A New Mouse Model for Carcinogenesis Testing of Medical Implants

Originally Published MDDI May 2005Testing

May 1, 2005

17 Min Read
A New Mouse Model for Carcinogenesis Testing of Medical Implants

Originally Published MDDI May 2005


The rasH2 mouse model can help device manufacturers and FDA make a safe and accurate decision regarding the carcinogenic potential of a device.

Glenda J. Moser, Michael A. Streicker, and William Wustenberg

Most devices require a series of tests for assessing biocompatibility. Some devices also require carcinogenicity testing in a live animal model. The standard biocompatibility tests found in ISO 10993 are fairly straightforward, and these tests are performed by many laboratories. However, with the advent of new technologies and new animal models for the evaluation of new medical devices, it is difficult to stay current on the available testing options.

One newly available test is the carcinogenicity assay using the transgenic rasH2 mouse model. FDA often recommends this test, but many manufacturers are still unfamiliar with this mouse model and its benefits. This article describes the rasH2 mouse model and explains how this new option in carcinogenicity testing can be used for evaluating the carcinogenic potential of medical device materials.

Carcinogenicity Testing

Carcinogenicity testing using experimental animal models is the most common means of prospectively identifying possible human carcinogens.1 Generally, preclinical tests to evaluate the potential carcinogenicity of devices are time-consuming, laborious, and expensive. To evaluate the carcinogenic potential of chemicals, radiation, biological effectors, and physical agents, a highly standardized carcinogenicity assay is employed; both sexes of mice and rats (50 per sex per species) are exposed to the agent of interest at multiple dose levels in a 2-year cancer bioassay. Nearly all known human carcinogens are carcinogenic in one or more rodent species.2

In the last few years, new transgenic models have been developed that are being accepted by FDA for use in characterizing the carcinogenicity of various compounds and medical device materials. One of these models is the rasH2 transgenic mouse. The rasH2 transgenic mouse model has been used to evaluate the carcinogenic potential of device materials. Such materials include those with positive results in a short-term genotoxicity test, without a long history of safe use in humans. Other materials tested include those having constituents that have been identified as carcinogens that may be released into the body. FDA has accepted data from this model for premarket approval and 510(k) clearance, and current studies using this model are ongoing.

Biomaterials Testing

The Medical Device Amendments of 1976 gave FDA authority to regulate medical devices. These amendments were established to ensure the safety of medical devices in humans. The ISO 10993 series of standards was created to provide methodologies for evaluating the local and systemic toxicity potential of medical devices and/or the leachable constituents of the device material. The types of preclinical laboratory studies conducted to detect potential toxic effects vary with the
tissues or organ systems exposed, the length of exposure, and the findings in previous studies.3,4

Methodologies for the genotoxicity and carcinogenicity evaluations of solid-state materials are addressed in ISO 10993-3. Genotoxicity and carcinogenicity studies are conducted to determine the potential for agents to alter DNA or the potential to increase the incidence or number of preneoplastic, neoplastic, and nonneoplastic toxic lesions.

Screening for potential toxic effects on the human genome is accomplished by the use of genotoxicity assays. Genotoxic agents interact with DNA, causing gene mutations or changes in chromosome structure or number. There are a variety of short-term in vitro and in vivo genotoxicity assays in mammalian and nonmammalian cells and species. A major advance in predictive toxicology came with the development of the Ames Salmonella bacterial reverse mutation assay. When strains of Salmonella with specific mutations are exposed to mutagenic agents, some mutations reverse back to normal and grow on minimal agar plates with glucose. The rate of reversions of the test agent is compared with the rate of negative controls. Short-term assays in mammalian cells with targeted genes may be used to complement the Salmonella assays.

A variety of other assays can also be used to identify genotoxic agents. These include tests for DNA adducts, damage, and methylation; alterations in chromosomal integrity such as chromosome breaks, aneuploidy, and sister chromatin exchange; and identification of specific mutations. Chemicals or other agents that produce a consistent response in short-term tests of mutagenicity are defined as acting through genetic mechanisms.

Although not all mutagenic agents cause cancer, it is often assumed that there is a human risk associated with exposure to genotoxic agents. Thus, genotoxic agents are often tested for their ability to induce cancer before regulatory agencies grant premarket approval (PMA) or 510(k) clearance. A typical genotoxicity test battery conducted for permanent implantable materials includes three assays: a bacterial reverse mutation assay, a mammalian cell forward mutation assay or mammalian cell chromosomal aberration assay, and an in vivo evaluation such as the mouse micronucleus assay.

Transgenic Animal Model

Proto-oncogenes play a role in normal cell growth and division. The mutation or altered expression of proto-oncogenes may confer a cancerous phenotype by producing normal signals at the wrong time or at variant expression levels. One oncogene protein that controls normal cell growth is p21 Ha-ras.5-7 A mutation in the p21 gene is believed to cause cells to grow uncontrollably and divide. Such growth has been shown to contribute to cancers of a variety of organs including the breast, bladder, lung, kidney, and colon in humans and rodents.8-10

Susceptible populations are at an increased risk of developing cancer as a consequence of a genetic predisposition.11 These individuals commonly have alterations in proto-oncogenes or other genes that predispose them to tumorigenesis. To identify potential human carcinogens, genetically altered animal models that are more susceptible to various carcinogens than normal, nongenetically altered animals have been constructed for various research approaches. In some of these animal models, foreign genes are introduced and are known as transgenes. Organisms that contain foreign DNA are known as transgenic.

Transgenes are integrated into apparently random sites in the host genome. Every cell of the host has the inserted gene, and stable expression of the gene has been confirmed in all major organ systems. The introduced gene can be transmitted through the germ line to subsequent generations in a Mendelian fashion. New strains, hemizygous and homozygous for the transgene, can be generated by selective breeding. Wild mice have two normal alleles for the gene of interest; hemi-zygous animals have one normal and one transgene; homozygous mice have transgenes at both alleles.

Although there are hundreds of transgenic models of human diseases, relatively few are widely accepted models to identify human carcinogens.12 Transgenic carcinogenic models generally use fewer animals for a shorter period of time at a lower cost than the traditional 2-year bioassay. Furthermore, transgenic models can contribute additional insights into the mechanisms of carcinogenesis. These models may be more relevant in certain situations for evaluating potential human risk. The ability to genetically manipulate the genome has enabled a targeted approach to the molecular study of toxicology and carcinogenesis at the level of the whole animal.

Solid-State Tumorigenesis

The predilection of rodents to exhibit solid-state tumorigenesis has confounded the ability to conduct standard 2-year rodent bioassays. Mice or rats implanted with smooth-surfaced device materials exhibit sarcoma development at the host implant interface with a latency period of 11–24 months.13

As detailed in the IARC monograph that covers surgical implants and other foreign bodies, all solid (low-porosity) polymeric materials evaluated in numerous rodent studies have demonstrated the solid-state effect, regardless of polymer chemistry.13 The carcinogenic effect of an implanted material in the rodent models (both 2-year studies on wild rodents and 6-month studies on transgenic mice) are influenced not only by the material composition but also by the size, continuity, pore size, and surface condition, and, for subcutaneous testing, implant thickness. It is therefore important to select appropriate controls and testing methodology that separate material-related carcinogenic effects from the ones that derive from the implant's physical characteristics.

Previous carcinogenicity testing in the rodent models using different-pore-sized Millipore cellulose filters have been reported.14 Millipore cellulose filters ranging in pore size from 0.025 to 8.0 µm induced tumor-development rates at the implant site ranging from 100% down to 0%, respectively. Filters 20 mm in diameter with a 0.22-µm or larger pore size were considered nontumorigenic in the lifetime rodent studies. ISO 10993-3 and ASTM 1439-92 recommend using controls of similar porosity to minimize the solid-state effect. However, many materials are of very low porosity, similar to the Millipore filter with a small pore size (0.02 µm). Therefore, it is problematic to embark on a study for which the test material is expected to have a response similar to that of a positive control.

This solid-state tumor effect is seen in rodents but does not seem to affect human studies. It poses an obvious problem concerning data interpretation. When the test material shows a tumor response similar to that of the positive control, the solid-state tumor response may be caused solely by the solid-state effect in rodents and not by the chemical entity or degradant leaching from the device. It would be difficult, if not impossible, for a pathologist to consistently differentiate tumors caused by the solid-state effect from those caused by the implant.

Based on the results of several studies, it has been proposed that the rasH2 transgenic model, which makes mice more sensitive to neoplastic induction following exposure to many chemical agents, does not shorten the latency period for solid-state tumorigenesis.15 This beneficial property allows these mice to demonstrate chemical carcinogenic response without the solid-state tumorigenesis response seen in the traditional 2-year rodent bioassays.

The rasH2 Mouse Model

Following the publication of the ICH guidance in 1998, international regulatory guidelines allow for a short-term alternative bioassay in transgenic mice as a substitute for a 2-year rodent bioassay.16 FDA sometimes proposes to a device company the use of a transgenic model in evaluating medical devices for carcinogenicity. One such transgenic model is the rasH2 (CB6F1TgN) mouse hemizygous for the H-ras proto-oncogene.17,18 The rasH2 transgenic mouse, established by Saitoh and coworkers at the Central Institute for Experimental Animals (Kawasaki, Japan), has been used in previous solid-state carcinogenicity studies at Integrated Laboratory Systems Inc. (ILS; Research Triangle Park, NC) and elsewhere.

Transgenic (Tg) rasH2 mice carry the prototype human c-Ha-ras gene with its own promoter region. In the rasH2 model, there are five or six copies of the human H-ras gene in tandem array with its endogenous promoter and enhancer produced by pronuclear injection.19,20 The p21 protein product of the c-H-ras gene is expressed in normal tissues and tumors at levels that are two to three times greater than those found in wild mice, suggesting point mutations. This also suggests that a two- to threefold increase in p21 is sufficient to cooperate with other carcinogen-induced genotoxic or nongenotoxic changes to develop tumors.

The rasH2 transgenic mouse model detects various types of genotoxic and nongenotoxic carcinogens within 26 weeks.21,22 In addition, the transgenic mice do not demonstrate significant tumor induction when exposed to noncarcinogens. Thus, the rasH2 model has been used as an animal model for the development of a rapid carcinogenicity test system. The rasH2 model continues to be genetically stable and elicits a stable response to a variety of chemicals. Pritchard et al. evaluated the sensitivity and specificity of the rasH2 model for the identification of human carcinogens and noncarcinogens from traditional 2-year rodent bioassay data. Of the 27 chemicals analyzed in 26-week studies, the rasH2 transgenic mouse model accurately predicted the carcinogenic response in 81% of the chemicals evaluated.23

To date, data demonstrate that the rasH2 transgenic model has the potential to play an important role in the identification of potential human carcinogens and can function as an alternative to the 2-year bioassay. Critical requirements for regulatory acceptance of data from this transgenic model and its use in human risk assessment require optimization of study design. Optimizing study design eliminates the potential for false negatives in carcinogen identification. Critical test parameters include the number of animals studied and group sizes, duration of exposure, number of tissues and the extent to which tissues are histopathologically evaluated, positive and negative controls, and the interpretation of results.

Methods for rasH2 Studies

In rasH2 carcinogenicity studies, the test materials and the control materials are surgically implanted into anesthetized mice through an incision on their backs. Implantation into subcutaneous tissue of the mouse simulates the implantation of the medical device into the tissue space of humans.

Although gender-related differences in tumor type or incidence are rare, both sexes are routinely evaluated for tumorigenicity at 26 weeks in the rasH2 model. Treatment groups include mice implanted with low and high doses of the test medical device, leachable chemicals, or particles. The appropriate dosage should be based on the size of the implant relative to the human usage, the amount of debris generated over the lifetime use of the device, or the amount of leachable chemicals or potential toxicants from the device.

Positive-control treatment groups are critical for correct data interpretation and demonstrate tumor responsiveness of the rasH2 model. We recommend the use of a positive control that is implanted in a manner similar to that of the test article. ILS uses a positive control that consistently produces an increased incidence and multiplicity of lung adenomas and carcinomas and hemangiosarcomas in the spleen; some metastasis to lymph nodes, subcutaneous tissue, and other organ systems may be found. Given that lung alveolar epithelial adenomas and splenic hemangiosarcomas develop spontaneously in rasH2 mice at 33–35 weeks of age, it appears that the positive-control chemical enhances the spontaneous tumorigenesis seen in the model. Implanted materials are used as solid-state controls. Some spontaneous tumors and a variety of nonneoplastic lesions are found in control mice at termination of 26-week studies (34–36 weeks of age). However, lesions are minimal at this point and to date have shown limited or no effect on study interpretation.

Full-screen necropsies should be carried out on all moribund mice and mice at the terminal necropsy after 26 weeks on study. Animals should be evaluated for the presence of grossly visible lesions, and all lesions should be collected for future phistopathological evaluation. Location and appearance of the implantation site should be recorded, and if the test material has moved from the site of implantation, an attempt should be made to locate the device, record its location, and take a sample of the test material in its new location. Organ weights may be recorded for select tissues to help better assess the overall health conditions of the mice being studied. Incidences of preneoplastic, neoplastic, and other toxic lesions should be recorded and compared with negative-control mice.

Histopathology experience and an historical database of neoplastic and nonneoplastic findings in control rasH2 mice are critical in the interpretation of lesions in study animals. Implanted materials at the test site are generally encapsulated with infiltration of inflammatory cells. The interpretation of a normal inflammatory response compared with a test-material-induced inflammatory response is critical to the interpretation of the study and to the evaluation of the test materials' carcinogenic potential. Dose-dependent nonneoplastic toxic effects and tumor induction enhance characterization of the medical device as a potential carcinogenic material. Because of the importance of proper experimental design and study conduct and of distinguishing between background tumors and test-article-related tumors, we strongly suggest working with a laboratory experienced in these types of studies.

Why Would FDA Ask for a rasH2 Study? ISO 10993-3 lists the situations that suggest a need for carcinogenicity testing. These include the use of

• Reabsorbable materials and devices, unless there are significant and adequate data on human use or exposure.
• Materials and devices where positive results have been obtained in genetic toxicity on mammalian cells.
• Materials and devices introduced in the body or its cavities with a permanent or cumulative contact of 30 days or longer, except when significant and adequate human-use history is available.


The rasH2 transgenic mouse model can be used to evaluate the carcinogenic potential of medical devices in lieu of the standard 2-year bioassay. Submittal of a study of this nature requires a high level of interaction between the sponsor, the contract laboratory, and the FDA reviewers who are working with the company on the testing program for the device.
The data from a rasH2 mouse model study can assist device manufacturers and FDA in making a safe and accurate decision regarding the toxic or carcinogenic potential of a device. The 26-week time frame allows a much faster time to market for a device than the standard 2-year rodent bioassay, and a well-designed and properly performed rasH2 study can answer many of FDA's questions about a given device.


1. JF Contrera and JJ DeGeorge, “In Vivo Transgenic Bioassays and Assessment of Carcinogenic Potential,” Environmental Health Perspectives 106 (1998): 71–80.
2. GA Boorman et al., “Rodent Carcinogenicity Bioassay: Past, Present, and Future,” Toxicologic Pathology 22, no. 2 (1994): 105–111.
3. DJ Ferguson, “Cellular Attachment to Implanted Foreign Bodies in Relation to Tumorigenesis,” Cancer Research 37 (1977): 4367–4371.
4. MM Iomhair et al., “Effects of 3 Growth Control Substances on Foreign Body Sarcomagenesis: IFN, IudR, MGBG,” Irish Journal of Medical Science 168, no. 1 (1999): 42–44.
5. T Satoh and Y Kaziro, “Ras in Signal Transduction,” Seminars in Cancer Biology 3, no. 4 (1992): 169–177.
6. PD Adams et al., “Identification of a
Cyclin-cdk2 Recognition Motif Present in Substrates and p21-Like Cyclin-
Dependent Kinase Inhibitors,” Molecular and Cellular Biology 12 (1996): 6623–6633.
7. AL Gartel and AL Tyner, “Transcriptional Regulation of the p21 (WAF1/CIP1) Gene,” Experimental Cell Research 246 (1999): 280–289.
8. JM Bishop, “The Molecular Genetics of Cancer,” Science 235 (1987): 305–311.
9. AP Albino et al., “Analysis of ras Oncogenes in Malignant Melanoma and Precursor Lesions: Correlation of Point Mutations with Differentiation Phenotype,” Oncogene 4, no. 11 (1989): 1363–1374.
10. RR Maronpot et al., “Mutations in the ras Proto-Oncogene: Clues to Etiology and Molecular Pathogenesis of Mouse Liver Tumors,” Toxicology 101 (1995): 125–156.
11. L Chin et al., “Cooperative Effects of INK4a and ras in Melanoma Susceptibility In Vivo,” Genes and Development 11 (1997): 2822–2834.
12. RW Tennant et al., “Evaluation of Transgenic Mouse Bioassays for Identifying Carcinogens and Noncarcinogens,” Mutatation Research 365 (1996): 119–127.
13. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 74, Surgical Implants and Other Foreign Bodies (Lyon, France: 1999).
14. Karp et al., “Tumorigenesis by Millipore Filters in Mice: Histology and Ultrastructure of Tissue Reactions as Related to Pore Size,” Journal of the National Cancer Institute 51 (1973): 1275–1285.
15. JS MacDonald, “Human Carcinogenic Risk Evaluation, Part IV: Assessment of Human Risk of Cancer from Chemical Exposure Using a Global Weight-of-
Evidence Approach,” Toxicological Sciences 82 (2004): 3–8.
16. International Conference on Harmonization (ICH) of Technical Requirements for Registration of Pharmaceuticals for Human Use, Step 4, 1998.
17. A Saitoh et al., “Most Tumors in Transgenic Mice with Human c-Ha-ras Gene Contained Somatically Activated Transgenes,” Oncogene 5 (1990): 1195–1200.
18. N Tamaoki, “The rasH2 Transgenic Mouse: Nature of the Model and Mechanistic Studies on Tumorigenesis,” Toxicologic Pathology 29, Supplement (2001): 81–89.
19. H Suemizu et al., “Transgene Stability and Features of rasH2 Mice as an Animal Model for Short-Term Carcinogenicity Testing,” Molecular Carcinogenesis 34, no. 1 (2002): 1–9.
20. M Katsuki et al., “Chemically Induced Tumors in Transgenic Mice Carrying Prototype Human c-Ha-ras genes,” Princess Takamatsu Symposium 22 (1991): 249–257.
21. S Yamamoto et al., “Rapid Induction of More Malignant Tumors by Various Genotoxic Carcinogens in Transgenic Mice Harboring a Human Prototype
c-Ha-ras Gene Than in Control Non-Transgenic Mice,” Carcinogenesis 17 (1996): 2455–2461.
22. S Yamamoto et al., “Validation of Transgenic Mice Carrying the Human Prototype c-Ha-ras Gene as a Bioassay Model for Rapid Carcinogenicity Testing,” Environmental Health Perspectives 106, no. 1 (1998): 57–69.
23. JB Pritchard et al., “The Role of Transgenic Mouse Models in Carcinogen Identification,” Environmental Health Perspectives 111, no. 4 (2003): 444–454.

Glenda J. Moser, PhD, is director of toxicology and Michael A. Streicker is study director in the health sciences division of ILS (Research Triangle Park, NC). William Wustenberg, DVM, is a consultant specializing in biomaterial quality systems.

Copyright ©2005 Medical Device & Diagnostic Industry

Sign up for the QMED & MD+DI Daily newsletter.

You May Also Like