Implant Design Affects Markers of Bone ... - Wiley Online Library

JOURNAL OF BONE AND MINERAL RESEARCH Volume 17, Number 5, 2002 © 2002 American Society for Bone and Mineral Research

Implant Design Affects Markers of Bone Resorption and Formation in Total Hip Replacement* ABID A. QURESHI,1 AMARJIT S. VIRDI,1,2 MICHAEL L. DIDONNA,1 JOSHUA J. JACOBS,1 KOICHI MASUDA,3 WAYNE P. PAPROSKY,1 EUGENE J.M.A. THONAR,1,3,4 and DALE R. SUMNER1,2 ABSTRACT Concentrations of the bone resorption markers pyridinoline and deoxypyridinoline and the bone formation marker osteocalcin were measured in 24-h urine collections from 30 subjects who underwent unilateral total hip replacements for monoarticular symptomatic osteoarthrosis and 10 controls. The patient groups were divided based on the femoral implant type (cemented cobalt alloy stem, cementless porous coated cobalt alloy stem, and cementless porous coated titanium alloy stem). Urine collections were performed before surgery and then at 3, 6, 12, 24, and 36 months. There were significant changes over time in the three patient groups for pyridinoline, deoxypyridinoline, and the ratio of osteocalcin to deoxypyridinoline (p < 0.01), but the control group values did not change over time. The resorption markers tended to peak at 3 months and the osteocalcin to deoxypyridinoline ratio was more variable, having depressed values in the cementless cobalt alloy group and elevated values in the other two groups compared with baseline. The cementless cobalt alloy group had higher resorption marker levels than the cemented cobalt alloy group at 6, 12, 24, and 36 months and higher levels than the cementless titanium alloy group at all postoperative times (p < 0.05). The osteocalcin to deoxypyridinoline ratio was lower in the cementless cobalt alloy group than in the cemented cobalt alloy group at 3, 6, and 24 months and the cementless titanium alloy group at 6, 12, and 24 months (p < 0.05). For the cemented cobalt chrome group, the baseline-normalized resorption marker values at 3 months and 6 months were correlated with the severity of radiographically assessed bone loss at 36 months (0.749 < r < 0.840; p < 0.05). For the cementless titanium alloy group, baseline-normalized osteocalcin/ deoxypyridinoline ratios at 3 months and 6 months were related inversely to radiographic bone loss at 36 months (0.687 < r < 0.749; p < 0.05). Thus, body fluid markers of bone metabolism change after total hip replacement. In addition, the changes in the marker concentrations were sensitive to implant design and were correlated with subsequent stress-shielding–induced bone loss. (J Bone Miner Res 2002;17:800 – 807) Key words:

markers, pyridinoline, deoxypyridinoline, osteocalcin, total hip replacement

INTRODUCTION he noninvasive assessment of bone turnover using biochemical markers has significantly improved in recent years.(1–3) These markers assess either the activities of os-

T

*Presented in part at the annual meeting of the Orthopedic Research Society, 1999, Anaheim, California. Dr. Jacobs has financial interests in the forms of serving as a consultant and receiving research support from Zimmer, Inc., Wright Medical, and Merck. Dr. Sumner has financial interest in the form of research support from Zimmer, Inc. All other authors have no conflict of interest.

teoclasts and osteoblasts or the breakdown products of collagen. Various markers of bone resorption include urinary calcium and hydroxyproline, urinary hydroxylysine, plasma tartrate-resistant acid phosphatase, and the more recent urinary and serum pyridinium cross-links and associated telopeptides. Presently, pyridinium cross-links and associated telopeptides are the most sensitive and specific markers of bone resorption.(2,4) Pyridinoline and deoxypyridinoline, also called hydroxylysylpyridinoline and lysylpyridinoline, respectively, are two nonreducible pyridinium cross-links that stabilize the collagen chains within the extracellular ma-

1

Department of Orthopedic Surgery, Rush Medical College, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL, USA. Department of Anatomy, Rush Medical College, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL, USA. 3 Department of Biochemistry, Rush Medical College, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL, USA. 4 Section of Rheumatology, Department of Medicine, Rush Medical College, Rush-Presbyterian-St. Luke’s Medical Center, Chicago, IL, USA. 2

800

URINARY MARKERS OF BONE TURNOVER AFTER THR

trix.(5,6) Pyridinoline, the most abundant of the cross-links, is found at high concentrations in bone and cartilage and at various levels in many other connective tissues. In contrast, deoxypyridinoline is found primarily in calcified tissue and is an excellent marker of the rate of bone resorption.(4) These cross-links are present in a minute amount in the tendon and aorta but absent from skin, an abundant source of type I collagen.(4) Urinary cross-link excretion, especially deoxypyridinoline, correlates with bone turnover in patients with vertebral osteoporosis(7,8) and fracture risk.(9,10) These markers are now being used to assess efficacy of treatment for metabolic bone diseases(11–15) and fracture risk.(9,10,16) Osteocalcin is a low molecular weight protein produced by active osteoblasts during bone formation. It is incorporated primarily into the mineralizing matrix, but a small amount is released into circulation. Its concentration is considered a sensitive and specific marker of osteoblastic activity and levels of osteocalcin are an accepted index in a variety of metabolic bone disorders and in osteoporosis management.(1) Loss of bone adjacent to the femoral component of total hip replacement (THR) is thought to be a major factor in determining the longevity of the joint reconstruction. These changes have been studied radiographically, most precisely with dual-energy X-ray absorptiometry.(17) These studies have yielded quantitative data on net changes in bone mass adjacent to the implant but are not informative about bone turnover. Recently, researchers have attempted to identify periprosthetic changes in bone turnover using various biochemical markers.(18 –21) In general, these studies lack longterm prospective data, provide little information on asymptomatic well-functioning implants, and have not assessed the potential role of implant design on bone turnover. The purpose of this prospective study was to determine if osteocalcin and pyridinium cross-link excretion in urine changed after THR in patients with well-functioning implants. Because such changes were observed, further analyses were performed to determine if the changes were sensitive to implant design, when and if a new steady state was reached, and if the changes correlated with long-term periprosthetic bone loss. In particular, because both theoretical and experimental studies(22–27) suggest that greater net femoral periprosthetic bone loss occurs as the stiffness of the femoral component increases, we included patients with THRs in which this design element varied.

MATERIALS AND METHODS Patient characteristics and groups Forty subjects, in three patient groups and one control group were included in this prospective study. Each group had 5 male and 5 female subjects. In the patient groups, all patients had symptomatic monoarticular osteoarthrosis for which they underwent unilateral THRs. The control consisted of age- and sex-matched subjects with no implants or systemic illness. In many of the cases, the control subjects were spouses of patients in the study. The protocol was approved by an Internal Review Board at our institution. Patient groups were based on the femoral implant type. Group 1 patients underwent “hybrid” primary THR with a

801

cemented modular cobalt alloy femoral component (5 Precoat, 5 Iowa; Zimmer, Warsaw, IN, USA) and a porous coated cementless titanium acetabular component (HG; Zimmer). Group 2 patients underwent cementless THR with a modular fully porous coated cobalt alloy tapered femoral component (5 AML, 5 Solution; Depuy, Warsaw, IN, USA) and a porous coated cementless titanium acetabular component (300 Series Cup; Depuy). Group 3 patients underwent cementless THR with a proximally porous coated titanium alloy femoral component (Anatomic; Zimmer) and a porous coated cementless titanium acetabular component (HG; Zimmer). Based on femoral stem geometry and material properties, one would expect that the group 2 femoral stems were the most stiff, followed by groups 3 and 1.(24,25,27) All patients in the study had clinically well-functioning implants. None of the subjects had metabolic bone disease or metallic implants (except, of course, for the hip replacement implants in the patient groups). The mean ages and ranges in group 1 (mean, 65 years; range, 54 –77 years), group 2 (mean, 64 years; range, 52– 82 years), and the control subjects in group 4 (mean, 60 years; range, 48 –73 years) were similar and the patients in group 3 were younger (mean, 52 years; range, 38 – 62 years) than patients in group 1 (p ⫽ 0.010) and group 2 (p ⫽ 0.015).

Urine collection and analysis Twenty-four– hour urine samples were collected as previously described,(28) at time 0 (preoperative in the patient groups), 3, 6, 12, 24 and 36 months postoperatively. These samples were stored at ⫺80°C. Urinary pyridinoline, deoxypyridinoline, and osteocalcin concentrations were determined for each interval as described in the following paragraphs and corrected for urinary creatinine excretion. The collagen-specific cross-links pyridinoline and deoxypyridinoline were quantified using fluorescent detection of the cross-linking amino acids after high-performance liquid chromatography (HPLC). The samples were hydrolyzed in 6N HCl for 24 h at 120°C. The samples were applied to fractionation columns (CF-1; Whatman, Clifton, NJ, USA); 1 ml of sample was mixed with the mobile phase (acetic acid, water, and 1-n-butanol [1:1:4]) and applied to the column. After three full column washes with mobile phase, the cross-linking amino acids were eluted from the column with distilled water and dried. The samples were then redissolved in 1% n-heptafluorobutyric acid (HBFA) and separated by HPLC as described by Uebelhart et al.(29) The reversedphase column was equilibrated in 95% eluant A (0.01 M of HBFA in 12% [vol/vol] acetonitrile) and 5% eluant B (0.01 M of HBFA in 100% acetonitrile). The samples were eluted with 5–12% B gradient at a flow rate of 1 ml/minute for 20 minutes. The column was stripped with 100% eluant B for 5 minutes and reequilibrated with 95% eluant A for 5 minutes. Fluorescence of the eluted peaks was monitored using a spectrofluorometer (excitation, 297 nm; emission, 395 nm). Under these chromatographic conditions, the peaks for pyridinoline and deoxypyridinoline were eluted at 15 minutes and 17 minutes, respectively. The concentrations of cross-links reported are equivalents of external standards injected every 10 samples. All samples were run in duplicate from acid hydrolysis to HPLC. Final concen-

802

QURESHI ET AL.

trations of both pyridinoline and deoxypyridinoline were corrected for the 24-h urinary creatinine excretion and expressed as picomoles per micromoles. Urine osteocalcin concentrations were determined by an immunoradiometric assay kit (Immutopics, San Clemente, CA, USA) with slight modifications. The kit is designed for specimens of serum or plasma and was adapted for osteocalcin determination in urine by consultation with the supplier. A volume of 200 ␮l of the centrifuged urine was used per tube and all samples and standards were run in triplicate. Final concentrations were corrected for the 24-h urinary creatinine excretion. The intra-assay precision and interassay precision, as reported by the manufacturer, varies from 3.9 to 5.2% and from 5.5 to 6.7%, respectively. In our hands, the intra-assay precision for serum is 9% and for urine is 12%. Correlation between serum and urine levels of osteocalcin was established by testing paired samples from 10 individuals (r ⫽ 0.857; p ⫽ 0.002).

Radiographic evaluation Radiographic evaluation was performed by reviewing preoperative and postoperative radiographs of the proximal femur. The X-rays were taken with the patient in the supine position. For the present analyses, we used anteroposterior femur and cross-table or frog-leg lateral views. The standard distance between the X-ray tube and cassette was 40 in for the anteroposterior femur and frog-leg lateral views and 30 in for the cross-table lateral view. However, the distance between the bone and the cassette varied based on the thickness of the soft tissue for each patient. The standard kilovolt was ⬃75– 80 and the standard milliampere was 25–35. The X-rays were reviewed independently by two readers and then the results were compared and a consensus was reached regarding the three parameters studied: bone type, implant stability, and degree of stress-shielding– related bone loss. All of the radiographic analyses were performed with the readers blinded to the marker levels. Qualitative assessment of bone type was performed on preoperative radiographs, with the femurs being classified as either type A, B, or C, as proposed by Dorr and coworkers.(30) Briefly, type A is defined as femurs in which “thick cortices are seen on the anteroposterior radiograph and a large posterior cortex is seen on the lateral view,” type B as femurs that exhibit “bone loss from the medial and especially posterior cortices,” and type C as femurs that have “virtually lost the medial and posterior cortices.” Cementless stem fixation was categorized as either stable (through bone ingrowth or fibrous ingrowth) or unstable using the criteria described by Engh and coworkers.(31) The stability of the cemented stems was assessed with the radiographic criteria described by Harris and coworkers.(32) For these analyses, the 6 week, 6 month, 12 month, and 36 month radiographs were used. Stress-shielding–related bone loss was examined also by comparing changes in cortical thickness and density between the 6-week postoperative radiograph and the 36-month radiograph. The degree of stress shielding was determined by dividing the femur into four levels on the anteroposterior and lateral radiographs. Each level was divided then into medial, lateral, anterior, and posterior sites, producing a total of 16 sites for exam-

FIG. 1. Radiographs from a patient that received a fully porous coated cobalt alloy tapered femoral component. Comparison of the 6-week and 3-year radiographs was scored as showing “severe” stressshielding–related bone loss (i.e., at least 8 of the 16 periprosthetic zones had evidence of bone loss). (A) Anteroposterior femoral radiograph at 6 weeks; (B) anteroposterior femoral radiograph at 36 months; (C) frog-leg lateral radiograph at 6 weeks; (D) frog-leg lateral radiograph at 36 months.

ination of bone resorption.(33) The criterion for resorption in each site was whether the bone appeared either darker, thinner, or osteopenic when compared with the initial postoperative radiograph.(33) Radiographs were classified as either having no stress shielding (no change in any of the sites except for rounding of the calcar), mild (one to four sites), moderate stress shielding (five to seven sites), or severe stress shielding (eight or more sites). An example of a case with severe stress-shielding bone loss is shown in Fig. 1.

Statistical analyses The ␹2 statistic was used to determine if bone type, implant stability classification, or stress-shielding–related bone loss differed among the groups. A one-way analysis of variance (ANOVA) with Bonferroni post hoc tests was used to compare subject age at entry into the study among the groups. The cross-link and osteocalcin data were analyzed with nonparametric tests. Specifically, Friedman tests were used to assess the effect of time on marker concentrations and Kruskall-Wallis and Mann-Whitney tests were used to analyze the effect of femoral component design (group). Rank order correlation analysis was used to determine if change in marker values from baseline were correlated with the severity of stress-shielding–related bone loss at 36 months. In all cases, the exact level of significance is given and the differences were considered to be significant when p ⬍ 0.05.


TABLE 1. DISTRIBUTION

(30)

803 (31)

OF BONE TYPE, IMPLANT FIXATION, AND STRESS-SHIELDING–INDUCED BONE LOSS(33) AMONG THE THREE PATIENT GROUPS

Patient group 1 (cemented cobalt alloy)

2 (cementless cobalt alloy)

3 (cementless titanium alloy)

Bone typea

A B C

10 0 0

9 1 0

8 1 0

Implant fixation

Stable—bone ingrowth Stable—fibrous ingrowth Unstable

10b 0 0

10 0 0

8 2 0

Stress-shielding–induced bone loss

None Mild Moderate Severe

2 5 2 1

0 5 3 2

0 6 3 1

p Value

0.566

0.136

0.560 The p values for the ␹ tests are given. a Type A—“thick cortices are seen on the anteroposterior radiograph and a large posterior cortex is seen on the lateral view.” Type B femurs exhibit “bone loss from the medial and especially posterior cortices.” Type C femurs have “virtually lost the medial and posterior cortices.”(30) b All 10 stems were judged to be stable. 2

FIG. 2. Median values for (A) pyridinoline cross-links, (B) deoxypyridinoline cross-links, and (C) the osteocalcin/deoxypyridinoline ratio, all expressed as a percentage of the baseline value. Note that error bars have been omitted for sake of clarity. The Appendix includes more information on variance within each group.

RESULTS The three patient groups had similar distributions of bone type and severity of stress-shielding–related bone loss (Table 1). In addition, all of the cementless femoral stems were judged stable and the proportion of patients with bone ingrowth and fibrous ingrowth did not differ between the two groups (Table 1). All of the cemented femoral stems were judged stable. There were significant overall time effects for both pyridinoline and deoxypyridinoline in all three patient groups

(p ⱕ 0.01; Friedman tests). Specifically, both cross-link concentrations increased from baseline levels to reach a peak at 3 months in the patient groups (Figs. 2A and 2B). At each postoperative time, there were significant overall between-group differences for both cross-links (p ⬍ 0.05; Kruskall-Wallis tests). Most of the specific differences involved the cementless cobalt alloy group. This group had higher pyridinoline marker values than the cemented cobalt alloy group at 6 months (p ⫽ 0.003), 12 months (p ⫽ 0.019), 24 months (p ⫽ 0.019), and 36 months (p ⫽ 0.035) and higher deoxypyridinoline values at 6 months (p ⫽

804

QURESHI ET AL.

FIG. 3. Individual values for the markers for the 10 patients that received a cementless cobalt alloy stem, expressed as a percentage of the baseline value for (A) pyridinoline cross-links, (B) deoxypyridinoline cross-links, (C) ostoecalcin, and (D) the osteocalcin/deoxypyridinoline ratio.

0.005), 12 months (p ⫽ 0.015), 24 months (p ⫽ 0.019), and 36 months (p ⫽ 0.005). The cementless cobalt alloy group also had higher pyridinoline marker values than the cementless titanium alloy group at 3 months (p ⫽ 0.011), 6 months (p ⫽ 0.001), 12 months (p ⫽ 0.002), 24 months (p ⫽ 0.019), and 36 months (p ⫽ 0.001) and higher deoxypyridinoline values at 3 months (p ⫽ 0.023), 6 months (p ⬍ 0.001), 12 months (p ⫽ 0.004), 24 months (p ⫽ 0.052), and 36 months (p ⫽ 0.002). The only other significant betweengroup differences were for the cemented cobalt alloy group, which had higher cross-link values at 3 months and lower values at 24 months than the control group (p ⬍ 0.05). The osteocalcin levels changed over time only in the cementless cobalt alloy group (p ⫽ 0.044; Friedman test, data not shown). There were two significant between-group differences—the cemented cobalt alloy group had a higher value than the control group at 6 months (p ⫽ 0.035) and a higher value than the cementless cobalt alloy group at 24 months (p ⫽ 0.035). The osteocalcin/deoxypyridinoline ratio varied over time in each of the patient groups (p ⱕ 0.01), with the ratio staying below baseline in the cementless cobalt alloy group and above baseline in the cemented cobalt alloy and cementless titanium alloy groups (Fig. 2C). The cementless cobalt alloy group had lower ratios than the cemented cobalt alloy group at 3 months (p ⫽ 0.004), 6 months (p ⫽ 0.003), and 24 months (p ⫽ 0.005) and lower ratios than the cementless titanium alloy group at 6 months (p ⫽ 0.002), 12 months (p ⫽ 0.010), and 24 months (p ⫽ 0.019). In addition, the cemented cobalt alloy group had higher ratios than the control group at 24 months and 36 months (p ⬍ 0.05). Further analysis of the 10 patients in the cementless cobalt alloy group showed that at 3 months postoperatively, there was a bimodal distribution of cross-link excretion (Figs. 3A and 3B). Four patients had an increase of 400% or more of the baseline value and 6 patients had smaller increases of ⬃200%. There was no correlation between this bimodal distribution and

age, specific implant type (AML or Solution), or implant size. However, 3 of the 4 patients with the highest elevation in cross-link excretion were men. The osteocalcin values consistently had a slight increase at 3 months followed by oscillations about baseline, with the exception of one case that stayed elevated, increasing to 600% of baseline by 36 months (Fig. 3C). The osteocalcin/deoxypyridinoline ratio was uniformly decreased at 3 months and had a bimodal distribution at 36 months, with four cases having an increased ratio and six cases having a decreased ratio (Fig. 3D). In the cemented cobalt alloy group, increased release of pyridinoline and deoxypyridinoline at 3 months and 6 months was associated with a greater degree of radiographic periprosthetic bone loss at 36 months (0.749 ⬍ r ⬍ 0.840; p ⬍ 0.05). In the cementless titanium alloy group, there was more periprosthetic bone loss at 36 months in individuals with increased osteocalcin/deoxypyridinoline ratios at 3 months and 6 months (r ⫽ 0.687 and p ⫽ 0.028 and r ⫽ 0.749 and p ⫽ 0.013, respectively) and in individuals with increased osteocalcin values at 36 months (r ⫽ 0.846 and p ⫽ 0.002). There were no significant associations between baseline-normalized marker levels and stress-shielding– related bone loss in the cementless cobalt alloy group.

DISCUSSION This study shows that urinary pyridinium cross-link values increased after THR. The time course of the changes varied among the implant types studied, with persistent elevation of the cross-link values in the cementless cobalt alloy group and return to baseline values in the other two groups. Change in osteocalcin levels was more variable; however, there was a persistent elevation in the cemented cobalt alloy group. In addition, the osteocalcin/deoxypyridinoline ratio was elevated postoperatively in patients with cemented cobalt alloy and cementless titanium alloy stems compared with patients with


cementless cobalt alloy stems, in whom the ratio remained below baseline values throughout the course of the study. Finally, there were correlations between the change in marker levels from baseline to 3 months or 6 months and the radiographic severity of stress-shielding–related bone loss at 36 months in the cemented cobalt alloy and cementless titanium alloy groups. The greater initial peak in the cross-link values in the cementless cobalt alloy femoral component group is most likely caused by the greater stiffness of these implants. Most experimental studies(23,34) and theoretical analyses(22) indicate that stem stiffness is a dominating variable in predicting periprosthetic bone loss. Stress-shielding depends on implant design (alloy composition and implant geometry), position of the implant within the host femur, and the geometry and material properties of the host femur.(17,26,27) All other factors being equal, the cementless cobalt chrome implants would be expected to induce the most stress shielding of any of the implants used in this study. Although our ability to control for other potentially important factors such as host bone geometry and material properties and implant stability was not optimal, the lack of differences among the groups in terms of bone type and implant stability suggests that these were not important confounding factors. Nevertheless, it is possible that more subtle differences in implant stability, as could be measured by roentgen stereophotogrammetric analysis,(35,36) or more precise studies of stress shielding, as can be assayed with individual-specific modeling,(37) might prove useful in parsing the variability in marker data. The measurement of markers of bone formation and bone resorption in patients with THRs may provide the clinician with valuable information regarding the metabolic status of the periprosthetic bone that cannot be obtained from radiographic analyses. In particular, this study provides an indication that rates of periprosthetic bone turnover may vary as a function of implant design. Specifically, the observation that urinary cross-link values in patients with cementless cobalt chrome stems did not return to baseline even at 3 years, when considered in conjunction with dual-energy X-ray absorptiometry studies showing that net periprosthetic bone loss with various implant designs is minimal after 1 year,(17,38,39) suggests that periprosthetic bone turnover may have remained elevated in the patients with cementless cobalt chrome implants. One must keep in mind the caveat that markers measure activity throughout the skeleton. Hence, inferences about differential rates of periprosthetic bone turnover as a function of implant design require the assumption that the differences in implant design affect only local and not remote bone turnover rates. Whether or not the increased rate of bone resorption (as assessed by the cross-link markers) is related directly to a net change in periprosthetic bone mass cannot be determined definitively from this study, although the correlation between marker levels and subsequent radiographic assessment of bone loss suggests that this may well be the case. The best type of data for this correlation, from dual-energy X-ray absorptiometry, was not available for this cohort. Despite the limitations of the radiographic method used to assess periprosthetic bone loss,(33) it is an accepted technique and indicated that baseline-normalized cross-link levels at 3 months and 6 months were positively correlated

805

with the severity of stress-shielding–induced bone loss at 36 months. If this finding is eventually confirmed with better quantitative data on bone loss, it would suggest that the rate of bone resorption in the early postoperative months could be a useful marker for predicting long-term periprosthetic bone loss. This, in turn, may be a means to identify prospectively individual patients or patient groups that are at risk for developing severe stress-shielding–induced bone remodeling. This tool would be useful in the evaluation of new technologies and designs of THRs and may aid the clinician in the identification of individuals who may benefit from pharmacologic interventions aimed at retarding the loss of periprosthetic bone stock. This prospective study will help establish an understanding of time-related changes in the concentration of urinary markers of bone metabolism in the early postoperative period in patients with well-functioning THRs in the absence of osteolysis. The observations that the concentrations of the resorption markers were elevated early, were apparently sensitive to femoral component stiffness, and may be predictive of longer-term stress-shielding–induced bone loss should be considered in the design and interpretation of marker studies of bone loss after THR, including particleinduced periprosthetic osteolysis.

ACKNOWLEDGMENTS The authors thank Anastasia Skipor for help with organizing the urine samples and Daniel Pietryla for the HPLC analyses. This work is supported by the National Institutes of Health (NIH) grants AR16485, AG04736, AR39310, and AR39239.

REFERENCES 1. Woitge HW, Seibel MJ 1999 Molecular markers of bone and cartilage metabolism. Curr Opin Rheum 11:218 –225. 2. Knott L, Bailey AJ 1998 Collagen cross-links in mineralizing tissues: A review of their chemistry, function, and clinical relevance. Bone 22:181–187. 3. Eyre DR 1997 Bone biomarkers as tools in osteoporosis management. Spine 22:17S–24S. 4. Garnero P, Delmas PD 1998 Biochemical markers of bone turnover. Applications for osteoporosis. Endocrinol Metab Clin North Am 27:303–323. 5. Eyre DR 1987 Collagen cross-linking amino acids. In: Cunningham LW (ed.) Methods in Enzymology. Academic Press, New York, NY, USA, pp. 115–139. 6. Eyre DR, Dickson IR, Van Ness K 1988 Collagen crosslinking in human bone and articular cartilage. Age-related changes in the content of mature hydroxypyridinium residues. Biochem J 252:495–500. 7. Eastell R, Colwell A, Hampton L, Reeve J 1997 Biochemical markers of bone resorption compared with estimates of bone resorption from radiotracer kinetic studies in osteoporosis. J Bone Miner Res 12:59 – 65. 8. Delmas PD, Schlemmer A, Gineyts E, Riis B, Christiansen C 1991 Urinary excretion of pyridinoline crosslinks correlates with bone turnover measured on iliac crest biopsy in patients with vertebral osteoporosis. J Bone Miner Res 6:639 – 644. 9. Garnero P, Hausherr E, Chapuy MC, Marcelli C, Grandjean, Muller C, Cormier C, Breart G, Meunier PJ, Delmas PD 1996

806

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

24.

QURESHI ET AL. Markers of bone resorption predict hip fracture in elderly women: The EPIDOS Prospective Study. J Bone Miner Res 11:1531–1538. Riis BJ, Hansen MA, Jensen AM, Overgaard K, Christiansen 1996 Low bone mass and fast rate of bone loss at menopause: Equal risk factors for future fracture: A 15-year follow-up study. Bone 19:9 –12. Chesnut CH, Bell NH, Clark GS, Drinkwater BL, English SC, Johnson CCJ, Notelovitz M, Rosen C, Cain DF, Flessland KA, Mallinak NJ 1997 Hormone replacement therapy in postmenopausal women: Urinary N-telopeptide of type I collagen monitors therapeutic effect and predicts response of bone mineral density. Am J Med 102:29 –37. Garnero P, Shih WJ, Gineyts E, Karpf DB, Delmas PD 1994 Comparison of new biochemical markers of bone turnover in late postmenopausal osteoporotic women in response to alendronate treatment. J Clin Endocrinol Metab 79:1693–1700. Harris ST, Gertz BJ, Genant HK, Eyre DR, Survill TT, Ventura JN, DeBrock J, Ricerca E, Chesnut CH 1993 The effect of short term treatment with alendronate on vertebral density and biochemical markers of bone remodeling in early postmenopausal women. J Clin Endocrinol Metab 76:1399 –1406. Riis BJ, Overgaard K, Christiansen C 1995 Biochemical markers of bone turnover to monitor the bone response to postmenopausal hormone replacement therapy. Osteoporos Int 5:276 –280. Thiebaud D, Kriegbaum H, Huss H 1996 Intravenous injections of ibandronate in the treatment of postmenopausal osteoporosis. In: Papapoulos SE, Lips P, Pols HAP (eds.) Osteoporosis 96. Elsevier Science, Amsterdam, The Netherlands, p. 32. Vergnaud P, Garnero P, Meunier PJ, Breart G, Kamihagi K, Delmas PD 1997 Undercarboxylated osteocalcin measured with a specific immunoassay predicts hip fracture in elderly women: The EPIDOS Study. J Clin Endocrinol Metab 82: 719 –724. Sumner DR 1998 Bone remodeling of the proximal femur. In: Callaghan JJ, Rosenberg AG, Rubash H (eds.) The Adult Hip. Lippincott-Raven Publishers, New York, NY, USA, pp. 211– 216. Tauber C, Noff M, Noff D, Malkin C 1989 Active metabolites of vitamin D in patients with coxarthrosis and with aseptic loosening of total hip endoprostheses. Arch Orthop Trauma Surg 108:176 –178. Granchi D, Verri E, Ciapetti G, Stea S, Savarino L, Sudanese A, Mieti M, Rotini R, Dallari D, Zinghi G, Montanaro L 1998 Bone-resorbing cytokines in serum of patients with aseptic loosening of hip prostheses. J Bone Joint Surg Br 80:912–917. Joerring S, Jensen LT 1993 Changes in collagen metabolites in serum after cemented hip and knee arthroplasty. Arch Orthop Trauma Surg 112:139 –141. Schneider U, Breusch SJ, Termath S, Thomsen M, Brocai DRC, Niethard F-U, Kasperk C 1998 Increased urinary crosslink levels in aseptic loosening of total hip arthroplasty. J Arthrop 13:687– 692. Weinans H, Sumner DR 1997 Finite element analyses to study periprosthetic bone adaptation. In: Lowet G, Ru¨ egsegger P, Weinans H, Meunier A (eds.) Bone Research in Biomechanics. IOS Press, Amsterdam, The Netherlands, pp. 3–16. Turner TM, Sumner DR, Urban RM, Igloria R, Galante JO 1997 Maintenance of proximal cortical bone with use of a less stiff femoral component in hemiarthroplasty of the hip without cement. J Bone Joint Surg Am 79:1381–1390. Weinans H, Huiskes R, Grootenboer HJ 1994 Effects of fit and bonding characteristics of femoral stems on adaptive boneremodeling. J Biomech Eng 116:393– 400.

25. Huiskes R, Weinans H, van Rietbergen B 1992 The relationship between stress shielding and bone resorption around total hip stems and the effects of flexible materials. Clin Orthop 274:124 –134. 26. Sumner DR, Galante JO 1992 Determinants of stress shielding: Design vs. materials vs. interface. Clin Orthop 274:202– 212. 27. Dujovne AR, Bobyn JD, Krygier JJ, Miller JE, Brooks CE 1993 Mechanical compatibility of noncemented hip prostheses with the human femur. J Arthrop 8:7–22. 28. Jacobs JJ, Skipor AK, Black J, Urban RM, Galante JO 1991 Release and excretion of metal in patients who have a total hip-replacement component made of titanium-base alloy. J Bone Joint Surg Am 73:1475–1486. 29. Uebelhart D, Thonar EJMA, Pietryla DW, Williams JM 1993 Elevation in urinary levels of pyridinium cross-links of collagen following chymopapain-induced degradation of articular cartilage in the rabbit knee provides evidence of metabolic changes in bone. Osteoarthritis Cartilage 1:185–192. 30. Dorr LD, Faugere MC, Mackel AM, Gruen TA, Bognar B, Malluche HH 1993 Structural and cellular assessment of bone quality of proximal femur. Bone 14:231–242. 31. 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 Br 69:45–55. 32. Harris WH, McCarthy JC, O’Neill DA 1982 Femoral component loosening using contemporary techniques of femoral cement fixation. J Bone Joint Surg Am 64:1063–1067. 33. 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 231:7–28. 34. Sumner DR, Turner TM, Urban RM 1999 Animal models of bone ingrowth and joint replacement. In: An YH, Friedman RJ (eds.) Animals Models in Orthopedic Research. CRC Press, Boca Raton, FL, USA, pp. 407– 425. 35. Mjo¨ berg B, Hansson LI, Selvik G 1984 Instability, migration and laxity of total hip prostheses: A roentgen stereophotogrammetric study. Acta Orthop Scand 55:141–145. ¨ nsten I, Berzins A, Shott S, Sumner DR 2001 Accuracy and 36. O precision of radiostereometric analysis in the measurement of THR femoral component translations: Human and canine in vitro models. J Orthop Res 19:1162–1169. 37. Weinans H, Sumner DR, Igloria R, Natarajan RN 2000 Sensitivity of periprosthetic stress-shielding to load and the density-modulus relationship in subject-specific finite element models. J Biomech 33:809 – 817. 38. 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 Arthrop 11:184 –193. 39. Nakamura K 1996 Measurement of periprosthetic bone mineral density after cementless hip arthroplasty by dual energy X-ray absorptiometry: Longitudinal and cross-sectional evaluation. J Orthop Sci 1:113–122.

Address reprint requests to: Dale R. Sumner, Ph.D. Department of Anatomy Rush Medical College Rush-Presbyterian-St. Luke’s Medical Center 600 South Paulina Chicago, IL 60612, USA Received in original form July 10, 2001; in revised form November 13, 2001; accepted December 19, 2001.


APPENDIX 1. DESCRIPTIVE STATISTICS

Pyridinoline Cemented CoCr

Cementless CoCr

Cementless Ti6A14V

Control

Deoxypyridinoline Cemented CoCr

Cementless CoCr

Cementless Ti6A14V

Control

Osteocalcin Cemented CoCr

Cementless CoCr

Cementless Ti6A14V

Control

Osteocalcin/deoxypyridinoline Cemented CoCr

Cementless CoCr

Cementless Ti6A14V

Control

FOR THE

807

BASELINE-NORMALIZED MARKER VALUES

Baseline

3 Months

6 Months

12 Months

24 Months

36 Months

Median Minimum Maximum Median Minimum Maximum Median Minimum Maximum Median Minimum Maximum

100 100 100 100 100 100 100 100 100 100 100 100

223 50 424 273 184 1078 140 43 370 119 84 137

122 37 231 235 125 544 82 41 199 104 68 186

108 45 204 211 72 373 59 22 226 75 24 192

87 43 196 156 66 378 79 15 215 118 60 299

86 49 318 159 86 323 64 16 223 123 47 178


100 100 100 100 100 100 100 100 100 100 100 100

160 50 608 278 125 792 158 63 250 103 82 179

95 26 565 257 127 459 77 47 146 88 53 165

83 42 512 201 76 283 83 27 179 65 21 132

69 36 297 178 67 346 92 42 225 141 60 169

74 33 210 179 82 379 63 26 169 115 46 179


100 100 100 100 100 100 100 100 100 100 100 100

152 59 1013 121 71 220 111 51 257 91 34 411

150 76 762 115 72 224 99 49 235 82 26 179

142 35 1430 94 51 187 91 54 151 103 22 282

136 79 1279 83 57 261 121 61 231 90 50 215

159 61 661 105 76 640 102 56 178 99 26 196


100 100 100 100 100 100 100 100 100 100 100 100

114 18 829 45 12 82 79 26 207 88 33 330

221 33 823 48 21 107 102 57 334 77 27 223

144 23 3370 50 19 146 128 69 330 156 82 313

183 47 2606 64 17 152 122 53 334 82 36 187

243 38 972 64 25 208 157 57 451 85 31 185

All values are presented as the follow-up value divided by the baseline value multiplied by 100.