A histologic and biochemical comparison of the interface tissues in cementless and cemented hip prostheses

Thirty-four membranes from 20 cementless and 14 cemented prostheses from 33 patients excised during revision arthroplasty were analyzed for foreign-body reactive processes. As a control, identical analyses were performed using adjacent pseudocapsular tissue taken from 12 patients during revision arthroplasty. Histologically, cementless membranes tended to contain more metal debris, and cemented membranes contained more foreign-body giant cells. Collagenase, gelatinase, prostaglandin E2, and interleukin-1 were measured in conditioned media taken from organ cultures of each interface membrane. Both cementless and cemented membranes released significant levels of gelatinase and collagenase, prostaglandin E2, and interleukin-1 into the conditioned media when compared with control tissues. However, the tissue concentrations were not significantly different between cementless and cemented membranes. Although histologic differences between cementless and cemented interface membranes exist, both share a similar capacity to release bioactive products that can induce bone resorption and cause aseptic loosening of the hip prostheses.     

Results of an operative policy in the treatment of periprosthetic femoral fracture

OBJECTIVE: To determine the clinical outcome of patients with periprosthetic femoral fractures treated operatively. DESIGN: Retrospective analysis from 1986 to 1993. SETTING: Edinburgh Orthopaedic Trauma Unit, Edinburgh, Scotland. PATIENTS: Forty-five patients identified from a computer database as being admitted to the Edinburgh Orthopaedic Trauma Unit with periprosthetic femoral fractures. MAIN OUTCOME MEASURES: Clinical outcome grade (good, fair, poor) dependent on integrity of fixation, refracture rate, and ability to perform activities of daily living analyzed against age, type of fracture, prosthetic alignment, loosening, and method of fixation. RESULTS: Type I fractures were more common in uncemented or loosely cemented prostheses, whereas type II fractures occurred predominantly in securely cemented prostheses. Type I fractures treated by revision had the poorest results. Outcome in type II fractures was equally good whether treated by internal fixation or by revision. Age, loosening, and prosthetic alignment did not influence outcome. The mortality rate in this series was 20%. CONCLUSIONS: If a prosthesis is loose, it should be revised, or treatment varies with fracture and prosthetic type. In type I fractures, an uncemented stem may be revised to a cemented one; however, a securely cemented prosthesis probably is better when fixed internally. Type II fractures should be fixed internally because there is less operative insult. Type III fractures probably are not related to the prosthesis and should be fixed internally according to normal practice. The results of an operative policy compare well with the results of conservative management and avoid the problems of long-term immobilization.     

Activation of periprosthetic connective tissue in aseptic loosening of total hip replacements

In aseptic loosening of initially well inserted total hip prostheses, implant wear debris and cyclic mechanical loading lead to a foreign body type of chronic inflammatory reaction, then to osteolysis, and finally to loosening of the implant. In the present work the reactive and adaptive changes of the periprosthetic tissues and pseudojoint were characterized by analysis of the local cell proliferation. Immunohistochemical demonstration of proliferating cells was performed by application of affinity purified rabbit antihuman Ki-67 antibodies to periprosthetic tissues obtained from revision operations for loose total hip prostheses. The fibrous areas and, in particular, the cell rich, vascular areas of the interface tissue (between implant and bone) and the pseudocapsule around aseptically loosened implants contained higher numbers of proliferating cells than the tissues around well fixed implants. In addition, the pseudosynovial lining occasionally contained some Ki-67 positive proliferating cells. Somewhat surprisingly, proliferating vascular endothelial cells were relatively rare. These findings suggest that reactive (interface tissues) and adaptive (pseudojoint and capsule formed around the artificial joint) tissue changes in loosening total hip prostheses comprise proliferation of local fibroblastlike cells. It is concluded that periprosthetic tissues of the loosened total hip prosthesis represent activated mesenchymal tissue.     

Somatostatin expression in TS16 mouse brain cultures

Regional periprosthetic bone resorption plays an important role of prosthesis loosening. In order to study the possible mechanisms of loosening, we investigated the presence of matrix proteolytic enzymes in the periprosthetic tissue by immunohistochemical technique in 72 patients undergoing revision operation of loosened joint prosthesis, including 22 males and 50 females and aged from 19 to 88 years (mean, 61.7 years). Thirty-nine patients had a loosened hip prosthesis (18 males and 21 females) whereas 33 patients had a loosened knee prosthesis (4 males and 29 females). Tissue specimens collected during revision surgery underwent thin slide sections and H & E staining, and were observed under light microscopy and polarized-light microscopy. The results showed many macrophages, histiocytes, fibroblasts, as well as many phagocytosed metal debris and polyethylene debris in the periprosthetic tissues, suggesting an active bone resorption. Furthermore, we used immunohistochemical techniques to detect the distribution of matrix proteolytic enzymes in periprosthetic tissue, including lysosome enzymes (cathepsin B, cathepsin D and cathepsin G), and matrix metalloproteinase (MMPs, MMP-1, MMP-2, MMP-3). The immunostaining were classified as strong positivity, > 70% positive cells; moderate positivity, 20-70% positive cells; weak/negative, < 20% positive cells. The results showed that cathepsin B, cathepsin D and cathepsin G were found in most fibroblasts and macrophage-like cells, including multinuclear giant cells and epithelioid cells. MMPs were found in most fibroblasts and macrophage-like cells, as well as a scant amount in the extracellular matrix. These enzymes were also found in or around blood vessels, the endothelial cells in the richly vascularized tissue. All negative controls showed no staining. The results of immunoreactive staining ranged from 61.1% to 68.1% of strong to moderate positivity. Since these enzymes were related to the degradation of matrix protein, they may be related to the periprosthetic bone resorption. The further clinical significance needs further investigation.     

Effects of design changes on cemented tapered femoral stem fixation

The effects of matte finish and modularity on loosening of tapered stems using the same cementing technique were studied prospectively. In 80 patients, 82 cemented Exeter primary stems were implanted at total hip revision by one surgeon using the same surgical and cementing technique throughout the series. The polished stems behaved differently than the matte surfaced stems behaved. Polished stems subsided in the cement mantle an average of 1 mm at 2 years after implantation, but without subsequent loosening of stems at as long as 12 years after implantation. Matte surfaced stems with metal centralizers had a higher loosening rate, and loss of fixation at the prosthesis to cement interface was identified as an early sign of loosening of these stems. At a mean 6-year followup, there were no revisions nor was there radiographic evidence of loosening of the polished modular stems. It is concluded that matte finish results in increased loosening of tapered stems but the introduction of modularity did not.     

Expression of gelatinase A and its activator MT1-MMP in the inflammatory periprosthetic response to polyethylene

Wear debris of polyethylene prosthetic components is known to induce a host granulomatous reaction which recruits numerous macrophages and multinucleated giant cells. By releasing cellular mediators of a nonspecific inflammatory reaction, activated phagocytic cells are thought to play a key role in osteolysis leading to aseptic loosening of the prosthesis. Matrix metalloproteinases (MMPs) have been implicated in this destructive process by their ability to degrade extracellular matrix components of bone and adjacent connective tissue. To investigate the roles of gelatinase A, its activator MT1-MMP, and the MMP inhibitors TIMP-1 and TIMP-2 in aseptic loosening of polyethylene prostheses, immunohistochemistry (IHC) and in situ hybridization (ISH) were performed on periprosthetic pseudosynovial interface tissues. Gelatinase A and MT1-MMP were strongly detected immunohistochemically in macrophages and multinucleated giant cells in contact with polyethylene wear debris. In contrast to MT1-MMP, gelatinase A mRNAs were not found in phagocytic cells but in surrounding fibroblasts, thereby suggesting cooperation between macrophages and fibroblasts in this process. While TIMP-1 was expressed essentially in hyperplastic pseudosynoviocytes as assessed by IHC and ISH, TIMP-2, MT1-MMP, and gelatinase A were colocalized in phagocytic cells. These data support the concept of progelatinase A activation involving a trimolecular complex (MT1-MMP-TIMP-2-gelatinase A) mechanism. Thus, this study demonstrated that gelatinase A and its activator might contribute to the aseptic loosening of polyethylene prostheses.     

Revision of failed cementless total knee implants with cement

Between 1986 and 1996, 268 revision total knee replacements were performed. Revisions for aseptic loosening were reviewed prospectively. Sixty-three patients had implants revised for aseptic loosening. Thirty-six of the patients had revisions of cemented constructs, whereas 27 of the patients had revisions of cementless constructs. The average time before revision for the cemented group was 86 months. The average time to revision for the cementless group was only 43 months. Fifty-two percent of the patients in the cementless group had revision of their implants within 2 years of their index arthroplasty. The average pain free interval for the cementless group was only 11 months with 63% of the patients having no pain relief after their index arthroplasty. The average postoperative score for the cementless group was 88, whereas the average postoperative score for the cemented group was 82. Revision of failed cementless total knee implants with cement is a reliable procedure. Results are similar to those of cemented knees revised with cement.     

Porous-coated total hip replacement

Fifteen years of clinical experience with porous-coated prostheses demonstrated the durability of this type of fixation. This experience was documented by clinical follow-up study of the 393 cases treated by the senior author before 1985. Only six of these femoral components have been revised: three for loosening, two for stem breakage, and one for infection. Thus, the revision rate for the porous-coated stems was 1.5%. Porous-coated acetabular components were used in 227 of the arthroplasties. Five of these porous-coated cups have been revised: four for malposition leading to dislocation and one for late loosening secondary to osteolysis. Thus, the revision rate for these porous-coated acetabular components was 2.2%. Twenty bipolar and 146 cemented acetabular components were used in the remaining 166 cases treated before 1985. Eleven (7.5%) of the cemented acetabular components were revised. Revisions of the porous-coated components were rare in the first ten postoperative years. The clinical data were supplemented with analysis of postmortem specimens from 15 patients. Mechanical testing of the femoral specimens showed the relative micromotion at the porous surface to be exceptionally small (less than 40 microns). Seven of these postmortem retrievals involved cases with unilateral arthroplasties. In these cases, the contralateral normal femur also was removed, and a prosthesis identical to that in the in vivo implanted side was inserted to simulate the immediate postoperative condition. Dual-energy X-ray absorptiometry (DEXA) of the seven paired femora demonstrated that bone remodeling can be expected to produce a 5%-52% loss of periprosthetic bone mineral content, with the greatest loss occurring in the more osteoporotic patients.(ABSTRACT TRUNCATED AT 250 WORDS)     

Matrix metalloproteinases and tissue inhibitors of metalloproteinases in loose artificial hip joints

The role of extracellular matrix metalloproteinase enzymes and the tissue inhibitors of metalloproteinase in the periprostetic connective tissue matrix of loose artificial hip joints is reviewed. In the periprosthetic granulomatous interface connective tissues between bone and implants and inner cellular regenerating pseudocapsular tissues, matrix metalloproteinase 1, matrix metalloproteinase 2, matrix metalloproteinase 3, matrix metalloproteinase-9, and membrane type 1 matrix metalloproteinase enzymes can be shown in the light of immunohistochemistry, enzyme activity analysis, and messenger ribonucleic acid levels. Tissue inhibitors of metalloproteinase 1 and tissue inhibitors of metalloproteinase 2 also are found in the corresponding tissues. Analysis of matrix metalloproteinase and tissue inhibitors of metalloproteinase interaction shows imbalance between the enzymes and the endogenous inhibitors in favor of matrix metalloproteinase. This induces pathologic connective tissue remodeling in the interface and pseudocapsule. The data suggest that matrix metalloproteinase and tissue inhibitors of metalloproteinase system participate in the extracellular matrix degradation and tissue remodeling in artificial hip joints, and may contribute to the periprosthetic weakening, implant loosening, and osteolysis around implants. More evidence for their active involvement is sought by intervention studies with type specific matrix metalloproteinase inhibitors.     

Static and dynamic function of the lower esophageal sphincter before and after laparoscopic Nissen fundoplication

SUBJECT: We analysed the periprosthetic bone mineral density (BMD) in a prospective longitudinal study over two years after operation and in a separate cross-sectional study more than five years after implantation of cementless total hip arthroplasty (CLS-stem, Mecron threaded acetabular cup) by using dual-energy X-ray absorptiometry (DEXA). MATERIALS AND METHODS: In n = 53 patients (29 women, 24 men) we analysed the periprosthetic BMD prospectively in certain periods. All patients had an uncomplicated postoperative course and good clinical outcome (Merle d'Aubigne score > 12). In the cross-sectional study we analysed 23 patients (13 women, 10 men). Regions of Interest (ROI) were defined according to Gruen et al. for the periprosthetic femur and to De Lee and Charnley for the periprosthetic acetabulum. RESULTS: BMD significantly decreases in the periprosthetic femur as well as in the periprosthetic acetabulum during the first three months. In men BMD reaches its lowest values between six months (femur) and one year (acetabulum) after operation and then increases to 96.2% at the femur and 93.8% at the acetabulum. In women BMD decreases during the entire follow-up to 89.4% at the periprosthetic femur and 80.0% at the periprosthetic acetabulum. In the proximal zones 1 and 7 of the femur and the cranial-medial zone II of the acetabulum we observed the highest decrease of mineralisation. More than five years after implantation of the prosthesis BMD in the femur showed only little changes. On the other side BMD around the threaded acetabular cup significantly decreased to 67.4% in women and 79.1% in men. CONCLUSION: The results reflect the different stress on the periprosthetic bone after implantation of the prosthesis and fit to earlier reported good clinical results of the CLS-stem and to increased loosening rate of threaded acetabular cups after five years. Analysing changes of mineralisation in cementless total hip arthroplasty DEXA together with the analization scheme according to Gruen at the femur and to De Lee and Charnley at the acetabulum is a useful method and has been reliable in clinical practise.     

The effect of hydroxyapatite coating on the fixation of hip prostheses. A comparison of clinical and radiographic results of hip replacement in a matched-pair study

The efficacy of a total hip replacement with a hydroxyapatite-coated hip prosthesis was compared with that of an uncoated, cementless prosthesis of the same type. Preoperatively, there was no difference in the patient's diagnosis, hip score, age, and sex. All operations were performed by one surgeon in a standardized manner. The choice of the implant was randomized, and the follow-up period was equal for both types. The implant used was associated with a poor outcome due to a high incidence of early aseptic loosening. Probably because of a poor initial fixation, there was a significant difference in the clinical results after a short follow-up period when an additional HA layer was used. According to the patients' pain, migration of the implant, and presence of a progressive radiolucent line, use of the HA-coated prosthesis led to a significantly better result; however, we also found an increased rate of heterotopic bone formation in the HA-coated group. It was concluded that the HA coating improves the initial fixation of a hip prosthesis.     

From the RSNA refresher courses. Total hip arthroplasty: radiographic evaluation

Expected appearances of total hip arthroplasty vary according to type of implant, its method of fixation (cemented, porous coating for bone ingrowth, press fit), and whether it is a revision. Cemented arthroplastic components normally may show 1-2-mm-wide radiolucent zones at cement interfaces. Definite loosening is diagnosed when progressive widening of the radiolucent zone, migration of a cemented component, or change in alignment is seen. In cementless arthroplasty, normal findings include calcar resorption, radiolucent zones up to 2 mm in width, cortical thickening, periosteal reaction, endosteal sclerosis, and even subsidence of the femoral component that stabilizes at less than 1 cm. The most reliable radiographic signs of loosening in cementless arthroplasty are progressive subsidence, migration, or tilt of the component. Because subsidence or change in alignment may be very subtle, serial radiography and measurement are often required for diagnosis. Other signs that indicate loosening include bead shedding (in porous-coated prostheses), extensive cortical hypertrophy, endosteal bone bridging at the stem tip, endosteal scalloping, and a radiolucent zone wider than 2 mm. In revision arthroplasty, wide radiolucent zones and subsidence are common. The diagnosis of revision failure is based on progressive widening of the radiolucent zones and change in component position after 12 months.     

Middle-term results of anatomic medullary locking total hip arthroplasty

We report the results of cementless, anatomic, medullary locking hip prosthesis application in our first consecutive series. We used the so-called Asian size of prosthesis with proportionally smaller stem sizes in both diameter and length. Forty-seven stems and sockets were analyzed with a mean follow-up of 69 months. The mean Merle d'Aubigné hip scores were 8 points preoperatively and 16 points at the final follow-up. Radiologically, the stems showed excellent stability without loosening. Stress shielding around the stems did occur in most cases but did not progress. Preoperative bone quality influenced the extent of stress shielding evaluated at the final follow-up: higher stress shielding was noted in poorer quality bones at the time of operation. There were problems with the sockets. The shallow socket and impingement at the protruded rim seemed to cause a high incidence of dislocation (13%). Massive polyethylene wear occurred in 5 sockets. These sockets were 48 and 46 mm in diameter with 26 mm heads. In conclusion, the stems of the anatomic medullary locking hip prostheses used in Japan showed satisfactory stability even in poor quality bones, but there were problems with the polyethylene liners. Our solution was to use larger sockets with 22 mm heads.     

Distribution of serum lipoprotein(a) levels--a non-parametric analysis

Persisting postoperative pain of the thigh is a common problem of cementless hip endoprostheses occurring in about 15-20% of the patients. We performed a comparative study including patients with (n = 40) and without (n = 45) pain of the thigh. 85 cementless porous-coated anatomic (PCA) hip endoprostheses in 74 patients were examined. All patients underwent clinical examination including a standardized questionnaire, x-ray, and 3-phase bone scintigraphy. Slight or moderate 99mTc-MDP uptake in the area of the greater and lesser trochanter as well as at the tip was a common finding in PCA prostheses in patients without pain and was not a sign of loosening of the hip. Radiologically, there was no difference between patients with and without pain. However, persisting pain of the thigh in patients with PCA prosthesis corresponded with an increased uptake at the tip and the medial and lateral femur, not being a sign of loosening even in this group. The special biomechanical conditions of cementless prostheses causing inhomogeneous intraosseous stress distribution are supposed to be the reason for that.     

Evaluation of periprosthetic bone using dual-energy x-ray absorptiometry: precision of the method and effect of operation on bone mineral density

To assess the perioperative bone loss of femur during total hip arthroplasty (THA), periprosthetic bone mineral density (BMD) of the seven regions of interests (Gruen zones) was determined with dual-energy x-ray absorptiometry (DXA) preoperatively in both proximal femurs and postoperatively in the involved side in 53 patients with degenerative hip osteoarthrosis. The mean (standard deviation, SD) precision error (coefficient of variation percent, CV%) in various regions of interest (ROIs) based on two consecutive measurements (n = 16) were 2.3 (0.8)%, 2.5 (1.5)%, and 2.8 (1.6)% for uncemented stems, cemented stems, and control sides, respectively. Furthermore, the mean variability caused by the rotation of femur was 3.5 (1.4)%. The most significant perioperative bone loss (13.5-19.2%) was found in the calcar area (zone 7) after noncemented THA. Zone 4, representing the bone below the prosthesis, also showed BMD decreases. These decreases suggest perioperative bone loss owing to rasping and reaming the calcar and bone canal. However, after cemented THA, highly significant BMD increases were found in all the lateral zones. The calcar area was the only site where significant perioperative bone loss was detected (12.8%). In conclusion, DXA is a precise method for quantifying bone mass and density changes in the follow-up of THA. However, when interpreting the results, the preoperative BMD, differences between the femurs and the effect of operation on bone mass should be taken into account. We suggest that the best reference for BMD follow-up is the periprosthetic BMD of the involved side measured soon after the THA.     

Ceramic endoprosthesis in total hip arthroplasty

Aluminia-on-aluminia hip prosthesis with titanium alloy stem was used in 255 cases, for 143 patients with cemented acetabular cup and 112 with an uncemented screw cup. The average age of the patients was 62 years. Of the patients, 35 were treated bilaterally. The indication for surgery was osteoarthritis in 186 cases, fractures and nonunions of the hip joint in 34 cases, and congenital dislocations of the hip joint in 16 cases. Previous surgery of the hip joint was recorded for 43 cases. All procedures were primary arthroplasties. Only personal clinical examinations together with radiographic studies were accepted as clinical data. Three patients failed to show up for routine follow-up evaluation and these were excluded from the series. Revision operation was classified as failure. The indications for revision were aseptic loosening, late infection, and fracture of the acrylic cement resulting in loosening of the acetabular or femoral component. The acetabular component was cemented in 143 patients and the mean follow-up period for these patients was 6.7 years (range, 1-12 years). In this series, a revision operation was undertaken for 16 patients (11%). In 12 cases, broken acrylic cement resulted in acetabular aseptic loosening. In the series of 112 patients with uncemented titanium screw cup, the mean follow-up period was 3.6 years (range, 1-7 years) and a revision operation was undertaken in seven cases (7%). In two, the indication was late infection; in one, technical failure; and in two, progression of Pigmented villonodular synovitis. Acetabular aseptic loosening resulted in revision in only two cases. With cementless acetabular component bone transplantation is indicated.(ABSTRACT TRUNCATED AT 250 WORDS)     

Long-term results of cemented femoral stems with roughened precoated surfaces

Several recent articles have raised the question that rough surfaces with or without precoating may contribute to the loosening of cemented femoral stems. In this article the author collates selected reports of several different cemented femoral stem designs that have rough surfaces which have shown excellent long term results. Femoral stem loosening is generally multifactorial. It often is difficult to separate the effect of stem design or other changes from the effect of surface finish. Many rough surfaced cemented femoral stems have served well for decades. It is important to analyze all the pertinent factors in cases of early loosening.     

Cemented versus cementless total hip arthroplasty. A comparative study of equivalent patient populations

Seventy-one cemented total hip arthroplasties (THAs) in 66 patients were compared with 70 cementless THAs in 61 patients. The cemented and cementless components were identical except for the presence of a porous proximal surface on the cementless component. The mean follow-up time was 4.3 years in the cemented group and 4.1 years in the cementless group, with a minimum follow-up of two years. Postoperative Charnley pain scores were identical for both groups, as was the incidence of limp. The final Charnley function score, however was significantly higher in the cementless group (5.6) than in the cemented group (5.1). One revision was performed in the cemented group, none in the cementless group. Complete demarcation of the bone-cement interface was presented in two cemented components and was associated with subsidence. Three cementless components had early subsidence, and two of these had complete radiopaque lines around the porous surface. None of these three patients, however, had unsatisfactory pain or function scores. The most important finding of this study is the equivalence of clinical results noted in both cemented and cementless THA as late as six years after operation.     

Biphasic spiral computed tomography versus digital subtraction angiography for evaluation of arterial thrombosis after orthotopic liver transplantation

Aseptic loosening of implant components is a common and important complication of both cemented and uncemented prosthetic joint replacements. Wear particles derived from organic polymer and metal implant biomaterials are commonly found within macrophages and macrophage polykaryons in the fibrous membrane between loose implant components and the host bone undergoing resorption. In order to determine whether biomaterial particle-containing, foreign-body macrophages may contribute to periprosthetic bone resorption, we cultured murine monocytes that had phagocytosed particles of biomaterials commonly employed in bone implant surgery [polymethylmethacrylate (PMMA), ultra-high molecular weight polyethylene (PE), titanium and chromium-cobalt] on bone slices and glass coverslips with UMR 106 osteoblast-like stromal cells in the presence of 1,25-dihydroxy-vitamin D3. Under these conditions, all biomaterial particle-containing, foreign-body macrophages differentiated into osteoclastic cells, i.e. tartrate-resistant acid phosphatase (TRAP)-positive multinucleated cells capable of extensive lacunar bone resorption. This study shows that particle phagocytosis by macrophages does not abrogate the ability of these cells to undergo osteoclast differentiation. These findings emphasise the importance of the foreign-body macrophage response to biomaterial wear particles in the pathogenesis of aseptic loosening.