3.1. Experimental implants in rodents

Isolated cases of sarcomas and carcinomas arising in association with metallic or plastic nondegradable foreign bodies and implants (Radio & McManus, 1996) have been reported at many anatomical sites (Table 4). An experimental model of foreign-body or solid-state carcinogenesis is implantation of smooth, nondegradable films subcutaneously in rats or mice (reviewed in Brand, 1982, 1987). Several physical factors are correlated with tumorigenicity, including surface area, surface continuity, size and shape, surface smoothness and erosion resistance (Moizhess & Vasiliev, 1989). Generally, powdered non-metallic materials are not tumorigenic. In various strains of mice, there are genetically determined differences in tumour latency (6–30 months) and frequency. These foreign-body tumours were classified as sarcomas and had a variety of histopathological appearances: fibromyxosarcoma, rhabdomyosarcoma, haemangiosarcoma, reticulosarcoma and osteogenic sarcoma. [The Working Group noted that, as with human sarcomas, these pathological diagnoses might be revised based on modern criteria.] The cell of origin has been postulated to be a pluripotential mesenchymal stem cell originating in the microvasculature (Johnson et al., 1973a,b, 1977, 1980). The sarcomas have been shown to be of clonal origin and transplantable. They usually grow rapidly and spread by local invasion; metastases are rare (Brand, 1982). The term ‘pluripotential cell’ is no longer used and it is currently the view that the cells of origin of mesenchymal tumours cannot be identified. Instead, the tumours are classified according to their direction of differentiation (i.e., where they are going) rather than their histogenesis (i.e., where they came from) (Gould, 1986).

Table 4

Foreign bodies and cancer.

The morphological sequence of tissue reactions to subcutaneous foreign bodies has been described (Brand et al., 1975a). There is an initial acute inflammatory reaction with infiltration of neutrophils and monocytes at the site. Proliferating macrophages and multinucleated giant cells accumulate focally and adhere to the surface. Elongated spindle cells and collagen fibres surround the implant and new capillaries grow in. Fibroblast proliferation and collagen deposition are active until the cells become dormant at about three months. At this time, a collagenous capsule of variable density encases the implant. During this dormant phase, preneoplastic cells have been recovered from the fibrous capsule. Several months later, preneoplastic cells can also be recovered from the implant surface (Buoen et al., 1975). Neoplastic cells have been hypothesized to arise from these adherent preneoplastic cells. Since these sarcomas are presumably derived from primitive vascular cells associated with capillary ingrowth, there is no direct physical interaction between the preneoplastic cells and the surface of the implant during the initial stages of tumour development. Sarcomas are also induced by intraperitoneal implants which have become surrounded by a fibrous capsule (Brand, 1982).

It is proposed that foreign bodies are nongenotoxic carcinogens that act as mitogens by stimulating proliferation of mesenchymal cells that surround the implant. Initiated cells have been postulated to arise spontaneously from microvascular precursors within this proliferating cell population (Johnson et al., 1973a,b, 1977, 1980). Promotion or maturation of these initiated cells was assumed to occur during the dormant phase of encapsulation by dense fibrous tissue (Brand, 1982). Some biochemical and molecular mechanisms that might be responsible for these initiation and promotion stages are discussed in Section 5 of this monograph.

Since the initial observation that cellophane films implanted subcutaneously in rats induced sarcomas (Oppenheimer et al., 1948), numerous plastics and polymers developed for use as medical and dental prosthetic devices have been tested for toxicity and carcinogenicity in animals (reviewed in Rigdon, 1975). In contrast to smooth nondegradable films, similar materials implanted subcutaneously or intraperitoneally in particulate or fibrous form do not induce the types of sarcoma that are typical of foreign-body tumours. Even high doses of particulates injected intraperitoneally in rats generally do not induce sarcomas or malignant mesotheliomas, although diffuse malignant mesotheliomas are readily induced by natural and man-made fibrous materials in rats (Pott et al., 1987). An important exception is the induction of both mesotheliomas and sarcomas in rats following intraperitoneal injection of nickel and nickel alloys containing more than 50% nickel (Pott et al., 1989, 1992). Direct intraperitoneal injection of particulates such as titanium dioxide induces a mild, transient inflammatory response followed by clearance of particles to mesenteric lymph nodes with no subsequent fibrotic or carcinogenic response, even after repeated weekly intraperitoneal injections in mice (Branchaud et al., 1993). The initial inflammatory reaction and subsequent fibrosis induced by implanted biomaterials in animals depend on the species and the anatomical site (Brand, 1982).

An acute inflammatory response of variable intensity occurs that involves oedema, leukocytes and erythrocytes, as in any inflammatory reaction. The involvement of specific cell types and the extent of fibrous encapsulation depend on the kind and physical form of material implanted (Rigdon, 1975). Implanted biomaterials become rapidly coated with host plasma proteins including albumin, immunoglobulin type G (IgG) and fibrinogen. Coating with albumin appears to down-regulate the inflammatory response, while adsorption of fibrinogen appears to trigger both coagulation and acute inflammation (Tang et al., 1993, 1996). Persistent inflammation at the surface of biomaterials with localized generation of oxidants not only accelerates autoxidation (Radio & McManus, 1996), but may also induce mutations indirectly in proliferating mesenchymal cells surrounding the implant. Deposition of haemosiderin following haemolysis of extravasated erythrocytes may provide a local source of iron that catalyses generation of hydroxyl radicals from oxidants released by inflammatory cells. So far, there is no evidence for or against this mechanism in rodents or humans, although there is not a good correlation between the intensity or persistence of the acute inflammatory response and the subsequent development of foreign-body tumours (Rigdon, 1975).

The rare cases of human sarcomas arising in association with metallic foreign bodies embedded in soft tissues or bone or in association with prosthetic devices could arise by a mechanism similar to that described for foreign-body tumorigenesis in mice and rats; however, using the same techniques developed in the mouse model, no preneoplastic cells were identified in 50 patients who had received a variety of surgical implants (Brand, 1982). Carcinomas may develop in response to exogenous foreign bodies such as fragments of bullets or splinters of bone. In these cases, prolonged physical damage to epithelial surfaces followed by regeneration of epithelial cells may provide the stimulus for persistent cell proliferation.