Medical Device Amendments of 1976

The 1976 Medical Device Amendments to the 1938 Food, Drug and Cosmetic Act (FDCA) gave FDA the responsibility for regulation of medical devices sold in the U.S. Not all devices are regulated in the same manner. Class I devices such as eyeglasses, tooth brushes, scalpels, prosthetic heart valve sizers and stomach pH electrodes are, for the most part, exempt from premarket review by FDA, although they can be subject to some or all quality system (QS) regulation manufacturing and development controls.

Preclinical Testing and IDEs

Preclinical testing is an integral part of the product development process. If required, much of this testing must be completed and reports available, prior to the start of human testing. 510(k) notifications for Class II devices with direct patient contact or contact with the blood supply will contain biocompatability testing data, according to ISO 10993, in order to demonstrate that the risks posed by the new device are “substantially equivalent” to the risks posed by the predicate device. Preclinical testing can also include animal and bench studies to demonstrate the effectiveness of the new device, confirm electrical safety or investigate electromagnetic interference. Although it is often prudent to discuss preclinical testing plans with FDA, no prior approvals are necessary in order to perform this testing. Both the Investigational Device Exemption (IDE)¹ and the PreMarket Approval (PMA)² regulations require that nonclinical laboratory studies be performed under the Good Laboratory Practices (GLP) regulations in 21CFR 58,³ or that an explanation be provided if a study was not performed in accordance with the GLPs. Once preclinical testing has progressed to the appropriate stage, the reports are written, overall summaries of related tests compiled, and overall conclusions drawn. For relatively simple devices, conclusions of substantial equivalence may be drawn from preclinical testing alone. When complex devices are considered, preclinical testing serves as the foundation for more extensive clinical testing.

A relatively small number of 510(k) notifications, perhaps 5%,⁴ also contain human clinical data. As shown in Figure 3.1, there are two routes available in order to initiate a human clinical trial. Non Significant Risk studies involve devices that pose little risk to the study population and can be initiated after local Institutional Review Board (IRB) review, while Significant Risk studies require submission of an IDE and IRB approval prior to study initiation. Informed Consent must be obtained from all study subjects, regardless of the risk status of the device, unless the emergency use provisions of the IDE regulations are invoked.

An IDE application contains a description of the device, theory of operation, indication for use and method of manufacture. Available preclinical data, including biocompatability information is also included in the IDE. A clinical study protocol, clinical investigator information and a clinical monitoring SOP must also be submitted. If the manufacturer plans to charge for the investigational device, this must be stated in the IDE. FDA permits recovery of research and development costs but will not allow commercialization during the investigation. Once the IDE is received, FDA has 30 days to review the application. The FDA reviewer's main concern is to assure the level of risk incurred by the patient population is appropriate given the benefits of the investigational device. The IDE can be approved, rejected or FDA can ask for additional information before making a final determination. Typically, about 70%⁵ of IDEs are approved during the first round of review. Once the IDE is approved, the manufacturer and clinical investigators must comply with a wide variety of regulations which are described in 21CFR 812. Approval of an IDE does not imply that marketing approval via the 510(k) or PMA routes is assured. It only signifies that the clinical research does not pose unreasonable risks to the study population. The manufacturer must periodically update FDA regarding the progress of the investigation.

510(k) and PMA Marketing Applications

More than 99% of all the medical devices that FDA reviews in a given year are cleared for marketing through the 510(k) PreMarket Notification process. (510(k) refers to Section 510(k) of the FDCA.) The goal of the process is to demonstrate to FDA that the new device is “substantially equivalent” to a “predicate device” which was already on the U.S. market prior to the May 28, 1976 effective date of the Medical Device Amendments. Historically, FDA has interpreted substantial equivalence in a broad sense. For an electronic device with no direct patient contact, a table containing a point-by-point comparison of key engineering features of the device is often sufficient. Class II devices such as some implants, many catheters, electrocardiograph systems, and ultrasound imaging systems are all included in an intermediate risk category where 510(k) notifications and adherence to FDA Quality system regulations are required. Under the Medical Device Amendments of 1976, the submitter of a 510(k) notification could market a device 90 days after submitting the notification to FDA. As we shall see later in this chapter, the Safe Medical devices Act of 1990 made several significant changes to the 510(k) process.

Approximately 40 medical devices are approved for marketing in the U.S. through the PMA process each year. All PMA devices are Class III. The PMA process is used for devices that cannot be shown to be “substantially equivalent” to a suitable predicate device. This commonly occurs because the device utilizes a novel technology or material, or is intended for a novel indication for use. The goal of the PMA process is for the manufacturer to provide “reasonable assurance” that the new device is safe and effective. This is a higher standard than the 510(k)'s “substantial equivalence” and nearly always requires manufacturers to perform human clinical trials. Biomaterial and biocompatability concerns, as with 510(k)s, are related to the type of device, materials used and the indication for use, rather than the application type. PMA review times generally extend from 9–18 months.

Safe Medical Devices Act of 1990

The Safe Medical devices Act of 1990 (SMDA) [http://www.fda.gov/cdrh/ode/655.pdf] modified the 1976 Medical Device Amendments in many key areas. SMDA permits applicants to utilize any device that has been cleared through the 510(k) process as a predicate device. It is no longer necessary to select a device that was on the U.S. market prior to May 28, 1976. With regard to PMAs, SMDA makes advisory panel review optional so that the panels can focus on the most innovative technology. SMDA also contains the “four of a kind” rule which could permit submitters of the fourth PMA for a particular device to reference data submitted by the first three applicants. These data, could potentially, contain biocompatability information. As with several other provisions of this law, FDA has not drafted enabling regulations for this section and it is not currently being used. Another provision of SMDA gives FDA the ability to regulate the design process for medical devices. After considerable public comment, FDA published final regulations, which implement this provision and significantly revise all the manufacturing regulations. These changes are discussed in the following section.

The FDA Modernization Act of 1997 (FDAMA97)

This wide-ranging law [http://www.fda.gov/cdrh/modact/modern.html] adds many features to the medical device regulatory landscape. One provision, Section 216 will permit FDA to use biocompatability data submitted in one PMA when reviewing another PMA, once the original data is at least six years old. There are also several provisions that detail various types of meetings that sponsors of IDEs or PMAs can request with FDA. This law also formalizes FDA's reliance on compliance with recognized national and international standards to expedite the 510(k) review process.

FDA Regulation of Biomaterials

ISO 10993¹²

The International Organization for Standardization's Technical Committee 194, comprised of members from the European Committee for Standardization (CEN) and the U.S.-based Association for the Advancement of Medical Instrumentation (AAMI), produced the ISO 10993 standard “Biological Evaluation of Medical devices Part 1: Evaluation and Testing” in 1994. FDA issued a memo in 1995 [http://www.fda.gov/cdrh/g951.html] that made compliance with a modified version of ISO 10993 mandatory after July 1, 1995. This document contains two tables (Refer to Figs. 3.2 and 3.3) and a flow chart (Refer to Fig. 3.4) to assist manufacturers in determining the appropriate testing plan for their device. Additional modifications to the ISO 10993 matrix which specifically relate to immunotoxicity were announced by FDA in the Immunotoxicity Testing Guidance dated May 6, 1999 [http://www.fda.gov/cdrh/ost/ostggp/immunotox.pdf]. The FDA modified ISO 10993 testing matrix (refer to Fig. 3.2 and 3.3) in several specific areas.^13,14 First, for surface devices contacting mucosal membranes, acute and chronic toxicology testing along with implantation testing is expected in most cases. Second, for externally communicating devices, those that contact tissue or bone are often expected to include irritation, systemic toxicology, subchronic and chronic toxicology testing. Another important feature of the ISO 10993 standard that FDA accepted is a change in the time periods for the three exposure categories. The ISO 10993 matrix uses the contact terms “limited “ (>24 hours), “prolonged” (24 hours to 30 days), and permanent (>30 days). The most important change here is that the limited term extends to <24 hours, rather than the Tripartite's <5 minutes. This permits a greater number of devices to be assessed under the simpler “limited” requirements. In documentation that accompanies the testing matrix, FDA makes the point that test selection should be based upon sound scientific reasoning. Specific biomaterials utilized for particular indications may require more or less testing than the matrix specifies. Additional, more specialized tests may also be necessary. It should be kept in mind that samples of devices used for this testing should resemble the actual marketed product as closely as possible. Devices sold as sterile should be sterilized using the method that will be used for commercial production prior to biocompatability testing. In cases where aging may affect biocompatability, devices should be aged to the end of their shelf life before testing. Organizations with internal biocompatability experts can often make these judgments independently, while other organizations frequently enlist the assistance of outside consultants and scientists at contract testing laboratories. Contacts with FDA prior to embarking on a time-consuming and expensive testing program are also a prudent course of action.

Figure 2

Initial evaluation tests for consideration. X = ISO Evaluation tests for consideration; O = Additional tests which may be applicable; Note + Tissue includes tissue fluids and subcutanous spaces; Note ^ For all devices used in extracorporial circuits; (more...)

Figure 3

Supplementary evaluation tests for consideration; X = ISO Evaluation tests for consideration; O = Additional Tests which may be applicable; *See Figure 3.2 for initial evaluation tests.

Figure 4

Biocompatibility flow chart for the selection of toxicity tests for 510(k)s.

Table 2

Principles for the evaluation of medical devices*.

FDA Product Specific Guidance Documents

FDA's Center for Devices and Radiological Health (CDRH) publishes a wide variety of guidance documents to assist medical device manufacturers in determining CDRH expectations for data in marketing applications. The three best sources for obtaining any of these documents are the FDA itself. FDA maintains a World-Wide Web Site at [http://www.fda.gov] which has extensive material from CDRH. The Division of Small Manufacturer's Assistance (DSMA), at CDRH, also maintains a Facts-On-Demand service at (800) 899-0381 (in USA) or (301) 827-0100 (international). The user calls this number from a touch-tone telephone and follows the automated instructions to receive a lengthy index of documents via FAX. Once the index is available, guidance documents can be ordered with a second telephone call. Finally, DSMA can be contacted directly at (800) 638-2041 (in USA) or (301) 443-6597 (international). A wide variety of private organizations provide FDA documentation for a fee. These organizations include newsletter publishers and CD-ROM publishers.

Table 3

Paradigm for biological evaluation and risk assessment of medical devices.

Biocompatibility Testing Program Examples

In the real world, it is difficult for any guidance document or international standard to anticipate all the issues that may arise during the device development process. Device developers must utilize sound professional engineering, scientific, medical, and regulatory judgment when planning a project. In some cases, going beyond the tests specified in a standard or guidance document may make both good scientific and business sense. In other cases, specific tests may be avoided if they are not scientifically justified or if the relevant data already exist. The following section contains examples of how two products that required biocompatability testing were handled. In the first case, a complicated product, used to support life, was subjected to extensive testing. The second case describes a less critical device where much of the data were obtained from previous testing programs.

FDA Regulation of the Biomaterials Testing Process

In the previous sections FDA data requirements have been discussed. We have answered the question, “What tests do we need to conduct?” In this section, we address the question, “How is the recommended testing conducted?”

Description of GLP Regulations

Safety testing data such as biocompatability data are vital components to the marketing application. In order to assure that these data actually represent the true experimental results, FDA regulates safety testing with the Good Laboratory Practices (GLP) regulations (21CFR 58).³ The regulation requires documented controls for the regulation of organization and personnel, facilities, equipment, testing facilities operation, test and control articles and study protocols. One key provision of the regulation is the establishment of a Quality Control Unit (QAU) which periodically and independently monitors GLP-regulated studies. Just as FDA investigators can inspect manufacturer's facilities, GLP laboratory facilities are also subject to on-site inspections by FDA. When choosing a contract testing laboratory it is important to select one with a good compliance history. In extreme cases, FDA has refused to review data from a particular laboratory because of a high level of noncompliance with the GLP regulations. Careful selection of contract testing laboratories, periodic auditing and maintenance of open communication channels will reduce the likelihood of compliance issues.

Examples of GLP Noncompliance Taken From FDA Warning Letters

From time to time, inspections of GLP-regulated testing laboratories result in the issuance of a warning letter to laboratory management detailing apparently serious violations of regulations. These publicly available documents list the major violations found and may include sanctions against the laboratory including disallowal of study data. One must always bear in mind that warning letters are written to prove a point and additional data may change initial FDA conclusions. The following excerpts from FDA warning letters, for medical device studies, serve to illustrate the GLP compliance areas that FDA considers important.

Training for Study Personnel

“Failure to have documentation to show that each individual participating or auditing this study had received any training in GLP's or that this facility provided GLP training as required by 21 CFR 58.29(a)(b).

The testing facility management failed to assure that the personnel followed the current written standard operating procedures entitled, “[purged text] Biomedical Research Standard Operating Procedures,” dated September 30, 1997. For example, in Study [purged word] the testing facility failed to have current training summaries for each person involved in the study and documentation defining the critical phases and associated QAU audit schedule.”

SOPs for Study Activities

“Failure to have written standard operating procedures established, during the period of this study, e.g., animal room preparation, test systems observations, data handling, storage and retrieval, laboratory test and the housing, feeding, handling and care of the animals as required by 21 CFR 58.81(b)(1–12).

There were no written procedures available for the receipt, security, storage, maintenance, disposition, or inventory control of test articles; for laboratory tests, such as blood chemistry, urinalysis, and histological analysis; for specimen collection and labeling requirements; for histopathology; and for specimen archiving.”

Key GLP Compliance Points to Consider When Reviewing GLP Studies

As the points included above illustrate, documentation is a key component of all GLP systems. All tasks required by regulations must be documented or FDA will assume that they were not performed.

The Quality Assurance Unit (QAU) receives a great deal of attention during an FDA inspection as the first few items in this list indicate. QAU personnel must be trained, have adequate time and be knowledgeable enough to perform their review functions assuring that study protocols and laboratory SOPs are followed. There must be enough people in the QAU so that each study can be controlled in an adequate manner. The Standard Operating Procedures (SOPs) for the QAU must describe an appropriate QA system in a sufficient level of detail. Audit reports and logs must also be carefully maintained. When testing follows internationally recognized standards, the laboratory must have current copies of those standards, and their procedures must accurately reflect requirements in the standards.

Study protocols must be carefully drafted to fully describe the research effort, all test article used, number and types of animals used and their diet and all other related study procedures. It is especially important in surgical implant studies to carefully describe the location and method for placing the device into position.

Data processing and data management usually occur during or at the end of most studies and can sometimes be given less attention than the portions of the study protocol that directly relate to handling or observation of the device. No matter how well these earlier parts of the study are executed, the raw data only gain full credibility after the analysis phase of the study. These concerns begin when raw data are initially entered into study notebooks or electronic files and continue when data are placed into databases for statistical analysis.

General laboratory systems are also regulated. In order to generate credible data, laboratory equipment from electronic balances to spectrophotometers must be properly maintained and regularly calibrated. Of course, all these activities must be documented.

Validation of laboratory methods and equipment can often become an issue with GLP studies. Assays performed exactly as they are specified in the USP or standard AOAC methods do not need to be validated. If, on the other hand, an assay from the research literature is utilized, or a special purpose assay is developed, then validation becomes necessary. The validation process provides assurance that the assay accurately and reproducibly measures the parameter of interest under all expected conditions and using defined test equipment.^15–17 Of course, there must also be assurance that the parameter measured does indeed reflect the condition monitored. If the analysis method should change, revalidation may be necessary. Calibration, as mentioned above, is an important part of laboratory operations, but it is not a substitute for validation. A properly calibrated instrument can generate highly accurate, but irrelevant, data if the overall assay method is flawed.

Electronic data capture, analysis and storage is becoming widespread in all areas of biomaterials testing. These types of software offer analytical capabilities that were previously unavailable and save considerable quantities of time. As with assay methodology, software that manipulates data must be validated. Often, vendors of software specifically designed for the GLP environment will perform validations and have a mechanism in place to make the results available to regulatory authorities, if requested. Software which is well known to the regulators such as Lotus 1-2-3 or Excel do not, in general, need to be validated. However, if a researcher writes a macro using a spreadsheet, the user-defined macro must be validated to assure that it processed data in the expected manner.

The Federal Register Notices can be found on [http://www.access.gpo.gov/su_docs/aces/aces140.html]

Electronic data storage presents other challenges. Well run paper based systems provide for audit trails when data must be corrected. These notations enable a reviewer to immediately learn who made the change, when it was made and what the previous value was. This audit trail capability must be present in electronic systems. The system must be validated to assure that data cannot be changed with an audit trail being created and that all other data remains undisturbed. System security issues such as assuring that only authorized personnel have data modification privileges must also be validated. The FDA Electronic Signature Regulation (21CFR11) contains the regulatory requirements for these systems. Electronic data management systems can save enormous quantities of time in the testing lab; however, measures must be take to assure that they perform only as intended.

Conclusion

Biomaterials involve a wide variety of regulatory issues throughout the product life cycle. Recent changes in regulations have focused an increased level of attention on the design phase of the device development process. This attention originates both from the U.S. FDA, in the form of the Quality system Regulations, and internationally from the ISO 10993 standards. Awareness of regulatory requirements and informal expectations early in the planning process can speed the entire development process, allowing the general population rapid access to innovative technology.