by Michael Potter, Chief of the Laboratory of Genetics, NCI
In January 1992, David A. Kessler, Commissioner of the Food and Drug Administration (FDA) requested a voluntary moratorium on the clinical use of silicone-gel-filled breast implants until more was known about the biological activity of the components (1). Our laboratory had been studying materials such as paraffin oils and solid plastics that induce the formation of plasmacytomas (PCTs) in genetically susceptible strains of mice, and the FDA moratorium rekindled our interest in exploring silicone as another possible plasmacytomagenic material. Our studies now show that injection of silicone gel into the peritonium of Balb/c mice induces plasmacytoma formation.
As a poorly digestible chemical, the physically stable silicone gel represented a new type of material that would not produce the mechanical effects of the solid plastics or have the liquidity of the paraffin oils, such as pristane, which we had studied extensively. The chemistry of silicones and, particularly, the complex silicone gels, could open new possibilities for studying the tumorigenic properties of these agents, which have similarities to solid-state carcinogens (2). In general, such agents evoke chronic inflammatory reactions.
Silicone gels were obtained from mammary implants purchased commercially. Manipulation of these gels was difficult because they are highly sticky, very elastic, and could not be cut into fragments and inserted into the peritoneum. Each time a small amount was pushed into the space and the instrument withdrawn, the gel came out with it. Susan Morrison in our lab, however, developed a simple method for injecting this gel into mice (3). BALB/c mice began developing PCTs around 5 months after the first of either one 0.4-mL or three 0.1-mL injections of the gel; in these experiments, the yield of PCTs in groups of mice was comparable to that obtained with single 1.0 -mL or three 0.5-mL injections of pristane (3). Further work is in progress to compare these two agents. The silicone gels have not yet produced PCTs in other strains of mice, nor have they induced PCTs when injected subcutaneously in BALB/c mice. We have carried out some very preliminary work with linear dimethylpolysiloxane (DMPS) polymer, but this liquid form of silicone has not yet produced PCTs, either. Thus far, we have results with gels from two implants, and one is only half as effective as the other in generating PCTs. We are now testing more gels, including preparations that can be made in the laboratory.
The injected silicone gel congeals into a single mass in the abdominal cavity. As long as 5 to 10 months after injection, the blob of gel remains discrete and can be lifted out of the peritoneum intact, but after 13 months, the gel is broken down, and all that remains is stringy, sticky material. Injected mice do not develop ascites, and the vast majority of the peritoneal surfaces appear normal. The gel blobs appear to be well tolerated and relatively noninflammatory, but later, cells infiltrate some of them, changing them from clear to cloudy. In contrast to the response to pristane, injection with silicone does not affect the diaphragm and upper abdominal connective tissues; however, the gel is a source of liquid material that seeps out from the blob into the peritoneal space. This liquid material provokes the formation and deposition of a granulomatous tissue that accumulates with time on the intestinal mesenteric surfaces and in the omentum. Small spheres of oily material become surrounded by inflammatory cells in the peritoneal space, and these aggregates adhere to peritoneal surfaces and then become organized into a silicone granulomatous tissue. Much of the process of plasmacytomagenesis takes place in this tissue.
The silicone-granulomatous tissue consists of highly refractive spaces or vacuoles where liquid silicone material has been deposited. During the first 6 to 8 months after injection,the silicone granulomatous tissue has a high content of inflammatory cells that accumulate between the vacuoles. These cells include macrophages, multinucleated giant cells, fibroblasts, neutrophils, lymphocytes, plasma cells, and probably other cell types. Later, the granuloma changes. In some, but not all mice, there is a dense deposition of collagenous material around the vacuoles. Quite late, after a year, the granulomatous tissue is dramatically different: the large vacuoles break down, giving way to many smaller ones, creating a foamy appearance. The inflammatory cells largely disappear, and at this stage, the silicone granuloma seems to be "burned out".
The Importance of the Peritoneal Site
The peritoneal connective tissues appear to be a required site for pristane- and silicone-induced PCT development in mice. There are several possible reasons for this. First, the peritoneal space has a cadre of resident macrophages, and when stimulated by materials such as pristane, large numbers of new cells of the monocyte and neutrophil series migrate into the space to engage and remove the oil droplets. Second, the intestinal mesenteric vessels appear to undergo angiogenesis and supply the newly developing tissue. Third, the vascular supply also serves as a route for circulating B lymphocytes to enter the granulomatous tissue. Histological sections taken at various times after the injection of silicone or pristane show various kinds of plasma cell proliferation that progressively becomes larger and contains more atypical cells that resemble those seen in fully developed PCTs.
Composition of Silicone Gels
Silicone gels are made by crosslinking liquid, linear silicone copolymer chains (4). The most commonly used linear silicone copolymers are methylhydrogenpolysiloxane and vinylmethylpolysiloxane:
- Si -O- Si - O - R = -H, -CH3, or -CH=CH2
The polymers are a mixture of molecules of different lengths. They include many low-molecular-weight chains and cyclic silicone intermediates that are side products during the purification of single strands of long polymers. These polymers are characterized by their viscosity, which reflects chain length; 1000 centiStokes DMPS has around 330 silicone atoms (pristane has only 15 linear carbons). To generate cross-linking sites, hydrogens or vinyl groups along the chains are covalently joined by adding the catalyst platinic chloride, which reduces the double bond of the vinyl group and links it covalently to the hydrogen on another chain. Platinum is known to be an immunologically active substance and appears to induce lymphocyte proliferation as well as allergic skin reactions (5). It is probably difficult to remove all of the platinum and many of the lower-molecular-weight silicone polymers from the gels, so we cannot rule out the possibility that some of the effects we observed were due to these materials. Also, cross-linking is never carried to completion, and as a result, the so-called silicone gels consist of a web of cross-linked silicones wrapped around liquid silicones. The incompletely cross-linked silicones are used for implants because they are soft, not hard and rubbery like completely crosslinked silicone. The incompletely cross-linked gels contain residual vinyl groups in linked polymer residues and, possibly, low-molecular-weight liquids as well, a potential source of highly reactive molecules.
Vinyl chloride (which is probably not present initially in the gel) has been shown to be epoxidized in cells to form DNA adducts with guanine (6), and vinyl acetate can be metabolized to acetaldehyde, which also cross-links DNA (7). We are now exploring whether the silicone gels can be broken down into low-molecular-weight, vinyl-containing molecules that can be taken into, and metabolized in B lymphocytes. Only fragmentary evidence so far suggests that silicone polymers can be biodegraded (8). Analyses of the silicone gels carried out by Xiaokui Zhang and Henry Fales of NHLBI show that low-molecular-weight linear and cyclic polymers with as few as four silicone units are present in the gels (3). The biological activity of these compounds, especially of the cyclic compounds octadimethyltetrasiloxane and vinylmethyltetrasiloxane, needs to be studied.
Genetic Predisposition of Plasmacytoma Induction in BALB/c Mice
A critical factor for the plasmacytomagenic process in mice is determined by genotype of inbred BALB/c strain of mice. This strain could be regarded as a "natural mutant" born with a predisposition to develop PCTs, but only when appropriately stimulated. Most other inbred strains that have been tested are resistant to developing PCTs after pristane or silicone-gel treatment. These strains carry PCT-resistance genes. Genetic analysis of first-generation backcross hybrids derived from suceptible BALB/c and resistant DBA/2 mice has been carried out by Beverly Mock of our laboratory (9). She has identified two PCT-susceptibility genes on mouse chromosome 4. Using a series of BALB/c.DBA/2 congenic strains constructed in our laboratory, we found 2 PCT-resistance genes, also on chromosome 4 (10). These susceptibility and resistance genes are probably alleles. Mice carrying PCT-resistance genes develop typical oil granulomas, and it is possible to find foci of proliferating atypical plasma cells in them; however, in most of these mice the number of foci is smaller than it is in susceptible mice.
Chromosomal Translocations that Activate the c-myc Protooncogene
The most important clue about the neoplastic phenotype in PCTs comes from cytogenetic studies carried out in collaboration with Francis Wiener at our laboratory (11). Over 95% of the PCTs induced by pristane or silicone carry chromosomal translocations that directly or indirectly involve the c-myc oncogene, such as the t(12;15) translocation, that deletes part of the c-myc gene and links it directly to an Ig heavy-chain switch-region gene. The Sa site is the preferential target among the seven switch sites occurring in 60% or more of the PCTs. Recently, Siegfried Janz and Jürgen Müller in our laboratory have developed a PCR assay for detecting c-myc-Sa illegitimate recombinations (12). We have been able with this PCR methodology to consistently detect illegitimate recombinations of c-myc and Sa in cells from pristane oil granulomas 30 days after the injection of pristane. Thus, the c-myc-Sa recombination is a very early event and is potentially the initiating mutation that leads to PCT development. We postulate that some of the cells bearing these c-myc-Sa recombinations are the clonal precursors of the PCTs.
Hypothetical Scheme of PCT Development
A speculative scenario that fits most of the facts is that BALB/c mice, for genetic reasons, have a high predilection to develop illegitimate recombinations between c-myc and Ig loci. This oncogenic mutation results in deregulation of c-myc transcription such that the c-myc gene cannot be turned off. Unregulated transcription is probably a critical change because c-myc is shut down when cells exit from the cell cycle. Plasma cells are thought to be an end stage of B-cell development and usually cease dividing. Cessation of c-myc transcription may drive the cell into a postmitotic state. Rearrangement of the c-myc gene in t(12;15) removes the normal negative control sites that govern transcription. The mutant plasma cells have a continuous supply of c-myc protein, which may make it difficult for the cell to exit from the cell cycle, yielding the paradoxical phenotype of a mitotically active, terminally differentiated plasma cell. Cells with these illegitimate recombinations are probably eliminated or are not a cause of tumor formation, but they survive in the chronic inflammatory tissue, possibly because of the high concentrations of various growth factors (13,14). During this survival period, late-acting progressor genes may play a crucial role in rescuing the cells from cell death and other forms of elimination. This allows further time for changes that permit the cells to adapt and proliferate without control.
What does this imply for women with silicone-gel implants? Leaky breast implants in humans release silicone materials into connective tissues, where it induces a granulomatous tissue quite similar histologically to that in the mouse peritoneum (15,16). Silicone gels are known to be immunological adjuvants in experimental animals (17). Leaky implants have been reported to occur in around 0.2% to 1.1% of cases (18), but the Council on Scientific Affairs of the American Medical Association suspects the incidence of leakage is higher. In unconfirmed studies, leaky implants have been reported to raise the levels of IgG molecules that react with silicone-like materials (19). Because silicone gels do have immunostimulatory properties, it may be useful to evaluate clinically how individuals respond by obtaining a quantitative analysis of the various classes of immunoglobulins and a serum protein electrophoresis, possibly using immunofixation.
It is reassuring, though, that PCT development in mice is highly dependent on a rare and unique genetic constitution -- one that may never occur in humans. In mice, only the peritoneal granulomatous tissue is important in this process, and this tissue site is probably not involved in granuloma formation in humans. Also, local PCTs have not been reported as a complication of an implant. Finally, mouse plasmacytomagenesis is dependent on c-myc activating chromosomal translocations. Although homologous translocations such as t(8;14) occur in Burkitt's lymphomas in humans, they very rarely occur in human plasma cell tumors.
1. D. A. Kessler. "The basis of the FDA's decision on breast implants." N. Eng. J. Med. 326, 1713 - 45 (1992).
2. K. G. Brand. "Solid state carcinogenesis." In: Banbury Report 25: Nongenotoxic Mechanisms in Carcinogenesis, Cold Spring Harbor, ed. New York: Cold Spring Harbor Laboratory, pp. 205 - 13 (1987).
3. M. Potter, S. Morrison, F. Wiener, X. K. Zhang, and F. W. Miller. "Induction of plasmacytomas with silicone gel in genetically susceptible strains of mice." J. Natl. Cancer Inst. (in press).
4. R. R. LeVier, M. L. Chandler, and S. R. Wendel. "The pharmacology of silanes and siloxanes." In: Biochemistry of silicone and related problems, G. Bendz, and I. Lindqvist, eds. New York: Plenum Publishing Corp., pp. 473 - 514 (1978).
5. H-C. Schuppe, D. Haas- Raida, J. Kulig, U. Bomer, E. Gleichmann, and P. Kind. "T- cell- dependent popliteal lymph node reactions to platinum compounds in mice." Int. Arch. Allergy. Immunol. 97, 308 - 14 (1992).
6. T. Green. "Chloroethylenes: a mechanistic approach to human risk evaluation." Annu.Rev. Pharmacol. Toxicol. 30, 73 - 89 (1990).
7. H. Norppa, F. Tursi, P. Pfaffli, J. Maki- Paakkanen, and H. Jarventaus. "Chromosome damage induced by vinyl acetate through in vitro formation of acetaldehyde in human lymphocytes and Chinese hamster ovary cells." Cancer Res. 45, 4816 - 21 (1985).
8. L. Garrido, B. Pfleiderer, M. Papisov, and J. L. Ackerman. "In vivo degradation of silicones." Magn. Reson. Med. 29, 839 - 43(1993).
9. B. A. Mock, M. M. Krall, and J. K. Dosik. "Genetic mapping of tumor susceptibility genes involved in mouse plasmacytomagenesis." Proc. Natl. Acad. Sci. USA 90, 9499 - 503 (1993).
10. M. Potter, E. B. Mushinski, J. S. Wax, J. Hartley, and B. A. Mock. "Identification of two genes on chromosome 4 that determine resistance to plasmacytoma induction in mice." Cancer Res. 54, 1 - 7 (1994).
11. M. Potter and F. Wiener. "Plasmacytomagenesis in mice: model of neoplastic development dependent upon chromosomal translocations," Carcinogenesis 13, 1681 - 97 (1992).
12. S. Janz, J. Müller, J. Shaughnessy, and M. Potter. "Detection of recombinations between c- myc and immunoglobulin switch alpha in murine plasma cell tumors and preneoplastic lesions by polymerase chain reaction." Proc. Natl. Acad. Sci. USA 90, 7361 - 65 (1993).
13. R. P. Nordan and M. Potter. "A macrophage- derived factor required by plasmacytomas for survival and proliferation in vitro." Science 233, 566 - 69 (1986).
14. E. Shacter, G. K. Arzadon, and J. A. Williams. "Stimulation of interleukin- 6 and prostaglandin E2 secretion from peritoneal macrophages by polymers of albumin." Blood 82, 2853 - 64 (1993).
15. W. D. Travis, K. Balogh, and J. L. Abraham. "Silicone granulomas: report of three cases and review of the literature." Hum. Pathol. 16, 19 - 27 (1985).
16. L. G. Dodd, N. Sneige, G. P. Reece, and B. Fornage. "Fine- needle aspiration cytology of silicone granulomas in the augmented breast." Diagn. Cytopathol. 9, 498 - 502 (1993).
17. J. O. Naim, R. J. Lanzafame, and C. J. van Oss. "The adjuvant effect of silicone- gel on antibody formation in rats." Immunol. Invest. 22, 151 - 61 (1993).
18. Council on Scientific Affairs, AMA. "Silicone gel breast implants." J. Am. Med. Assoc. 270, 2602 - 8 (1993).
19. L. E. Wolf, M. Lappe, R. D. Peterson, and E. G. Ezrailson. "Human immune response to polydimethylsiloxane (silicone): screening studies in a breast implant population." FASEB J. 7, 1265 - 68 (1993).