The potential role of fibroblasts in periprosthetic osteolysis: fibroblast response to titanium particles

Effect of polymer molecular weight and addition of calcium stearate on response of MG63 osteoblast-like cells to UHMWPE particles

Periprosthetic osteolysis and implant loosening is associated with the presence of ultrahigh molecular weight polyethylene (UHMWPE) wear debris particles. Osteoblast phenotypic expression in vitro is affected by UHMWPE particles, suggesting that bone formation may also be affected by wear debris. Here we tested the hypothesis that the response of osteoblasts to UHMWPE can be modified by changes in UHMWPE particle chemistry. We used four different commercially available preparations of GUR UHMWPE particles to determine if chemical composition (+/- Ca-stearate) or polymer molecular weight (3.1-4.2 million or 5.4-6.5 million g/mol) modulates osteoblast response. Particles were characterized by size distribution, morphology, and number of particles added to the culture medium. They had an average equivalent circle diameter ranging from 0.46-1.26 microm. MG63 cell response was assessed by measuring cell number, cellular and cell layer alkaline phosphatase, and prostaglandin E2 (PGE2) production. There were dose-dependent effects of the particles on cell response. Cell number and PGE, production were increased, while alkaline phosphatase specific activity was decreased. In addition, there was a marked difference between cultures treated with particles containing Ca-stearate and as a function of polymer molecular weight. Particles of higher molecular weight caused a greater stimulation of proliferation and inhibition of alkaline phosphatase than particles of lower molecular weight. The presence of Castearate exerted a more pronounced depression of osteoblast phenotype as well as a significantly greater increase in PGE2 release by the cells. The present study shows that chemical composition and polymer molecular weight of UHMWPE are capable of modulating osteoblast response to particles. The results suggest that osteoblast differentiation is inhibited by UHMWPE particles, whereas cell proliferation and PGE2 production are stimulated. This may have direct effects on osteoblasts and bone formation, but also paracrine effects on cells of the monocytic lineage inducing bone resorption and promoting inflammation which may lead to aseptic loosening. The present results suggest that the cellular events in aseptic loosening may be modulated or even accelerated by changes in the composition of the UHMWPE used to fabricate implants.

Alternate bearing surfaces in total joint arthroplasty: biologic considerations

The problem of periprosthetic osteolysis is currently the major limiting factor in joint arthroplasty longevity. Because this process has been shown to be primarily a biologic response to wear particles, corrosion products, or both, efforts to reduce particle generation are being undertaken. These efforts include the development of modified polyethylene and alternative articulating surfaces. These alternate bearing surfaces currently include ceramic-on-polyethylene, ceramic-on-ceramic, and metal-on-metal. Although these alternate bearings diminish or eliminate the generation of polyethylene particles, ceramic and metal particles are produced. The purpose of the current review is to discuss the literature that addresses the biologic response to these particles, locally and systemically.

G-protein activity requirement for polymethylmethacrylate and titanium particle-induced fibroblast interleukin-6 and monocyte chemoattractant protein-1 release in vitro

Periprosthetic granulomatous membranes consisting of fibroblasts, macrophages, lymphocytes, foreign body giant cells, and abundant particulate debris occur at sites of implant loosening. Previous studies demonstrate that fibroblasts respond to particulate debris through the release of interleukin-6 (IL-6), prostaglandin E(2), and matrix metalloproteinases in vitro. C-C chemokines are observed in granulomatous tissue surrounding loosened prosthetic implants and are released by macrophages and fibroblasts in response to particle challenge in vitro. This study tested the hypothesis that G protein activity is required for fibroblast activation by titanium and polymethylmethacrylate (PMMA) particles, and that inhibition of G protein activity would alter IL-6 and and monocyte chemoattractant protein-1 (MCP-1) release from activated fibroblasts. The specific inhibitor of G protein activity, pertussis toxin, was added to the fibroblasts to examine the effects of G protein activity with respect to the production of IL-6 and MCP-1 by orthopedic biomaterial-challenged fibroblasts in vitro. Interleukin-1beta (IL-1beta), a proven activator of MCP-1 and interleukin-6, was used as a positive control. Exposure of fibroblasts to titanium and polymethylmethacrylate (PMMA) particles resulted in a dose-dependent release of MCP-1 and IL-6. Challenge with PMMA particles at doses of 0.150%, 0.300%, and 0.600% vol/vol increased the release of interleukin-6 by 7-, 19-, and 22-fold, respectively, compared to fibroblasts exposed to serum-free culture medium alone at 24 h. Challenge with PMMA particles at doses of 0.075%, 0.150%, 0.300%, and 0.600% vol/vol increased the release of MCP-1-6 by 2.5-, 3.6-, 4. 3-, and 4.5-fold, respectively, compared to fibroblasts exposed to serum-free culture medium alone. Challenge with titanium particles at concentrations of 0.075%, 0.150%, 0.300%, and 0.600% vol/vol increased the release of interleukin-6 by 2.6-, 6.4-, 9.6-, and 10. 0-fold, respectively, compared to fibroblasts exposed to serum-free culture medium alone at 24 h. Challenge with titanium particles at concentrations of 0.038%, 0.075%, 0.150%, 0.300%, and 0.600% vol/vol increased the release of MCP-1 by 2.9-, 3.1-, 5.8-, 5.4-, and 5. 8-fold, respectively, compared to fibroblasts exposed to serum-free culture medium alone. Pretreatment of fibroblasts with pertussis toxin inhibited the release of interleukin-6 and MCP-1 from PMMA and titanium particle challenged fibroblasts in a dose-dependent manner. PMMA particle induced fibroblast IL-6 release was inhibited by 23.6% and 35.3% with 20- and 200-ng/mL doses of pertussis toxin, respectively. Titanium particle induced fibroblast IL-6 release was inhibited by 48.2% and 56.3% with 20- and 200-ng/mL doses of pertussis toxin, respectively. PMMA particle-induced fibroblast MCP-1 release was inhibited by 36.0%, 50.4%, and 60.1% with 2-, 20- and 200-ng/mL doses of pertussis toxin, respectively. Titanium particle-induced fibroblast MCP-1 release was inhibited by 15.5%, 53.2%, and 64.6% with 2-, 20-, and 200-ng/mL doses of pertussis toxin, respectively. This study suggests that fibroblasts localized in periprosthetic membranes are a source of macrophage chemoattractant factors and proinflammatory mediators that may influence granuloma formation and lead to periprosthetic bone resorption. Furthermore, this study shows that G proteins are involved in particle-induced fibroblast activation, as evidenced by decrease levels of particle induced IL-6 and MCP-1 release following pertussis toxin treatment. (c) 2000 John Wiley & Sons, Inc.

Particulate wear debris activates protein tyrosine kinases and nuclear factor kappaB, which down-regulates type I collagen synthesis in human osteoblasts

Particulate wear debris generated mechanically from prosthetic materials is phagocytosed by a variety of cell types within the periprosthetic space including osteoblasts, which cells with an altered function may contribute to periprosthetic osteolysis. Exposure of osteoblast-like osteosarcoma cells or bone marrow-derived primary osteoblasts to either metallic or polymeric particles of phagocytosable sizes resulted in a marked decrease in the steady-state messenger RNA (mRNA) levels of procollagen alpha1[I] and procollagen alpha1[III]. In contrast, no significant effect was observed for the osteoblast-specific genes, such as osteonectin and osteocalcin (OC). In kinetic studies, particles once phagocytosed, maintained a significant suppressive effect on collagen gene expression and type I collagen synthesis for up to five passages. Large particles of a size that cannot be phagocytosed also down-regulated collagen gene expression suggesting that an initial contact between cells and particles can generate gene responsive signals independently of the phagocytosis process. Concerning such signaling, titanium particles rapidly increased protein tyrosine phosphorylation and nuclear transcription factor kappaB (NF-kappaB) binding activity before the phagocytosis of particles. Protein tyrosine kinase (PTK) inhibitors such as genistein and the NF-kappaB inhibitor pyrrolidine dithiocarbamate (PDTC) significantly reduced the suppressive effect of titanium on collagen gene expression suggesting particles suppress collagen gene expression through the NF-kappaB signaling pathway. These results provide a mechanism by which particulate wear debris can antagonize the transcription of the procollagen alpha1[I] gene in osteoblasts, which may contribute to reduced bone formation and progressive periprosthetic osteolysis.

The relative influence of the topography and chemistry of TiAl6V4 surfaces on osteoblastic cell behaviour

Proliferation and adhesion of mouse (MC3T3-E1) osteoblastic cells and primary human osteoblastic cells were carried out on Ti6Al4V titanium alloy samples with varied surface roughnesses. Mechanically or manually polished surfaces were prepared to produce respectively non-oriented or oriented residual polishing grooves. Sand-blasted surfaces were prepared using 500 microm or 3 mm alumina particles. Surface roughness parameters showed a negative correlation in comparison to proliferation and adhesion parameters. X-ray microprobe chemical surface microanalysis showed complete disturbance of the surface element composition of the Ti6Al4V alloy following sand-blasting treatment. An AlOx-enriched layer was observed on sample surfaces. This may lead to the suspicion that the concomittant effect of surface roughness amplitude and AlOx surface concentration has an effect on osteoblastic cell proliferation and adhesion. These findings show the significance of chemical surface analysis after any surface treatment of titanium-based implants before any biological use.

In vitro and in vivo activation of polymorphonuclear leukocytes in response to particulate debris

The host inflammatory response to particulate wear debris has been implicated as a principal cause of osteolysis and aseptic loosening following total joint arthroplasty. While it has long been assumed that this inflammatory response is mediated solely by a chronic process, there has been evidence to suggest that an acute response to particulate debris may be important in initiating the chronic response. We studied the in vitro and in vivo acute inflammatory responses mediated by polymorphonuclear leukocytes (PMNs) to both retrieved particulate from a catastrophically failed uncemented metal-backed acetabular component and to commercially pure particulate (polyethylene, cobalt-chrome, and titanium). Isolated, nonactivated human PMNs in vitro exhibited both a dose- and time-dependent degranulation response to opsonized particulate debris, as evidenced by release of both specific (increased lysozyme activity) and azurophilic (increased beta-glucuronidase activity) granule contents. In the rat subcutaneous pouch model in vivo, PMNs were recruited within 3-6 h after exposure to particulate debris and were noted to phagocytize particulate and subsequently degranulate, as evidenced by increased beta-glucuronidase and PMN-specific myeloperoxidase (azurophilic granule enzymes) activities. This response peaked within the first 6 h and gradually declined by 24 h. The results of this study demonstrate the presence of an acute inflammatory response mediated by PMNs both in vitro and in vivo to particulate debris, which may be important in the sequence of events that lead to the macrophage-dominated chronic inflammatory process culminating in osteolysis and aseptic loosening of total joint arthroplasties.

Phagocytosis of wear debris by osteoblasts affects differentiation and local factor production in a manner dependent on particle composition

Wear debris is considered to be one of the main factors responsible for aseptic loosening of orthopaedic endoprostheses. Whereas the response of cells in the monocytic lineage to foreign materials has been extensively studied, little is known about cells at the bone formation site. In the present study, we examined the hypothesis that the response of osteoblasts to wear debris depends on the chemical composition of the particles. We produced particles from commercially pure titanium (cpTi), Ti-6Al-4V (Ti-A), and cobalt-chrome (CoCr) and obtained ultrahigh molecular weight polyethylene (UHMWPE; GUR 4150) particles from a commercial source. The equivalent circle diameters of the particles were comparable: 1.0 +/- 0.96 microm for UHMWPE; 0.84 +/- 0.12 microm for cpTi; 1.35 +/- 0.09 microm for Ti-A, and 1.21 +/- 0.16 microm for CoCr. Confluent primary human osteoblasts and MG63 osteoblast-like cells were incubated in the presence of particles for 24 h. Harvested cultures were examined by transmission electron microscopy to determine if the cells had phagocytosed the particles. Particles were found intracellularly, primarily in the cytosol, in both the primary osteoblasts and MG63 cells. The chemical composition of the particles inside the cells was confirmed by energy-dispersive X-ray analysis. Morphologically, both cell types had extensive ruffled cell membranes, less-developed endoplasmic reticulum, swollen mitochondria, and vacuolic inclusions compared with untreated cells. CpTi, Ti-A, and CoCr particles were also added to cultures of MG63 cells to assess their effect on proliferation (cell number) and differentiation (alkaline phosphatase activity), and PGE2 production. All three types of particles had effects on the cells. The effect on cell number was dependent on the chemical composition of the particles; Ti-A and CoCr caused a dose-dependent increase, while cpTi particles had a biphasic effect with a maximal increase in cell number observed at the 1:10 dilution. Alkaline phosphatase specific activity was also affected and cpTi was more inhibitory than Ti-A or CoCr. PGE2 production was increased by all particles, but the magnitude of the effect was particle-dependent: CoCr > cpTi > Ti-A. This study demonstrates clearly that human osteoblast-like cells and MG63 cells can phagocytose small UHMWPE, CoCr, Ti-A, and cpTi particles. Phagocytosis of the particles is correlated with changes in morphology, and analysis of MG63 response shows that cell proliferation, differentiation, and prostanoid production are affected. This may have negative effects on bone formation adjacent to an orthopaedic implant and may initiate or contribute to the cellular events that cause aseptic loosening by inhibiting bone formation. The effects on alkaline phosphatase and PGE2 release are dependent on the chemical composition of the particles, suggesting that both the type and concentration of wear debris at an implant site may be important in determining clinical outcome.

In vitro response of human fibroblasts to commercially pure titanium

The generation of metal particles through surface wear of prosthetic joints has been associated with biological reactions that may lead to prosthetic component loosening. The role of the macrophage in these reactions has been studied extensively, but that of the fibroblast has not. The few fibroblast studies that there have been have shown that particles of several metals, with sizes over a wide range, can promote cytokine release and may cause cell necrosis. The intent of this study was to determine if there are metal particle exposure threshold levels that result in morphological changes and cell necrosis of fibroblasts in peri-articular tissues. Retrieved human fibroblasts (superior medial plica) were cultured in standard fashion and then were exposed to various particle dosages of commercially pure Titanium (cpTi). Cell morphological changes and necrosis were observed to occur when the total mass of the particle dosage exceeded a threshold level. These data imply that these cell responses occur at threshold levels of wear particle exposure.

Interleukin-11 (IL-11) in aseptic loosening of total hip replacement (THR)

The chronic inflammatory response to abrasion particles from total hip replacement (THR) is believed to cause osteolysis and to contribute to prosthetic loosening. The expression of interleukin-11(IL-11) and its major cellular sources in the interface and pseudocapsular tissues obtained from total hip revisions performed for aseptic loosening were investigated. The avidin-biotin-peroxidase complex (ABC) and alkaline phosphatase-anti-alkaline phosphatase (APAAP) methods were used for staining and VIDAS image analysis for quantification. IL-11 was found in the interface and pseudocapsular tissues in the aseptic loosening of THR. IL-11 containing cells were more numerous in the interface (760 +/- 171 cells) and pseudocapsular tissues (684 +/- 171 cells) than in the control synovial tissue (235 +/- 68 cells). Because IL-11 is an important component of cytokine network mediating osteoblast-osteoclast communication in normal and pathological bone remodeling, the current findings suggest that IL-11 may contribute to periprosthetic osteolysis and to the loosening of THR.

Hyperresponsiveness of vitamin D receptor gene expression to 1,25-dihydroxyvitamin D3. A new characteristic of genetic hypercalciuric stone-forming rats

Hypercalciuria in genetic hypercalciuric stone-forming (GHS) rats is accompanied by intestinal Ca hyperabsorption with normal serum 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] levels, elevation of intestinal, kidney, and bone vitamin D receptor (VDR) content, and greater 1,25(OH)2D3-induced bone resorption in vitro. To test the hypothesis that hyperresponsiveness of VDR gene expression to 1,25(OH)2D3 may mediate these observations, male GHS and wild-type Sprague- Dawley normocalciuric control rats were fed a normal Ca diet (0.6% Ca) and received a single intraperitoneal injection of either 1,25(OH)2D3 (10-200 ng/100 g body wt) or vehicle. Total RNAs were isolated from both duodenum and kidney cortex, and the VDR and calbindin mRNA levels were determined by Northern blot hybridization using specific cDNA probes. Under basal conditions, VDR mRNA levels in GHS rats were lower in duodenum and higher in kidney compared with wild-type controls. Administration of 1,25(OH)2D3 increased VDR gene expression significantly in GHS but not normocalciuric animals, in a time- and dose-dependent manner. In vivo half-life of VDR mRNA was similar in GHS and control rats in both duodenum and kidney, and was prolonged significantly (from 4-5 to > 8 h) by 1,25(OH)2D3 administration. Neither inhibition of gene transcription by actinomycin D nor inhibition of de novo protein synthesis with cycloheximide blocked the upregulation of VDR gene expression stimulated by 1,25(OH)2D3 administration. No alteration or mutation was detected in the sequence of duodenal VDR mRNA from GHS rats compared with wild-type animals. Furthermore, 1,25(OH)2D3 administration also led to an increase in duodenal and renal calbindin mRNA levels in GHS rats, whereas they were either suppressed or unchanged in wild-type animals. The results suggest that GHS rats hyperrespond to minimal doses of 1,25(OH)2D3 by an upregulation of VDR gene expression. This hyperresponsiveness of GHS rats to 1,25(OH)2D3 (a) occurs through an increase in VDR mRNA stability without involving alteration in gene transcription, de novo protein synthesis, or mRNA sequence; and (b) is likely of functional significance, and affects VDR-responsive genes in 1, 25(OH)2D3 target tissues. This unique characteristic suggests that GHS rats may be susceptible to minimal fluctuations in serum 1, 25(OH)2D3, resulting in increased VDR and VDR-responsive events, which in turn may pathologically amplify the actions of 1,25(OH)2D3 on Ca metabolism that thus contribute to the hypercalciuria and stone formation.

Histologic and biochemical differences between osteolytic and nonosteolytic membranes around femoral components of an uncemented total hip arthroplasty

The unique design characteristics of an uncemented femoral component were used to study differences in the periprosthetic membranes in the presence and absence of osteolysis. A component (Omniflex, Osteonics, Allendale, NJ) that has a large midportion of the stem that does not contact bone was studied. A membrane forms, even in the absence of bone resorption, in Gruen zones 3 and 5 of this component. Analysis of this membrane showed noninflammatory fibrous tissue, no or minimal particulate debris, and few macrophages. In contrast, analysis of membranes from osteolytic regions around the same prosthesis demonstrated typical inflammatory characteristics. Collagenase, gelatinase, and stromelysin expression was high in osteolytic membranes but was low in tissues from noninflammatory regions without osteolysis. The data suggest a sequence of events in periprosthetic membrane formation. A noninflammatory membrane initially forms between the bone and the prosthesis. This membrane is transformed into an inflammatory membrane by the influx of particulate debris. Matrix metalloproteinases are selectively expressed in regions of osteolysis, implicating these enzymes in periprosthetic bone loss and suggesting a possible level for pharmaceutical intervention to prevent or treat osteolysis. Formation of the noninflammatory membrane around the distal part of the prosthesis of this or similar designs in the absence of bone resorption implies that these tissue samples could serve as useful negative control tissues to identify factors present in osteolytic periprosthetic membranes.

16- to 25-year follow-up study of cemented arthroplasty of the hip in patients aged 50 years or younger

The clinical outcome of the cemented Charnley hip arthroplasty at follow-up periods of 16-25 years was evaluated retrospectively in patients aged 50 years or younger. One hundred thirty-two of a total of 167 hip arthroplasties were studied. The overall probability of survival of the implant at 20 years was 75% Survival of those with rheumatoid disease was 80% compared with 64% for those with osteoarthritis. Female sex was associated with a better prognosis. Accelerated wear was associated with decreased survival of the prosthesis. Varus orientation of the femoral component significantly influenced failure (P < .01). Radiographic loosening of the acetabular component was well tolerated, but loosening of the femoral component was significantly associated with pain (P = .01).

Changes in messenger RNA and protein levels of proteoglycans and link protein in human osteoarthritic cartilage samples

OBJECTIVE

To determine the steady-state messenger RNA (mRNA) levels and corresponding protein contents of major matrix components in osteoarthritic (OA) cartilage.

METHODS

Steady-state levels of gene-specific mRNA (relative to GAPDH) were measured by quantitative polymerase chain reaction (PCR), and the relative levels of the corresponding proteins were determined by Western blotting.

RESULTS

All mRNA levels and corresponding protein contents of aggrecan and versican (hyaluronan-binding large proteoglycans), decorin, biglycan, fibromodulin, and lumican (small proteoglycans), and link protein were higher in OA cartilage samples than in age-matched normal samples. The ratio of increase, however, was different for each component. The mRNA and protein levels of biglycan, decorin, and fibromodulin increased synchronously, whereas message for link protein and lumican were several-fold higher than expected by their measured protein contents. Versican was also detected in OA cartilage; however, the versican protein content was associated with a relatively low mRNA level.

CONCLUSION

The expression of matrix components was increased in chondrocytes of OA cartilage, especially the expression of small proteoglycans, most likely due to the repair processes. A discoordinate gene expression accompanied with imbalanced accumulation of noncollagenous matrix components may contribute to the disorganization of the cartilage and the development of OA processes.

Human monocyte/macrophage response to cobalt-chromium corrosion products and titanium particles in patients with total joint replacements

The responses of human peripheral blood monocytes of 10 normal volunteers and 14 patients with total hip replacements to particles of commercially pure titanium and chromium orthophosphate (a corrosion product from cobalt-chromium alloy implants) were studied. In addition, these phagocytosable particles were added to cultured mononuclear cells isolated from the interfacial membrane of 14 patients with failed implants. Peripheral blood monocytes from patients who had had a total hip replacement produced significantly higher levels of interleukin-1 (both interleukin-1 alpha and interleukin-1 beta) and prostaglandin E2 following particulate stimulation than those from normal volunteers. Supernatants from both titanium and chromium orthophosphate-stimulated peripheral blood monocytes from the volunteers and patients with total hip replacement induced bone resorption (assayed in organ cultures of newborn mouse calvariae) and the proliferation of human fibroblasts. The levels of bone resorption were significantly higher (p < 0.05) in patients with implants than in normal volunteers. There were no significant differences in the responses of cells between patients with focal osteolysis and those without osteolysis. Interfacial membrane mononuclear cells also produced high levels of interleukin-1 alpha, interleukin-1 beta, and prostaglandin E2 and expressed bone resorptive activities following stimulation with either titanium or chromium orthophosphate. More importantly, interfacial membrane mononuclear cells "spontaneously" produced high levels of prostaglandin E2 that were comparable with the response of peripheral blood monocytes stimulated with particulate wear debris. The clinical relevance of this study is 2-fold. First, mononuclear cells from patients with total hip replacement were some-how "sensitized" to metal particles in comparison with mononuclear cells from individuals without an implant. Second, the chromium orthophosphate corrosion product was a potent macrophage/monocyte activator and may contribute to macrophage-mediated osteolysis and aseptic loosening.

Suppression of osteoblast function by titanium particles

In order to understand the effect of particulate debris on osteoblast function, we studied the effect of different particles, including titanium and polystyrene, on bone collagen mRNA (messenger RNA) with the use of Northern blot hybridization analysis, and we studied the effect of the particles on the biosynthesis of bone collagen with analysis of 3H-proline incorporation and with the Western blot technique. The steady-state levels of mRNA for procollagens alpha1(I) and alpha1(III) were markedly suppressed in human MG-63 osteoblast-like cells exposed to phagocytosable titanium particles that were smaller than three micrometers. Both titanium and polystyrene particles smaller than three micrometers suppressed the expression of the gene that codes for collagen, and the suppression of the expression of the gene was related to the size but not to the composition of the particles. The biosynthesis of both type-I and type-III collagen also was decreased in cells that had been treated with titanium particles. Neither the viability nor the proliferation of cells was affected by particulate debris. These data indicate that phagocytosable titanium particles can significantly suppress the expression of the gene that codes for collagen in osteoblast-like cells (p < 0.05).

In vitro activation of human fibroblasts by retrieved titanium alloy wear debris

Titanium-aluminum-vanadium wear particles isolated from the soft-issue membrane of a failed total hip arthroplasty were added to human fibroblasts in cell culture. The cellular response to particle challenge was determined by assaying for levels of interleukin-1 beta, interleukin-6, tumor necrosis factor-alpha, prostaglandin E2, basic fibroblast growth factor, platelet-derived growth factor-AB, and transforming growth factor-beta. Collagenase and gelatinase activities were analyzed by zymography and [3H]collagen degradation. Cell viability was assessed by measuring the uptake of [3H]thymidine. Over the range of particle concentrations tested, cell viability, as demonstrated by [3H]thymidine uptake, remained unaffected. Fibroblasts exhibited a dose-dependent release of interleukin-6 in response to exposure to titanium-aluminum-vanadium particles. At 6 and 48 hours, the highest concentration of titanium alloy particles (0.189% [vol/vol]) resulted in 7-fold and 16-fold increases in interleukin-6 release, respectively, when compared with negative controls. Neither interleukin-1 beta nor tumor necrosis factor-alpha was detected in the culture medium at any particle concentration tested for both dermal and foreskin fibroblasts. The pattern of prostaglandin E2 release by fibroblasts mirrored the pattern of interleukin-6 release. Fibroblasts exposed to the highest concentration of titanium alloy particles showed an increase in collagenase activity, starting at 12 hours. When medium samples were treated with amino phenylmercuric acetate to activate latent enzymes, a statistically significant increase in collagenase activity was observed as early as 6 hours (p < 0.001). Substrate gel analysis of medium from fibroblasts stimulated by high particle concentrations also showed an increase in gelatinolytic activity when compared with unstimulated controls. Analysis of medium samples for growth factors showed an increase in basic fibroblast growth factor at low particle concentrations, beginning at 12 hours. Levels of platelet-derived growth factor-AB and transforming growth factor-beta were not detectable in the controls or at any particle concentration tested. The results of this study showed that fibroblasts exposed to titanium alloy wear particles become activated and release proinflammatory mediators that influence bone metabolism. These data support the hypothesis that direct activation of fibroblasts by particulate wear may play a role in particle-mediated osteolysis. Fibroblast activation coupled with the biologic response of macrophages to wear debris in the loosening membrane may have a synergistic effect on pathologic bone resorption.