Introduction Total hip arthroplasty (THA) is the most common reconstructive hip procedure in adult patients with an arthritic hip joint. Cementless THA has become pop- ular in many countries, showing very promising long-term stability (1,2). After THA, the femoral component bears a share of the body weight which was formerly carried by the femur alone and thus the normal level of weight-bearing stress within the proximal femur is reduced. The femur subsequently adapts to this phenomenon by proximal bone resorption. This phenomenon is so-called âstress shieldingâ (3). One of the major complications of Predictive Value of a Preoperative Biochemical Bone Marker in Relation to Bone Remodeling After Cementless Total Hip Arthroplasty Katsuyuki Yamaguchi, MD,*,1 Kensaku Masuhara, MD,2 Satoshi Yamasaki, MD,2 Tsuyoshi Nakai, MD,3 and Takeshi Fuji, MD2 1Department of Orthopaedic Surgery, Kaizuka City Hospital; 2Department of Orthopaedic Surgery, Osaka Koseinenkin Hospital; 3Department of Orthopaedic Surgery, Osaka Police Hospital, Osaka, Japan Abstract To identify the factors predicting proximal bone resorption of the femur after cementless total hip arthro- plasty (THA), we studied 24 postmenopausal women with osteoarthritis. Periprosthetic bone mineral density (BMD) measurements of the seven Gruen zones were determined with dual-energy X-ray absorptiometry at 3 wk, 6, 12, and 18 mo after operation. The greatest decrease in BMD (13.2%) was found in zone 7 at 18 mo. At 18 mo, preoperative serum bone alkaline phosphatase was associated univariately with BMD loss in zones 1 (r = 0.407, p = 0.048) and 7 (r = 0.543, p = 0.006) and urinary N-telopeptide cross-linked collagen type I (NTX) was associated univariately with that in zone 7 (r = 0.520, p = 0.009). Patient age correlated with BMD loss in zone 7 (r = 0.425, p = 0.039). Multiple regression analysis identified a significant relationship between the BMD loss and patient age and NTX in zone 7 at 18 mo (r2 = 0.422, p = 0.001). We conclude that preop- erative bone markers are significant predictors of bone remodeling after THA with the particular implants used in our study. Key Words: Bone mineral density; cementless total hip arthroplasty; DXA; markers; bone alkaline phos- phatase; N-telopeptide cross-linked collagen type I. Received 02/03/03; Revised 04/30/03; Accepted 04/30/03. *Address correspondence to Dr. Katsuyuki Yamaguchi, Department of Orthopaedic Surgery, Kaizuka City Hospital, 3-10-20, Hori, Kaizuka, Osaka 597-0015, Japan. E-mail:
[email protected] 259 Original Article Journal of Clinical Densitometry, vol. 6, no. 3, 259â265, 2003 © Copyright 2003 by Humana Press Inc. All rights of any nature whatsoever reserved. 1094-6950/03/6:259â265/$25.00 cementless THAs is proximal femoral bone resorp- tion due to stress shielding (4â6). Extensive resorp- tion may weaken the bone bed and promote bone, implant, or fixation failure. It is thus important to be able to predict bone loss in proximal periprosthetic femurs after cementless THA. There are several studies that have examined the factors influential on femoral bone remodeling after cementless THA. Nishii et al. (7) reported that stem size and the initial bone density in the distal portion around the stem were important factors to predict the proximal bone density decrease after implanting a fully coated Lübeck I stem (ESKA, Lübeck, Germany). Kiratli et al. (8) reported that body weight was the only variable that influences bone remodel- ing in patients receiving the Wisconsin Modular Femoral Component (Biomet, Warsaw, IN, USA). Venesmaa et al. (9) reported that the magnitude of the absolute BMD loss was greatest in subjects with the lowest preoperative BMD, but inconsistent relation- ships were found between age, stem size, and BMD change after implantation of one-third proximal porous-coated bimetric titanium alloy (Biomet, Inc., Warsaw, IN, USA). However, the predictive value of a preoperative biochemical bone marker in relation to bone remodeling after cementless THA by using dual-energy X-ray absorptiometry (DXA) has not been reported to date, to our knowledge. The purpose of this study was to examine the potential utility of a bone marker to identify periprosthetic BMD loss after cementless THA. Materials and Methods Subjects Twenty-four hips in 24 postmenopausal women who had undergone cementless THA were selected for this study (Table 1). The following patients were excluded from this trial: those with metabolic changes causing a reduction of BMD such as malig- nant disease, primary hyperthyroidism, active thy- roid disease, Pagetâs disease of the bones, and osteomalacia; those who had previously taken drugs affecting bone metabolism, such as steroids; those who had been receiving osteoporosis therapies, including calcitonin, bisphosphonates, or estrogen replacement therapy. Informed consent was obtained from all patients. Surgery The prosthesis used was the Spongiosa Metal® II Hip Prosthesis (S&G; ESKA, Lübeck, Germany), an improvement of the Lübeck I prosthesis (10â12). The femoral stem is made of a cobalt-chrome- molybdenum alloy with a spongiosa metal structure (Fig. 1). The implants used were all proximally coated stems with a porous surface area applied cir- cumferentially to the proximal 60%. The stem sizes used for the patients ranged from 100 mm to 260 Yamaguchi et al. Journal of Clinical Densitometry Volume 6, 2003 Table 1 Characteristics of the Study Population Variable Mean ± SD Range Age (yr) 61 ± 8 51â81 Weight (kg) 53.6 ± 7.8 35â62 Serum BAP (U/l) 28.5 ± 8.9 15.8â55.6 Urinary NTX 74.4 ± 33.0 34.0â152.5 (n mol/m mol Cr) Femoral stem size n 100 mm 11 110 mm 12 120 mm 1 Fig. 1. Photograph showing Spongiosa Metal® II Hip Prosthesis. 120 mm. The surgery was performed using stan- dardized techniques by a single hip surgeon. Partial weight-bearing was started at 1 wk postoperatively, and full weight-bearing was allowed at 3 wk. Laboratory Measurements For each woman, serum bone alkaline phos- phatase (BAP, enzyme immunoassay; SRL, Inc., Osaka, Japan) and urinary N-telopeptide cross- linked collagen type I (NTX, enzyme-linked immunosorbent assay; SRL, Inc., Osaka, Japan) were measured preoperatively. Blood samples were collected between 8 AM and 9 PM after an overnight fast. Second morning void urine samples were col- lected and urine creatinine was measured at the time of sampling. DXA Measurements Postoperative BMD measurements were per- formed at 3 wk after surgery, and subsequently at 6, 12, and 18 mo by DXA (QDR-2000; Hologic Inc., Waltham, MA, USA). The software (pros- thetic scanning software version 5.73; Hologic Inc., Waltham, MA, USA) was designed to mea- sure the periprosthetic BMD in 7 regions of inter- est based on the zones of Gruen et al. (13) (Fig. 2). The zones were positioned based on the level of the medial neck resection and the distal tip of the stem, dividing the femur into 3 equal zones. Zone 4 was defined just distal to the stem tip with equal height to the other zones. In our institution, the precision error of the calculation by DXA ranged from 1% in zone 6 to 3.8% in zone 7 for peripros- thetic zones, with an overall precision for Gruen zones of 2.4%. Statistical Analysis It is well known that periprosthetic BMD decreases in all regions of interest based on Gruen zones. Periprosthetic BMD may be affected by stress shielding as well as external factors such as immediate postoperative disuse atrophy and subse- quent physical activity (9). Zone 4, representing the diaphyseal area below the stem, also showed post- operative BMD decreases. These decreases are probably partly due to reduced activity in the early period after surgery. Furthermore, weight-bearing stress is transferred to the distal femur through the stem and the femur remodels adjacent the stem by subsequent proximal bone resorption. Huiskes (3) demonstrated progressive decrease of stress shield- ing from proximal to distal in the periprosthetic femoral bone and the normal stress below the tip of the stem in finite-element methods. The degree of bone density reduction in the proximal area as com- pared with the distal area seems to be important to consider the genuine effect of stress shielding alone. Therefore, we selected zone 4 to serve as a refer- ence baseline region to evaluate stress shielding for each patient. The R value (%) is the ratio calculated by dividing the BMD in each zone by that in zone 4 on the same scan during the follow-up period and multiplying by 100. The stress shielding value (SSV) in each zone is estimated as SSV = R at 3 wkâR in the follow-up period. The Wilcoxon test was used to compare the R value in the follow-up Predictive Value of Preoperative Bone Markers 261 Journal of Clinical Densitometry Volume 6, 2003 Fig. 2. Schematic drawings of the seven regions of interest based on Gruen zones. period with that 3 wk postoperatively in each zone. P values less than 0.05 were considered statistically significant. In zones where R values changed signif- icantly from 3 wk to 18 mo, we determined univari- ate relationships between the SSV at the final follow-up (18 mo) in those zones and patient age, weight, stem size, serum BAP, and urinary NTX with the Spearman correlation test. Variables that were significant at p < 0.05 in the univariate analy- sis were then included in a forwardâbackward step- wise regression model to identify significant predictors of stress shielding 18 mo after surgery. All statistical analyses were performed using SPSS (version 11.0 J; SPSS Japan Inc., Tokyo, Japan). Results The main characteristics of predictor variables are shown in Table 1. There was little change in BMD in zones 3, 4, and 5, which represented the distal portion of the stem (Table 2). R values decreased significantly at 18 mo in zones 1, 6, and 7 compared with the base- line at 3 wk after surgery (Table 3). During 18 mo of follow-up, there were decreases of 4.5% and 4.8% in BMD in zones 1 and 6, respectively (Table 4). The greatest loss of BMD (13.2%) was found in zone 7 (Table 4). In the proximal and middle areas around the stem (zones 1, 2, 6, and 7), BMD decreased rapidly during the first 6 mo and little further change was 262 Yamaguchi et al. Journal of Clinical Densitometry Volume 6, 2003 Table 2 Raw Data of Periprosthetic Bone Mineral Density (g/cm2) Postoperative period 3 wk 6 mo 12 mo 18 mo Zone 1 0.701 ± 0.101 0.609 ± 0.122 0.607 ± 0.125 0.617 ± 0.117 Zone 2 1.094 ± 0.172 1.018 ± 0.163 1.014 ± 0.167 1.029 ± 0.179 Zone 3 1.197 ± 0.170 1.163 ± 0.157 1.175 ± 0.164 1.178 ± 0.167 Zone 4 1.316 ± 0.168 1.285 ± 0.167 1.269 ± 0.162 1.261 ± 0.167 Zone 5 1.190 ± 0.196 1.169 ± 0.186 1.156 ± 0.190 1.165 ± 0.180 Zone 6 1.123 ± 0.193 1.036 ± 0.199 1.020 ± 0.168 1.021 ± 0.161 Zone 7 0.796 ± 0.156 0.636 ± 0.193 0.612 ± 0.162 0.605 ± 0.167 The values are expressed as the average bone mineral density (and standard deviation). Table 3 The R Value (%) in Each Zone During 18 Mo Postoperative period 3 wk 6 mo 12 mo 18 mo Zone 1 53.6 ± 7.4 47.6 ± 8.3** 48.0 ± 8.2** 49.1 ± 7.9** Zone 2 84.0 ± 13.2 79.7 ± 10.4** 80.3 ± 11.4* 82.5 ± 13.2 Zone 3 91.9 ± 14.4 91.2 ± 11.4 93.1 ± 10.6 94.4 ± 13.8 Zone 4 100 ± 0 100 ± 0 100 ± 0 100 ± 0 Zone 5 91.0 ± 15.1 91.8 ± 14.6 91.8 ± 14.3 93.4 ± 15.4 Zone 6 86.3 ± 15.6 81.5 ± 15.6** 81.1 ± 13.2** 81.5 ± 11.5** Zone 7 60.9 ± 11.2 49.3 ± 12.0** 48.2 ± 10.3** 47.8 ± 10.5** The values are expressed as the average percentage (and standard deviation). * p < 0.05; ** p < 0.01; differs from 3 wk (Wilcoxon test). observed thereafter (Table 2). In SSV in zone 6 at 18 mo, no significant predictor was found in univariate analysis (Table 5). SSV in zone 1 at 18 mo was asso- ciated univariately with serum BAP (r = 0.407 and p = 0.048). SSV in zone 7 at 18 mo was associated with patient age (r = 0.425 and p = 0.039), serum BAP (r = 0.543 and p = 0.006), and urinary NTX (r = 0.520 and p = 0.009). In a multiple regression analysis, patient age and urinary NTX were signifi- cant predictors of stress shielding in zone 7 (r2 = 0.422 and p = 0.001) (Table 6). No significant rela- tionship existed between patient age and preopera- tive BAP and NTX (r = 0.233, p = 0.272; and r = 0.159, p = 0.458, respectively). Discussion The potential utility of preoperative biochemical markers to precisely predict periprosthetic bone loss after THA has not been reported to date, to our knowledge. It is generally known that biochemical markers of bone turnover may be used to identify fast bone losers. Garnero et al. (14) reported that increased levels of the biochemical markers of bone turnover including urinary NTX were significantly associated with forearm bone loss over 4 yr in a cohort of 305 postmenopausal women. Qureshi et al. (15) demonstrated that the changes in bone resorption markers in the early postoperative months after THA were correlated with peripros- thetic bone loss at 3 yr in radiographic assessment. In their study, the increased release of pyridinoline and deoxypyridinoline at 3 and 6 mo from the pre- operative baseline was associated with the severity of radiographically assessed bone loss at 36 mo in the cemented cobalt alloy group (0.749 < r < 0.840, p < 0.05). Furthermore, in their cementless titanium alloy group, increased osteocalcin/deoxypyridino- line ratios at 3 and 6 mo were related inversely to radiographic bone loss at 36 mo (0.687 < r < 0.749, p < 0.05). Our study demonstrated that preoperative biochemical bone markers might predict peripros- thetic bone loss after cementless THA. Schneider et al. (16) reported that bone formation markers were slightly elevated after 3 mo and bone resorption markers were elevated during the first 6 mo after cemented and uncemented THA. Therefore, turnover may remain elevated in the immediate post- operative phase in women with high preoperative turnover. We did not show longitudinal assessment of biochemical bone markers after operation in the âResultsâ section. However, 10 patients underwent Predictive Value of Preoperative Bone Markers 263 Journal of Clinical Densitometry Volume 6, 2003 Table 4 The Mean SSV (%) in Each Zone Where R Values Changed Significantly From Baseline to the Final Follow-Up Postoperative period 6 mo 12 mo 18 mo Zone 1 6.0 ± 6.6 5.7 ± 6.7 4.5 ± 5.1 Zone 6 4.8 ± 7.2 5.2 ± 7.5 4.8 ± 8.2 Zone 7 11.6 ± 8.3 12.7 ± 9.3 13.2 ± 9.0 Table 5 Multiple Correlation Analysisa Between the Variables and SSV in Zones 1, 6, and 7 at 18 Mo SSV1 SSV6 SSV7 Age 0.230b 0.062 0.425* (0.279)c (0.772) (0.039) Weight â0.326 0.105 0.157 (0.120) (0.627) (0.465) Stem size â0.124 0.128 0.229 (0.565) (0.553) (0.281) BAP 0.407* 0.140 0.543** (0.048) (0.515) (0.006) NTX 0.296 0.265 0.520** (0.160) (0.211) (0.009) a Multiple correlation analysis were performed with the Spearman correlation test. b Correlation coefficient. c Significant probability * p < 0.05, ** p < 0.01. Table 6 Predictors of SSV in Zone 7 at 18 Mo Variables Estimate SE p Value SSV7 (full model, adjusted r2 = 0.422, p = 0.001) Intercept â31.532 11.618 0.013 Age 0.580 0.182 0.005 NTX 0.123 0.043 0.010 laboratory examinations preoperatively and at 6 and 12 mo postoperatively. Mean serum BAP within this group was 28.6 ± 6.1, 32.7 ± 9.2, and 30.5 ± 9.5 U/L preoperatively, at 6 mo, and at 12 mo, respectively. Mean urinary NTX was 65.4 ± 14.4, 72.5 ± 13.5, and 59.2 ± 15.1 nmol/m mol Cr preoperatively, at 6 mo, and at 12 mo, respectively. Both biochemical bone markers were slightly elevated at 6 mo after opera- tion. Previous studies have confirmed that maximal bone loss is seen in the first 6 mo after THA (8,17). The value of bone resorption markers at this early postoperative period seems to be important to periprosthetic BMD loss. Kröger et al. (18) demon- strated that increased periprosthetic bone activity at 6 mo was associated with the decrease in BMD (6â12 mo) in the lesser trochanter region (r = â0.589) alone by using single emission computed tomography. Interestingly, this study confirms that preoperative biochemical bone markers such as serum BAP and urinary NTX may be important predictors of stress shielding in zone 7, including the lesser trochanter. This study has several limitations. First, these results apply only to postmenopausal women and the population assessed is quite small (n = 24). In Japan, the incidence of osteoarthritis secondary to acetabular dysplasia is much higher than that of primary osteoarthritis and most patients with osteo- arthritis who had undergone THA were female patients at our hospital. Second, the results of this study may be specific to the particular implants that we used. The differences in implant material and design would have some effect on bone remodeling. Third, we could find no significant association between the stem size and bone remodeling. Engh et al. (19) reported that the use of large stems resulted in the increased occurrence of marked bone resorption in their roentogenographic analysis of 411 cases of primary cementless THA. Increased stem size is related to increased bending stem stiff- ness. The weight bearing load is transmitted between the stem and the femur in proportion to the respective stiffness. Consequently, a stiffer stem carries a greater portion of the load, thus including stress protection in the proximal femur and proxi- mal bone resorption (20). All the stems used in this study were relatively uniform and had relatively small diameters and less stiffness. Therefore, a study with larger stems might reveal some signifi- cant effects of stem size on stress shielding. Fourth, the follow-up was too short, although it has been reported that the most significant reduction of periprosthetic BMD occurs within 6 mo after THA. In our previous study of the Lübeck I prosthe- sis, BMD decreased rapidly during the first 3 mo in all zones and then a slow decline in BMD appeared to occur (21). Finally, there are at best modest (but significant) predictive correlations between biochemical bone markers and SSV in zone 7 at 18 mo. The r2 calcu- lations of the correlations indicate that only about 25â30% of the variance of periprosthetic BMD change in zone 7 was due to preoperative baseline biochemical bone markers. We conclude that increased levels of biochemical markers of bone turnover seem to be predictive of periprosthetic bone loss after cementless THA. Venesmaa et al. (22) reported that alendronate ther- apy led to a significant reduction in periprosthetic bone loss 6 mo after cementless THA. If the patient is at risk for developing severe periprosthetic bone resorption after operation, it seems useful to begin osteoactive drugs preoperatively or soon after surgery as maximal bone loss occurs within 6 mo postoperatively. Acknowledgments The authors wish to thank Kazuhiko Nukui for his assistance in statistical analyses. This study is based on data collected from Osaka Koseinenkin Hospital. References 1. Engh CA Jr, Culpepper WJII, Engh CA. 1997 Long-term results of use of the anatomic medullary locking prosthesis in total hip arthroplasty. J Bone Joint Surg 79A:177â184. 2. Xenos JS, Callaghan JJ, Heekin RD, Hopkinson WJ, Savory CG, Moore MS. 1999 The porous-coated anatomic total hip prosthesis, inserted without cement. A prospective study with a minimum of ten years of follow-up. J Bone Joint Surg 81A:74â82. 3. Huiskes R. 1993 Stress shielding and bone resorption in THA: clinical versus computer-simulation studies. Acta Orthop Belg 59:118â129. 4. Bobyn JD, Mortimer ES, Glassman AH, Engh CA, Miller JE, Brooks CE. 1992 Producing and avoiding stress shield- ing. laboratory and clinical observations of noncemented total hip arthroplasty. Clin Orthop 274:79â96. 264 Yamaguchi et al. Journal of Clinical Densitometry Volume 6, 2003 5. Lord G, Marotte JH, Blanchard JP, et al. 1988 Cementless Madreporic and polarised total hip prostheses. a ten-year review of 2688 cases. French J Orthop Surg 2:82â92. 6. Maloney WJ, Sychterz C, Bragdon C, et al. 1996 Skeletal response to well fixed femoral components inserted with and without cement. Clin Orthop 333:15â26. 7. Nishii T, Sugano N, Masuhara K, Shibuya T, Ochi T, Tamura S. 1997 Longitudinal evaluation of time related bone remodeling after cementless total hip arthroplasty. Clin Orthop 339:121â131. 8. Kiratli BJ, Checovich MM, McBeath AA, Wilson MA, Heiner JP. 1996 Measurement of bone mineral density by dual-energy X-ray absorptiometry in patients with the Wisconsin hip, an uncemented femoral stem. J Arthroplasty 11:184â193. 9. Venesmaa PK, Kröger HPJ, Miettinen HJA, Jurvelin JS, Suomalainen OT, Alhava EM. 2001 Monitoring of peripros- thetic BMD after uncemented total hip arthroplasty with dual-energy X-ray absorptiometryâa 3-year follow-up study. J Bone Miner Res 16:1056â1061. 10. Plötz W, Gradinger R, Rechl H, Ascherl R, Wicke-Wittenius S, Hipp E. 1992 Cementless prosthesis of the hip joint with âspongy metalâ surface. a prospective study. Arch Orthop Trauma Surg 111:102â109. 11. Sugano N, Saito S, Takaoka K, et al. 1994 Spongy metal Lübeck hip prostheses for osteoarthritis secondary to hip dysplasia. a 2â6 year follow-up study. J Arthroplasty 9:253â262. 12. Matsui M, Nakata K, Masuhara K, Ohzono K, Sugano N, Ochi T. 1998 The Metal-Cancellous Cementless Lübeck total hip arthroplasty. five-to-nine-year results. J Bone Joint Surg 80B:404â410. 13. Gruen TA, McNeice GM, Amstutz HC. 1979 âModes of failureâ of cemented stem-type femoral components. a radi- ographic analysis of loosening. Clin Orthop 141:17â27. 14. Garnero P, Sornay-Rendu E, Duboeuf F, Delmas PD. 1999 Markers of bone turnover predict postmenopausal forearm bone loss over 4 years: the OFELY study. J Bone Miner Res 14:1614â1621. 15. Qureshi AA, Virdi AS, Didonna ML, et al. 2002 Implant design affects markers of bone resorption and formation in total hip replacement. J Bone Miner Res 17:800â807. 16. Schneider U, Schmidt-Rohlfing B, Knopf U, Breusch SJ. 2002 Effects upon bone metabolism following total hip and total knee arthroplasty. Pathobiology 70:26â33. 17. Massari L, Bagni B, Biscione R, Traina GC. 1996 Periprosthetic bone density in uncemented femoral hip implants with proximal hydroxyapatite coating. Bull Hosp Jt Dis 54:206â210. 18. Kröger H, Vanninen E, Overmyer M, Miettinen H, Rushton N, Suomalainen O. 1997 Periprosthetic bone loss and regional bone turnover in uncemented total hip arthroplasty: a prospective study using high resolution single photon emission tomography and dual-energy X-ray absoptiometry. J Bone Miner Res 12:487â492. 19. Engh CA, Bobyn JD. 1988 The influence of stem size and extent of porous coating on femoral bone resorption after primary cementless hip arthroplasty. Clin Orthop 231:7â28. 20. Engh CA, Bobyn JD, Glassman AH. 1987 Porous-coated hip replacement. the factors governing bone ingrowth, stress shielding, and clinical results. J Bone Joint Surg 69B:45â55. 21. Yamaguchi K, Masuhara K, Ohzono K, Sugano N, Nishii T, Ochi T. 2000 Evaluation of periprosthetic bone-remodeling after cementless total hip arthroplasty: the influence of the extent of porous coating. J Bone Joint Surg 82A:1426â1431. 22. Venesmaa PK, Kröger HPJ, Miettinen HJA, Jurvelin JS, Suomalainen OT, Alhava EM. 2001 Alendronate reduces periprosthetic bone loss after uncemented primary total hip arthroplasty: a prospective randomized study. J Bone Miner Res 16:2126â2131. Predictive Value of Preoperative Bone Markers 265 Journal of Clinical Densitometry Volume 6, 2003