J. BIOMED. MATER. RES.
VOL. 9, PP. 645-659 (1975)
Reaction Associated with a Silicone Rubber Gel: An Experimental Study R. H. RIGDON, Department of Pathology, and ALFRED DRICKS, Plastic Surgery Division, Department of Surgery, University of Texas Medical Branch, Galveston, Texas 77650
Summary A blend of Silastic 382 (Room Temperature Vulcanizing, RTV) Medical Grade silicone oil and a catalyst was prepared in vitro, in both the catalyzed and noncatalyzed state, and injected subcutaneously in mice, rats, and rabbits. When properly blended, this catalyzed material, referred to as “silicone gel,” formed a soft rubbery mass that remained at the site of injection. Properly catalyzed silicone rubber gel produces no macroscopic inflammatory reaction, attracts few polymorphonuclear leucocytes, and after 5-6 days a thin fibrous capsule begins to form around the gel. No degeneration of the silicone gel was observed during the 62 days of this experiment. Additional rats with this silicone gel have been under observation for 8 months and clinically have shown no changes in the local mass of silicone. If the catalyst is partly oxidized when added to the silicone fluid, the degree of gelling is much less. A local mass usually forms at the site of injection with some of the fluid diffusing into the tissue, forming minute cysts. The inflammatory reaction is characterized by polymorphonuclear leucocytes, associated with many macrophages and giant cells phagocytizing oil droplets and particles of the diatomaceous earth. The pathogenesis of the inflammatory reaction is discussed, referring to the ionic change and the emigration of polymorphonuclear leucocytes to particles of plastics embedded in tissue.
INTRODUCTION There is a wide variety of synthetic polymers used in medicine today. Silicone rubber, of which there are many types, has been reviewed by B r a l e ~ . ’ - ~The tissue reaction associated with certain and silicone has been reported in kinds of silicone man and animal. These have been reviewed recently by Lee and Nevelle.’O Greater emphasis has been placed upon the fibrous capsule than upon the acute reaction associated wit,h silicones. 645 @ 1975 by John Wiley & Sons, Inc.
646
RIGDON AND DRICKS
Silicone rubber, a nonwettable material, produces a local, thin, fibrous capsule and drops of silicone liquid are encapsulated by a similar capsule.1° The histological reaction following implantation of Silastic silicone rubber was interpreted by Spiers and Blacksmas as “a normal healing response in a sterile wound.” Silicone fluid (dimethylpolysiloxane) , considered to be chemically and physically inert, produced minimal tissue reaction9J1 I n mice, AndrewsIz observed that most of the silicone fluid injected subcutaneously disappeared by the third day. The tissue reaction, from the first to the third day, was characterized by an infiltration of leucocytes, few plasma cells and macrophages. By the fourth to the tenth day, the number of leucocytes had decreased and lymphocytes, plasma cells, and fibroblasts had appeared.12 Freeman et aL8concluded that a mixture of a room temperature vulcanizing rubber, (Silastic S-5392), silicone fluid, and stannous octoate was totally inert and seemed to be incapable of eliciting inflammation or a reparative reaction in the subcutaneous tissue of mice. Granulomatous reactions have been observed to follow the injection of some silicone fluids in the human breast.13-16 It has been suggested that in all other cases, the nature of the silicone fluid was unknown and there is every reason to believe that the reactions resulted from either industrial grade silicone fluid, or to adulterants. Recently, a granulomatous mass in which there was silicone fluid developed in the inguinal region of a 22 year old woman, following the injection in her breasts of this fluid. This inguinal lesion was considered to have developed by gravitational migrati0n.l’ Electrical charges in relation to plastics were recently discussed by Murphy et al.ls who observed that many polymers when negatively charged do show an improved blood compatibility. The rapid adsorption of plasma protein on foreign surfaces presumably modifies platelet-surface interaction, and fewer platelets are adsorbed on a negative surface. It was their opinion that materials exhibiting strong in vivo thromboresistance were virtually all strongly electronegative.’s The mechanism by which polymorphonuclear leucocytes are attracted to embedded plastics in experimental animals is being investigated by one of the author^.^^-^^ Our observations would indicate that the electrochemical character of the material is a significant factor in chemotaxis. Materials with a positive ionic
REACTION WITH A SILICONE RUBBER GEL
647
charge on the surface will attract negatively charged polymorphonuclear leucocytes. Leucocytes have been observed t o emigrate t o polyethylene embedded subcutaneously in the mouse for 232 days and t o a preparation of polyurethane for only 6 days.2I I n the rabbit, leucocytes emigrate t o these same plastics for a much shorter time. This difference apparently is related t o the chemical composition of the plastic and to the adsorption of plasma protein by the plastics. I n this experimental study of the tissue reaction to plastics, a blend of Silastic 382, silicone oil and stannous octoate was injected subcutaneously in mice, rats, and rabbits. This “silicone gel” is not t o be confused with “silicone oil.” Silastic 382 when vulcanized with the catalyst becomes a firm rubbery mass. Our observations are reported.
MATERIALS AND METHODS Materials were medical grade Silastic 382, medical grade silicone fluid, and stannous octoate obtained from DoU Corning, Midland, Michigan. The Silastic 382 contains a filler, diatomaceous earth. These were blended into a preparation referred to in this study as (‘silicone rubber blend.” The following technique was used for its preparation: Two drops of fresh catalyst were added to 35 ml of warm (95” - 96” F) silicone rubber blend in a small glass bottle and stirred with a glass rod. Five to 10 minutes later the liquid was poured into a 10 ml syringe with an attached 15 gauge needle and immediately injected subcutaneously into mice, rats, and rabbits from which local areas of hairs had been removed and alcohol applied. The needle wound was closed with a metal clip. One milliliter of this silicone rubber blend was injected in mice, and multiple injections were given t o the rats and rabbits. The number of animals and the interval between injection and death are given in each experiment. Animals were sacrificed by ether inhalation. The skin was reflected, and the local mass of gelled silicone with adjacent stroma was removed. I n a second experiment, the silicone rubber blend after routine preparation was permitted t o gel then cut into blocks approximately 1mm and autoclaved. A few of these blocks were put into the lumen of a 13 gauge needle, injected subcutaneously, and expelled with a
648
RIGDON AND DRICKS
plunger into mice which were examined a t varying intervals, and specimens were removed for histological study. I n the third experiment the same preparation of silicone rubber blend was mixed with a slightly oxidized* preparation of stannous octoate and after 5 t o 20 min was injected subcutaneously in mice, rats, and rabbits. They were examined the same as the other animals. This preparation of silicone rubber blend gelled in the glass container in which it was mixed. I n the fourth experiment, a group of mice was injected subcutaneously with the same preparation of silicone rubber blend but with no catalyst added. The number of mice and the interval between injection and examination are given in the specific experiment. The specimens of silicone gel and the adjacent tissue obtained from these animals in each of the 4 experiments were fixed immediately in a 40/, solution of formaldehyde. Paraffin sections were prepared and stained routinely with hematoxylin and eosin. A few select specimens were stained for fat with the Oil Red 0 stain.
RESULTS The silicone rubber blend with fresh catalyst added gelled when injected subcutaneously. However, there were some variations in the degree of gelling of different preparations. When a perfect chemical reaction occurred, a large, solid mass of silicone gel formed locally a t the site of the injection (Fig. 1). However, when the gel was less firm, a similar local mass developed a t the site of injection and some of the liquid silicone blend diffused into the surrounding stroma, producing a varying number of microscopic cysts. Macroscopically, the local mass of silicone rubber blend was the same in each situation; only in the histological reaction was a difference observed. Histologically, in these masses of gelled silicone there were clear thread-like fibers that varied in size (Fig. 2a). Particles of the diatomaceous earth filler (Fig. 2b) usually were surrounded by these fibers. I n tissues stained with Oil Red 0, these silicone fibers had granules and short narrow blocks of particles in a chain-like pattern that stained red with the lipid stain and were surrounded by a thin, narrow transparent border (Figs. 3a and b). *Slightly oxidized is indicated by a change in color from a clear to a faintly brown color.
REACTION WITH A SILICONE RUBBER GEL
649
Fig. 1. The silicone rubber blend is firmly gelled. It is surrounded by a thin fibrous capsule. The silicone was injected subcutaneously 30 days previously in this mouse.
Reaction Associated with Silicone Rubber Blend that Gelled Following Subcutaneous Injection I n these experiments, fresh catalyst was added to warm silicone rubber blend and after 5 min was subcutaneously injected in 10 mice. The silicone rubber blend was only partly gelled; 5 mice were examined a t 72 hr and the silicone was only partlygelled. The same preparation of silicone rubber blend was injected into 7 mice 20 min after the catalyst was added. The silicone rubber blend in these mice was completely gelled a t 24 hr. Silicone rubber blend was injected subcutaneously in eighteen mice 15 min after the catalyst was added. A local mass developed at the site of injection and remained for 24 hr a t which time the mice were examined. A firm mass of the gel, 1.0 t o 2.0 cm in diameter, was present in each mouse. It was easily removed from the surrounding tissue. I n another experiment, 6 mice were injected 20 min after the catalyst was added to the silicone rubber blend; three of these were examined a t 7 days and three a t 25 days. The gel was localized, firm, andeasily removed. There
650
RIGDON AND DRICKS
Fig. 2(a) In the gelled silicone rubber blend are thread-like fibers. Particles of the filler are entrapped by these fibers. (H & E stain x432). (b) Rabbits 153-4A. Particles of the diatomaceous earth filler used in this silicone rubber blend. (H & E stain X 688.)
Fig. 3(a) Thread-like fibers in the gelled silicone rubber blend have lipid staining particles in a chain-like pattern. (Oil Red 0 stain X688). (b) A higher magnification of the lipid particles shown in (a). (Oil Red 0 stain X1836.)
REACTION WITH A SILICONE RUBBER GEL
651
was no macroscopic evidence of inflammation in the subcutaneous tissue of any of these 6 mice. Six mice were injected 40 min after the catalyst was added; three of these were examined a t 72 hr and three a t 30 days. I n each, the silicone was firmly gelled. I n the latter three mice, a fine membrane that was easily removed surrounded the silicone. There was no evidence of inflammation (Fig. 1). Silicone rubber blend after catalyzing for 10-20 min when injected subcutaneously in rats and rabbits, gelled the same as in mice. Nine rats, after subcutaneous injection were examined a t the following intervals: three a t 24 hr, two a t 48 hr, one a t 6 days, one a t 18 days, and two on the 62nd experimental day. A local, firm mass of silicone, 1.0 t o 2.0 cm in diameter, was present in each rat. There was no gross evidence of inflammation. A mass of firm, silicone rubber gel, 1.0 X 1.5 cm, was present in the subcutaneous tissue of a rabbit 50 days after i t had been injected. A minimal amount of fibrous tissue was present. Histologically, 24 hr after the silicone rubber blend was injected subcutaneously in mice, many polymorphonuclear leucocytes were present in the surrounding stroma (Fig. 4). The number of leuco-
Fig. 4. Polymorphonuclear leucocytes in the stroma around a local mass of silicone rubber blend 24 hr after it was injected in a mouse. The silicone is absent in this photograph. (H I% E stain X144.)
652
RIGDON AND DRICKS
cytes significantly decreased by the 7th day and a band of fibroblasts surrounded the gelled silicone. A few leucocytes and mononuclear cells were present in the stroma. This same type of fibrous capsule was present on the 25th and 30th experimental day (Fig. 5). I n some of the mice and rats w-ith the large mass of gelled siLicone, there were also multiple small cysts in the adjacent stroma, usually microscopic in size. Particles of the filler were present in the lumen of some of these cysts, and the silicone apparently was not gelled. More leucocytes and mononuclear cells were present in the stroma around these small cysts than around the large mass of gelled silicone. Mononuclear cells phagocytized particles of the filler and oil droplets in the stroma. The local reactions t o gelled silicone in the rat and rabbits was very similar to that in the mouse. There were many polymorphonuclear leucocytes a t 24 and 48 hr in the stroma around the silicone. The number of leucoeytes however significantly decreased by the 6th day. At this time, there was a moderate number of mononuclear cells and fibroblasts. The reaction on the 62nd experimental day
Fig. 5. Fibrous capsule around a local mass of gelled silicone rubber blend in a mouse on the 30th experimental day. The gross lesion is shown in Fig. 1. (H & E stain x210.)
REACTION WITH A SITJCONE RUBBER GEL
653
was similar to that on the 18th day except for the presence of some giant cells in and around the smaller cysts.
Reaction Associated with Silicone Gel Particles Implanted Subcutaneously Silicone gel was cut into pieces 1 mm in size and autoclaved. Several of these were put into the lumen of a 13 gauge needle, injected subcutaneously, and expelled by a plunger into 25 mice. Specimens for pathologic study were removed a t the following intervals: six a t 48 hr, five a t 3 days, three a t 8 days, seven a t 21 days, and four on the 34th experimental day. These particles of silicone gel in the subuctaneous tissue formed a local mass 5-8 mm in diameter. Specimens examined a t 48 hr and 5 days were a loose group of particles that readily fell apart, while those removed on the 8th day and thereafter were surrounded by a thin fibrous membrane. There was no gross reaction associated with these particles of silicone gel other than this fibrous capsule.
Fig. 6. A group of gelled silicone rubber blend particles were injected in this mouse 34 days previously. The particles are absent in this photograph. Fibrous tissue, with a minimal cellular reaction is present in the surrounding stroma. (H &Estain X32.)
654
RIGDON AND DRICKS
Histologically, the specimens examined a t 48 hr had many leucocytes, some mononuclear cells and fibroblasts in the surrounding stroma. This silicone that was gelled when implanted was morphologically the same as that injected subcutaneously and gelled in the tissue. The inflammatory reaction on the 5th and 8th days was similar to that a t 48 hr except there was a n increase in the number of fibroblasts and a decrease in the number of leucocytes. Variations did occur in the reaction in different parts of a section. The reaction was more extensive when the silicone was adjacent t o muscle than to fatty tissue. Few leucocytes and mononuclear cells and a narrow capsule of fibrous tissue were present around the silicone particles on the 21st and 34th experimental day (Fig. 6).
Reaction Associated with Silicone Rubber Blend with Catalyst Added Without Gelling I n this experiment, the catalyst was added to the silicone blend and stirred for approximately 1min. The silicone in the bottle gelled within 1 t o 4% hr in 4 of these experiments but did not gel in 24 hr in the 5th experiment. Subcutaneous injections in mice were started 1 t o 6 min after the catalyst was added and each group was completed within 15-30 min. The silicone escaped through the needle hole of a few mice. One hundred and sixteen mice were used and examined a t the following intervals: 32 between 3 and 8 hr, 31 at 24 hr, five at 48 hr, 24 on the 4th and 5th day, five on the 10th day, four on the 18th, four on the 28th, three on the 34th, and eight on the 60th or 62 experimental day. The silicone was localized in a cyst, approximately 1.5 cm in diameter, in the subcutaneous tissue of a majority of the mice. I n some, there were multiple smaller cysts in the subcutaneous tissue a t the site of injections. The silicone was liquid in the subcutaneous tissue in each of these mice except for one in which the silicone was inadvertently injected intraperitoneally and it was partly gelled. I n another experiment, 21 mice were given a n intraperitoneal injection of the same silicone fluid and it did not solidify in any mouse. There was no local edema or hyperemia associated with the silicone in any of these mice injected subcutaneously and intraperitoneally. The response of rats and rabbits t o the subcutaneous injection of this preparation of silicone fluid was the same as that in mice. However, in rabbits, the silicone fluid frequently gravitated from the site
REACTION WITH A SILICONE RUBBER GEL
655
of injection on the lateral side of the abdomen t o the midline of the abdomen. There was no macroscopic evidence of inflammation in the subcutaneous tissue of rats and rabbits receiving this preparation of silicone. Histologically, in mice there were many microscopic cysts of varying size in the stroma around the macroscopic cysts. Many of theses cysts contained particles of the filler, while others were empty. Since the silicone was liquid it escaped when the paraffin sections were processed (Fig. 7). Four hours after the silicone fluid was injected a few polymorphonuclear leucocytes were present in the stroma separating these cysts. Leucocytes progressively increased, reaching their maximum number of 3+ t o 4f between 24 and 48 hr. Leucocytes progressively emigrated from the surrounding stroma t o concentrate at the margin of these cysts. The number of leucocytes in this reaction had begun t o decrease by the 4th experimental day. At this time, the number of mononuclear cells had increased and a few fibroblasts were present in the stroma around some of the cysts. Phagocytosis of oil droplets and particles of the filler, by mononuclear cells,
Fig. 7. Multiple cysts with particles of the filler in the stroma 4 days after the silicone rubber blend that did not gel was injected subcutaneously in this mouse. (H & E stain X543.)
656
RIGDON AND DRICKS
Fig. %(a) A cyst resulting from the subcutaneous injection of the silicone blend 62 days previously. The contents of this cyst were liquid since the catalyst used was partly oxidized and gelling did not occur. (b) Wall of the cyst shown in (a) with many of the particles of the filler phagocytized by mononuclear cells. (H & E stain X140.)
was conspicuous a t this time. By the 5th day, the number of leucocytes had decreased conspicuously but a few were present between the 10th and 43rd day, however, none were found in the specimens examined on the 62nd day. With the decrease in the number of the leucocytes, there was an increase in the number of mononuclear cells
REACTION WITH A SILICONE RUBBER GEL
657
and fibroblasts. Phagocytosis by the mononuclear cells was conspicuous in the older lesions. Giant cells were present. A macroscopic cyst showing where the liquid silicone had been before the slides were processed on the 62nd experimental day is shown in Figure 8a and b. The cells in the wall of this cyst were primarily mononuclears with particles of the filler in their cytoplasm. There were only a few fibroblasts present in this reaction. The histological reaction in the subcutaneous tissue of rats and rabbits following the injection of a preparation of silicone in which gelling did not occur was essentially the same as that observed in mice.
Reaction Associated with Silicone Rubber Blend without Catalyst Added The silicone rubber blend with no catalyst added was injected subcutaneously in 27 mice which were examined a t the following intervals: one a t 3 hr, five a t 6 hr, ten a t 24 hr, six a t 4 days, and five a t 13 days. A cyst, 1-1.5 cm, was present a t the site of injection in a majority of the mice. The contents of the cysts were liquid. There was no gross evidence of inflammation in any of the mice. The histological reaction in this group of mice was the same as that observed in mice given the same silicone with the catalyst added in which gelling did not occur.
DISCUSSION There is no macroscopic evidence of inflammation in any of the animals injected subcutaneously with any of these preparations of silicone. The silicone rubber blend (silicone gel) when properly prepared forms a local, soft mass a t the site of injection. However, when not properly prepared the silicone blend remains liquid and either forms local cysts or diffuses into the tissue. I n the preparation of this silicone gel, it is essential to use a nonoxidized (fresh) preparation of stannous octoate, heat for 5-10 min a t 95" to 96' F, and thoroughly mix with a glass rod. It may then be poured into a syringe and expelled through a 15 gauge needle. Solidification occurs within 10-15 min. When improperly prepared only a portion of the silicone solidifies and the accompanying liquid diffuses into the adjacent tissue. Rats are now being observed in which the silicone gel has remained localized a t the site of subcutaneous injection for 8 months.
658
RIGDON AND DRICKS
The silicone gel when injected subcutaneously solidified within 10-15 min. It remained clear and soft during the 62 days in which these mice were followed. Microscopically, there were thread-like fibers (Figures 2 and 3) in this silicone gel that encapsulated particles of the oil. There was no cellular infiltration into the mass of silicone, only a minimal fibrous capsule formed around the gelled silicone. Histologically, polymorphonuclear leucocytes were present in the stroma around the silicone preparations within 24 hr and progressively increased in number for 72-96 hr, a t which time they began t o progressively decrease in number in those animals in which the silicone was firmly gelled. By the 6th experimental day, a moderate number of mononuclear cells and fibroblasts and only a few polymorphonuclear leucocytes were present. A fibrous capsule progressively developed around the gelled silicone by the 25th day. The histological reaction was essentially the same when the silicone gel, formed in vivo, and particles of the silicone gel formed in vitro, were implanted subcutaneously. The local inflammatory reaction, however, was more extensive and persisted for a longer time when the silicone was poorly gelled. Mononuclear cells and giant cells were conspicuous in this tissue reaction, having phagocytized the oil droplets and particles of diatomaceous earth. It has been suggested that during the processing of paraffin sections, the solvent used to carry the paraffin into the sample and again t o remove it subsequent to microtoming are excellent solvents for dimethylpolysiloxane. If this silicone polymer is not crosslinked (as is the case with the silicone fluid and the unvulcanized RTV silicone rubber), the processing will very effectively remove all of the silicone present. If the polymer is vulcanized, the solvent will swell it (about 200p/,), and ill extract all uncrosslinked polymer. During the swelling process, artifacts will be formed. Since silicones will adhere to very few- materials, any swelling of the section on the slide will almost invariably pop the rubber off the slide. Therefore, it is very unusual t o have a piece of silicone rubber remain on a slide, and it is impossible to retain any silicone fluid. These facts should very seriously be considered when interpreting slides made from tissue in contact with any type of silicones. The ionic charge on plastics used in cardiovascular surgery today is recognized as a possible cause for the development of mural t h r ~ r n b i . ~ It ~ -is~ suggested ~ that a n ionic charge on silicone could
REACTION WITH A SILICONE RUBBER GEL
659
be a basic factor in the attraction of leucocytes when it is embedded in tissue. The authors acknowledge the assistance of Bettie Joe Williams and Ava Stevenson, Laboratory Technicians, for preparation of the sections, and John Ellis for the photographs. This study was supported by Grant 3430118-NSF GB 31240 from the National Science Foundation.
References S. Braley, J. Macromol. Sci-Chem. A , 4 (3), 529 (1970). S. Braley, Arch. Otolaryngol., 78, 669 (1963). S. Braley, Med. Electron. Biol. Eng., 3, 127 (1965). W. F. Agnew, E. M. Todd, H. Richmond, and W. 9. Chronister, J . Surg. Res., 2, 357 (1962). 5. A. G. Spiers and R. Blacksma, Plast. Reconstr. Surg., 31, 166 (1963). 6. T. 0. Rees, D. L. Ballantyne, I. Seidmon, and G. A. Hawthorne, Plast. Reconstr. Surg., 39,402 (1967). 7. D. L. Ballantyne, T. D. Rees, and I. Seidman, Plast. and Reconstr. Surg., 1. 2. 3. 4.
36, 330 (1965). 8. B. S. Freeman, E. L. Bigelow, and S. A. Braley, Amer. J. Surg., 112, 534 (1966). 9. F. L. Ashley, S. Braley, T. D. €tees, 1). Goulian, and D. L. Ballantyne, Plast. Reconstr. Surg., 39, 411 (1967). 10. H. Lee and K. Neville, Handbook of Biomedical Plastics, Pasadena Technology Press, Pasadena, California, 1971. 11. F. L. Ashley, S. Braley, and E. G. McNall, Surg. Clin. North Amer., 51, 501 (1971). 12. J. M. Andrews, Plast. Reconstr. Surg., 38, 581 (1966). 13. L.H. Winer, T. H. Sternberg, R. Lehmann, and F. L. Ashley, Arch. Derm., 90, 588 (1964). 14. W. St. C. Symmers, Brit. Med. J., 3, 19 (1968). 15. J. S. Nosanchuk, Arch. Surg., 97, 583 (1968). 16. C. H. Chaplin, Plast. Reconstr. Surg., 44, 447 (1969). 17. C. Delage, J. J. Shane, and F. B. Johnson, Arch Derm., 108. 104 (1973). 18. P. V. Murphy, A. LaCroix, S. Marchant, and W. Berhard, Biomed. Mater. Symp. No. 1, 59 (1971). 19. R. H. Rigdon, J . Biomed. Mater. Res., 4, 57 (1970). 20. R. H. Rigdon, J. Riomed. Mater. Res., 7, 79 (1973). 21. R. H. Rigdon, J. Riomed. Mater. Res., 8, 97 (1974). 22. R. H.Rigdon, Texas Rep. Biol. Med., 32, 535 (1974). 23. P. N. Sawyer and S. Sainivason, J . Riomed. Mater. Res., 1, 83 101 (1967). 24. V. L. Bott and A. Furuse, Fed. Proc. Fed. A m . SOC.Exp. Biol., 39,1679 (1971). 25. P. W. Madras, W A Morton, and H. E. Petschek, Fed. Proc. Fed. Am. SOC. Exp. Biol., 30, 1665 (1971). 26. R. I. Leininger, C R C C r i . Rev. Bioeng. 1, 333 (1971-73).
Received December 13, 1974 Revised January 21, 1975
VOL. 9, PP. 645-659 (1975)
Reaction Associated with a Silicone Rubber Gel: An Experimental Study R. H. RIGDON, Department of Pathology, and ALFRED DRICKS, Plastic Surgery Division, Department of Surgery, University of Texas Medical Branch, Galveston, Texas 77650
Summary A blend of Silastic 382 (Room Temperature Vulcanizing, RTV) Medical Grade silicone oil and a catalyst was prepared in vitro, in both the catalyzed and noncatalyzed state, and injected subcutaneously in mice, rats, and rabbits. When properly blended, this catalyzed material, referred to as “silicone gel,” formed a soft rubbery mass that remained at the site of injection. Properly catalyzed silicone rubber gel produces no macroscopic inflammatory reaction, attracts few polymorphonuclear leucocytes, and after 5-6 days a thin fibrous capsule begins to form around the gel. No degeneration of the silicone gel was observed during the 62 days of this experiment. Additional rats with this silicone gel have been under observation for 8 months and clinically have shown no changes in the local mass of silicone. If the catalyst is partly oxidized when added to the silicone fluid, the degree of gelling is much less. A local mass usually forms at the site of injection with some of the fluid diffusing into the tissue, forming minute cysts. The inflammatory reaction is characterized by polymorphonuclear leucocytes, associated with many macrophages and giant cells phagocytizing oil droplets and particles of the diatomaceous earth. The pathogenesis of the inflammatory reaction is discussed, referring to the ionic change and the emigration of polymorphonuclear leucocytes to particles of plastics embedded in tissue.
INTRODUCTION There is a wide variety of synthetic polymers used in medicine today. Silicone rubber, of which there are many types, has been reviewed by B r a l e ~ . ’ - ~The tissue reaction associated with certain and silicone has been reported in kinds of silicone man and animal. These have been reviewed recently by Lee and Nevelle.’O Greater emphasis has been placed upon the fibrous capsule than upon the acute reaction associated wit,h silicones. 645 @ 1975 by John Wiley & Sons, Inc.
646
RIGDON AND DRICKS
Silicone rubber, a nonwettable material, produces a local, thin, fibrous capsule and drops of silicone liquid are encapsulated by a similar capsule.1° The histological reaction following implantation of Silastic silicone rubber was interpreted by Spiers and Blacksmas as “a normal healing response in a sterile wound.” Silicone fluid (dimethylpolysiloxane) , considered to be chemically and physically inert, produced minimal tissue reaction9J1 I n mice, AndrewsIz observed that most of the silicone fluid injected subcutaneously disappeared by the third day. The tissue reaction, from the first to the third day, was characterized by an infiltration of leucocytes, few plasma cells and macrophages. By the fourth to the tenth day, the number of leucocytes had decreased and lymphocytes, plasma cells, and fibroblasts had appeared.12 Freeman et aL8concluded that a mixture of a room temperature vulcanizing rubber, (Silastic S-5392), silicone fluid, and stannous octoate was totally inert and seemed to be incapable of eliciting inflammation or a reparative reaction in the subcutaneous tissue of mice. Granulomatous reactions have been observed to follow the injection of some silicone fluids in the human breast.13-16 It has been suggested that in all other cases, the nature of the silicone fluid was unknown and there is every reason to believe that the reactions resulted from either industrial grade silicone fluid, or to adulterants. Recently, a granulomatous mass in which there was silicone fluid developed in the inguinal region of a 22 year old woman, following the injection in her breasts of this fluid. This inguinal lesion was considered to have developed by gravitational migrati0n.l’ Electrical charges in relation to plastics were recently discussed by Murphy et al.ls who observed that many polymers when negatively charged do show an improved blood compatibility. The rapid adsorption of plasma protein on foreign surfaces presumably modifies platelet-surface interaction, and fewer platelets are adsorbed on a negative surface. It was their opinion that materials exhibiting strong in vivo thromboresistance were virtually all strongly electronegative.’s The mechanism by which polymorphonuclear leucocytes are attracted to embedded plastics in experimental animals is being investigated by one of the author^.^^-^^ Our observations would indicate that the electrochemical character of the material is a significant factor in chemotaxis. Materials with a positive ionic
REACTION WITH A SILICONE RUBBER GEL
647
charge on the surface will attract negatively charged polymorphonuclear leucocytes. Leucocytes have been observed t o emigrate t o polyethylene embedded subcutaneously in the mouse for 232 days and t o a preparation of polyurethane for only 6 days.2I I n the rabbit, leucocytes emigrate t o these same plastics for a much shorter time. This difference apparently is related t o the chemical composition of the plastic and to the adsorption of plasma protein by the plastics. I n this experimental study of the tissue reaction to plastics, a blend of Silastic 382, silicone oil and stannous octoate was injected subcutaneously in mice, rats, and rabbits. This “silicone gel” is not t o be confused with “silicone oil.” Silastic 382 when vulcanized with the catalyst becomes a firm rubbery mass. Our observations are reported.
MATERIALS AND METHODS Materials were medical grade Silastic 382, medical grade silicone fluid, and stannous octoate obtained from DoU Corning, Midland, Michigan. The Silastic 382 contains a filler, diatomaceous earth. These were blended into a preparation referred to in this study as (‘silicone rubber blend.” The following technique was used for its preparation: Two drops of fresh catalyst were added to 35 ml of warm (95” - 96” F) silicone rubber blend in a small glass bottle and stirred with a glass rod. Five to 10 minutes later the liquid was poured into a 10 ml syringe with an attached 15 gauge needle and immediately injected subcutaneously into mice, rats, and rabbits from which local areas of hairs had been removed and alcohol applied. The needle wound was closed with a metal clip. One milliliter of this silicone rubber blend was injected in mice, and multiple injections were given t o the rats and rabbits. The number of animals and the interval between injection and death are given in each experiment. Animals were sacrificed by ether inhalation. The skin was reflected, and the local mass of gelled silicone with adjacent stroma was removed. I n a second experiment, the silicone rubber blend after routine preparation was permitted t o gel then cut into blocks approximately 1mm and autoclaved. A few of these blocks were put into the lumen of a 13 gauge needle, injected subcutaneously, and expelled with a
648
RIGDON AND DRICKS
plunger into mice which were examined a t varying intervals, and specimens were removed for histological study. I n the third experiment the same preparation of silicone rubber blend was mixed with a slightly oxidized* preparation of stannous octoate and after 5 t o 20 min was injected subcutaneously in mice, rats, and rabbits. They were examined the same as the other animals. This preparation of silicone rubber blend gelled in the glass container in which it was mixed. I n the fourth experiment, a group of mice was injected subcutaneously with the same preparation of silicone rubber blend but with no catalyst added. The number of mice and the interval between injection and examination are given in the specific experiment. The specimens of silicone gel and the adjacent tissue obtained from these animals in each of the 4 experiments were fixed immediately in a 40/, solution of formaldehyde. Paraffin sections were prepared and stained routinely with hematoxylin and eosin. A few select specimens were stained for fat with the Oil Red 0 stain.
RESULTS The silicone rubber blend with fresh catalyst added gelled when injected subcutaneously. However, there were some variations in the degree of gelling of different preparations. When a perfect chemical reaction occurred, a large, solid mass of silicone gel formed locally a t the site of the injection (Fig. 1). However, when the gel was less firm, a similar local mass developed a t the site of injection and some of the liquid silicone blend diffused into the surrounding stroma, producing a varying number of microscopic cysts. Macroscopically, the local mass of silicone rubber blend was the same in each situation; only in the histological reaction was a difference observed. Histologically, in these masses of gelled silicone there were clear thread-like fibers that varied in size (Fig. 2a). Particles of the diatomaceous earth filler (Fig. 2b) usually were surrounded by these fibers. I n tissues stained with Oil Red 0, these silicone fibers had granules and short narrow blocks of particles in a chain-like pattern that stained red with the lipid stain and were surrounded by a thin, narrow transparent border (Figs. 3a and b). *Slightly oxidized is indicated by a change in color from a clear to a faintly brown color.
REACTION WITH A SILICONE RUBBER GEL
649
Fig. 1. The silicone rubber blend is firmly gelled. It is surrounded by a thin fibrous capsule. The silicone was injected subcutaneously 30 days previously in this mouse.
Reaction Associated with Silicone Rubber Blend that Gelled Following Subcutaneous Injection I n these experiments, fresh catalyst was added to warm silicone rubber blend and after 5 min was subcutaneously injected in 10 mice. The silicone rubber blend was only partly gelled; 5 mice were examined a t 72 hr and the silicone was only partlygelled. The same preparation of silicone rubber blend was injected into 7 mice 20 min after the catalyst was added. The silicone rubber blend in these mice was completely gelled a t 24 hr. Silicone rubber blend was injected subcutaneously in eighteen mice 15 min after the catalyst was added. A local mass developed at the site of injection and remained for 24 hr a t which time the mice were examined. A firm mass of the gel, 1.0 t o 2.0 cm in diameter, was present in each mouse. It was easily removed from the surrounding tissue. I n another experiment, 6 mice were injected 20 min after the catalyst was added to the silicone rubber blend; three of these were examined a t 7 days and three a t 25 days. The gel was localized, firm, andeasily removed. There
650
RIGDON AND DRICKS
Fig. 2(a) In the gelled silicone rubber blend are thread-like fibers. Particles of the filler are entrapped by these fibers. (H & E stain x432). (b) Rabbits 153-4A. Particles of the diatomaceous earth filler used in this silicone rubber blend. (H & E stain X 688.)
Fig. 3(a) Thread-like fibers in the gelled silicone rubber blend have lipid staining particles in a chain-like pattern. (Oil Red 0 stain X688). (b) A higher magnification of the lipid particles shown in (a). (Oil Red 0 stain X1836.)
REACTION WITH A SILICONE RUBBER GEL
651
was no macroscopic evidence of inflammation in the subcutaneous tissue of any of these 6 mice. Six mice were injected 40 min after the catalyst was added; three of these were examined a t 72 hr and three a t 30 days. I n each, the silicone was firmly gelled. I n the latter three mice, a fine membrane that was easily removed surrounded the silicone. There was no evidence of inflammation (Fig. 1). Silicone rubber blend after catalyzing for 10-20 min when injected subcutaneously in rats and rabbits, gelled the same as in mice. Nine rats, after subcutaneous injection were examined a t the following intervals: three a t 24 hr, two a t 48 hr, one a t 6 days, one a t 18 days, and two on the 62nd experimental day. A local, firm mass of silicone, 1.0 t o 2.0 cm in diameter, was present in each rat. There was no gross evidence of inflammation. A mass of firm, silicone rubber gel, 1.0 X 1.5 cm, was present in the subcutaneous tissue of a rabbit 50 days after i t had been injected. A minimal amount of fibrous tissue was present. Histologically, 24 hr after the silicone rubber blend was injected subcutaneously in mice, many polymorphonuclear leucocytes were present in the surrounding stroma (Fig. 4). The number of leuco-
Fig. 4. Polymorphonuclear leucocytes in the stroma around a local mass of silicone rubber blend 24 hr after it was injected in a mouse. The silicone is absent in this photograph. (H I% E stain X144.)
652
RIGDON AND DRICKS
cytes significantly decreased by the 7th day and a band of fibroblasts surrounded the gelled silicone. A few leucocytes and mononuclear cells were present in the stroma. This same type of fibrous capsule was present on the 25th and 30th experimental day (Fig. 5). I n some of the mice and rats w-ith the large mass of gelled siLicone, there were also multiple small cysts in the adjacent stroma, usually microscopic in size. Particles of the filler were present in the lumen of some of these cysts, and the silicone apparently was not gelled. More leucocytes and mononuclear cells were present in the stroma around these small cysts than around the large mass of gelled silicone. Mononuclear cells phagocytized particles of the filler and oil droplets in the stroma. The local reactions t o gelled silicone in the rat and rabbits was very similar to that in the mouse. There were many polymorphonuclear leucocytes a t 24 and 48 hr in the stroma around the silicone. The number of leucoeytes however significantly decreased by the 6th day. At this time, there was a moderate number of mononuclear cells and fibroblasts. The reaction on the 62nd experimental day
Fig. 5. Fibrous capsule around a local mass of gelled silicone rubber blend in a mouse on the 30th experimental day. The gross lesion is shown in Fig. 1. (H & E stain x210.)
REACTION WITH A SITJCONE RUBBER GEL
653
was similar to that on the 18th day except for the presence of some giant cells in and around the smaller cysts.
Reaction Associated with Silicone Gel Particles Implanted Subcutaneously Silicone gel was cut into pieces 1 mm in size and autoclaved. Several of these were put into the lumen of a 13 gauge needle, injected subcutaneously, and expelled by a plunger into 25 mice. Specimens for pathologic study were removed a t the following intervals: six a t 48 hr, five a t 3 days, three a t 8 days, seven a t 21 days, and four on the 34th experimental day. These particles of silicone gel in the subuctaneous tissue formed a local mass 5-8 mm in diameter. Specimens examined a t 48 hr and 5 days were a loose group of particles that readily fell apart, while those removed on the 8th day and thereafter were surrounded by a thin fibrous membrane. There was no gross reaction associated with these particles of silicone gel other than this fibrous capsule.
Fig. 6. A group of gelled silicone rubber blend particles were injected in this mouse 34 days previously. The particles are absent in this photograph. Fibrous tissue, with a minimal cellular reaction is present in the surrounding stroma. (H &Estain X32.)
654
RIGDON AND DRICKS
Histologically, the specimens examined a t 48 hr had many leucocytes, some mononuclear cells and fibroblasts in the surrounding stroma. This silicone that was gelled when implanted was morphologically the same as that injected subcutaneously and gelled in the tissue. The inflammatory reaction on the 5th and 8th days was similar to that a t 48 hr except there was a n increase in the number of fibroblasts and a decrease in the number of leucocytes. Variations did occur in the reaction in different parts of a section. The reaction was more extensive when the silicone was adjacent t o muscle than to fatty tissue. Few leucocytes and mononuclear cells and a narrow capsule of fibrous tissue were present around the silicone particles on the 21st and 34th experimental day (Fig. 6).
Reaction Associated with Silicone Rubber Blend with Catalyst Added Without Gelling I n this experiment, the catalyst was added to the silicone blend and stirred for approximately 1min. The silicone in the bottle gelled within 1 t o 4% hr in 4 of these experiments but did not gel in 24 hr in the 5th experiment. Subcutaneous injections in mice were started 1 t o 6 min after the catalyst was added and each group was completed within 15-30 min. The silicone escaped through the needle hole of a few mice. One hundred and sixteen mice were used and examined a t the following intervals: 32 between 3 and 8 hr, 31 at 24 hr, five at 48 hr, 24 on the 4th and 5th day, five on the 10th day, four on the 18th, four on the 28th, three on the 34th, and eight on the 60th or 62 experimental day. The silicone was localized in a cyst, approximately 1.5 cm in diameter, in the subcutaneous tissue of a majority of the mice. I n some, there were multiple smaller cysts in the subcutaneous tissue a t the site of injections. The silicone was liquid in the subcutaneous tissue in each of these mice except for one in which the silicone was inadvertently injected intraperitoneally and it was partly gelled. I n another experiment, 21 mice were given a n intraperitoneal injection of the same silicone fluid and it did not solidify in any mouse. There was no local edema or hyperemia associated with the silicone in any of these mice injected subcutaneously and intraperitoneally. The response of rats and rabbits t o the subcutaneous injection of this preparation of silicone fluid was the same as that in mice. However, in rabbits, the silicone fluid frequently gravitated from the site
REACTION WITH A SILICONE RUBBER GEL
655
of injection on the lateral side of the abdomen t o the midline of the abdomen. There was no macroscopic evidence of inflammation in the subcutaneous tissue of rats and rabbits receiving this preparation of silicone. Histologically, in mice there were many microscopic cysts of varying size in the stroma around the macroscopic cysts. Many of theses cysts contained particles of the filler, while others were empty. Since the silicone was liquid it escaped when the paraffin sections were processed (Fig. 7). Four hours after the silicone fluid was injected a few polymorphonuclear leucocytes were present in the stroma separating these cysts. Leucocytes progressively increased, reaching their maximum number of 3+ t o 4f between 24 and 48 hr. Leucocytes progressively emigrated from the surrounding stroma t o concentrate at the margin of these cysts. The number of leucocytes in this reaction had begun t o decrease by the 4th experimental day. At this time, the number of mononuclear cells had increased and a few fibroblasts were present in the stroma around some of the cysts. Phagocytosis of oil droplets and particles of the filler, by mononuclear cells,
Fig. 7. Multiple cysts with particles of the filler in the stroma 4 days after the silicone rubber blend that did not gel was injected subcutaneously in this mouse. (H & E stain X543.)
656
RIGDON AND DRICKS
Fig. %(a) A cyst resulting from the subcutaneous injection of the silicone blend 62 days previously. The contents of this cyst were liquid since the catalyst used was partly oxidized and gelling did not occur. (b) Wall of the cyst shown in (a) with many of the particles of the filler phagocytized by mononuclear cells. (H & E stain X140.)
was conspicuous a t this time. By the 5th day, the number of leucocytes had decreased conspicuously but a few were present between the 10th and 43rd day, however, none were found in the specimens examined on the 62nd day. With the decrease in the number of the leucocytes, there was an increase in the number of mononuclear cells
REACTION WITH A SILICONE RUBBER GEL
657
and fibroblasts. Phagocytosis by the mononuclear cells was conspicuous in the older lesions. Giant cells were present. A macroscopic cyst showing where the liquid silicone had been before the slides were processed on the 62nd experimental day is shown in Figure 8a and b. The cells in the wall of this cyst were primarily mononuclears with particles of the filler in their cytoplasm. There were only a few fibroblasts present in this reaction. The histological reaction in the subcutaneous tissue of rats and rabbits following the injection of a preparation of silicone in which gelling did not occur was essentially the same as that observed in mice.
Reaction Associated with Silicone Rubber Blend without Catalyst Added The silicone rubber blend with no catalyst added was injected subcutaneously in 27 mice which were examined a t the following intervals: one a t 3 hr, five a t 6 hr, ten a t 24 hr, six a t 4 days, and five a t 13 days. A cyst, 1-1.5 cm, was present a t the site of injection in a majority of the mice. The contents of the cysts were liquid. There was no gross evidence of inflammation in any of the mice. The histological reaction in this group of mice was the same as that observed in mice given the same silicone with the catalyst added in which gelling did not occur.
DISCUSSION There is no macroscopic evidence of inflammation in any of the animals injected subcutaneously with any of these preparations of silicone. The silicone rubber blend (silicone gel) when properly prepared forms a local, soft mass a t the site of injection. However, when not properly prepared the silicone blend remains liquid and either forms local cysts or diffuses into the tissue. I n the preparation of this silicone gel, it is essential to use a nonoxidized (fresh) preparation of stannous octoate, heat for 5-10 min a t 95" to 96' F, and thoroughly mix with a glass rod. It may then be poured into a syringe and expelled through a 15 gauge needle. Solidification occurs within 10-15 min. When improperly prepared only a portion of the silicone solidifies and the accompanying liquid diffuses into the adjacent tissue. Rats are now being observed in which the silicone gel has remained localized a t the site of subcutaneous injection for 8 months.
658
RIGDON AND DRICKS
The silicone gel when injected subcutaneously solidified within 10-15 min. It remained clear and soft during the 62 days in which these mice were followed. Microscopically, there were thread-like fibers (Figures 2 and 3) in this silicone gel that encapsulated particles of the oil. There was no cellular infiltration into the mass of silicone, only a minimal fibrous capsule formed around the gelled silicone. Histologically, polymorphonuclear leucocytes were present in the stroma around the silicone preparations within 24 hr and progressively increased in number for 72-96 hr, a t which time they began t o progressively decrease in number in those animals in which the silicone was firmly gelled. By the 6th experimental day, a moderate number of mononuclear cells and fibroblasts and only a few polymorphonuclear leucocytes were present. A fibrous capsule progressively developed around the gelled silicone by the 25th day. The histological reaction was essentially the same when the silicone gel, formed in vivo, and particles of the silicone gel formed in vitro, were implanted subcutaneously. The local inflammatory reaction, however, was more extensive and persisted for a longer time when the silicone was poorly gelled. Mononuclear cells and giant cells were conspicuous in this tissue reaction, having phagocytized the oil droplets and particles of diatomaceous earth. It has been suggested that during the processing of paraffin sections, the solvent used to carry the paraffin into the sample and again t o remove it subsequent to microtoming are excellent solvents for dimethylpolysiloxane. If this silicone polymer is not crosslinked (as is the case with the silicone fluid and the unvulcanized RTV silicone rubber), the processing will very effectively remove all of the silicone present. If the polymer is vulcanized, the solvent will swell it (about 200p/,), and ill extract all uncrosslinked polymer. During the swelling process, artifacts will be formed. Since silicones will adhere to very few- materials, any swelling of the section on the slide will almost invariably pop the rubber off the slide. Therefore, it is very unusual t o have a piece of silicone rubber remain on a slide, and it is impossible to retain any silicone fluid. These facts should very seriously be considered when interpreting slides made from tissue in contact with any type of silicones. The ionic charge on plastics used in cardiovascular surgery today is recognized as a possible cause for the development of mural t h r ~ r n b i . ~ It ~ -is~ suggested ~ that a n ionic charge on silicone could
REACTION WITH A SILICONE RUBBER GEL
659
be a basic factor in the attraction of leucocytes when it is embedded in tissue. The authors acknowledge the assistance of Bettie Joe Williams and Ava Stevenson, Laboratory Technicians, for preparation of the sections, and John Ellis for the photographs. This study was supported by Grant 3430118-NSF GB 31240 from the National Science Foundation.
References S. Braley, J. Macromol. Sci-Chem. A , 4 (3), 529 (1970). S. Braley, Arch. Otolaryngol., 78, 669 (1963). S. Braley, Med. Electron. Biol. Eng., 3, 127 (1965). W. F. Agnew, E. M. Todd, H. Richmond, and W. 9. Chronister, J . Surg. Res., 2, 357 (1962). 5. A. G. Spiers and R. Blacksma, Plast. Reconstr. Surg., 31, 166 (1963). 6. T. 0. Rees, D. L. Ballantyne, I. Seidmon, and G. A. Hawthorne, Plast. Reconstr. Surg., 39,402 (1967). 7. D. L. Ballantyne, T. D. Rees, and I. Seidman, Plast. and Reconstr. Surg., 1. 2. 3. 4.
36, 330 (1965). 8. B. S. Freeman, E. L. Bigelow, and S. A. Braley, Amer. J. Surg., 112, 534 (1966). 9. F. L. Ashley, S. Braley, T. D. €tees, 1). Goulian, and D. L. Ballantyne, Plast. Reconstr. Surg., 39, 411 (1967). 10. H. Lee and K. Neville, Handbook of Biomedical Plastics, Pasadena Technology Press, Pasadena, California, 1971. 11. F. L. Ashley, S. Braley, and E. G. McNall, Surg. Clin. North Amer., 51, 501 (1971). 12. J. M. Andrews, Plast. Reconstr. Surg., 38, 581 (1966). 13. L.H. Winer, T. H. Sternberg, R. Lehmann, and F. L. Ashley, Arch. Derm., 90, 588 (1964). 14. W. St. C. Symmers, Brit. Med. J., 3, 19 (1968). 15. J. S. Nosanchuk, Arch. Surg., 97, 583 (1968). 16. C. H. Chaplin, Plast. Reconstr. Surg., 44, 447 (1969). 17. C. Delage, J. J. Shane, and F. B. Johnson, Arch Derm., 108. 104 (1973). 18. P. V. Murphy, A. LaCroix, S. Marchant, and W. Berhard, Biomed. Mater. Symp. No. 1, 59 (1971). 19. R. H. Rigdon, J . Biomed. Mater. Res., 4, 57 (1970). 20. R. H. Rigdon, J. Riomed. Mater. Res., 7, 79 (1973). 21. R. H. Rigdon, J. Riomed. Mater. Res., 8, 97 (1974). 22. R. H.Rigdon, Texas Rep. Biol. Med., 32, 535 (1974). 23. P. N. Sawyer and S. Sainivason, J . Riomed. Mater. Res., 1, 83 101 (1967). 24. V. L. Bott and A. Furuse, Fed. Proc. Fed. A m . SOC.Exp. Biol., 39,1679 (1971). 25. P. W. Madras, W A Morton, and H. E. Petschek, Fed. Proc. Fed. Am. SOC. Exp. Biol., 30, 1665 (1971). 26. R. I. Leininger, C R C C r i . Rev. Bioeng. 1, 333 (1971-73).
Received December 13, 1974 Revised January 21, 1975
