FUNDAMENTAL AND APPLIED TOXICOLOGY 36, 1 - 1 4 ( 1 9 9 7 ) ARTICLE NO. FA962279 SYMPOSIUM OVERVIEW Immunotoxicity of Medical Devices1 KATHLEEN RODGERS,* PAAL KLYKKEN,| JOSHUA JACOBS,$ CARMELITA FRONDOZA,§ VESNA TOMAZIC,' AND JUDITH ZELIKOFF|| *Livingston Research Center, University of Southern California School of Medicine, 1321 North Mission Road, Los Angeles, California 90033; Corning Corporation, Midland, Michigan 48686; ^.Department of Orthopedic Surgery, Rush Medical College, Rush Arthritis and Orthopedic Institute, Chicago, Illinois 60612; %Johns Hopkins Orthopedics, Good Samaritan Hospital, 5601 Loch Raven Boulevard G-l, Baltimore, Maryland 21239; 'Center for Medical Devices, Food and Drug Administration, Rockville, Maryland 20857; and \\New York University Medical Center, Nelson Institute of Environmental Medicine, Long Meadow Road, Tuxedo, New York 10987 Received December 5, 1996; accepted December 9, 1996 Immunotoxicity of Medical Devices. RODGERS, K., KLYKKEN, P., JACOBS, J., FRONDOZA, C , TOMAZIC, V., AND ZELIKOFF, J. (1997). Fundam. Appl. Toxicol. 36, 1-14. INTRODUCTION Judith T. Zelikoff Since 1988, when the last symposium on medical devices was presented at the Society of Toxicology meeting, concern for the influence of these devices on the immune system has reached the forefront in the media, thereby increasing the awareness and interest of the general public and scientists alike regarding the potential immunotoxicity of medical devices. The statutory definition of a medical device is broad ' Symposium held at the 35th Annual Meeting of the Society of Toxicology (SOT), Anaheim, CA. Sponsored by the Immunotoxicology Specialty Section of the SOT. 0272-059(V97 $25.00 Copyright C 1997 by the Society of Toxicology. All rights of reproduction in any form reserved. Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 Determination of the ability of a medical device to interact with the immune system currently involves assessment of the immunogenic potential and biocompatibility of the device or an extract of the device. However, implants are often in the body for extended periods of time and/or are placed by a surgical procedure that in and of itself will generate an acute inflammatory response. This symposium discussed studies that have been performed to evaluate the immunogenicity of various devices consisting of several different compositions (i.e., silicone, metals, and latex) in contact with different anatomical sites, the ability of a device to modulate an inflammatory response generated by a surgical procedure or trauma, and the response of the body to a material left in place for extended periods of time. This symposium brought together scientists from many different disciplines to begin to identify and fill in the gaps in this area, c 1997 society of Toxicology. and includes "Any item promoted for a medical purpose that does not rely on chemical action to achieve its intended effect" (21 U.S.C., Section 321b, 1982). Many of these devices, such as implants, drug delivery, and extracorporeal devices, have intimate and prolonged in vivo or ex vivo contact with biological tissues. Due to this device/tissue interface, adverse reactions and/or toxicities from medical device materials potentially represent serious health effects. In addition, rapid advances in medical device and polymer technology have led to the development of materials whose biological properties may not be well characterized. Prior to 1976, medical devices were essentially unregulated. In 1976, the medical device amendments to the Food, Drug, and Cosmetic Act established procedures for obtaining marketing approval. One provision of the amendment establishes three categories or classes of devices depending on the regulatory controls necessary to provide reasonable assurance of safety and effectiveness. In addition, the medical device amendments specifically require that risk-benefit analysis be done to show that the benefit to the patient outweighs the risks. If the toxicity of medical devices may be implied from assessment of their constituent parts, as many potentially toxic materials are used in the manufacture of medical devices, the question arises as to why toxic responses are not frequently observed. Likely explanations for this include: (1) only minute amounts of these compounds are present in the device; and (2) exposure to these chemicals is minimized by the physical or chemical nature of the polymer. The scope of the potential problem is quite varied, including corrosion of metallic implants, leaching of chemicals from plastics and elastomers, tissue reactions to implants, and particle generation. The biocompatibility of a given material with tissue is described in terms of acute and chronic inflammatory re- RODGERS ET AL. sponses and fibrous capsule formation. The implantation of any foreign material in soft tissue initiates this normal response. The intensity and duration of the response are related to a variety of factors, including the size and nature of the material, the site of implantation, and the reactive capacity of the host. The purpose of the symposium was to heighten the awareness of the need to examine the potential immunotoxicity of new and existing medical device materials. In order to accomplish this goal, individuals from a variety of disciplines were invited to participate in these discussions. Dr. Kathleen Rodgers outlined the potential for immunotoxicity with devices implanted intraperitoneally after surgery. Dr. Paal Klykken outlined extensive research into the potential interactions between implanted silicone devices and the immune system. Drs. Joshua Jacobs and Carmelita Frondoza presented information on the problem of the failure of orthopedic implants and the possible contribution of the immune system to the failure of these materials. Finally, Dr. Vesna Tomazic discussed the development of hypersensitivity to latex. Kathleen Rodgers Adhesions are the source of a great deal of postoperative morbidity. Recent studies have shown that a majority of the small bowel obstructions which occur after abdominal surgery are the result of adhesion formation (Strickler et al, 1994). In addition, adhesion formation has been associated with infertility, chronic pain, and prolonged surgery time upon reoperation (Stout et al, 1991; Peters et al, 1992; DeCherney et al, 1984; Trimbos-Kemper, 1985). Because of the great opportunity for benefit, many attempts have been made to develop materials or Pharmaceuticals to reduce adhesion formation. This presentation summarized the limited studies that have been conducted to evaluate the potential immunotoxic effects of devices used clinically to reduce adhesion formation. Interceed (TC7; oxidized regenerated cellulose), a knit fabric that gels in vivo to form a solid barrier, is approved for use as a barrier to prevent adhesion formation (Adhesion Barrier Study Group, 1989; Sekiba et al., 1992; Diamond et al., 1987). This material was shown to degrade 5-7 days after implantation (Dimitrijevich et al., 1990). Another device, Preclude (Goretex), is made of expanded polytetrafluoroethylene (Boyers et al., 1988; Surgical Membrane Group, 1992). Goretex surgical membrane is approved for use as a pericardia! replacement. Unlike Interceed, which interacts with the tissue and is held in place without sutures, Preclude requires suturing. In addition, in contrast with Interceed, Host Resistance to Infection Of utmost importance is that these materials should not reduce the ability of the cells present in the abdomen after surgery to respond to a bacterial infection. Bacteria may be introduced during surgery, and modification of the ability to clear bacteria may have adverse effects. Unpublished studies have shown that HA reduced abscess formation and mortality after infection. On the other hand, Hyskon increased mortality after infection, did not affect or reduce abscess formation, and increased bacterial growth in vitro (Bernstein et al, 1982). Hyskon has been shown to produce an oncotic effect resulting in a transient ascites (Krinsky et al, 1984). Studies have shown that a large volume of fluid in the peritoneal cavity after surgery may substantially reduce the ability Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 IMMUNOTOXICITY OF MEDICAL DEVICES USED IN ADHESION PREVENTION Preclude is not biodegradable and may, therefore, act as a foreign body. If the physician desires to remove the material, a second operation is required. Hyskon (32% Dextran 70), a liquid barrier used routinely to reduce adhesion formation (diZerega and Rodgers, 1992), is approved for use as a distention agent in hysteroscopy. Dextran is used as a plasma expander and anaphylaxis has been associated with its use in a small number of patients (Bailey et al., 1967; Borten et al., 1983; Trimbos-Kemper and Veering, 1989; Stangel et al., 1984; Ahmed et al., 1991). Common side effects associated with intraperitoneal use of Hyskon include ascites formation, transient weight gain, vulvar edema, and pleural effusion (Cleary et al., 1985; Magyar et al., 1985; Adhesion Study Group, 1983; Tulandi, 1987). Hyaluronic acid (HA) can be prepared in a viscoelastic solution or film that is currently undergoing clinical trials and FDA review for use in reduction of adhesion formation. A variety of barrier or device products were found effective in animals and may be developed for clinical use in adhesion prevention. However, no systematic assessment of immunotoxic potential (other than histopathological assessment at implantation site and assessment of sensitization potential) is ongoing for these devices. There are several reasons to consider the immunotoxic potential of adhesion prevention devices. These materials are placed at the time of surgery and may modify inflammatory response and healing. In addition, alterations in inflammatory response by foreign bodies (e.g., talc) and ischemia may contribute to adhesion formation. Last, bacteria may be introduced at surgery and devices may modify the antibacterial activity of peritoneal cells. The purpose of this discussion was to: (1) overview the published information on the immunotoxicology of barriers used in adhesion prevention and (2) suggest avenues of further inquiry into interactions between the medical devices used to reduce adhesion formation and the inflammatory and immune responses. IMMUNOTOXICTTY OF MEDICAL DEVICES of the host to eliminate a bacterial inoculum (Dunn et al., 1984). Therefore, increased susceptibility to intraperitoneal infection may be due to the increase in peritoneal fluid after Hyskon administration. In addition, studies involving in vitro exposure to Hyskon showed that this material reduced the phagocytic capability of peritoneal macrophages and reduced the basal and mitogen-stimulated proliferation of peritoneal lymphocytes from women undergoing pelvic surgery (Rein and Hill, 1989). This immunomodulatory effect may also contribute to the reduced host resistance to bacterial infection observed in animal studies. In summary, Hyskon may be immunosuppressive and may increase the risk of bacterial infection. Alterations in Cellular Function Summary Immunotoxicology studies of devices implanted to reduce postoperative adhesion formation should be concerned with effects on host resistance to infection. In addition, immunotoxicologic evaluations should examine the response in postsurgical environments and monitor the effects of the device on the postoperative inflammatory response. The implantation studies currently being performed give insight into the effects of the material on the mesothelial surface. In the future, these evaluations may also include determination of cell number and cell function. IMMUNOLOGICAL INVESTIGATIONS OF SILICONE IMPLANT MATERIALS Paal Klykken and Kimber White Spearheaded by the implant controversy, the biocompatibility of silicone implant materials is being actively investi- Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 Peritoneal surgery has been shown to modulate the activity of peritoneal cells. Leukocytes were shown to increase in number and functional activity as measured by respiratory burst, phagocytosis, cytokine release, and arachidonic acid metabolism (Rodgers et al., 1988; Abe et al, 1989, 1991; Shimanuki et al., 1986). Preclude was shown to increase polymorphonuclear neutrophil (PMN) numbers (slightly), cause a sustained increase in leukocyte number, and increase the respiratory burst function of peritoneal leukocytes (Haney and Doty, 1992). This prolonged increase in peritoneal leukocyte number after implantation of Preclude may be due to its action as a foreign body. On the other hand, Interceed did not increase the inflammatory response (as determined by PMN number) when implanted and was degraded in vivo by macrophages (Dimitrijevich et al., 1990). As a result of this degradation by leukocytes, Interceed caused a transient increase in peritoneal leukocyte numbers during the period of ongoing degradation. As with Preclude, Interceed increased the respiratory burst function of peritoneal leukocytes, but, in contrast, increased leukocyte adherance. gated at the clinical and nonclinical levels. The primary focal points of these studies have been the potential interactions of silicones and the immune system and whether immunemediated diseases are linked to systemic silicone exposure. The principal areas of laboratory investigation have addressed the following four questions: Can silicones alter immune competence? Can silicones trigger specific immune responses? Can silicones serve as adjuvants? Can silicones elicit or amplify autoimmune-like diseases? The silicone materials which have been primarily studied are the three basic components of the breast implant device: the outer silicone rubber or elastomer envelope, the resinous gel, and the highviscosity silicone fluid used to swell the gel to the appropriate physical consistency. To address the question whether silicone implant materials can alter the immune competence of the test animal, a number of studies have been conducted at the Medical College of Virginia/Virginia Commonwealth University under the auspices of the National Toxicology Program (NTP) as well as Dow Corning Corporation. In the NTP studies, two exposure periods were used. A 10-day exposure period was chosen to allow for characterization of immune competence during the peak of the acute inflammatory response. To minimize the Likelihood of a solid state tumorigenesis response, 180 days was selected as the time period to measure possible effects due to chronic silicone exposure. Extending beyond the traditional Tier I and Tier II batteries, 25 immunotoxicological parameters were evaluated at both the 10-day and 6-month time points. The results of this comprehensive research effort demonstrated that no silicone test material produced an adverse effect, with the exception of mammary gel, which suppressed natural killer cell activity when measured in a 4-hr chromium release assay (Bradley et ai, 1994a,b). Under similar exposure conditions, no effects (either augmentation or suppression) were observed in any of the host resistance studies, including the Listeria monocytogenes, Streptococcus pneumoniae, or the B6F10 melanoma tumor model. L monocytogenes studies conducted at Dow Corning after 14, 45, or 90 days of exposure to silicone fluid, gel, or elastomer also failed to alter host resistance (Klykken et al, 1994). A second area of investigation which has received considerable attention is whether silicone implant materials can trigger specific immune responses, either because the silicone material is inherently antigenic or because it generates a neoantigen secondary to alterations in protein conformation. An early clinical observation which suggested a specific anti-silicone immune response was published by Goldblum et al. in Lancet in 1992. In these studies, an ELISA-based system using silicone tubing as the solid-phase substrate was used to evaluate sera from silicone ventriculoperitoneal (VP) shunt-implanted patients and non-silicone-implanted con- RODGERS ET AL. highly hydrophilic polymers, are not immunogenic. These data are also consistent with Dow Corning studies which failed to detect heightened histological responses to silicone implant materials (fluid, gel, and elastomer) in mice immunized with Freund's complete adjuvant-silicone (fluid, gel, and elastomer) mixtures. Indices of capsule thickness, maturity, cellularity, and composition were equivalent in the sensitized mice and the unsensitized controls for all three silicone implant materials at all three time points (2, 6, and 13 weeks) examined. Alternatively, it has also been hypothesized that systemic silicone exposure can trigger a specific immune response(s) by causing conformational changes in protein structure and subsequent presentation of cryptic antigens at the siliconetissue interface. This in turn would increase the likelihood of a cross-reactive immune response(s) against normal, unaltered self-proteins. In a study reported by Kossovsky et al. (1993), 9 of the 249 women with silicone breast implants had significantly higher titers directed against laminin and fibronectin than the nonimplanted 47 healthy women or 39 rheumatological patients tested. Based on this differential response of 9 women, the authors concluded that breast implant exposure can induce an autoimmune response to self proteins. Whether this interpretation is valid is at present unknown. The 9 responders of the 249 women represent less than 4% of the test population. As the referenced agedmatched healthy and rheumatologic disease control populations had comparatively few women, a value of 0 instead of 1 or 2 for the number of responders may not be unexpected from a statistical perspective. Other investigators have also examined the possibility of silicone inducing conformational changes in proteins. In an in vitro setting, Butler et al. (1996) examined the antigenic behavior of two proteins (IgG2a and IgGl) on two different surfaces (Immulon2 and silicone elastomer). The ELISA determinations indicated that the immunoglobulins will adsorb onto the silicone surface and the antigenicity of the proteins will change upon adsorption. However, consistent with the biomaterials literature, this phenomenon is not unique to silicone surfaces. As reported, the proteins adsorbed onto the Immulon2 surface were more antigenically detectable than when the proteins were adsorbed onto the silicone surface. Importantly, these studies failed to demonstrate that adsorption of proteins onto a silicone surface produces a new antigenic specificity. The inability to detect an IgG antibody with specificity for the silicone substrate is in agreement with the research findings of Nairn and van Oss (1992), who investigated the effects of hydrophilicity, hydrophobicity, and water solubility on the immunogenicity of some natural and synthetic polymers. Their structure activity studies indicated that highly hydrophobic polymers such as silicones, as well as To address the (third) question of whether silicones can serve as adjuvants and nonspecifically amplify immune responses, two different types of silicone exposure regimens have been utilized. Under conditions in which the hydrophobic silicone gel is coadministered with an aqueous solution of bovine serum albumin antigen in an emulsion-like preparation, a robust adjuvant response is noted in rats and mice Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 trols. Based on the increased adsorption of IgG immunoglobulin to the silicone tubing in two of the VP shunt patients, it was concluded that the immune system can recognize and mount a specific immune response to silicone. Importantly, follow-up studies by these investigators have indicated that the differential adsorption of the positive and negative VP shunt sera are largely lost when the whole serum samples are enriched for immunoglobulins. Consistent with this observation, Goldblum further demonstrated that albumin can competitively compete for and displace adsorbed IgG molecules from the silicone surface (Goldblum et al., 1995). Clinical samples analyzed at Emerald Biomedical Sciences, Inc. (Houston, TX) have also suggested that silicone breast implant exposure can elicit anti-silicone antibodies (Wolf et al., 1993). In their ELISA-based system, sera from patients with intact implants and ruptured implants recorded optical density readouts higher than those of their nonimplanted healthy controls, suggesting a correlation between patient history and silicone-specific IgG levels. However, an independent examination of the Emerald test procedure, using materials and coded serum samples provided by Emerald, generated a different data profile at Johns Hopkins University (Rose et al., 1996). In the Johns Hopkins study, not only were the positive control sera provided by Emerald positive, but patients with connective tissue diseases and no history of silicone implants were also positive. Accordingly, the Johns Hopkins' investigators concluded that this test system should not be used as a diagnostic probe for the presence of a silicone implant or as clinical evidence of an immune-mediated reaction to a silicone implant. The issue of a specific immune response to silicones has also been addressed in controlled animal studies. Sera derived from long-term (180 days) silicone elastomer or gel implant studies in mice were analyzed for the presence of anti-silicone antibody using the ELISA procedure based on methodology described by Goldblum et al. (1992). While there is a concentration-dependent adsorption of IgG to the solid-phase silicone substrate, there is no difference between the silicone (elastomer and gel)-implanted mice and their appropriate vehicle controls. Similar results were generated when female Fischer 344 rats were evaluated for the presence of anti-silicone antibodies following implantation with 15 ml of mammary gel for 84 days. Again, no difference in adsorption of IgG to the solid phase silicone substrate was observed between gel-implanted and vehicle control rats. IMMUNOTOXICITY OF MEDICAL DEVICES BIOLOGY OF DEGRADATION PRODUCTS FROM ORTHOPEDIC BIOMATERIALS Joshua Jacobs Paniculate wear debris from prosthetic materials or bone cement are phagocytosed by tissue macrophages and these activated cells release various mediators. Among many cellular mediators, interleukin-1 (IL-1), IL-6, and prostaglandin E2 (PGE2) are believed to be the most important components capable of inducing cell proliferation, generating osteoclast formation, and stimulating osteoclasts to resorb adjacent bone (Giant et al., 1993, 1994a,b,c; Goldring et ai, 1992; Goodman et ai, 1989; Gowen et ai, 1992; Shanbhag et ai, 1995). An aggressive granulomatous tissue composed of fibroblasts, macrophages, and foreign body giant cells develops at the bone/cement or bone/prosthesis interface (interfacial membrane) and replaces the resorbed bone (Goldring et ai, 1983). Many factors influence the extent of pathological bone resorption (Gowen, 1992; Goldring et ai, 1992; Harvey, 1988). Cytokines are produced transiently, usually have multiple overlapping activities, regulate each other, interfere with receptor expression on target cells, and induce gene expression of many other "bone-resorbing" agents. Several potent cytokines and enzymes are involved in pathological bone resorption and fibrous tissue formation at the bone/ cement or bone/prosthesis interface. "Simplified" in vitro systems are needed to reveal certain aspects of more complex in vivo events. Many new cell and molecular biology approaches were adopted in orthopedic research to understand the mechanisms of particulate-induced osteolysis. These studies include measurements of particulate-induced cytokine release, organ culture experiments performed to simulate pathological bone resorption taking place in vivo, and detection of gene expression for various cellular mediators, metalloproteinases, and tissue inhibitor of metalloproteinases (TIMP) in interfacial membranes obtained from failed total hip arthroplasties (THAs). Paniculate Wear Debris in Cells of Interfacial Membranes Bright-field and polarized light microscopy consistently have identified abundant intracellular birefringent, presumptive polyethylene (PE) and dark metal particles. This is the case for the monocytic infiltrations present in the joint capsule, femoral pseudomembrane, and tissues within periprosthetic osteolytic lesions in both stable and loose implants. The most prominent paniculate species is PE derived from the acetabular component in either cemented (Lee et ai, 1992) or uncemented (Shanbhag et ai, 1994) implants. The size of the PE particulates retrieved from tissue samples was found to be in a narrow range; >90% of wear debris were smaller than 1 (im (Shanbhag et ai, 1994). Steady-State mRNA Ejcpression for Cytokines in Interfacial Membranes of Failed THAs Certain areas of the tissue specimens were found to be more "active" than others in terms of the expression of genes coding for IL-ls, while other cytokines (e.g., TNF-a) were expressed in only a few cells. IL-1 a was not detected in all samples but, when present, was localized adjacent to the bone. The steady-state mRNA levels for various cytokines were Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 in comparison to that of the appropriate FCA control (Nairn et ai, 1993a; Klykken and White, 1996). In agreement with the published literature which indicates that a wide variety of polymers and hydrophobic substances can serve as adjuvants (Warren, 1992), personal care ingredients including lanolin, white mineral oil, and isopropyl palmitate were also active under these test conditions. Under more relevant exposure conditions (i.e., antigen not premixed with the breast implant material) no adjuvant response is noted when silicone-exposed animals are subsequently challenged with a known antigen and evaluated at the time of peak response (Bradley et ai, 1994a,b). Laboratory efforts have also been directed to determine whether silicone gel can elicit or amplify autoimmune-like diseases. In a series of rat studies, coadministration of gel and antigen has been shown to enhance the immune response to the administered antigen, but failed to induce thyroiditis in a Wistar thyroglobulin model (Nairn et al., 1993b), arthritis in a Lewis type II collagen model (Nairn et al., 1995), and arthritis in a Dark Agouti collagen II model (Nairn et ai, 1996). Our preliminary investigations with silicone gel implants in a Brown Norway lupus model have yielded similar conclusions. Exposure to silicone gel for 26 weeks failed to induce the characteristic alterations in serum autoantibody levels. Furthermore, there was no difference in the IgE response of control and gel-implanted animals following challenge with mercuric chloride. Silicone gel and 1000 cs PDMS fluid have also been shown to be inactive in the Tight Skin mouse model for scleroderma, where a number of parameters including histopathology and circulating autoantibody profiles to topoisomerase and RNA polymerase were monitored (Frondoza et ai, 1996b). In summary, the controversy still exists, and it exists in part because of anecdotal reports and associations. Neither should be considered scientific data, as they lack essential control groups for comparison. From a scientific perspective, there is little support at present for a cause and effect relationship between silicone exposure and immune-mediated disease. RODGERS ET AL. different with TGF-/3 > IL-1/9 > PDGF-A > PDGF-B > TNF-a (Giant et al, 1994b). There were large variations in cytokine mRNA levels among patients. However, the observed levels of cytokine expression noted above were characteristic for all samples in either focal or diffuse osteolytic lesions in both cemented or cementless THAs. The steady-state mRNA level for IL-l a, as mentioned above, was very low in fresh tissue samples and detectable in only a few cases. However, when the samples of interfacial membranes were cultured as explants, cells expressed high levels of IL-l (Giant et al, 1994c) and released this protein into the culture media (Shanbhag et al, 1995). In a 24-hr culture, cells of the same tissue sample expressed many hundred times more mRNA for IL-1 (and IL-6 as well) than that found in fresh tissue (Giant et al., 1994c), whereas there was only a slight (10-fold) increase in the expression of IL1 (Giant et al., 1994c). Thus, although IL-la is believed to be a less important bone-resorbing ' 'agent'' than IL-1/9, cells of interfacial membranes have a high, albeit latent, capacity to respond to changes in their microenvironment (e.g., phagocytosis of particulates) by IL-la secretion. PGE2 is believed to be the major regulatory component of IL-l- and TNF-a-mediated bone resorption (Giant et al., 1993; Akatsu et al, 1991; Yoneda et al., 1978); however, a direct correlation between the amounts of PGE2 and other cytokines (TNF-a, IL-l, and IL-6) was not found in interfacial membranes (Shanbhag et al., 1995). IL-la was found to be the only cytokine which was detected in high amounts in explant cultures of THA membranes. The cemented THA membranes demonstrated a 3- to 16-fold increase compared to control synovial tissues from either femoral neck fractures or autopsies. Despite similar biological effects of IL-l and TNF-a (both recognized as potent inducers of inflammation), levels of TNF-a were not increased compared to those of IL-la in membranes from uncemented arthroplasties. IL-6 is considered to be an "anti-inflammatory" agent. In spite of its ability to cause bone resorption (Ishimi et al., 1990), IL-6 protects against the harmful effects of IL-l by suppressing secretion of IL-l and TNF-a at the level of gene transcription (Schindler et al., 1990; Dinarello, 1991). On the other hand, IL-la, IL-1/3, and TNF-a induce the expression of mRNA for IL-6 (Ishimi et al, 1990). However, while IL-6 levels were elevated in control synovial tissues (from cadavers and patients with femoral neck fractures), they were significantly lower in THA membranes retrieved at revision surgery. It is possible that continuous stimulation of macrophages by wear debris may overstimulate and/or ' 'exhaust'' the IL-6-generating mechanism and contribute to the uncontrolled elevation of IL-l secretion. This, in turn, could contribute to the process of aseptic loosening. The effects of various particulate species upon the expression of mRNAs coding for metalloproteinases and TTMP have been studied (Giant et al, 1994a; Yao et al, 1995). It was found that human fibroblasts of various origins (normal skin, rheumatoid synovial tissue, and periprosthetic granulomatous tissues of failed THAs) could respond differently to direct exposure to the same particulate species, although all responded to titanium (Ti) by increasing metalloproteinase and, to a lesser extent, TEMP expression. These findings suggested a direct pathologic effect of Ti particles on fibroblast functions, which might be responsible for the failure of bone remodeling in osteolytic lesions (Yao et al, 1995). In addition, Ti particles blocked the expression of mRNAs for procollagen at(J) and a1(Tn), thus abolishing collagen synthesis in the human osteoblastoid cell line MG-63. As osteoblasts play crucial roles in the maintenance and remodeling of bone tissue, a particulate-induced suppression of collagen synthesis in these cells may inhibit osteogenesis and die repair of bone matrix in osteolytic lesions (Yao et al, 1995). Bone Resorption in Calvarial Bone Organ Culture Phagocytosable particulates induced PGE2 secretion by peritoneal macrophages (Giant et al, 1993). The PGE2 production was even higher in cocultures of bone and macrophages and in bone organ cultures exposed to conditioned medium (CM) from particulate-stimulated macrophage culture. CM of particulate-stimulated macrophages increased the active 45Ca release from bones. CM from nonstimulated monocyte/macrophage cell lines exhibited bone-resorbing capacity, but this was not the case with CM from peritoneal macrophages, although it contained high levels of PGE2 and approximately the same amount of IL-l as measured in nonstimulated IC-21 cultures. CM of Ti, polystyrene, and PMMA particulate-stimulated macrophages (peritoneal, IC21, and P388D, cell lines) increased the 43Ca release by bones. In bone organ cultures, the major source of PGE2 is the calvarial bone itself, which most likely mediates a "spontaneous" efflux of calcium (Giant et al, 1993). This passive efflux was inhibited significantly, but not completely, by either indomethacin (IM, cyclooxygenase inhibitor) or misoprostol (prostaglandin E] analogue). Since increased 45Ca release generally correlated with increased PGE2 and IL1 levels and since exogenous PGE2 and recombinant IL-l increased 43Ca release from calvarial bones, one might conclude that increased bone resorption was a result of the ac- Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 Mediators Released in Explant Cultures of Interfacial Membranes Expression of mRNAs Coding for Bone-Resorbing Factors/Mediators in Human Fibroblast Cell Lines Exposed to Particulates IMMUNOTOXICITY OF MEDICAL DEVICES EFFECT OF PROSTHETIC WEAR DEBRIS ON MACROPHAGES, FIBROBLASTS, AND BONE CELLS Carmelita Frondoza Aseptic loosening of implant components is a major complication after total joint replacement (Barrack et al, 1992). Loosening is characterized by foci of radiolucent zones at the implant—bone interphase pointing to bone resorption or osteolysis (Lombardi et al., 1989; Willert etal, 1990). Analysis of the periprosthetic tissue shows the presence of wear debris derived from implant materials and a granulomatous, "foreign body-like" response (Goldring et al., 1983; Maguire et al, 1987; Pazzaglia et al, 1987; Goodman et al, 1989; Santavirta et al, 1990, 1991; Dorr et al, 1990). The periprosthetic tissue appears thickened, fibrous, and infiltrated with inflammatory cells and has been referred to as "pseudomembrane" (Goldring et al., 1983). These observations led to the supposition that wear debris generated from the articulating surfaces of joint prostheses plays a key role in aseptic loosening of implant components (Charnley et al, 1975; Agins et al, 1988; Salvati et al, 1993). Multiple factors that are mechanical or biological in nature have been implicated in the cascade of events leading to aseptic loosening (Maloney et al, 1989; Galanteef al, 1991; Amstutz et al, 1992). Mechanical factors have been cited as causing prosthesis instability and the loosening process induces osteolysis. In contrast, others propose that biological factors trigger loosening. Wear debris induces osteolysis and progressive osteolysis results in loosening of prosthetic components. While mechanical factors are initially involved in production of wear debris, a strong case can be made that the adverse biological response to particulates is the ultimate cause of prosthesis loosening. This notion is supported by the evidence that wear debris distributed in the periprosthetic tissue is associated with infiltrates of activated macrophages, giant cells, fibroblasts, and lymphocytes (Revell et al, 1975; Kozinn et al, 1986; Howie et al, 1988; Dorr et al, 1992; Salvati et al., 1991; Kim et al, 1993). The appearance of these inflammatory cells at the periprosthetic tissue could trigger osteolysis through release of soluble mediators such as cytokines, prostaglandins, and degradative enzymes (Herman et al, 1989; Santavirta et al, 1991; Quin et al, 1992; Jiranek et al, 1993). Bone loss could eventually result in aseptic loosening. Whether wear debris causes loosening or is the consequence of the loosening process is still unresolved. Most likely, mechanical and biological factors are equally involved in the loosening and eventual failure of the implant. To help define the role of prosthetic wear debris on aseptic loosening, in vitro tissue culture and animal models have been used. In vitro tissue culture models have the advantage in that they facilitate analysis at the cellular and molecular levels. For example, changes in transcription or translation of cellular products in response to wear debris can be analyzed. Cell to cell communication and signal transduction events can be traced as cells are exposed to wear debris. In vitro techniques are also less costly and less time consuming. However, in vitro models are of limited value in detecting crucial physiological systemic effects. In using in vitro models, several important questions must be considered: (1) the cell type to be used; (2) the appropriate assays and their optimum endpoints; and (3) the debris preparation to be evaluated. Several cell prototypes have been studied to model or mimic tissue response to biomaterials. Examples of these are prototypes of inflammatory cells such as peripheral blood monocytes, lymphocytes, and fibroblastic cells (Rae et al, 1979; Goldring et al, 1990; Giant et al., 1993; Horowitz et al, 1991, 1995; Haynes etal, 1993; Frondoza etal, 1994; Shanbhag etal, 1995; Gonzales etal, 1996). Cells retrieved from tissues or blood which are maintained in short-term culture for immediate use are thought to most closely express their original tissue characteristics or "phenotype." However, the number of cells available for analysis is usually limited and further expansion to increase cell numbers may be required. In this case, cells propagated in culture for extended passage need to be assessed if they still retain their Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 tions of PGE2 and IL-1. However, the use of anti-IL-1 antibodies and IM further demonstrated that: (1) PGE2 alone was not necessarily responsible for active 45Ca release; and (2) factors other than PGE2 and IL-1 also might participate in bone resorption in vitro. Although PGE2 and IL-1 have been shown to regulate one another (Goldring et al, 1992; Boyce et al, 1989; Harvey, 1988a), this mutual regulatory effect is not evident in calvarial organ cultures. Since the seminal report of Willert et al. (1977), attention has focused upon the role of particulate release in the etiology of bone loss and aseptic loosening and subsequent clinical failure of total joint replacement prostheses. It is now clear that macrophages, fibroblasts, and osteoblasts respond differently to size, dose, composition, and perhaps surface energy of different particulate species which promote the release of various levels of bone-resorbing agents. The cellular heterogeneity and the aggressive behavior of the tissue in osteolytic areas are most likely the consequences of multiple cellular interactions and the effects of various products of cells exposed to particulate biomaterials. The tissue heterogeneity may determine how local events proceed in osteolytic lesions. Although considerable limitations exist when comparing in vitro and in vivo events, new cell biology and molecular biology techniques offer the opportunity to correlate particulate-related in vitro experiments with in vivo events taking place in the periprosthetic tissue environment. RODGERS ET AL. 50 o Tritiated Thymidine Incorporation 40 •H 4-1 (0 u o 30 o u 20 c 10 Control l)jg/ml lO^g/ml FIG. 1. Effect of metallic wear debris on human synoviocyte proliferation. Metallic wear debris from periprosthetic tissue obtained during revision arthroplasty was isolated by sequential enzyme digestion. Different concentrations of the wear particles were added to cultures of human synoviocytes and proliferative indices were determined by [3H]thymidine incorporation. Effects of Wear Debris on Cellular Function Metallic wear debris (Ti, CoCr) at the lower concentrations tested (1 /ig/ml) decreased the proliferative capacity of human synoviocytes (Fig. 1). Similarly, cells exposed to PMMA beads showed reduction in proliferative indices. The decrease in proliferative indices was more profound when cells were incubated with higher concentrations of particulates. In contrast, higher concentrations of particulates induced release of lower concentrations of cytokines TNFa (Fig. 2). A similar inverse dose-dependent pattern in cytokine release was observed when synoviocytes were incubated with PMMA particles. Machine-generated Bioglass particles were not cytotoxic for synoviocytes but induced release of TNFa similar to the pattern shown in Fig. 2. Control (Oug/ml) 1|jg/ml 10ug/ml 100ug/ml FIG. 2. Effect of metallic wear debris on human synoviocyte cytokine TNFa release. Metallic wear debris from periprosthetic tissue obtained during revision arthroplasty was isolated by sequential enzyme digestion. Different concentrations of the wear particles were added to cultures of human synoviocytes and levels of secreted TNFa were assayed in the supernatant medium ELISA. Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 original characteristics. Alternatively, tumor cell lines which continue to display certain features of their normal tissue of origin have been used. To evaluate cellular response to wear debris, various techniques are available to determine cytoxicity and cellular activation. Cytotoxicity is commonly determined by conventional vital dye uptake and enumeration of cell numbers. Perturbation in cell proliferation can be detected using more sensitive radioisotopic and immunocytochemical techniques. Cellular activation may be assessed by measurement of inflammatory mediator release such as the interleukins or prostaglandins using ELISA or radioimmunoassay. Another important consideration is the wear debris preparation for testing. Over the past 20 years, the most common prosthetic materials available clinically have been cobaltchrome (Co-Cr) alloys, titanium, titanium alloys, ultrahighmolecular-weight polyethylene (UHMWPE) and polymethyl methacrylate cement (PMMA). Other less commonly used materials or those which are currently being tested are bioceramics such as hydroxyapatite and Bioglass. Numerous studies identified cell and tissue response to machine-generated particulates from prosthetic materials of different sizes, shapes, and concentrations. Although more easily obtained, there is concern that these debris preparations may not be representative of those generated in the patient's tissues. To address this concern, several investigators attempted to isolate, characterize, and test wear debris retrieved from periprosthetic tissue. Yet these efforts have encountered technical difficulties due to the limitation of tissue particulates that can be retrieved and the possibility that they are altered by the isolation procedures. There is also wider variability in sizes and shapes in the wear debris retrieved from tissue which could contribute to variation in in vitro results. Cur- rent studies to compare cell response to wear debris that are machine generated or isolated from tissue are underway (Schmiedbergefa/., 1994; Campbell etal, 1995; Shandbhag et al, 1995). The major problem however, is to compare matched preparations with similar size, shape, and chemical structure in order to draw valid conclusions. Taking all these factors into consideration, we have proceeded to characterize cellular response to wear debris. Our study focused on synovial tissue as a source of cells for use in in vitro models. The anatomical accessibility of synovia! tissue to the bearing surfaces of artificial knee joint implant facilitates exposure to wear debris. Synovial tissue consists of phagocytic macrophage-like and fibroblastic cells capable of producing inflammatory cytokines and degradative enzymes (Hirsch et al, 1985; Allard etal, 1990; Wilkinson etal, 1992). Synoviocytes also produce collagen type I, which is the major constituent of the thickened periprosthetic membrane of loosened implants. These factors may contribute to the active participation of synoviocytes in adverse tissue response to wear debris. IMMUNOTOXICITY OF MEDICAL DEVICES Exposure of synoviocytes to UHMWPE resulted in a dosedependent decrease in cell numbers. By Day 5 significant reduction in cell numbers was noted at 100 /ig/ml of debris. Light microscopy revealed that fewer cells remained adherent on microcarriers upon exposure to higher concentrations of UHMWPE. Microcarriers containing cells cultured in medium alone exhibited confluent cells on their surface. Ultrastructural analysis indicated that synoviocytes phagocytosed debris. At the higher concentrations of UHMWPE, cells showed signs of lysis and internal structures were not discernible. The few cells that remained adherent to the surface of the microcarriers appeared pyknotic and had lost their cellular integrity. Synoviocytes exposed to UHMWPE stained more intensely for IL-1/3 compared to the untreated control, suggesting higher levels of intracellular IL-1/3. Summary Vesna Tomazic Natural rubber latex is a source material for a variety of medical devices such as surgical and examination gloves, condoms, and catheters. For years, the only known problem associated with rubber was occasional irritation and dermatitis caused by residual chemical additives in the finished products. During 1989/1990, however, there was a sudden surge of reports of the type I, IgE-mediated hypersensitivity to latex. Although contact with latex devices is mainly dermal and only occasionally through the mucous membranes, frequent exposures to latex products may result in the development of the IgE-mediated hypersensitivity induced by proteins that are constituents of the natural rubber latex. The clinical manifestation of immediate or type I hypersensitivity is highly variable, from relatively mild symptoms of local contact urticaria to more severe asthma-like respiratory symptoms to the most severe and life-threatening systemic anaphylactic reactions. The most obvious reason for a sudden appearance and such a high frequency of type I allergic reactions is a 10-fold increase in the use of latex gloves as a result of preventive measures against HIV infection. In addition to the increased exposure to latex, other factors contributed to the increased frequency and severity of the reactions. First, due to such a high demand, a large portion of gloves marketed in this country were imported from various small manufacturers around the world, where the quality control may have been suboptimal. Also, a faster turnover of the raw latex may have resulted in the higher amounts of latex protein on the finished products. Another change that occurred at the same time was replacement of talc with cornstarch as a donning powder on gloves. Cornstarch, however, has a strong propensity to bind latex proteins and creates an airborne allergen and an additional route of exposure. Finally, sterilization with ethylene oxide was also introduced a few years ago; it was recently shown that ethylene oxide residues can sensitize individuals and also potentiate sensitization to latex proteins. The occupational exposure to latex products, therefore, presents the highest risk for sensitization to latex proteins. Several epidemiological studies of occupationally exposed groups published since 1990 confirmed that the highest prevalence of type I allergy was found among hospital personnel, especially among operating room physicians and nurses, dentists, and rubber plant workers. The prevalence levels in these groups range from 9 to 17%. The most affected group, however, is spina bifida children, in whom the prevalence of type I allergy is as high as 70%. Due to the frequent surgeries and other minor medical procedures associated with their congenital malformation, these children are con- Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 The major finding of the present study is that metallic wear debris retrieved from periprosthetic tissue is cytotoxic to human synoviocytes. As the cells die, inflammatory cytokines which are also known to mediate bone resorption are released. Our study supports the hypothesis that retrieved wear debris from periprostatic tissue can directly activate synoviocytes in vitro. These observations suggest that synovial cells surrounding the implant may be susceptible to activation by wear debris in the knee joint. Thus, metallic wear debris in the periprosthetic tissue may induce synoviocytes to release cytokines capable of mediating osteolysis. A novel method to evaluate the effect of buoyant UHMWPE on tissue cells was successfully used for the results presented above (Frondoza et al., 1996a). By allowing synoviocytes to adhere to microcarriers, they can be readily exposed to buoyant UHMWPE debris in spinner cultures. Biocompatibility studies of UHMWPE on a monolayer cell culture model have been complicated by its low density and its tendency to float. The spinner culture method facilitates contact between cells and buoyant particulates and makes it possible to determine whether phagocytosed UHMWPE debris affects cell functions. Since the spinner culture system mimics some features of the biomechanical environment in the joint, cell response to UHMWPE in spinner culture may yield insights into the in vivo behavior of synoviocytes during the pathogenesis of aseptic loosening. We also demonstrated that paniculate bioactive glassBioglass was not cytotoxic to synoviocytes but elicited release of inflammatory cytokines. Bioglass has been reported to have physicochemical properties suitable for use as implant coating. However, our observation that Bioglass elicits the release of inflammatory cytokines suggests the need to further examine and develop methods of applying Bioglass to an implant which will minimize production of wear particles. HYPERSENSITIVITY TO LATEX PROTEINS: ETIOLOGY, DIAGNOSIS, AND PREVENTION 10 RODGERS ET AL. These findings suggest that NAL extract may have the most complete presentation of all potential allergenic proteins. Further confirmation of this finding came from our studies performed with rabbit anti-latex sera. Rabbits were immunized with NAL, AL, and GL protein extracts, and their sera were analyzed in immunoblots with corresponding protein extracts. In comparison with the SDS-PAGE profile of three latex extracts, the immunoblots with rabbit antisera revealed that NAL-immunized rabbit sera reacted with all proteins present in all three extracts. Sera from AL- or GLimmunized rabbits failed to recognize some of the proteins present in NAL extract. We have not, however, observed any new positive proteins in AL and GL extracts that were not present in NAL extract. These data supported our previous conclusion that NAL extract appears to contain the most complete display of antigenic epitopes of latex proteins. Since all of the latex proteins are not present on every finished product, the antibody specificities in individual human sera usually reflect proteins present on the specific device that caused sensitization. A marked variety in the reaction patterns of human sera was repeatedly observed, especially when the adult population was compared to the pediatric population. Therefore, to have a reliable and predictive test to evaluate the potential allergenicity of latex products, it is important to prepare a reference pool of sera that will have the capacity to react with all potentially allergenic proteins. In these studies, we used a pool of adult immune sera with diversified responses, prepared after careful screening of a large number of individuals from various geographic locations and with various sensitization histories (Akasawa, 1995). When this serum pool was evaluated in immunoblots against NAL, AL, and GL protein extracts, the largest number of positive proteins was observed in the NAL extract. The human serum pool reacted with some proteins in AL and GL extracts that were not recognized by rabbit sera immunized with analogous protein preparations. Currently available diagnostic tests for allergies include the in vitro tests for serum IgE levels and allergen-specific IgE levels and the in vivo gold standard of skin testing. It would be advantageous to be able to replace skin testing with an adequate in vitro counterpart. The skin reaction, which is a direct measure of the potential allergic reaction, does not necessarily correlate with the level of serum IgE. Usually, the presence of IgE antibodies in the serum is a clear indication that the individual has or is in the process of developing allergy. The presence of anti-latex IgE antibodies in the serum may precede the manifestation of clinical symptoms for an extended time and, therefore, may indicate an ongoing process of sensitization. Although a correlation between these two tests may not always exist, detecting the sensitization process at an early stage, before the fully blown allergy develops, may be of great value for occupationally exposed individuals. Predictiveness of an in vitro test for Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 tinuously exposed to latex devices starting at the earliest days of their life. Another high-risk factor is the genetic predisposition to type I allergy. The study of a randomly selected population of atopic individuals showed 6% prevalence of latex allergy in this group. In light of such a high and still-increasing prevalence of sensitivity to latex proteins, effective preventive measures and good diagnostic tests are urgently needed. The immediate prevention efforts were focused on the increasing awareness of healthcare providers and consumers and establishing measures to reduce the incidence of reactions in already sensitized individuals. The diagnostic procedures, generally performed as a follow-up to the clinical manifestation of an allergic reaction, include skin testing as a "gold standard" test and in vitro measurement of the total serum IgE and of the allergen-specific IgE antibodies. The specific applications of these approaches and procedures to the evaluation of latex allergy appear to be more cumbersome than anticipated. The most critical point is the definition of the latex allergen(s) in relation to all proteins in natural rubber latex. A defined allergen serves as a basis for either development of diagnostic tests or the manufacture of safe latex products. In addition, a representative anti-latex antiserum is equally important for a reliable evaluation of the potential allergenicity of latex products. The question of the specific allergen is complicated by the fact that latex proteins represent a group of about 2 5 30 proteins differing in their structure and size, with apparent molecular weights ranging from 5 to 200 kDa. Published reports show a significant diversity in the size of proteins identified as allergenic proteins, indicating that a large number of latex proteins may be potential allergens, depending on the dose and the route of exposure. Therefore, the best standard protein should contain all latex proteins that may be potentially allergenic. Sources of latex proteins include fresh unprocessed latex sap (NAL), raw latex containing ammonia (AL), and finished latex products (GL), where proteins were exposed to chemicals and high temperatures. Since the processing of latex may, on the one hand, result in the loss of some proteins or antigenic specificities and, on the other hand, expose or create epitopes that were not present in the native proteins, the question was which of these sources would be the closest to including all antigenic specificities to which users may be sensitized. Our studies compared immunoblots of NAL, AL, and GL protein extracts with a variety of human sera. These data demonstrated that sensitized individuals present various antibody specificities and respond differently to various sources of latex proteins. However, it was uniformly confirmed that all human sera evaluated reacted with the largest number of proteins in NAL extracts (Tomazic, 1995). IMMUNOTOXICITY OF MEDICAL DEVICES REFERENCES Abe, H., Rodgers, K. E., Ellefson, D., and diZerega, G. S. (1989). Kinetics of interleukin 1 secretion by murine macrophages recovered from the peritoneal cavity after surgery. J. Surg. Res. 47, 178-182. Abe, H., Rodgers, K. E., Ellefson, D., and diZerega, G. S. (1991). Kinetics of interleukin 1 and tumor necrosis factor secretion by rabbit macrophages recovered from the peritoneal cavity. J. Invest. Surg. 4, 141-151. Adhesion Barrier Study Group. (1989). Prevention of postsurgical adhesions by Interceed (TC7), an absorbable adhesion barrier A prospective, randomized multicenter clinical trial. Fertil. Steril. 51, 933-938. Adhesion Study Group (1983). Reduction of postoperative pelvic adhesions with intraperitoneal 32% dextran 70: A prospective, randomized clinical trial. Fertil. Steril. 40, 612-619. Agins, H. J., Alcock, N. W., Bansal, M., Salvati, E. A., Wilson, P. D., Pellicci, P. M., and Bullough, P. G. (1988). Metallic wear in failed titaniumalloy total hip replacements. J. Bone Joint Surg. A 70, 347-356. Ahmed, N., Falcone, T., Tulandi, T., and Houle, G. (1991). Anaphylactic reaction because of intrauterine 32% dextran-70 instillation. Fertil. Steril. 55, 1014-1016. Akasawa, A., Hsieh, L., and Lin, Y. (1995). Serum reactivities to latex proteins (Hevea brasiliensis). J. Allergy Clin. Immunol. 95, 1196-2005. Akatsu, T., Takahashi, N., Udagawa, N., Imamura, K., Yamaguchi, A., Sato, K., Nagata, N., and Suda, T. (1991). Role of prostaglandins in interleukin-1-induced bone resorption in mice in vitro. J. Bone Miner. Res. 6, 183-190. Allard, A., Bayliss, M. T., and Maini, R. N. (1990). The synovium-cartilage junction of the normal human knee. Arthritis Rheum. 33, 1170-1179. Amstutz, H. C , Campbell, P., Kossovsky, N., and Clarke, I. C. (1992). Mechanism and clinical significance of wear debris-induced osteolysis. Clin. Orthop. Relat. Res. 276, 7-18. Bailey, G., Strub, R., Klein, R. C , and Salvaggio, F. (1967). Dextraninduced anaphylaxis. JAMA 200, 185-189. Barrack, R. L., Mulroy, J., RD, and Harris, W. H. (1992). Improved cementing techniques and femoral component loosening in young patients with hip arthroplasty. A 12-year radiographic review. J. Bone Joint Surg. B 74, 385-389. Bernstein, J., Mattox, J., and Ulnch, J. (1982). The potential for bacterial growth with dextran. J. Reprod. Med. 27, 77-80. Borten, M., Seibert, C. P., and Taymor, M. L. (1983). Recurrent anaphylactic reaction to intraperitoneal dextran 75 used for prevention of postsurgical adhesions. Obstet. Gynecol. 61, 755-757. Boyce, B. F., Aufdemorte, T. B., Garrett, I. R., Yates, A. J. P., and Mundy, G. R. (1989). Effects of interleukin-1 on bone turnover in normal mice. Endocrinology 125, 1142-1150. Boyers, S. P., Diamond, M. P., and DeCherney, A. H. (1988). Reduction of postoperative pelvic adhesions in the rabbit with Gore-Tex surgical membrane. Fertil. Steril. 49, 1066-1070. Bradley, S. G., Munson, A. E., McCay, J. A., Brown, R. D., Musgrove, D. L., Wilson, S., Stern, M., Luster, M. I., and White, K. L., Jr. (1994a). Subchronic 10 day immunotoxicity of polydimethylsiloxane (silicone) fluid, gel and elastomer and polyurethane disks in female B6C3F1 mice. Drug Chem Toxicol. 17, 175-220. Bradley, S. G., White, K. L., Jr., McCay, J. A., Brown, R. D., Musgrove, D. L., Wilson, S., Stern, M., Luster, M. I., and Munson, A. E. (1994b). Immunotoxicity of 180 day exposure to polydimethylsiloxane (silicone) fluid, gel and elastomer and polyurethane disks in female B6C3F1 mice. Drug Chem. Toxicol. 17, 221-269. Butler, J. E., Lu, E. P., Navarro, P., and Christiansen, B. (1996). The adsorption on a polydimethylsiloxane elastomer (PEP) and their antigenic behavior. Curr. Top. Microbiol. Immunol. 210, 75—84. Campbell, P., Ma, S., Yeom, B., McKellop, H., Schmalzried, T. P., and Amstutz, H. C. (1995). Isolation of predominantly submicron-sized Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 the potential reaction in patients, however, can only be established by direct comparison with skin testing. An ongoing clinical trial, designed for that purpose, indicates that an in vitro test is very reliable for identifying sensitized individuals, but the quantitative correlation between the two tests is not always observed (Hamilton, 1996). Finally, for the prevention of further sensitization, it is necessary to reduce or eliminate all or the majority of allergens from latex products. The uncertainty associated with this approach is the definition of the latex allergen. Many investigators describe findings of proteins that represent the "major allergens" in latex, but their results vary markedly as to the number and molecular weight of allergenic proteins. The determining factors that make a specific protein allergenic (i.e., triggering production of IgE antibodies) are the route of exposure, the dose and the duration of the exposure to the protein, and the genetic profile of the individual. Protein should also be in the "allergenic" molecular weight range, which is from 10 to approximately 100 kDa. It is not known whether any other physicochemical properties of proteins besides molecular weight are decisive in the induction of allergy. If these facts are taken into consideration, it could be expected that any latex protein that meets the above requirements may be a potential allergen. Based on the data available at the present time, and knowing that the majority of latex proteins are in the allergenic molecular weight range, the level of allergenic proteins should correlate with the total amount of protein present on latex products. In other words, total level of extractable protein on the latex device should be an indirect indicator of its potential allergenicity. Although there are differences in opinions regarding the correlation between total amount of protein and allergenicity, it seems that recent clinical studies support this relationship. When latex-sensitive individuals were skin tested with extracts from NAL, AL, and GL, with dosing equalized on the basis of total protein content, the intensity of skin reactions was equal for all three extracts within the same dose and same individual (Hamilton, 1996). In summary, the question of a reference latex protein and reference immune serum are the crucial points in the development of diagnostic methods and prevention strategies for latex allergy. A large body of new data concerning latex proteins and the efforts to identify the specific allergens constitute a base for the development of diagnostic tests and tests for the potential allergenicity of finished medical devices. The availability of both methodologies will prevent allergic reactions and reduce or stop further sensitization of latex product users. 11 12 RODGERS ET AL. UHMWPE wear particles from periprosthetic tissues. J. Biomed. Mater. Res. 29, 127-131. intervention. In Cytokines and Bone Metabolism (M. Gowen, Ed.), pp. 383-406. CRC Press, Boca Raton, FL. Charnley, J., and Halley, D. K. (1975). Rate of wear in total hip replacement Clin. Orthop. Relat. Res. 112, 170-179. Goldring, S. R., Schiller, A. L., Roelke, M., Rourke, C. M., O'Neill, D. A., and Harris, W. H. (1983). The synovial-like membrane at the bone-cement interface in loose total hip replacements and its proposed role in bone lysis. J. Bone Joint Surg. A 65, 575-584. Cleary, R. E., Howard, T., and diZerega, G. S. (1985). Plasma dextran levels after abdominal instillation of 32% dextran 70: Evidence for prolonged intraperitoneal retention. Am. J. Obstet. Gynecol. 152, 78-79. DeChemey, A. H. (1987). A model for sidewall adhesions in the rabbit: Reduction by an absorbable barrier. Microsurgery 8, 197-200. DeChemey, A. H., and Mezer, H. C. (1984). The nature of post-tuboplasty pelvic adhesions determined by early and late laparoscopy. Fend. Steril. 41, 643-647. Diamond, M. P., Linsky, C. B., Cunningham, T., Constanune, B., diZerega, G. S., Dimitrijevich, S. D., Tatarko, M., Gracy, R. W., Wise, G. E., and Oakford, L. X. (1990). In vivo degradation of oxidized regenerated cellulose. Carbohydr. Res. 198, 331-341. Dinarello, C. A. (1991). Interleukin-1 and interleukin-1 antagonism. Blood 77(8), 1627-1652. diZerega, G., and Rodgers, K. E. (1992). Prevention of postoperative adhesions. In 77K; Peritoneum, pp. 307-370. Springer-Verlag, New York. Dunn, D. L., Barke, R. A., Ahrenholz, D. H., el al. (1984). The adjuvant effect of peritoneal fluid in experimental peritonitis. Ann. Surg. 199, 3 7 43. Goodman, S. B., Chin, R. C , Chiou, S. S., Schurman, D. J., Woolson, S. T., and Masada, M. P. (1989). A clinical-pathologic-biochemical study of the membrane surrounding loosened and nonloosened total hip arthroplasties. Clin. Orthop. 244, 182-187. Gowen, M. (1992). Interleukin 1 and tumor necrosis factor. In Cytokines and Bone Metabolism (M. Gowen, Ed.), pp. 71-91. CRC Press, Boca Raton, FL. Hamilton, R. G., and Adkinson, N. F. (1966). Natural rubber latex skin testing reagents: Safety and diagnostic accuracy of non-ammoniate, ammoniated and latex rubber glove extracts. J. Allergy Clin. Immunol. 97 (in press). Haney, A. F., and Doty, E. (1992). Comparison of the peritoneal cells elicited by oxidized regenerated cellulose (Interceed) and expanded polytetrafluoroethylene (Gore-Tex Surgical Membrane) in a murine model. Am. J. Obstet. Gynecol. 166, 1137-1149. Harvey, W. (1988a). Methodology. In Prostaglandins in Bone Resorption (W. Harvey and A. Bennett, Eds.), pp. 11-25. CRC Press, Boca Raton, FL. Frondoza, C. G., Sohrabi, A., and Hungerford, D. S. (1996a). Human chondrocytes proliferate and produce matrix components in microcarrier suspension culture. Biomaterials 17, 879-888. Harvey, W. (1988b). Inflammation, cytokines, and prostaglandins. In Prostaglandins in Bone Resorption (W. Harvey and A. Bennett, Eds.), pp. 57-72. CRC Press, Boca Raton, FL. Galante, J. O., Lemons, J., Spector, M., Wilson, P. D., Jr., and Wright, T. M. (1991). The biologic effects of implant materials. J. Orthop. Res. 9, 760-775. Haynes, D. R., Rogers, S. D., Hay, S., App, B., Pearcy, M. J., and Howie, W. (1993). The differences in toxicity and release of bone-resorbing mediators induced by titanium and cobalt-chromium-alloy wear particles. J. Bone Joint Surg. 75, 825-834. Giant, T. T., Jacobs, J. J., Molnar, G., Shanbhag, A. S., Valyon, M., and Galante, J. O. (1993). Bone resorption activity of paniculate-stimulated macrophages. J. Bone Miner. Res. 8, 1071-1079. Giant, T. T., Mikecz, K., Cole, A., Bittar, T., Lark, M. W., Sandell, L., Jacobs, J. J., and Kuettner, K. E. (1994a). In situ localization of metalloproteinases and TTMP in interfacial membranes of loosened total hip arthroplasties. Trans. Orthop. Res. Soc. 19, 138. Herman, J. H., Sowder, W. G., Anderson, D., Appel, A. M., and Hopson, C. N. (1989). Polymethylmethacrylate-induced release of bone-resorbing factors. J. Bone Joint Surg. A 71, 1530-1541. Hirsch, R. S., Smith, K., and Vernon-Roberts, B. (1985). A morphological study of macrophage and synovial cell interactions with hydroxyapatite crystals. Ann. Rheum. Dis. 44, 844-851. Giant, T. T., Valyon, M., Dellacqua, D., Walcz, E., Mikecz, K., Chubinskaja, S., Jacobs, J. J., Rosenberg, A., Gitelis, S., Sheinkop, M., and Galante, J. O. (1994b). Quantitative measurement of genes coding for "bone-resorbing" cytokines in periprosthetic inflammatory tissues of failed total hip arthroplasties. Trans. Orthop. Res. Soc. 19, 152. Horowitz, S. M., Gautsch, T. L., Frondoza, C. G., and Riley, L. (1991). Macrophage exposure to polymethyl methacrylate leads to mediator release and injury. J. Orthop. Res. 9, 406-413. Giant, T. T., Valyon, M., Mikecz, K., Cole, A., Jacobs, J. J., Yao, J., and Kuettner, K. E. (1994c). Discoordinate expression of IL-1 and IL-1 in interfacial membranes of failed joint prosthesis. Agents Actions 41, CI72-173. Howie, D. W., and Vernon-Roberts, B. (1988). The synovial response to intraarticular cobalt-chrome wear particles. Clin. Orthop. Relat. Res. 232, 244-254. Goldblum, R. M., Pelley, R. P., O'Donell, A. A., Pyron, D., and Heggars, J. P. (1992). Antibodies to silicone elastomers and reactions to ventriculoperitoneal shunts. Lancet 340, 510-513. Goldblum, R. M., Pyron, D., and Shenoy, M. (1995). Modulation of IgG binding to silicone by human serum albumin. FASEB J., A914. Goldring, S. R., Flannery, M. S., Petrison, K. K., Evins, A. E., and Jasty, M. J. (1990). Evaluation of connective tissue cell responses to orthopaedic implant materials. Connect. Tissue Res. 24, 7 7 - 8 1 . Goldring, S. R., and Goldring, M. B. (1992). Prospects for pharmacological Horowitz, S. M., and Purdon, M. A. (1995). Mediator interactions in macrophage/particulate bone resorption. J. Biomed. Mater. Res. 29, 477—484. Ishimi, Y., Miyaura, C , Jin, C. H., Akatsu, T., Abe, E., Nakamura, Y., Yamaguchi, A., Yoshiki, S., Matsuda, T., Hirano, T., Kishimoto, T., and Suda, T. (1990). IL-6 is produced by osteoblasts and induces bone resorption. J. Immunol. 145, 3297-3303. Jiranek, W. A., Machado, M., Murali, J., Jevsevar, D., Wolfe, H. J., Goldring, S. R., Goldberg, M. J., and Harris, W. H. (1993). Production of cytokines around loosened cemented acetabular components. /. Bone Joint Surg. A 75, 863-879. Kim, K. J., Rubash, H. E., Wilson, S. C , D'Antonio, J. A., and McClain, E. J. (1993). A histologic and biochemical comparison of the interface Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 Frondoza, C , Jones, L., Rose, N. R., Hatakeyama, A., Phelps, R., and Bona, C. (1996b). Development of scleroderma-like syndrome in Tsk/+ mice is not enhanced by silicone administration. Curr. Top. Microbiol. Immunol. 210, 299-306. Gonzalez, O., Smith, R. L., and Goodman, S. B. (1996). Effect of size, concentration, surface area, and volume of polymethylmemacrylate particles on human macrophages in vitro. /. Biomed. Mater. Res. 30, 4 6 3 473. IMMUNOTOX1CITY OF MEDICAL DEVICES 13 tissues in cementless and cemented hip prostheses. Clin. Orthop. Relax. Res. 287, 142-152. intraperitoneal adhesions and chronic pelvic pain. Br. J. Obstet. Gynaecol. 99, 59-67. Klykken, P. C , LeVier, R. R., and Mast, R. W. (1994). Immunological investigations of silicone gel in mice and rats. Toxicologist 14, 29. Quinn, J., Joyner, C , Triffitt, J. T., and Athanasou, N. A. (1992). Polymethylmethacrylate-induced inflammatory macrophages resorb bone. J. Bone Joint Surg. B 74, 652-658. Klykken, P. C , and White, K. L., Jr. (1996). The adjuvancy of sihcones: Dependency on compartmentalization. Curr. Top. Microbiol. Immunol. 210, 113-121. Kossovsky, N., Zeidler, M., Chun, G., Papasian, N., Nguyen, A., Rajguru, S., Stassi, J., Gelman, A., and Sponsier, E. (1993). Surface dependent antigens identified by high binding avidity of serum antibodies lha subpopulation of patients with breast prostheses. J. Appl. Biomater. 4, 281 288. Kozmn, S. C , Johanson, N. A., and Bullough, P. G. (1986). The biologic interface between bone and cementless femoral endoprostheses. J. Alhroplasty 1, 249-259. Krinsky, A. H., Haseltine, F. P., and DeCherney, A. (1984). Peritoneal fluid accumulation with dextran 70 instilled at time of laparoscopy. Fertil. Steril. 41, 647-653. Lee, J-M., Salvati, E. A., Berts, F., DiCarlo, E. F., Doty, S. B., and Bullough, P. G. (1992) Size of metallic and polyethylene debris in failed cemented total hip replacements. J. Bone Joint Surg. B 74, 380-384. Maguire, J. K., Coscia, M. F., and Lynch, M. H. (1987). Foreign body reaction to polymeric debris following total hip arthroplasty. Clin. Orthop. Relal. Res. 216, 213-223. Magyar, D. M., Hayes, M. F., Spirtos, N. J., Hull, M. E., and Moghissi, K. S. (1985). Is intraperitoneal dextran 70 safe for routine gynecologic use? Am. J. Obstet. Cynecol. 152, 198-204. Maloney, W. J., Jasty, M., Burke, D. W., O'Connor, D. O., Zalenski, E. B., Bragdon, C , and Hams, W. H. (1989). Biomechanical and histologic investigation of cemented total hip arthroplasties. Clin, Orthop. Relat Res. 249, 129-140. Manlapaz, M., Maloney, W. J., and Smith, R. L. (1996). In vitro activation of human fibroblasts by retrieved titanium alloy wear debris. J. Onhoped. Res. 14, 465-472. Nairn, J. O., Ippolito, K. M. L., Lanzafame, R. J., and van Oss, C. J. (1996). Induction of type II collagen arthritis in the DA rat using silicone gel as adjuvant. Curr. Top. Microbiol. Immunol. 210, 103-111. Nairn, J. O., Lanzafame, R. J., and van Oss, C. J. (1995). The effect of silicone gel on the immune response. J. Biomater. Sci. Polymer Ed. 7, 123-132. Nairn, J. O., Lanzafame, R. J., and van Oss, C. J. (1993a). The effect of silicone gel on antibody formation in rats. Immunol. Invest. 22, 1 5 1 161. Nairn, J. O., and van Oss, C. J. (1992). The effect of hydrophilicity-hydrophobicity and solubility on the immunogenicity of some natural and synthetic polymers. Immunol. Invest. 21, 649-662. Nairn, J. O., van Oss, C. J., and Lanzafame, R. J. (1993b). The induction of auto-antibodies to thyroglobulin in rats with silicone gel as adjuvant. Surg. Forum, XLIV, 676-678. Pazzaglia, U. E., Dell'Orbo, C , and Wilkinson, M. J. (1987). The foreign body reaction in total hip arthroplasties. A correlated light-microscopy, SEM, and TEM study. Arch. Orthop. Trauma Surg. 106, 209-219. Peters, A. A. W., Trimbos-Kemper, T. C. M., and Admiral, C. (1992). A randomized clinical trial on the benefit of adhesiolysis in patients with Rein, M.S., and Hill, J. A. (1989). 32% dextran 70 (Hyskon)* inhibits lymphocyte and macrophage function in vitro: A potential new mechanism for adhesion prevention. Fertil. Steril. 52, 953-957. Revell, P. A., Weightman, B., Freeman, M. A. R., and Roberts, B. V. (1978). The production and biology of polyethylene wear debris. Arch. Orthop. Trauma Surg. 91, 167-181. Rose, N. R., Landavere, M., and Kuppers, R. C. (1996). Silicone binding immunoglobulins in human sera. Curr. Top. Microbiol. Immunol. 210, 269-276. Salvati, E. A., Berts, F., and Doty, S. B. (1993). Paniculate metallic debris in cemented total hip Arthroplasty. Clin. Orthop. Relat. Res. 293, 160173. Santavirta, S., Konttinen, Y. T., Bergroth, V., Eskkola, A., Tallroth, K., and Lindholm, T. S. (1990). Aggressive granulomatous lesions associated with hip arthroplasty. J. Bone Joint Surg. 72, 252-258. Schindler, R., Mancilla, J., Endres, S., Ghorbani, R., Clarke, S. C , and Dinarello, C. A. (1990). Correlations and interactions in the production of interleukin-6 (IL-6), IL-1 and tumor necrosis factor (TNF) in human blood mononuclear cells: IL-1 suppresses IL-1 and TNF. Blood 75, 4 0 47. Schmiedberg, S. K., Chang, D. H., Frondoza, C. G., Valdevit, A. D. C , and Kostuik, J. P. (1994). Isolation and characterization of metallic wear debris from a dynamic intervertebral disc prosthesis. J. Biomed. Mater. Res. 28, 1277-1288. Sekiba, K., and the Obstetrics and Gynecology Adhesion Prevention Committee. (1992). Use of Interceed (TC7) absorbable adhesion barrier to reduce postoperative adhesions in infertility and endometriosis surgery. Obstet. Gynecol. 79, 518-522. Shanbhag, A. S., Jacobs, J. J., Black, J., Galante, J. O., and Giant, T. T. (1995). Cellular mediators secreted by interfacial membranes obtained at revision total hip arthroplasty. J. Arthroplasty 10, 498-506. Shanbhag, A. S., Jacobs, J. J., Giant, T. T., Gilbert, J. L., Black, J., and Galante, J. O. (1994). Composition and morphology of wear debris in failed uncemented total hip replacement arthroplasty. J. Bone Joint Surg. B 76, 60-67. Shaw Warren, H. (1992). Adjuvants. In Encyclopedia of Immunology (I. Roitt and P. Delves, Eds.), pp. 28-30. Academic Press, New York. Shimanuki, T., Nakamura, R. M., and diZerega, G. S. (1986). A kinetic analysis of peritoneal fluid cytology and arachidonic acid metabolism after abrasion and reabrasion of the rabbit peritoneum. J. Surg. Res. 41, 245-251. Stangel, J. J., Nisbet, J. D., and Settles, H. (1984). Formation and prevention of postoperative abdominal adhesions. J. Reprod. Med. 29, 143—156. Stout, A. L., Steege, J. F., Dodson, W. C , and Hughes, C. L. (1991). Relationship of laparoscopic findings to self-report of pelvic pain. Am. J. Obstet. Gynecol. 146, 73-77. Strickler, B., Blanco, J., and Fox, H. E. (1994). The gynecologic contribution to intestinal obstruction in females. J. Am. Coll. Surg. 178, 6 1 7 622. Surgical Membrane Study Group. (1992). Prophylaxis of pelvic sidewall Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 Lombardi, A. V., Jr., Mallory, T. H., Vaughn, B. K., and Drouillard, P. (1989). Aseptic loosening in total hip arthroplasty secondary to osteolysis induced by wear debris from titanium-alloy modular femoral heads. J. Bone Joint Surg. A 71, 1337-1342. Rae, T. (1979). Comparative laboratory studies on the production of soluble and paniculate metal by total joint prostheses. Arch. Orthop. Trauma Surg. 95, 71-79. 14 RODGERS ET AL. adhesions with Gore-Tex surgical membrane: A multicenter clinical investigation. Fertil. Steril. 57, 921-923. Thomas, G. A., Kilgus, B. J., Cracchiolo, D. J., and Allameh, V. (1991). Periprosthetic chronic inflammation characterized through the measurement of superoxide anion production by synovial-derived macrophages. Clin. Onhop. Relat. Res. 263, 263-271. Tomazic, V. J., Withrow, T. J., and Hamilton, R. G. (1995). Characterization of the allergen(s) in latex protein extracts. J. Allergy Clin. Immunol. 96, 635-642. Trimbos-Kemper, T. C. M., Trimbos, J. B., and van Hall, E. V. (1985). Adhesion formation after tubal surgery results of the eight-day laparoscopy in 188 patients. Fertil. Steril. 43, 396-401. Trimbos-Kemper, T. C , and Veering, B. T. (1989). Anaphylactic shock from intracavitary 32% Dextran-70 during hysterectomy. Fertil. Steril. 51, 1053-1054. Tulandi, T. (1987). Adhesion formation after reproductive surgery with and without the carbon dioxide laser. Feriil. Steril. 47, 704-706. Wilkinson, L. S., Pitsillides, A. A., Worrell, J. G., and Edwards, J. C. W. (1992). Light microscopic characterization of the fibroblast-like synovial intimal cell (synoviocyte). Arthritis Rheum. 35, 1179-1184. Willert, H. G., and Semlitsch, M. (1976). Tissue reactions to plastic and metallic wear products of joint endoprostheses. In Total Hip Prosthesis (N. Gschwend and H. U. Debrunner, Eds.), pp. 205-239. Williams & Wilkins, Baltimore. Willert, H. G., and Semlitsch, M. (1977). Reactions of the articular capsule to wear products of artificial joint prostheses. J. Biomed. Mater. Res. 11, 157-164. Wolf, L. E., Lappe, M., Peterson, R. D., and Ezailson, E. G. (1993). Human immune response to polydimethylsiloxane (silicone): Screening studies in a breast implant population. FASEB J. 7, 1265-1268. Yao, J., Giant, T. T., Lark, M. W., Mikecz, K., Jacobs, J. J., Hutchinson, N. I., Hoermer, L. A., Kuettner, K. E., and Galante, J. O. (1995). The potential role of fibroblasts in periprosthetic osteolysis. Fibroblast response to titanium particulates. J. Bone Miner. Res. 10, 1417-1427. Yoneda, T., and Mundy, G. R. (1978). Prostaglandins are necessary for osteoclast-activating factor production by activated peripheral blood leukocytes. J. Exp. Med. 149, 279-283. Downloaded from toxsci.oxfordjournals.org by guest on July 6, 2011 View publication stats