- Email: [email protected]
Serum Interleukin 8 Levels Correlate With Synovial Fluid Levels in Patients With Aseptic Loosening of Hip Prosthesis
The Journal of Arthroplasty Vol. 20 No. 8 2005
Serum Interleukin 8 Levels Correlate With Synovial Fluid Levels in Patients With Aseptic Loosening of Hip Prosthesis Ryuji Tanaka, MD,* Yuji Yasunaga, MD, PhD,* Takashi Hisatome, MD, PhD,* Takuma Yamasaki, MD,* Hiroshi Iwamori, MD,y and Mitsuo Ochi, MD, PhD*
Abstract: Levels of inflammatory cytokines (tumor necrosis factor a, interleukin [IL] 6, and IL-8) in serum from patients with osteolysis on radiographs after hip arthroplasty [osteolysis(+), n = 28], patients without osteolysis after hip arthroplasty [osteolysis(), n = 24], and nonoperated healthy subjects [controls, n = 20] were determined. In addition, cytokine levels in synovial fluid from patients undergoing revision total hip arthroplasty (n = 14) for loosening were measured and compared with each other and with the area of osteolysis on radiographs. Serum IL-6 and IL8 levels were significantly higher in the osteolysis(+) group than in the osteolysis() or the control groups. Furthermore, a significant correlation was found between the serum and synovial fluid IL-8 levels and between synovial fluid IL-8 levels and the area of osteolysis in patients undergoing revision total hip arthroplasty. Therefore, serum IL-8 levels could be a useful periprosthetic osteolysis marker. Key words: aseptic loosening, periprosthetic osteolysis, serum, synovial fluid, cytokines, interleukin 8. n 2005 Elsevier Inc. All rights reserved.
Currently, the extent of periprosthetic osteolysis can only be estimated by its appearance on plain radiographs, and another useful diagnostic method has not yet been established. Although there have been several reports which have dealt with the relationship of cytokine levels either in serum [4,5] or in synovial fluid [6-10] and periprosthetic osteolysis, use of their levels as a marker for monitoring osteolysis has not been documented, and to the best of our knowledge, there has been no report concerned with the relationship between cytokine levels in serum and those in synovial fluid in patients with aseptic loosening of hip prostheses. We have therefore conducted the current study to determine levels of inflammatory cytokines (ie, tumor necrosis factor a [TNF-a], interleukin [IL] 6, and IL-8) in serum obtained from patients diagnosed with osteolysis on plain radiograph after THA or bipolar hip arthroplasty, patients without osteolysis after THA, and healthy subjects.
With improved cementing technique and implant design, the longevity of total hip arthroplasty (THA) has been prolonged. However, periprosthetic osteolysis with resulting aseptic loosening has been considered to be the most important factor that affects the long-term success of THA, and periprosthetic osteolysis, in turn, has been linked to the action of several inflammatory cytokines, prostaglandins, soluble adhesion molecules, and growth factors [1-3]. From the *Department of Orthopaedic Surgery, Hiroshima University Graduate School of Biomedical Sciences, Hiroshima, Japan; and yDepartment of Orthopaedic Surgery, Chuden Hospital, Hiroshima, Japan. No benefits or funds were received in support of the study. Reprint requests: R. Tanaka, MD, Department of Orthopaedic Surgery, Hiroshima University Graduate School of Biomedical Sciences, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. n 2005 Elsevier Inc. All rights reserved. 0883-5403/05/1906-0004$30.00/0 doi:10.1016/j.arth.2005.03.034
1049
1050 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 The results were compared among the relevant groups. In addition, we have analyzed the cytokine levels in synovial fluid in patients undergoing revision THA for aseptic loosening due to osteolysis and have investigated the correlation between serum, synovial fluid levels, and the area of osteolysis.
Table 2. Mean Serum Levels of Cytokines
TNF-a IL-6 IL-8
OL(+) (n = 28)
OL() (n = 24)
Control (n = 20)
2.51 F 0.32 (0.49-7.66) 3.31 F 0.70 (0-16.9)* 21.7 F 1.91 (6.13-49.6)*
2.15 F 0.26 (0.71-6.18) 1.11 F 0.28 (0-4.97) 11.2 F 1.08 (0-21.7)
2.01 F 0.17 (0.92-3.15) 0.67 F 0.23 (0-3.01) 11.2 F 1.34 (0-21.6)
Materials and Methods
Values are expressed as pg/mL F SEM (range). *P b .01 (Mann-Whitney U test) compared with OL() group and the control group.
Osteolysis was defined as radiolucent areas of a bone more than 2 mm in width and progressive on the anteroposterior or on the lateral radiographs. Nonprogressive and linear radiolucent areas less than 2 mm in width were not considered osteolysis. After the radiograph was digitized, the area of osteolysis was outlined and measured using an image analysis program (National Institutes of Health Image program, Bethesda, Md). The area of radiolucency around the acetabular component or the femoral component, or around both, was measured on anteroposterior or on lateral radiographs, and the total amount was considered as the area value of osteolysis.
Details of the patients are given in Table 1. Patients with autoimmune diseases, allergies, infectious diseases, or cancer were excluded from this study. All patients or their relatives gave informed consent for blood and synovial fluid samples to be obtained for this study. The procedure was carried out after approval by the Hiroshima University School of Medicine Ethics Committee.
Osteolysis
Patients A total of 72 patients were divided into 3 groups as follows: ! An osteolysis(+) [OL(+)] group consisted of 28 patients with arthroplasties who had osteolysis around the acetabular or femoral component, or both. ! An osteolysis() [OL()] group consisted of 24 patients with arthroplasties who had no osteolysis around the acetabular or femoral component. ! A control group consisted of 20 nonoperated healthy subjects.
Sample Preparation Peripheral venous blood samples were collected from all patients in serum separator tubes and then centrifuged at 3000 rpm for 15 minutes. The separated serum was aliquoted and stored at 708C until assayed. Similarly, synovial fluid samples were collected at revision THA and then centrifuged at 3000 rpm for 15 minutes to remove cells and other debris. The supernatant was aliquoted and stored at 708C until assayed. Cytokine Assays Inflammatory cytokine levels (TNF-a, IL-6, and IL-8) in serum and synovial fluid were measured using quantitative, noncompetitive, sandwich enzyme-linked immunosorbent assays (Quantikine, R&D Systems, Minneapolis, Minn). The
Table 1. Details of the Patients in Each Group
Total Samples Sex (male/female) Mean age (range) Mean implant duration, mo (range) Cemented/uncemented THA/BHA
OL(+)
OL()
Control
28 4/24 65.6 (39-82) 147.1 (60-264) 11/17 15/13
24 1/23 66.1 (49-84) 65.3 (23-148) 3/21 24/0
20 6/14 63.0 (54-77) –
BHA indicates bipolar hip arthroplasty.
– –
synovial fluid TNF– α level
(pg / ml) 18 16 14 12 10 8 6 4 2 0 0
N.S.
0.5
1
1.5
2
2.5
3
serum TNF– α level
3.5
4
4.5
(pg / ml)
Fig. 1. Comparison between serum TNF-a level and synovial fluid TNF-a level.
1051
Serum IL-8 and Synovial Fluid Levels ! Tanaka et al
Statistical Analysis The serum cytokine levels of the groups were compared using the Mann-Whitney U test, and in the OL(+) group, the correlation between serum levels, synovial fluid levels, and the areas of osteolysis was tested with each other using the Spearman rank correlation method. A P value of less than .05 was considered statistically significant.
(pg / ml) 6000
synovial fluid IL–8 level
detection limits of the assays were 0.12 pg/mL for TNF-a, 0.70 pg/mL for IL-6, and 3.5 pg/mL for IL-8. Levels of the cytokines lower than the detection limit were considered to be 0 pg/mL.
P = .01 R = 0.71 N = 14
5000 4000 3000 2000 1000 0 0
5
10
15
20
25
30
35
40
(pg / ml)
serum IL-8 level
Fig. 3. Comparison between serum IL-8 level and synovial fluid IL-8 level.
Results Serum Cytokine Levels in OL(+), OL(–-), and Control Groups
Serum and Synovial Fluid Levels vs the Area of Osteolysis in OL(+)
There were no significant differences in the serum TNF-a levels among the 3 groups, whereas the OL(+) group had a significantly increased serum IL-6 level when compared with both the OL() group and the control group. The OL(+) serum IL-8 levels were also significantly increased when compared with both the OL() group and the control group (Table 2).
Furthermore, in these 14 patients, a significant correlation was found between the synovial fluid IL-8 level and the area of osteolysis (Fig. 4). No evident correlation was noted between the area of osteolysis and any serum cytokine level (Fig. 5) or synovial fluid TNF-a or IL-6 levels.
Serum vs Synovial Fluid in OL(+) In 14 patients of the OL(+) group from whom synovial fluids were collected at revision THA, no evident correlation regarding TNF-a and IL-6 levels was noted between serum and synovial fluid, whereas a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level (Figs. 1-3).
(pg / ml)
1400
1000
1200
N.S.
area of osteolysis
synovial fluid IL–6 level
No evident correlation was noted between any serum cytokine level and the implant duration in either the OL() group or the OL(+) group. No evident correlation was noted between any serum cytokine level and the area of osteolysis in the OL(+) group. There was no significant difference in any cytokines between cemented and uncemented stems.
(mm2)
1200
800 600 400 200 0 0
Comparison Between Serum Cytokine Levels and the Other Variables
P = .02 R = 0.64 N = 14
1000 800 600 400 200 0
1
2
3
4
5
6
serum IL-6 level
7
8
9
10
(pg / ml)
Fig. 2. Comparison between serum IL-6 level and synovial fluid IL-6 level.
0
1000
2000
3000
4000
synovial fluid IL-8 level
5000
6000
(pg / ml)
Fig. 4. Comparison between synovial fluid IL-8 level and area of osteolysis.
1052 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 (mm2)
area of osteolysis
1400 1200
N.S.
1000 800 600 400 200 0 0
5
10
15
20
25
serum IL–8 level
30
35
40
(pg / ml)
Fig. 5. Comparison between serum IL-8 level and area of osteolysis.
Discussion Periprosthetic osteolysis with the result of aseptic loosening has been considered to be the most important problem that affects the long-term results of THA. It has been understood to be a consequence of a biologic response to wear particles, and analysis of the periprosthetic tissue samples retrieved from patients with osteolysis at revision surgery has demonstrated that periprosthetic osteolysis involves the inflammatory mediators of bone resorption, such as prostaglandin E2 (PGE2), TNF-a, IL-1b, and IL-6 [11-15]. Analysis of serum and synovial fluid from patients with a loose hip prosthesis has demonstrated the presence of similar inflammatory mediators. Granchi et al [4] found an elevation in the level of serum granulocyte-macrophage colony-stimulating factor (GM-CSF), whereas the levels of IL-6 and TNF-a were not significantly elevated in the serum of patients with aseptic loosening of hip prosthesis. Hernigou et al [5] found that only increased levels of IL-6 were detected in patients having had THA more than 10 years ago. They each stated the possibility that serum GM-CSF and IL-6 levels could be a marker for periprosthetic osteolysis. However, it is difficult to distinguish between cytokine release from the osteolytic region and from other organs. Moreover, although it is speculated that cytokines flow out of the joint into the bloodstream through the blood vessels in the synovium of the pseudocapsule, clearance rate from the joint is unknown and the significance of serum cytokine levels as an osteolysis marker has not been established [16]. If serum cytokine levels can be established as a marker for osteolysis after THA, it would provide not only a simple and noninvasive assessment method but also the assessment of time-related change. These may
supplement plain radiographs for the assessment of the presence and progression of osteolysis. Sabokbar and Rushton [9] found an elevation in the levels of PGE 2, IL-6, IL-8, and multiple adhesion molecules in the synovial fluid around loose implants and compared them with the levels in the synovial fluid from controls. Kim et al [17] found elevation in the levels of IL-6, soluble IL-6 receptor, and tartrate-resistant acid phosphate in joint fluid from failed total hip arthroplasties. Takei et al [3] found that macrophage colony-stimulating factor levels in loose hip joint fluid were significantly higher than in active rheumatoid arthritis and mild osteoarthritis fluid. It is thought that synovial fluid is forced along the implant interface, particularly if there is osteolysis or if there are errors in fixation during surgery [3,18]. Therefore, as cells in the synovial capsule and those in the bone-implant interface may have the opportunity both to be affected by cytokine levels in the synovial fluid and to secrete cytokines into the fluid themselves [3,7,10,18], it appears that cytokine levels in synovial fluid may reflect the local condition better than cytokine levels in serum. However, if a significant correlation is found between the cytokine levels in synovial fluid and those in serum, it would seem probable that cytokine levels in serum could be a marker for periprosthetic osteolysis. However, to the best of our knowledge, there has been no previous report concerned with the relationship between cytokine levels in serum and those in synovial fluid in patients with aseptic loosening of hip prostheses. The present study confirmed that the OL(+) group had significantly increased serum IL-6 and IL-8 levels when compared with both the OL() group and the control group. Furthermore, a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level. Moreover, a significant correlation was found between the IL-8 level of synovial fluid and the area of osteolysis. Clarke et al [7] found no relationship between cytokine levels in synovial fluid and the degree of radiolucency around the implant that was graded based on the method of Harris et al [19]. Nivbrant et al [10] showed that higher levels of TNF-a were found in hips with bone resorption of type II and type III, compared with those with type I loosening, according to the classification by Gustilo and Pasternak. However, both of these studies were qualitative evaluations and, thus, probably less accurate when compared with our method. In our study, IL-8 levels in synovial fluid correlated
Serum IL-8 and Synovial Fluid Levels ! Tanaka et al
with the severity of osteolysis and, so, might be considered to be a more reliable parameter. Although it has been reported that Hernigou et al [5] accepted a correlation between serum IL-6 levels and the area of osteolysis by the same method as we used, the comparison between the cytokine levels in serum and those in synovial fluid was not made. In our study, there was no correlation between serum IL-6 and synovial fluid IL-6, indicating that serum IL-6 level reflects the local condition poorly. With regard to IL-8, Ishiguro et al [20] reported increased messenger RNA expression of IL-8 in periprosthetic granulomatous tissue when compared with fibrous scar tissue. Lassus et al [21] showed high count levels of IL-8–producing cells in both the synovial-like interface membrane and pseudocapsular tissue of aseptically loosened THAs when compared with controls. Analysis of synovial fluid from patients with a loose hip prosthesis has demonstrated that IL-8 levels in loose hip joint fluid were significantly higher than in controls [6,7,9]. Interleukin 8 contributes to the release of proteinases, PGE2, adhesion molecules, and superoxide molecules into joint fluid that may play an essential role in prosthesis loosening [9,21]. Our results show a significant rise in serum IL8 level in patients with aseptic loosening of hip prosthesis, and a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level. This suggests that the serum IL-8 level could be a useful marker for periprosthetic osteolysis. However, there was a small number of patients in the OL() group who had an increased serum IL-8 level, and continued follow-up is needed to see whether periprosthetic osteolysis appears on plain radiograph in the near future. If osteolysis does appear, it will add more evidence to suggest the usefulness of the IL-8 level as a marker for periprosthetic osteolysis. Pharmaceutical interventions are currently being developed in an attempt to treat or prevent osteolysis, including the use of bisphosphonates [22], osteoprotegerin [23,24], or an anti–TNF-a antibody [25,26]. To monitor and predict the effect of the drugs on osteolysis, a marker for periprosthetic osteolysis will be indispensable for the accumulation of data required for their clinical use.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
References 16. 1. Chomarat P, Vannier E, Dechanet J, et al. Balance of IL-1Ra/IL-1b in rheumatoid synovium and its
1053
regulation by IL-4 and IL-10. J Immunol 1995; 154:1432. Nabae M, Inoue K, Ushiyama T, et al. Gene expression of anti-inflammatory mediators in THR retrieved interfacial membranes. Acta Orthop Scand 1999;70:149. Takei I, Takagi M, Ida H, et al. High macrophagecolony stimulating factor levels in synovial fluid of loose artificial hip joints. J Rheumatol 2000;27:894. Granchi D, Verri E, Ciapetti G, et al. Bone-resorbing cytokines in serum of patients with aseptic loosening of hip prosthesis. J Bone Joint Surg Br 1998; 80:912. Hernigou P, Intrator L, Bahrami T, et al. Interleukin6 in the blood of patients with total hip arthroplasty without loosening. Clin Orthop 1999;366:147. Sypniewska G, Lis K, Bilinski PJ. Bone turnover markes and cytokines in joint fluid. Acta Orthop Scand 2002;73:518. Clarke SA, Brooks RA, Hobby JL, et al. Correlation of synovial fluid cytokine levels with histological and clinical parameters of primary and revision total hip and total knee replacements. Acta Orthop Scand 2001;72:491. Inomoto M, Miyakawa S, Mishima H, et al. Elevated interleukin-12 in pseudosynovial fluid in patients with aseptic loosening of hip prosthesis. J Orthop Science 2000;5:369. Sabokbar A, Rushton N. Role of inflammatory mediators and adhesion molecules in the pathogenesis of aseptic loosening in total hip arthroplasties. J Arthroplasty 1995;10:810. Nivbrant B, Karlsson K, Karrholm J. Cytokine levels in synovial fluid from hips with well-functioning or loose prostheses. J Bone Joint Surg Br 1999;81:163. Goldring SR, Schiller AL, Roelke M, et al. 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 Am 1983; 65:575. Dorr LD, Bloebaum R, Emmanual J, et al. Histologic, biochemical, and ion analysis of tissue and fluids retrieved during total hip arthroplasty. Clin Orthop 1990;261:82. Goodman SB, Chin RC, Chiou SS, et al. A clinicalpathologic-biochemical study of the membrane surrounding loosened and nonloosened total hip arthroplasties. Clin Orthop 1989;244:182. Kim KJ, Chiba J, Rubash HE. In vitro and in vitro analysis of membranes from hip prosthesis inserted without cement. J Bone Joint Surg Am 1994; 76:172. Chiba J, Rubash H, Kim KJ, et al. The characterization of cytokines in the interface tissue obtained from failed cementless total hip arthroplasty with and without femoral osteolysis. Clin Orthop 1994; 300:304. Takakuwa M, Matsuno T, Gotoh E. The serum bone resorbing cytokines with osteolysis of total hip arthroplasty (in Japanese). Seikeigeka 2002;53:1709.
1054 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 17. Kim KJ, Hijikata H, Itoh T, et al. Joint fluid from patients with failed total hip arthroplasty stimulates pit formation by mouse osteoclasts on dentin slices. J Biomed Mater Res 1998;43:234. 18. Schmalzried T, Jasty M, Harris W. Periprosthetic bone loss in total hip arthroplasty. Polyethylene wear debris and the concept of the effective joint space. J Bone Joint Surg Am 1992;74:849. 19. Harris WH, McCarthy Jr JC, O’Neill DA. Femoral component loosening using contemporary techniques of femoral cement fixation. J Bone Joint Surg Am 1982;64:1063. 20. Ishiguro N, Kojima T, Ito T, et al. Macrophage activation and migration in interface tissue around loosening total hip arthroplasty components. J Biomed Mater Res 1997;35:399. 21. Lassus J, Waris V, Xu JW, et al. Increased interleukin-8 (IL-8) expression is related to aseptic loosening of total hip replacement. Arch Orthop Trauma Surg 2000;120:328.
22. Millett PJ, Allen MJ, Bostorom MPG. Effects of alendronate on particle-induced osteolysis in a rat model. J Bone Joint Surg Am 2002;84:236. 23. Ulrich-Vinther M, Carmody EE, Goater JJ, et al. Recombinant adeno-associated virus-mediated osteoprotegerin gene therapy inhibits wear debrisinduced osteolysis. J Bone Joint Surg Am 2002;84: 1405. 24. Kim KJ, Kotake S, Udagawa N, et al. Osteoprotegerin inhibits in vitro mouse osteoclast formation induced by joint fluid from failed total hip arthroplasty. J Biomed Mater Res 2001;58:393. 25. Schwarz EM, Looney RJ, O’Keefe RJ. Anti–TNF-a therapy as a clinical intervention for periprosthetic osteolysis. Arthritis Res 2002;2:165. 26. Childs LM, Goater JJ, O’Keefe RJ, et al. Efficacy of etanercept for wear debris-induced osteolysis. J Bone Miner Res 2001;16:338.
Serum Interleukin 8 Levels Correlate With Synovial Fluid Levels in Patients With Aseptic Loosening of Hip Prosthesis Ryuji Tanaka, MD,* Yuji Yasunaga, MD, PhD,* Takashi Hisatome, MD, PhD,* Takuma Yamasaki, MD,* Hiroshi Iwamori, MD,y and Mitsuo Ochi, MD, PhD*
Abstract: Levels of inflammatory cytokines (tumor necrosis factor a, interleukin [IL] 6, and IL-8) in serum from patients with osteolysis on radiographs after hip arthroplasty [osteolysis(+), n = 28], patients without osteolysis after hip arthroplasty [osteolysis(), n = 24], and nonoperated healthy subjects [controls, n = 20] were determined. In addition, cytokine levels in synovial fluid from patients undergoing revision total hip arthroplasty (n = 14) for loosening were measured and compared with each other and with the area of osteolysis on radiographs. Serum IL-6 and IL8 levels were significantly higher in the osteolysis(+) group than in the osteolysis() or the control groups. Furthermore, a significant correlation was found between the serum and synovial fluid IL-8 levels and between synovial fluid IL-8 levels and the area of osteolysis in patients undergoing revision total hip arthroplasty. Therefore, serum IL-8 levels could be a useful periprosthetic osteolysis marker. Key words: aseptic loosening, periprosthetic osteolysis, serum, synovial fluid, cytokines, interleukin 8. n 2005 Elsevier Inc. All rights reserved.
Currently, the extent of periprosthetic osteolysis can only be estimated by its appearance on plain radiographs, and another useful diagnostic method has not yet been established. Although there have been several reports which have dealt with the relationship of cytokine levels either in serum [4,5] or in synovial fluid [6-10] and periprosthetic osteolysis, use of their levels as a marker for monitoring osteolysis has not been documented, and to the best of our knowledge, there has been no report concerned with the relationship between cytokine levels in serum and those in synovial fluid in patients with aseptic loosening of hip prostheses. We have therefore conducted the current study to determine levels of inflammatory cytokines (ie, tumor necrosis factor a [TNF-a], interleukin [IL] 6, and IL-8) in serum obtained from patients diagnosed with osteolysis on plain radiograph after THA or bipolar hip arthroplasty, patients without osteolysis after THA, and healthy subjects.
With improved cementing technique and implant design, the longevity of total hip arthroplasty (THA) has been prolonged. However, periprosthetic osteolysis with resulting aseptic loosening has been considered to be the most important factor that affects the long-term success of THA, and periprosthetic osteolysis, in turn, has been linked to the action of several inflammatory cytokines, prostaglandins, soluble adhesion molecules, and growth factors [1-3]. From the *Department of Orthopaedic Surgery, Hiroshima University Graduate School of Biomedical Sciences, Hiroshima, Japan; and yDepartment of Orthopaedic Surgery, Chuden Hospital, Hiroshima, Japan. No benefits or funds were received in support of the study. Reprint requests: R. Tanaka, MD, Department of Orthopaedic Surgery, Hiroshima University Graduate School of Biomedical Sciences, 1-2-3 Kasumi, Minami-ku, Hiroshima 734-8551, Japan. n 2005 Elsevier Inc. All rights reserved. 0883-5403/05/1906-0004$30.00/0 doi:10.1016/j.arth.2005.03.034
1049
1050 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 The results were compared among the relevant groups. In addition, we have analyzed the cytokine levels in synovial fluid in patients undergoing revision THA for aseptic loosening due to osteolysis and have investigated the correlation between serum, synovial fluid levels, and the area of osteolysis.
Table 2. Mean Serum Levels of Cytokines
TNF-a IL-6 IL-8
OL(+) (n = 28)
OL() (n = 24)
Control (n = 20)
2.51 F 0.32 (0.49-7.66) 3.31 F 0.70 (0-16.9)* 21.7 F 1.91 (6.13-49.6)*
2.15 F 0.26 (0.71-6.18) 1.11 F 0.28 (0-4.97) 11.2 F 1.08 (0-21.7)
2.01 F 0.17 (0.92-3.15) 0.67 F 0.23 (0-3.01) 11.2 F 1.34 (0-21.6)
Materials and Methods
Values are expressed as pg/mL F SEM (range). *P b .01 (Mann-Whitney U test) compared with OL() group and the control group.
Osteolysis was defined as radiolucent areas of a bone more than 2 mm in width and progressive on the anteroposterior or on the lateral radiographs. Nonprogressive and linear radiolucent areas less than 2 mm in width were not considered osteolysis. After the radiograph was digitized, the area of osteolysis was outlined and measured using an image analysis program (National Institutes of Health Image program, Bethesda, Md). The area of radiolucency around the acetabular component or the femoral component, or around both, was measured on anteroposterior or on lateral radiographs, and the total amount was considered as the area value of osteolysis.
Details of the patients are given in Table 1. Patients with autoimmune diseases, allergies, infectious diseases, or cancer were excluded from this study. All patients or their relatives gave informed consent for blood and synovial fluid samples to be obtained for this study. The procedure was carried out after approval by the Hiroshima University School of Medicine Ethics Committee.
Osteolysis
Patients A total of 72 patients were divided into 3 groups as follows: ! An osteolysis(+) [OL(+)] group consisted of 28 patients with arthroplasties who had osteolysis around the acetabular or femoral component, or both. ! An osteolysis() [OL()] group consisted of 24 patients with arthroplasties who had no osteolysis around the acetabular or femoral component. ! A control group consisted of 20 nonoperated healthy subjects.
Sample Preparation Peripheral venous blood samples were collected from all patients in serum separator tubes and then centrifuged at 3000 rpm for 15 minutes. The separated serum was aliquoted and stored at 708C until assayed. Similarly, synovial fluid samples were collected at revision THA and then centrifuged at 3000 rpm for 15 minutes to remove cells and other debris. The supernatant was aliquoted and stored at 708C until assayed. Cytokine Assays Inflammatory cytokine levels (TNF-a, IL-6, and IL-8) in serum and synovial fluid were measured using quantitative, noncompetitive, sandwich enzyme-linked immunosorbent assays (Quantikine, R&D Systems, Minneapolis, Minn). The
Table 1. Details of the Patients in Each Group
Total Samples Sex (male/female) Mean age (range) Mean implant duration, mo (range) Cemented/uncemented THA/BHA
OL(+)
OL()
Control
28 4/24 65.6 (39-82) 147.1 (60-264) 11/17 15/13
24 1/23 66.1 (49-84) 65.3 (23-148) 3/21 24/0
20 6/14 63.0 (54-77) –
BHA indicates bipolar hip arthroplasty.
– –
synovial fluid TNF– α level
(pg / ml) 18 16 14 12 10 8 6 4 2 0 0
N.S.
0.5
1
1.5
2
2.5
3
serum TNF– α level
3.5
4
4.5
(pg / ml)
Fig. 1. Comparison between serum TNF-a level and synovial fluid TNF-a level.
1051
Serum IL-8 and Synovial Fluid Levels ! Tanaka et al
Statistical Analysis The serum cytokine levels of the groups were compared using the Mann-Whitney U test, and in the OL(+) group, the correlation between serum levels, synovial fluid levels, and the areas of osteolysis was tested with each other using the Spearman rank correlation method. A P value of less than .05 was considered statistically significant.
(pg / ml) 6000
synovial fluid IL–8 level
detection limits of the assays were 0.12 pg/mL for TNF-a, 0.70 pg/mL for IL-6, and 3.5 pg/mL for IL-8. Levels of the cytokines lower than the detection limit were considered to be 0 pg/mL.
P = .01 R = 0.71 N = 14
5000 4000 3000 2000 1000 0 0
5
10
15
20
25
30
35
40
(pg / ml)
serum IL-8 level
Fig. 3. Comparison between serum IL-8 level and synovial fluid IL-8 level.
Results Serum Cytokine Levels in OL(+), OL(–-), and Control Groups
Serum and Synovial Fluid Levels vs the Area of Osteolysis in OL(+)
There were no significant differences in the serum TNF-a levels among the 3 groups, whereas the OL(+) group had a significantly increased serum IL-6 level when compared with both the OL() group and the control group. The OL(+) serum IL-8 levels were also significantly increased when compared with both the OL() group and the control group (Table 2).
Furthermore, in these 14 patients, a significant correlation was found between the synovial fluid IL-8 level and the area of osteolysis (Fig. 4). No evident correlation was noted between the area of osteolysis and any serum cytokine level (Fig. 5) or synovial fluid TNF-a or IL-6 levels.
Serum vs Synovial Fluid in OL(+) In 14 patients of the OL(+) group from whom synovial fluids were collected at revision THA, no evident correlation regarding TNF-a and IL-6 levels was noted between serum and synovial fluid, whereas a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level (Figs. 1-3).
(pg / ml)
1400
1000
1200
N.S.
area of osteolysis
synovial fluid IL–6 level
No evident correlation was noted between any serum cytokine level and the implant duration in either the OL() group or the OL(+) group. No evident correlation was noted between any serum cytokine level and the area of osteolysis in the OL(+) group. There was no significant difference in any cytokines between cemented and uncemented stems.
(mm2)
1200
800 600 400 200 0 0
Comparison Between Serum Cytokine Levels and the Other Variables
P = .02 R = 0.64 N = 14
1000 800 600 400 200 0
1
2
3
4
5
6
serum IL-6 level
7
8
9
10
(pg / ml)
Fig. 2. Comparison between serum IL-6 level and synovial fluid IL-6 level.
0
1000
2000
3000
4000
synovial fluid IL-8 level
5000
6000
(pg / ml)
Fig. 4. Comparison between synovial fluid IL-8 level and area of osteolysis.
1052 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 (mm2)
area of osteolysis
1400 1200
N.S.
1000 800 600 400 200 0 0
5
10
15
20
25
serum IL–8 level
30
35
40
(pg / ml)
Fig. 5. Comparison between serum IL-8 level and area of osteolysis.
Discussion Periprosthetic osteolysis with the result of aseptic loosening has been considered to be the most important problem that affects the long-term results of THA. It has been understood to be a consequence of a biologic response to wear particles, and analysis of the periprosthetic tissue samples retrieved from patients with osteolysis at revision surgery has demonstrated that periprosthetic osteolysis involves the inflammatory mediators of bone resorption, such as prostaglandin E2 (PGE2), TNF-a, IL-1b, and IL-6 [11-15]. Analysis of serum and synovial fluid from patients with a loose hip prosthesis has demonstrated the presence of similar inflammatory mediators. Granchi et al [4] found an elevation in the level of serum granulocyte-macrophage colony-stimulating factor (GM-CSF), whereas the levels of IL-6 and TNF-a were not significantly elevated in the serum of patients with aseptic loosening of hip prosthesis. Hernigou et al [5] found that only increased levels of IL-6 were detected in patients having had THA more than 10 years ago. They each stated the possibility that serum GM-CSF and IL-6 levels could be a marker for periprosthetic osteolysis. However, it is difficult to distinguish between cytokine release from the osteolytic region and from other organs. Moreover, although it is speculated that cytokines flow out of the joint into the bloodstream through the blood vessels in the synovium of the pseudocapsule, clearance rate from the joint is unknown and the significance of serum cytokine levels as an osteolysis marker has not been established [16]. If serum cytokine levels can be established as a marker for osteolysis after THA, it would provide not only a simple and noninvasive assessment method but also the assessment of time-related change. These may
supplement plain radiographs for the assessment of the presence and progression of osteolysis. Sabokbar and Rushton [9] found an elevation in the levels of PGE 2, IL-6, IL-8, and multiple adhesion molecules in the synovial fluid around loose implants and compared them with the levels in the synovial fluid from controls. Kim et al [17] found elevation in the levels of IL-6, soluble IL-6 receptor, and tartrate-resistant acid phosphate in joint fluid from failed total hip arthroplasties. Takei et al [3] found that macrophage colony-stimulating factor levels in loose hip joint fluid were significantly higher than in active rheumatoid arthritis and mild osteoarthritis fluid. It is thought that synovial fluid is forced along the implant interface, particularly if there is osteolysis or if there are errors in fixation during surgery [3,18]. Therefore, as cells in the synovial capsule and those in the bone-implant interface may have the opportunity both to be affected by cytokine levels in the synovial fluid and to secrete cytokines into the fluid themselves [3,7,10,18], it appears that cytokine levels in synovial fluid may reflect the local condition better than cytokine levels in serum. However, if a significant correlation is found between the cytokine levels in synovial fluid and those in serum, it would seem probable that cytokine levels in serum could be a marker for periprosthetic osteolysis. However, to the best of our knowledge, there has been no previous report concerned with the relationship between cytokine levels in serum and those in synovial fluid in patients with aseptic loosening of hip prostheses. The present study confirmed that the OL(+) group had significantly increased serum IL-6 and IL-8 levels when compared with both the OL() group and the control group. Furthermore, a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level. Moreover, a significant correlation was found between the IL-8 level of synovial fluid and the area of osteolysis. Clarke et al [7] found no relationship between cytokine levels in synovial fluid and the degree of radiolucency around the implant that was graded based on the method of Harris et al [19]. Nivbrant et al [10] showed that higher levels of TNF-a were found in hips with bone resorption of type II and type III, compared with those with type I loosening, according to the classification by Gustilo and Pasternak. However, both of these studies were qualitative evaluations and, thus, probably less accurate when compared with our method. In our study, IL-8 levels in synovial fluid correlated
Serum IL-8 and Synovial Fluid Levels ! Tanaka et al
with the severity of osteolysis and, so, might be considered to be a more reliable parameter. Although it has been reported that Hernigou et al [5] accepted a correlation between serum IL-6 levels and the area of osteolysis by the same method as we used, the comparison between the cytokine levels in serum and those in synovial fluid was not made. In our study, there was no correlation between serum IL-6 and synovial fluid IL-6, indicating that serum IL-6 level reflects the local condition poorly. With regard to IL-8, Ishiguro et al [20] reported increased messenger RNA expression of IL-8 in periprosthetic granulomatous tissue when compared with fibrous scar tissue. Lassus et al [21] showed high count levels of IL-8–producing cells in both the synovial-like interface membrane and pseudocapsular tissue of aseptically loosened THAs when compared with controls. Analysis of synovial fluid from patients with a loose hip prosthesis has demonstrated that IL-8 levels in loose hip joint fluid were significantly higher than in controls [6,7,9]. Interleukin 8 contributes to the release of proteinases, PGE2, adhesion molecules, and superoxide molecules into joint fluid that may play an essential role in prosthesis loosening [9,21]. Our results show a significant rise in serum IL8 level in patients with aseptic loosening of hip prosthesis, and a significant correlation was found between the serum IL-8 level and the synovial fluid IL-8 level. This suggests that the serum IL-8 level could be a useful marker for periprosthetic osteolysis. However, there was a small number of patients in the OL() group who had an increased serum IL-8 level, and continued follow-up is needed to see whether periprosthetic osteolysis appears on plain radiograph in the near future. If osteolysis does appear, it will add more evidence to suggest the usefulness of the IL-8 level as a marker for periprosthetic osteolysis. Pharmaceutical interventions are currently being developed in an attempt to treat or prevent osteolysis, including the use of bisphosphonates [22], osteoprotegerin [23,24], or an anti–TNF-a antibody [25,26]. To monitor and predict the effect of the drugs on osteolysis, a marker for periprosthetic osteolysis will be indispensable for the accumulation of data required for their clinical use.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
References 16. 1. Chomarat P, Vannier E, Dechanet J, et al. Balance of IL-1Ra/IL-1b in rheumatoid synovium and its
1053
regulation by IL-4 and IL-10. J Immunol 1995; 154:1432. Nabae M, Inoue K, Ushiyama T, et al. Gene expression of anti-inflammatory mediators in THR retrieved interfacial membranes. Acta Orthop Scand 1999;70:149. Takei I, Takagi M, Ida H, et al. High macrophagecolony stimulating factor levels in synovial fluid of loose artificial hip joints. J Rheumatol 2000;27:894. Granchi D, Verri E, Ciapetti G, et al. Bone-resorbing cytokines in serum of patients with aseptic loosening of hip prosthesis. J Bone Joint Surg Br 1998; 80:912. Hernigou P, Intrator L, Bahrami T, et al. Interleukin6 in the blood of patients with total hip arthroplasty without loosening. Clin Orthop 1999;366:147. Sypniewska G, Lis K, Bilinski PJ. Bone turnover markes and cytokines in joint fluid. Acta Orthop Scand 2002;73:518. Clarke SA, Brooks RA, Hobby JL, et al. Correlation of synovial fluid cytokine levels with histological and clinical parameters of primary and revision total hip and total knee replacements. Acta Orthop Scand 2001;72:491. Inomoto M, Miyakawa S, Mishima H, et al. Elevated interleukin-12 in pseudosynovial fluid in patients with aseptic loosening of hip prosthesis. J Orthop Science 2000;5:369. Sabokbar A, Rushton N. Role of inflammatory mediators and adhesion molecules in the pathogenesis of aseptic loosening in total hip arthroplasties. J Arthroplasty 1995;10:810. Nivbrant B, Karlsson K, Karrholm J. Cytokine levels in synovial fluid from hips with well-functioning or loose prostheses. J Bone Joint Surg Br 1999;81:163. Goldring SR, Schiller AL, Roelke M, et al. 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 Am 1983; 65:575. Dorr LD, Bloebaum R, Emmanual J, et al. Histologic, biochemical, and ion analysis of tissue and fluids retrieved during total hip arthroplasty. Clin Orthop 1990;261:82. Goodman SB, Chin RC, Chiou SS, et al. A clinicalpathologic-biochemical study of the membrane surrounding loosened and nonloosened total hip arthroplasties. Clin Orthop 1989;244:182. Kim KJ, Chiba J, Rubash HE. In vitro and in vitro analysis of membranes from hip prosthesis inserted without cement. J Bone Joint Surg Am 1994; 76:172. Chiba J, Rubash H, Kim KJ, et al. The characterization of cytokines in the interface tissue obtained from failed cementless total hip arthroplasty with and without femoral osteolysis. Clin Orthop 1994; 300:304. Takakuwa M, Matsuno T, Gotoh E. The serum bone resorbing cytokines with osteolysis of total hip arthroplasty (in Japanese). Seikeigeka 2002;53:1709.
1054 The Journal of Arthroplasty Vol. 20 No. 8 December 2005 17. Kim KJ, Hijikata H, Itoh T, et al. Joint fluid from patients with failed total hip arthroplasty stimulates pit formation by mouse osteoclasts on dentin slices. J Biomed Mater Res 1998;43:234. 18. Schmalzried T, Jasty M, Harris W. Periprosthetic bone loss in total hip arthroplasty. Polyethylene wear debris and the concept of the effective joint space. J Bone Joint Surg Am 1992;74:849. 19. Harris WH, McCarthy Jr JC, O’Neill DA. Femoral component loosening using contemporary techniques of femoral cement fixation. J Bone Joint Surg Am 1982;64:1063. 20. Ishiguro N, Kojima T, Ito T, et al. Macrophage activation and migration in interface tissue around loosening total hip arthroplasty components. J Biomed Mater Res 1997;35:399. 21. Lassus J, Waris V, Xu JW, et al. Increased interleukin-8 (IL-8) expression is related to aseptic loosening of total hip replacement. Arch Orthop Trauma Surg 2000;120:328.
22. Millett PJ, Allen MJ, Bostorom MPG. Effects of alendronate on particle-induced osteolysis in a rat model. J Bone Joint Surg Am 2002;84:236. 23. Ulrich-Vinther M, Carmody EE, Goater JJ, et al. Recombinant adeno-associated virus-mediated osteoprotegerin gene therapy inhibits wear debrisinduced osteolysis. J Bone Joint Surg Am 2002;84: 1405. 24. Kim KJ, Kotake S, Udagawa N, et al. Osteoprotegerin inhibits in vitro mouse osteoclast formation induced by joint fluid from failed total hip arthroplasty. J Biomed Mater Res 2001;58:393. 25. Schwarz EM, Looney RJ, O’Keefe RJ. Anti–TNF-a therapy as a clinical intervention for periprosthetic osteolysis. Arthritis Res 2002;2:165. 26. Childs LM, Goater JJ, O’Keefe RJ, et al. Efficacy of etanercept for wear debris-induced osteolysis. J Bone Miner Res 2001;16:338.