manufacturer is also urged to characterize and justify the method for calculating mean and alternating strain as well.

Other sections of the draft that were modified or explained in greater detail are those pertaining to:

• Corrosion testing (effect of manufacturing processes on surface finish);
• Pitting corrosion (impact of device geometry/size on surface finish); and
• Nickel-ion release (test duration for implanted devices).

Biocompatibility Draft Would Ease Testing Requirements

The draft guidance for biocompatibility testing for devices that come into contact with the skin offers a series of updates to be applied toward a separate guidance on the subject, which itself provided a risk-based assessment of whether biocompatibility testing is required. The FDA explained that part of the intent is to reduce, refine and replace animal use in testing when feasible. However, the agency made clear that it is amenable to further revision of the skin-contacting policy as circumstances dictate or permit. The agency is taking comment through Dec. 14, 2020.

This select updates draft takes up biocompatibility for devices and fabrics that come into contact with the skin, and of course relies on the ISO 10993 family of standards. The scope of the updates is limited to devices that come into contact with the skin only, and for any duration. The scope of materials includes 17 materials – such as polyethylenes and polymethylmethacrylate – and four fabrics, including cotton and silk fabrics.

Among the exclusions for materials are those intended to contact breached or compromised surfaces, including skin that has been abraded or shaved. The justification for this set of exclusions is the prospect that leachables can be transferred through the skin that is breached or otherwise compromised. Reprocessed single-use devices are also outside the scope of the draft, as are devices that employ an adhesive for fixation to the skin.

The FDA said that several elements of a manufacturer’s quality system and postmarket controls can aid in identifying biocompatibility issues. Among these is a review of complaint data for issues such as erythema, edema, and delayed Type IV hypersensitivity under Title 21, Part 820.198. Purchasing controls (Part 820.50) and production and process controls (Part 820.70) are other elements of the Quality Systems Regulations that should be well documented in the manufacturer’s device history record (DHR).

All premarket filing types are eligible for the use of this approach to biocompatibility, but the sponsor would have to provide additional information in these filings. Among the additional data points are: 

• A list of all materials that have direct or indirect contact with the skin;
• A statement confirming that none of the exclusions apply; and
• A statement confirming that the materials used have a history of safe use in devices legally marketed in the U.S. The sources of confirmatory data include medical device reports (MDRs) and the results of a literature search.

Applicants for investigational device exemptions (IDEs) would have to provide information on any adverse reactions in periodic IDE progress reports, such as any observed erythema, edema and Type IV hypersensitivity.