Part 2

Date: Tue, 14 Sep 1999 16:25:40 -0500 (CDT)

1.3.3 (continuity)

??quantification and identification of all chemicals below a molecular weight of 1500, including the monomer and their characterization;

??the trace metal/heavy metal analysis; ??the crosslink density (if it is a synthetic and cured material); and

??stability data. Long-term stability and accelerated aging studies (at least to 45?C) to demonstrate the effects of time and temperature on the physical properties and chemical composition of the device (filler and shell) should be provided. Key physical parameters such as viscosity and cohesivity should be measured at each time point. If there are no mechanical changes, the levels of the major components should be measured, but if there are mechanical changes, complete chemical analyses should be conducted to explain the physical changes.

1.3.4 Alternative Filler

??the rationale for the use of the specific alternative material; ??composition of the non-polymer, including characterization of smaller-molecular weight components; ??the method of purification of the non-polymer;

??the source of the non-polymer;

??the structural analyses of the non-polymer, including molecular weight distribution; and

??stability data. Long-term stability and accelerated aging studies (at least to 45

?C) to demonstrate the effects of time and temperature on the physical properties and chemical composition of the device (filler and shell) should be provided. Key physical parameters such as viscosity and cohesivity should be measured at each time point. If there are no mechanical changes, the levels of the major components should be measured, but if there are mechanical changes, complete chemical analyses should be conducted to explain the physical changes.

1.4 Bleed Tests

Bleed testing of saline-filled implants is not suggested.

1.4.1 Bleed Rates of Silicone Gel and Alternative Filler

Because gel or fluid can permeate or bleed through an intact shell, FDA believes that the bleed rate of a silicone gel-filled or of an alternative filler device should be determined. One possible method is described in ASTM F703 ("Implantable Breast Prostheses"). If this method is used, the average shell thickness and surface contact areas with test disks should be provided. Along with the data specified in ASTM F703, the results should include weight measurements of the breast implant along with those of the test and control disks for a better assurance that the disks are absorbing all of the bleed material. A description of the bleed.

1.4.2 Bleed Material Analysis of Alternative Filler

The FDA is concerned with the changes in composition in the filler resulting from long-term chronic bleed of alternative fillers, for which there is little known information. Because of the large number of possibilities of components for alternative filler materials, there is no existing test standard. Sponsors are advised to submit a protocol to PRSB before initiating this testing.

The sponsor should provide a complete description of the testing methodology with a schematic of the setup and a rationale. The rationale for the set-up should be based on the specific chemical make-up of the alternative filler device.

A stirred receptacle medium of physiological saline, phosphate buffered saline, or protein/lipids may be the best means of emulating actual in vivo bleed rates. The solution should be stirred so that the diffusion rate is not artificially slowed by poor mixing. If a sponsor believes that a different solvent is more suitable to the bleed experiment, the rationale should be provided. A knowledge of the specific chemical components of the alternative filler will dictate much of the testing methodology.

The testing should address the possibility that components or degradation products of the filler will diffuse out of the implant, changing the composition of the filler and perhaps the performance of the implant over time. The toxicology of the filler and degradation products should be tested as described below. The simultaneous rupture of two implants should not deliver a toxic dose to the body.

2. Preclinical Data - Toxicology

2.1 General Information

The level of potential local and systemic toxicity of any substance introduced into the body by the breast implant should be assessed. Breast implants contain not only the major polymeric materials (e.g., polymerized polydimethylsiloxane), but also low molecular weight components, such as monomers, oligomers, catalysts, and residues from the sterilization process that can leach out into the patient's body.

The toxicological safety assessment is based on information from two sources, from the chemical composition of the device and from a standard battery of toxicological tests. Knowledge of the total levels of all polymers and residues or components of alternative materials in the final sterilized device provides an upper limit of the amounts of these chemicals that can be delivered locally to the implant site or into the systemic circulation. The identification and quantification of the chemicals present in the device (described in the chemistry section) may enable the usage of available information in the scientific literature about the toxicity and pharmacokinetics of these compounds. Chemicals without adequate safety data in the literature should be subjected to specific testing to obtain the required information. Individual compounds should be tested at or above the local or systemic worst-case concentrations. In some cases, innovative delivery vehicles may be necessary to present chemicals to the toxicological test systems, and solutions mimicking physiological fluids should be considered. For example, an insoluble oil could be delivered in an emulsified form in an isotonic vehicle for cytotoxicity testing. The standard toxicological testing for the elastomer and the filler is described below, in sections 2.3 and 2.4.

The toxicity assessment should initially be based on the worst case levels of toxicants that would result if all the leachable compounds were released from the implant to the body at once. For some chemicals shown to be toxic at the worst-case concentration, however, it may be possible to demonstrate in vivo safety by demonstrating that the actual in vivo concentration levels of the compound will be considerably below the toxic level because of rapid rates of excretion and/or metabolism of the toxic compound.

2.2 Pharmacokinetic Studies

Knowledge of the pharmacokinetic behavior of potentially toxic chemicals is essential for the scientific assessment of the potential human health risks resulting from the implant. Pharmacokinetics may be used to determine the rates of clearance of chemicals from the blood, the distribution in the body, and the routes and rates of metabolism and excretion of device-associated chemicals. The pharmacokinetic study designs chosen are determined by the information needed. When radiolabeling is used, the device should be radiolabeled in ways that will reflect the fates of all of the components of interest. Of toxicological concern are questions regarding the ultimate fate, quantities, sites/organs of deposition, and the rates and routes of excretion or deposition of potentially toxic compounds. For known toxic compounds, e.g., the low molecular weight siloxanes contained in silicone implants, well-supported estimates of the maximal serum concentration and tissue accumulation levels should be estimated to determine if the compounds will produce significant adverse effects. These levels should be compared to the "no observed effect level" or "lowest observed effect level" determined from the literature or from studies using the isolated materials.

2.3 Biocompatibility Testing

A standard battery of toxicological tests is recommended in the ISO-10993 "Biological Evaluation of Medical Devices - Part 1: Evaluation and Testing." This guidance suggests short-term and long-term biological tests that might be applied to evaluate the safety of implanted medical devices. Both the shell and the filler material should be tested separately. The recommended tests include cytotoxicity, short and intermediate-term implantation tests, acute systemic toxicity, hemocompatibility, immunotoxicity, reproductive toxicity, teratogenicity, genotoxicity, and carcinogenicity. Additionally, the sponsor may refer to the guidance, "Required Biocompatibility Training and Toxicology Profiles for Evaluation of Medical Devices - 5/1/95 - (G95-1)" which can be obtained at

This guidance provides an overview of the general types of toxicity testing that should be considered for a medical device. The Special Considerations section 2.4 below provides information to consider related to these tests.

2.4 Special Considerations

The level of immunotoxicity of the shell and any leachable compounds from the shell or the gel should be assessed. Sponsor should refer to the CDRH "Immunotoxicity Testing Guidance" for additional information; this guidance can be obtained at

Reproductive and teratogenicity studies should measure rates of conception as well as record the numbers of fetal deaths and malformations. The studies should include at least two generations. The dose of shell or filler material tested should be at least double the worst-case exposure level, and higher if possible. Individual compounds should be tested at the highest possible exposure that does not produce non-reproductive systemic toxicity.

Subchronic and chronic toxicity testing are essential because the leaching process may be slow, even when the material is in pulverized form, exposing the animals or cells to very small quantities of the leachable toxicants or carcinogens. Implanted material may also degrade over time, producing toxic degradation products. These cases can only be detected by subchronic or chronic implantation tests. The subchronic implantation test reports should include gross and histopathology examinations of the tissue surrounding the implanted material and at appropriate sites remote from the implantation site. Genotoxicity and carcinogenicity information is essential because the potential to cause cancer is an important concern for any implanted device. This potential may arise from leachable compounds and/or degradation products of the device. Therefore, adequate long-term studies with implantation of device materials should be conducted to evaluate the long-term toxic and carcinogenic potential. Complete reports from the genotoxicity testing of shell and filler from the finished sterilized device should be provided. The testing should, at minimum, consist of bacterial mutagenicity (including point and frame shift mutations) and a mammalian cell forward mutation assay. Mammalian cells should also be tested for cell transformation and for genetic damage in tests such as unscheduled DNA synthesis, sister chromatid exchange, or chromosome aberration assays. Widely used, validated assays should be selected. Isolated compounds, mixtures, or extracts that are positive in any of the in vitro genetic toxicology tests should be tested in animals. For silicone implants made using the current silicone chemistry, neither implantation tests nor clinical experience have, thus far, revealed carcinogenic effects. Therefore, FDA may consider approval of an IDE before the carcinogenicity tests are completed if the material(s) are similar and the in vitro genetic toxicology and clinical carcinogenic experience with the materials continues to be negative.

3. Preclinical Data - Mechanical Properties

3.1 General Information

Breast implants are comprised of different designs. The basic components or design features of any breast implant are the shell, filler, and patch (or seal); optional components may include valves and/or adhesives. Breast implants may consist of single, double, or triple lumens. Preclinical testing is necessary to evaluate the material and mechanical properties of the specific breast implant under review.

Because the morphology and integrity of the materials and of the design features can be affected by processing, it is imperative that all testing be performed on finished, sterilized total devices or components (e.g., shell, gel, and valve). If the device is to be sterilized by different methods (e.g., ethylene oxide, gamma radiation, etc.), then preclinical testing should be performed on samples sterilized by the different methods unless an adequate rationale is provided that the change in sterilization method does not negatively impact the mechanical characteristics. Additionally, for samples prepared from silicone-gel implants, there may be difficulty in performing the testing without cleaning, particularly with respect to testing jig grip areas. FDA suggests cleaning only the grip contact areas as described in ASTM F703 to remove the presence of silicone gel and oils.

Testing should be performed on representative models and/or sizes. For example, the sponsor may choose to test the worst case implant model and size or test a range of sizes within a given model; however, the rationale for the model(s) and size(s) tested should be provided. Additionally, when determining what is the worst case implant to test, the sponsor should use implants manufactured with the thinnest shells allowed by the design release criteria. All testing should be performed to a pre-defined failure of the component.

A statistically valid number of samples should be used in each test performed. Complete reports of the preclinical testing should include, at minimum, the following elements:

??identification of the components/devices tested including model and size, sample dimensions, etc. (again note that all testing should be conducted on the final, sterilized version);

??the test set-up and methods including schematic drawings; ??the rationale that testing involved the worst case design and size or, at least, that it involved one that was representative of the other implants under review;

??an explanation of how or why the results are relevant if there are differences between the proposed and tested implants in terms of material, design, or sterilization method;

??the results with standard deviations, as well as the raw data and failure modes/analysis; and

??a discussion of the results in terms of its expected clinical performance, including a discussion of any safety factors.

3.2 Tensile Strength and Ultimate Elongation

Tensile strength and ultimate elongation represent the largest sustainable stress and stretching deformation on a test specimen before rupture occurs, respectively. The testing should be performed on material specimens taken from the thinnest location of the prosthesis shell. FDA suggests following the methodology described in ASTM D412 ("Vulcanized Rubber and Thermoplastic Rubbers and Thermoplastic Elastomers - Tension").

3.3 Tear Resistance

Tear resistance is a measure of the capability of the implant against catastrophic propagation of a puncture or small tear. The testing should be performed on material specimens taken from the thinnest location of the prosthesis shell. FDA suggests following the methodology described in ASTM D624 ("Tear Strength of Conventional Vulcanized Rubber and Thermoplastic Elastomers").

3.4 Integrity of Fused or Adhered Joints

Failure of a fused or adhered joint represents a potential source of leakage of the filler from the device. This testing provides a measure of the resistance of the device to such failures. Each type of patch/shell joint and valve/shell joint should be tested. FDA suggests following the methodology described in ASTM F703. However, unlike ASTM F703, destructive testing should be conducted (i.e., test samples to failure). The force at failure should be reported.

3.5 Fold Flaw

Folds or creases in a breast implant may be induced during the operative procedure, occur as the result of loss of material, or occur if, in the case of saline-filled implants, insufficient filler material was used. These folds or creases in the elastomer shell may compromise the integrity of the device making it more prone to mechanically induced trauma. There is no existing standard methodology to address fold flaw, so a sponsor should provide a test method with an adequate rationale. However, FDA believes that the following test method may adequately address fold flaw issues. An underfilled implant is folded and then placed inside a sealed saline-filled pouch to simulate an in vivo surgical pocket and to hold the folded sample in place. This total construct should be cyclically loaded in compression to runout (X million number of cycles) or failure. The runout value and applied load should be based on expected in vivo cycles and loading subjected to the implant in its lifetime. Adequate rationale for the runout value and applied load should be provided. For all samples, the sponsor should clearly identify the mode of failure (e.g., wear, tears, etc.) and location of failure. Depending on the test method used, abrasion (Section 3.6 below) and fold flaw may be evaluated together. If the sponsor chooses to perform testing that addresses both of these issues together, then the quantity and particle size distribution of the abraded material should be provided, with particular focus on the percentage of particles less than 100 micrometers, including photomicrographic documentation of the particles present in the debris field.

3.6 Abrasion

As stated above, depending on the testing performed, abrasion may be addressed as part of the fold flaw testing. However, a sponsor may choose to develop a specific test to address only wear/abrasion and particulate generation. If the sponsor chooses to do so, then a complete description of the testing method with an adequate rationale should be provided. One suggested specific method of addressing wear/abrasion is as follows. Shell samples are taken from the top of the finished, sterilized devices and loaded with a 1000g mass. An abrading surface of silicone elastomer should be considered because it is more reflective of the in vivo situation. However, if a sponsor chooses to use a different abrading surface because it is not possible to obtain particulate matter using a silicone elastomer surface, then a rationale for the abrading surface chosen should be provided. FDA suggests a mildly abrasive surface be used. In order to better simulate in vivo conditions and to adequately collect particulate matter, the testing should be conducted with the samples immersed in deionized water. The test should be conducted to determine the maximum number of cycles to failure. While failure may be defined as a tear, a sponsor should consider defining failure as a percentage reduction in implant thickness in order to prevent potential ruin of the abrading wheel. After every 10,000 cycles, the abraded particles should be removed and collected and the abraded area of the specimen examined. The testing is stopped after failure or after a justified number of cycles has been reached. The quantity and particle size distribution of the abraded material should be provided, with particular focus on the percentage of particles less than 100 micrometers, including photomicrographic documentation of the particles present in the debris field.

3.7 Static Rupture Testing of Total Device

A compressive force to the breast implant may be applied during daily activities, as well as during mammography or sleeping on the chest. Static rupture testing should be performed to capture the compressive static force required to rupture a total finished, sterilized device. The static loads as well as the mode and location of failure should be reported.

3.8 Fatigue Rupture Testing of Total Device

Most materials are subject to a finite fatigue life when repeatedly stressed or flexed. Repeated compression, folding, bending, or flexing of the device will, with time, weaken the material of the shell and eventually lead to shell failure. These failure mechanisms are addressed by compressive fatigue testing in which a constant compressive force is cyclically applied to an intact breast implant until the device ruptures. The samples should be cyclically loaded in compression to runout (X million number of cycles) or failure at varying loads to generate an applied force versus number of cycles (AF/N) curve. The runout value should be based on expected in vivo cycles subjected to the implant in its lifetime. Adequate rationale for the runout value should be provided. An adequate number of samples should be tested to construct the curve, including the "elbow point", i.e., the location of the maximum change in curvature of the plot. The load at runout, with no failure, should be determined, as well as the mode and location of failure.

3.9 Static Impact Testing of Total Device

Static impact testing should be designed to address a range of worst case trauma to the breast implant, such as in car accidents. No standard methodology exists regarding impact testing, so the sponsor should provide a test method with an adequate rationale. However, the FDA suggests the following. The test set-up should involve a range of weighted strikers and drop heights to simulate the range of worst case trauma. The contact area of the striker, the range of weights, the drop height, and speed should be justified. The impact energy, determined from the total area under a generated stress-strain curve, should be provided, as well as the mode and location of failure.

3.10 Valve Competence

This testing pertains to saline-filled implants with valves, as well as any alternative breast implants with valves. Valve competence tests conducted on saline-filled breast prostheses should demonstrate the resealing capabilities of the valve. The devices can be subjected to hydrostatic forces that tend to force fluid out of the device, causing a deflation and change in size and shape. The most likely source for increased pressure inside the devices would be from patients reclining with various body elements (head, arm, trunk, etc.) pressing on their prostheses. The maximum expected pressures exerted on the device during typical service loading should be defined, and the devices should be tested in a pressure regime that allows for a margin of safety.

FDA suggests the methodology described in ASTM F703. ASTM F703 states that there shall be no leakage observable for five minutes after a normally closed valve is subjected to a retrograde pressure equivalent to 30cm H2O and then to a retrograde pressure equivalent to 3cm H2O. However, FDA does not believe that ASTM F703 tests the efficacy of the device under actual in vivo load conditions. Therefore, the sponsor should predefine a pressure that adequately defines in vivo conditions, with a rationale, and provide testing at that pressure. Thus, sponsors should demonstrate that valve integrity is maintained at actual anticipated maximum in vivo loads, well in excess of those stipulated by the F703 standard. To accomplish this, the devices should be gradually loaded until valve failure occurs and a maximum service pressure can be defined for the device. Whether the failed test valves reseal upon removal of the excess failure-inducing pressures should also be reported. In addition, valve integrity testing should be performed on devices that were used in the fatigue testing described in section 3.8 above. This will provide data on the performance of the valve after simulated use. Pressure at failure of the fatigue-subjected samples can then be compared to those that were not subjected to prior fatigue loads.

3.11 Cohesivity of Silicone Gel or Alternative Filler

This testing pertains to silicone gel-filled and alternative filler implants. Cohesivity testing should be performed to measure both the rheological (flow) properties and the integrity (connectivity) of the gel. Testing should be conducted on gel-fill material obtained from finished, sterilized devices. Two suggested methods are briefly described in ASTM F703. However, because the methods are not completely described, the sponsor should provide a complete description of the test method used, including the pass-fail criteria, with an adequate rationale. The results reported should be appropriate for the testing methodology (e.g., length of pendant gel, level of gel slump, etc.).

CLINICAL DATA

4.1 General Information

FDA believes that a PMA may be filed with a minimum of 2 years of patient follow-up on a sufficient cohort of patients to evaluate the safety and effectiveness of the product. This is based on additional post-PMA filing follow-up for a total of a minimum of 10 years of prospective patient experience. Sample size estimates should be based on the precision of safety and effectiveness outcomes or detecting a clinically meaningful difference at two years but with consideration to lost-to follow-up rates estimated for 10 years of patient follow-up. Safety and effectiveness and risk/benefit assessments must be based on valid scientific evidence as defined in 21 CFR 860.7(c)(2) from well-controlled studies as described in 21 CFR 860.7(f). Studies may include the separate patient cohorts of primary augmentation, primary reconstruction, and/or revision. Because these studies are complicated by the fact that some patients receive implants for different reasons (e.g., a woman may receive one implant for reconstruction and one for augmentation) data should be recorded and analyzed on both a per patient and per device basis. The patient/device is classified by her indication at study entry. The following should be considered when classifying a patient/device:

??If a reconstruction patient undergoes contralateral augmentation, that patient is classified as reconstruction. The device classification is one reconstruction and one augmentation. ??If a revision patient (i.e., the patient entered the study due to replacement of an existing implant, irregardless of the type/manufacturer of the original implant), undergoes contralateral augmentation, that patient is classified as a revision patient. The device classification is one revision and one augmentation. ??If a revision (removal with replacement) occurs during the study (i.e., after initial implantation), the patient/device is classified based on the indication at original implantation at study entry. If patients who undergo removal and replacement with the same manufacturer's implant, then continued follow-up is expected. For patients who undergo removal without replacement or removal with replacement with another manufacturer's implant, then the FDA still encourages sponsors to continue follow-up evaluations. Full patient accounting and adequate and appropriate safety and effectiveness data presentations are essential. Please refer to Appendix I for general suggestions on the minimal type(s) of data and type(s) of data presentation for breast implants.

4.2 Study Design / Statistical Issues

A complete description of the protocol should be provided. This includes explanations of the study objectives, descriptions of primary and ancillary hypotheses, definitions of the study population (i.e., inclusion and exclusion criteria), methods of randomization (if used), number and locations of investigational sites, enrollment procedures, descriptions of surgical techniques, and lists of allowable ancillary interventions and/or drugs (e.g., use of closed capsulotomy, use of intraluminal corticosteroids or antibiotics). In addition, an explanation of how a control group, if utilized for comparison, was selected, or adequate justification for lack of use of a concurrent control group should be provided.