Medical Device & Diagnostic Industry Magazine | MDDI Article Index

Originally published June 1996

Brenda Seidman

Last July, the Office of Device Evaluation (ODE) in FDA's Center for Devices and Radiological Health (CDRH) officially replaced its longstanding use of the 1987 "Tripartite Biocompatibility Guidance for Medical Devices" with a new policy designed to bring the agency's requirements into accord with newer international standards for evaluating the biocompatibility of medical devices.¹ In most particulars, ODE's new policy adopts the practices recommended by the International Organization for Standardization (ISO) in its standard on the selection of biocompatibility testing, ISO 10993-1.² However, the agency's blue book memorandum initiating the policy change also includes a fair amount of information about how the agency intends to modify the ISO standard for use in the United States.³ To make the best use of FDA's new policy, manufacturers should familiarize themselves with the details of ODE's memo, and especially with those areas in which the agency's practice differs from the international standard.

Altogether, ODE's memorandum consists of three parts: a guidance letter from ODE director Susan Alpert, a flowchart to help reviewers determine when testing is necessary as well as what kinds of test data are required, and a two-part matrix of rec- ommended testing categories. Alpert's guidance letter provides general background about FDA's previous use of the Tripartite Guidance and the agency's reasons for adopting the ISO 10993-1 biocompatibility standard. She notes that the ISO standard is intended to be a flexible guidance, and that the agency, consistent with the standard, may recommend several tests that are not included in the ISO document. She emphasizes, however, that FDA's use of the ISO standard is also intended to be flexible. "Although several tests were added to the matrix," Alpert writes, reviewers should note that some tests are commonly requested while other tests are to be considered and only asked for on a case-by-case basis. Thus, the modified matrix is only a framework for the selection of tests and not a checklist of every required test. Reviewers should avoid proscriptive interpretation of the matrix. If a reviewer is uncertain about the applicability of a specific type of test for a specific device, the reviewer should consult toxicologists in ODE.³

Alpert's letter also announces that the agency intends to develop device-specific "toxicology profiles" that will assist reviewers in determining appropriate toxicology tests for specific devices. The memorandum provides little information about what FDA expects to be included in these new reviewer guidance materials, or how they might be made available to manufacturers. FDA has been preparing the toxicology profiles since the blue book memorandum was issued last May, and reportedly expects to issue several profiles at once. It is anticipated that the first profiles will be released in early 1997.

As discussed below, the flowchart functions as a decision tree, enabling reviewers of a particular device submission to determine whether it includes sufficient biocompatibility information to meet the agency's requirements. The matrix, which is divided into two tables, provides guidance about what types of testing the agency expects for particular categories of devices. Although it is based on the matrix of ISO 10993, the agency's version incorporates a number of modifications that may change the amount or type of testing required for a particular device. These differences are discussed below.

THE FLOWCHART

As its title implies, the "Biocompatibility Flowchart for the Selection of Toxicity Tests for 510(k)s" is intended to apply primarily to devices undergoing 510(k) review (see Figure 1). For this reason, many of the points considered in the flowchart are designed to enable reviewers to compare the biocompatibility data for a device under review to similar data for its predicate device. FDA's memorandum suggests that the chart may also be applicable to some devices undergoing premarket approval (PMA) review. However, it is not clear from any of the materials included with the memorandum how the agency expects reviewers to apply the flowchart to PMA devices.

In general, the purpose of the flowchart is to enable ODE reviewers to determine whether existing data are sufficient to meet biocompatibility requirements for the device under review or, if they are not sufficient, what additional test data should be submitted by the manufacturer. The chart prompts reviewers of 510(k)s to ask specific questions about a device's materials, manufacturing process, chemical composition, body contact, and sterilization method. Along the way, the chart also directs reviewers to consult device-specific guidance documents, to refer to toxicology profiles, and to seek the assistance of an ODE toxicologist if necessary.

In its final steps, the flowchart prompts reviewers to consider alternative sources of toxicology data (such as device master files and data from previous testing), which may enable a device submission to meet the biocompatibility requirements. Reviewers are also directed to consider whether the submission has included adequate justifications or risk-assessment data for not conducting certain tests. If these sources of information or justifications sufficiently resolve all outstanding questions, the chart advises reviewers to seek concurrence from an ODE toxicologist before definitively concluding that the submission has met its biocompatibility requirements. If the information or justifications do not resolve all questions, the chart advises reviewers that additional toxicology testing is required. Such testing can include some or all of the test categories identified in the FDA matrix, in toxicology profiles, or by toxicologists.

Since this chart was intended for use by ODE reviewers, it may be confusing to many manufacturers. In the absence of a similar guidance document designed specifically for industry, however, this internal document can be a great help to manufacturers. In effect, the flowchart informs manufacturers about ODE's newly adopted policy on biocompatibility and indirectly suggests a method that manufacturers can use for considering biocompatibility issues related to their product submissions. Manufacturers should use the flowchart to help them determine what testing they need to perform on a device or its materials.

To manufacturers, perhaps one of the most useful aspects of the flowchart is the guidance it provides on how to satisfy ODE's biocompatibility requirements without necessarily performing extensive biocompatibility testing. The flowchart presents a variety of methods for demonstrating biocompatibility, not all of which involve testing. For example, the chart leads its user into a series of questions about the similarity of the device and its materials to its predicate:

* Are the materials the same?

* Are the manufacturing processes the same?

* Are the materials' compositions chemically the same?

* Do the devices' materials have the same body contacts?

* Are the devices' sterilization methods the same?

Answering yes to all of these questions is tantamount to meeting biocompatibility requirements. A negative response to any of these questions leads the chart user to further consider material differences, reasons why the differences might not be meaningful, and bench testing data on the device's materials. If significant material differences are identified and appear meaningful, the chart assists the user in identifying the categories of biocompatibility testing that should be addressed.

Manufacturers should study the flowchart carefully to determine areas in which they may legitimately avoid testing. With the assistance of a toxicologist, if necessary, they should then justify in writing why certain tests recommended in the matrix or toxicology profile are unnecessary.

ISO 10993 AND THE FDA MATRIX

FDA's new matrix is very similar to the ISO 10993-1 testing matrix and reflects the agency's desire to harmonize its testing recommendations with international standards (see Tables I and II). FDA has adopted the body contact and contact duration categories used in ISO 10993-1, and also recommends all of the tests recommended by ISO. However, the similarity of the FDA matrix to that of ISO 10993-1 should not be interpreted as an unqualified adoption of the ISO standard. For certain device categories, FDA has identified additional testing that it recommends manufacturers should consider "if applicable."

With regard to the test-selection matrix, FDA's memorandum observes that the agency's recommendations conform to the spirit of the ISO standard. Under the heading of "guidance on selection of biological evaluation tests," ISO 10993-1 states that "Due to the diversity of medical devices, it is recognized that not all tests identified in a category will be necessary or practical for any given device. It is indispensable for testing that each device shall be considered on its own merits: additional tests not indicated in the table may be necessary." By adopting this strategy, FDA in effect acknowledges that relative to the ISO matrix, additional tests may be necessary for some devices and fewer tests needed for others.

Although FDA's memorandum and its attachments do not explicitly recommend protocols, the agency's unwritten policy has been, and continues to be, that manufacturers should use such widely accepted methods as those identified in ISO 10993 and the United States Pharmacopoeia. ODE still expects manufacturers to carefully consider testing methodology and to ensure that testing is designed to evaluate the potential toxicity related to the specific device and its materials.

The FDA matrix uses Xs and Os to distinguish testing recommendations that are identical to ISO's (Xs) from those additional tests that FDA considers possibly applicable (Os). In spite of what may appear as a broadening of ISO's core recommendations, FDA's new matrix has eliminated many tests from those previously recommended by the Tripartite Guidance. Hemocompatibility (including hemolysis), mutagenicity, and implantation testing are no longer recommended for many contact type and duration categories. If one considers "if applicable" testing recommendations, in addition to the ISO core recommendations, the greatest potential reduction in testing is for those devices in the surface devices, mucosal membrane category. As might be expected, the least-affected device category is implant devices involving blood; ISO's testing recommendations were already extensive and were in general agreement with the Tripartite Guidance for these devices.

Most manufacturers are by now familiar with the testing matrix and recommendations of ISO 10993-1, and are aware that the ISO standard includes pyrogenicity testing in the systemic toxicity category (unlike the Tripartite Guidance). They also know whether the ISO standard classifies their devices differently from the way they were classified under the Tripartite Guidance. Therefore, the following overview of the FDA-modified version of ISO 10993-1 considers only the most salient differences between it and the Tripartite Guidance and, more importantly, between the ISO and FDA-modified matrices.

Surface Devices. ISO 10993-1 divides this category into three subcategories of devices according to the nature of their body contact: skin, mucosal membranes, and breached or compromised surfaces. For the skin subcategory, FDA's modified matrix introduces no additional testing recommendations, and has in fact struck acute systemic toxicity from the battery of tests previously recommended by the Tripartite Guidance.

In keeping with the ISO categorization scheme, FDA has added a subcategory for mucosal membrane devices, which the Tripartite Guidance termed "externally communicating, intact natural channels" devices. Relative to the Tripartite Guidance, the FDA-modified matrix significantly reduces the testing load for such devices, particularly for those with limited contact duration (¾24 hours). However, FDA recommends that manufacturers consider performing additional tests, if applicable, for devices with prolonged exposure (24 hours to 30 days) and permanent exposure (> 30 days). Manufacturers should be especially thoughtful in considering these extra tests. Alpert's guidance letter explicitly mentions extra testing for surface devices with permanent exposure to mucosal membranes, suggesting that FDA is especially interested in additional safety information for these devices.

For devices in contact with breached or compromised surfaces, regardless of their contact duration, FDA may require additional testing beyond that listed in ISO 10993-1.

If it is unclear whether additional testing is expected, manufacturers would be well-advised to contact the relevant ODE office before planning testing or gathering other safety data for these or any other device categories.

Externally Communicating Devices. The ISO 10993-1 matrix divides this category into the three subcategories of blood path, indirect devices; tissue/bone/ dentin communicating devices; and circulating blood devices. Adopting this structure has resulted in some changes in the way FDA categorizes devices. For instance, some devices previously considered in the Tripartite Guidance's "intact natural channels" exposure subcategory, such as laparoscopes, are now categorized as tissue/bone/dentin communicating devices. Other devices formerly in the Tripartite Guidance's "internal device, tissue and tissue fluids" subcategory--particularly those in contact with tissue fluids and subcutaneous spaces--are also now included in the tissue/bone/ dentin communicating device subcategory.

For blood path, indirect devices with limited contact duration, the FDA-modified matrix introduces no testing recommendations beyond those of ISO 10993-1. For such devices with prolonged or permanent contact durations, however, FDA recommends that manufacturers consider some additional testing, if applicable. For the subcategory of tissue/bone/dentin devices, FDA recommends that manufacturers consider additional testing for all products regardless of their contact duration, but particularly for devices with prolonged or permanent contact. Alpert's guidance letter suggests that ODE is more than casually interested in reviewing testing or other documentation to address these contact durations and that such testing might be considered "applicable." Again, this does not mean that original testing need be performed; however, manufacturers should contact ODE to determine whether it indeed considers such testing necessary.

Manufacturers of circulating blood devices may need to perform additional tests for prolonged and permanent exposure durations. Among limited exposure devices, a genotoxicity test should be considered only for those used in extracorporeal circuits.

Implant Devices. As noted above, the Tripartite Guidance's subcategories "internal devices, bone" and "internal devices, tissue and tissue fluids" have been redistributed into the new subcategories of the FDA-modified matrix. For devices remaining in these Tripartite subcategories, the FDA matrix recommends additional testing regardless of contact duration.

The testing recommendations of ISO 10993-1 for implant devices in contact with blood are quite extensive, especially for devices in the prolonged and permanent exposure categories. For this reason, FDA identified only one additional test (a subchronic toxicity test), for devices with prolonged contact duration exposures.

ELIMINATING UNNECESSARY TESTING

Consistent with ISO 10993-1, specialized testing not identified in FDA's new matrix may also be expected, depending on the device and its materials. ODE has indicated that it intends to use the planned series of device-specific toxicology profiles to identify additional tests that it considers relevant, which may include immunological and neurotoxicological evaluations for certain devices. According to Alpert's guidance letter, the toxicology profiles will be generated first for devices in prolonged and permanent contact duration categories and for those that represent a large share of all submissions. In the interim--and for devices for which toxicology profiles will probably not be generated--ODE expects manufacturers to consult with appropriate individuals in the relevant review division or to obtain expert advice from other knowledgeable individuals. Such consultations can be invaluable in helping manufacturers determine what testing they must perform, and especially in enabling them to avoid expensive, long-term testing whenever possible.

Even if identified among the ISO core testing recommendations, certain areas do not necessarily need to be addressed with testing. The FDA memorandum advises both reviewers and manufacturers to avoid using the matrix as an inflexible checklist and to seek assistance in identifying and justifying testing that is unnecessary for a device or material. Taken together, the three parts of the blue book memorandum suggest several instances when testing might not be needed:

* For well-characterized materials with long histories of safe use.

* For materials acceptably evaluated and documented in device master files.

* For materials and devices for which appropriate risk assessments have been developed.

* When other acceptable justifications are provided for not performing tests on a specific material or device.

Such exceptions are not new; depending on the device and its perceived risk--and the reviewing division or individual reviewer--FDA has often accepted well-developed justifications for not performing testing or for testing materials from something other than the finished device.

CONCLUSION

Although ODE has many competent individuals who review biocompatibility documentation, few are toxicologists. PMA reviews almost always involve one or more toxicologists, but 510(k) reviews may not. Under these circumstances, nontoxicologist reviewers of 510(k)s may defer to ODE toxicologists when they have questions. Nevertheless, other reviewers may genuinely not recognize when the input of an ODE toxicologist is necessary. Should a reviewer's request for additional testing appear inappropriate, the submitter should not hesitate to question the reviewer about the basis of the request.

FDA is attempting to pursue a more informed and rational approach to its reviews of biocompatibility. Manufacturers should consider this an opportunity to avoid doing unnecessary testing. However, to take full advantage of this opportunity, manufacturers must themselves seek appropriate advice from knowledgeable individuals, and must approach their evaluation of testing requirements in a sophisticated manner. It is always to the advantage of the manufacturer to develop a well-written, scientifically sound justification for any decision not to conduct testing that is recommended in FDA's matrix.

REFERENCES

1. Toxicology Subgroup, Tripartite Subcommittee on Medical Devices, "Tripartite Biocompatibility Guidance for Medical Devices," Rockville, MD, FDA, Center for Devices and Radiological Health (CDRH), 1987.

2. "Biological Evaluation of Medical Devices, Part 1: Evaluation and Testing," ISO 10993-1, Geneva, Switzerland, International Organization for Standardization, 1994.

3. "Use of International Standard ISO-10993, 'Biological Evaluation of Medical Devices, Part 1: Evaluation and Testing'," Blue Book Memorandum G95-1, Rockville, MD, FDA, CDRH, Office of Device Evaluation, 1 May 1995.

Brenda Seidman is president of Seidman Toxicology, a consulting firm located in Falls Church, VA.

Copies of ODE's "Use of International Standard ISO-10993, 'Biological Evaluation of Medical Devices, Part 1: Evaluation and Testing'," including the flowchart and matrix, are available from CDRH's automated "Facts on Demand" system by calling 800/899-0381, and requesting shelf num-ber 164.