Bone 40 (2007) 218 – 222 www.elsevier.com/locate/bone
Strontium ranelate reduces the urinary level of cartilage degradation biomarker CTX-II in postmenopausal women Peter Alexandersen a,⁎, M.A. Karsdal b , Per Qvist c , J-Y. Reginster d , Claus Christiansen a a
Center for Clinical and Basic Research a/s, Ballerup Byvej 222, DK-2750 Ballerup, Denmark b Pharmos Bioscience, Herlev, Denmark c Nordic Bioscience, Herlev, Denmark d Service de Santé Publique, d’Epidémiologie et Économie de la Santé, Liège, Belgium Received 29 September 2005; revised 3 July 2006; accepted 24 July 2006 Available online 28 September 2006
Abstract Objective: Strontium ranelate significantly decreases the risk of osteoporotic fractures. The objective of the present study was to investigate whether strontium ranelate (2 g/day) also affects cartilage brakedown as measured by urinary marker of cartilage degradation, designated CTX-II. Methods: A subgroup of 2617 postmenopausal osteoporotic women (aged 75.7 ± 4.4 years) were selected from the TROPOS phase III study on the basis of a urinary sampling reported at each visit during the first three years of the study. When included in TROPOS, they were randomized to strontium ranelate or placebo in a double-blind fashion for 3 years. A calcium and vitamin D supplement was also provided to the subjects during the study. A marker of collagen type II degradation (CTX-II) corrected for urinary creatinine (CTX-II/cr.) was assessed at regular intervals throughout the study in 1310 patients in strontium ranelate group and 1307 patients in placebo group. Results: The response in CTX-II depended on time (p < 0.0001), and this time dependency differed statistically significantly between groups (time × treatment) (p < 0.0003). In addition, there was a statistically significant difference between treatments (p < 0.0001). The difference in the response of CTX-II/cr. appeared already after three months, with the strontium ranelate-treated subjects having approximately 15–20% lower values than the placebo-treated subjects for the remaining study period (p < 0.0001). Conclusion: Treatment with strontium ranelate significantly decreases urinary excretion of CTX-II, a marker of cartilage destruction. Further studies are warranted to investigate an effect on cartilage formation and symptoms of osteoarthritis. © 2006 Elsevier Inc. All rights reserved. Keywords: Strontium ranelate; Postmenopausal women; CTX-II; Type II collagen degradation; RCT
Introduction Structural damage of articular cartilage constitutes a hallmark in osteoarthritis (OA) resulting from an increased imbalance of cartilage degradation and synthesis. OA is a major health problem affecting the majority of the population over the age of 50, with varying degree of progression, typically involving the spine, hip, knee, or finger joints, although many other joints may also be affected [1]. Even though the initiation of OA is still debated, experimental animal studies have indicated that progression of OA involves both bone and cartilage alterations. These studies indicate that subchondral ⁎ Corresponding author. Fax: +45 44 68 42 20. E-mail address: pa@ccbr.dk (P. Alexandersen). 8756-3282/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2006.07.028
bone turnover may be augmented in OA and, conversely, attenuation of subchondral bone remodeling results in decreased progression of OA [2–4]. Current gold standard for cartilage measurements is radiographic examination of the affected joints, for instance by determining joint space width (JSW) of the knee. However, a surrogate biochemical marker of type II collagen degradation (CTX-II, see below) has been shown to correlate with JSN progression in OA [5], and in another study increased urinary levels of CTX-II were associated with rapidly progressive hip OA [6]. CTX-II is a type II collagen C-telopeptide neoepitope, which is generated by matrix metalloproteases (MMPs) through cleavage of intact type II collagen resulting in a smaller fragment that is specific of degradation [7]. In a previous study of patients with subchondral changes (visualized by magnetic
P. Alexandersen et al. / Bone 40 (2007) 218–222
resonance imaging (MRI)) interpreted as bone marrow edema as a consequence of knee OA, these changes were found to correlate directly with CTX-II excretion in the urine [8], further supporting a pathophysiologic role of CTX-II involvement in the progression of OA. Strontium ranelate can prevent osteoporotic vertebral and nonvertebral fractures including hip fractures in postmenopausal women with low bone mass [9,10]. Strontium ranelate simultaneously increases bone formation and inhibits bone resorption. Thereby, it displays potentially the same effects on the subchondral remodeling compartment and thus may indirectly affect cartilage degradation in the development of OA. As a consequence of that, strontium ranelate might attenuate the progression of OA by inhibition of bone resorption. In support, a recent preliminary in vitro study of isolated human chondrocytes demonstrating stimulation of proteoglycan production by strontium ranelate via an ionic effect, suggested a favorable effect of strontium on cartilage formation [11], similar to what is observed in bone. However, because the effect of strontium ranelate on systemic bone resorption is approximately 12% compared to placebo, p < 0.001 [9], the expected effect on cartilage through subchondral bone remodeling is likely to be lower. The objective of this study was to investigate the effect of strontium ranelate on the urinary excretion of CTX-II in postmenopausal women. Subjects and methods Study population The study population consisted of a subgroup of 2617 postmenopausal women from the randomized, phase III clinical study evaluating the effects of strontium ranelate on nonvertebral osteoporotic fractures (the TROPOS study, with a total number of 5091 patients included) [10], the subgroup being selected on the basis of a reported complete patient series of urinary samples covering all visits between baseline and 36 months. When checking the number of available samples, a total of 126 samples were missing (samples never received from the investigational sites) and another 52 tubes had insufficient volume. Therefore, of the 2617 women selected for the study, 2456 (93.8%) had measurable urinary samples at all the visits during the three years and are therefore referred to as “valid completers” [10].
Study protocol Inclusion criteria were those of the population selected in the TROPOS phase III study, designed for assessing the antifracture efficacy of strontium ranelate at nonvertebral sites: – ambulatory Caucasian postmenopausal women, ≥74 years old, or between 70 and 74 years old but with at least one additional fracture risk factor such as personal history of osteoporosis-related fractures after the menopause, or resident in retirement homes, or frequent falls, or maternal history of osteoporosis-related fractures (hip, vertebrae, wrist), – femoral neck BMD ≤0.600 g/cm2 (T-score <− 2.5 according to the reference population used in the study). Furthermore, the subjects did not take any medication known to affect bone metabolism prior to the study for periods of time depending on the medications considered. All women gave their written informed consent at baseline. The local Ethics Committee approved the study, and the study was conducted according to the rules of Good Clinical Practice (GCP).
219
Participants were randomly assigned to receive orally either strontium ranelate (2 g/day) or identically appearing placebo in a 1:1 ratio [10]. Before inclusion in the TROPOS study, patients were subjected to a run-in study in order to initiate normalization of their calcium and vitamin D status. The duration of this run-in study was 2 weeks to 6 months, depending on the severity of calcium and 25-OH vitamin D (25-OH D) deficiency. All enrolled women received daily supplements of up to 1000 mg of elemental calcium adapted to their needs according to their dietary intake (0, 500, or 1000 mg to reach a total daily intake above 1000 mg), and vitamin D according to their serum 25-OH D levels (800 IU for patients having serum 25-OHD lower than 45 nmol/L and 400 IU for all the others). For patients with severe vitamin D deficiency (25-OH D <30 nmol/L), the duration of the run-in period was at least 3 months. At the end of the run-in period, patients were included in the TROPOS study, which lasted for 5 years with the main statistical analysis done at 3 years. Both investigators and study participants were kept blinded to treatment code during the entire study period, and the code was not broken and treatment revealed to investigator and participants until the study was completed and the database checked and approved. At baseline, all participants were individually interviewed thus providing a status on previous and current medical and surgical history, and information on medication, time since menopause, and underwent a thorough physical examination. Body height and weight were assessed with the subject wearing light indoor clothing and no shoes to the nearest 0.5 cm and 0.1 kg, respectively, and body mass index (BMI) was calculated as the weight divided by the height squared (kg/m2). Blood and urinary samples were taken in the morning after overnight fasting (between 8 and 9 A.M.) to assess efficacy and safety parameters.
Biochemical markers Urinary collagen type II C-telopeptide fragments were assessed using a competitive enzyme immunoassay (ELISA) for measuring urine samples, and adjusted for urinary creatinine (CTX-II/cr., Nordic Bioscience a/s; Denmark) [5]. Urinary CTX-II/cr. was determined with an inter-assay and intra-assay CV% of 10.1% and 4.7%, respectively.
Statistical analysis The analysis concerning the effect of strontium ranelate on urinary CTX-II/ cr. during the study was performed on all the 2617 patients selected, representing the full analysis set population (FAS) that included all randomized patients who took at least one sachet of the study treatment, with at least one piece of postbaseline information about peripheral fracture occurrence after the inclusion visit. The valid completers group represents the 2456 patients subset from the FAS who had one urinary sampling available and measurable at each visit during the first three years (i.e., months 0, 3, 6, 12, 24, and 36). To avoid contribution of biological variation, changes were calculated as percent changes from baseline values (set to 100%). Primary data analyses included repeated measures analysis of variance (ANOVA) taking into account all measurements at the various time points (and thus the repetitive measurements in the same individual). In this analysis, time at baseline (t0) was eliminated because it was used for normalization of consecutive measurements. Adjustment for multiple comparisons was made by Bonferroni method. Secondly, with respect to baseline characteristics and changes from baseline between groups, the mean values of the two study groups were compared using two-sided Student's t test, whereas changes from baseline within a treatment group at each subsequent visit were evaluated by paired t tests. The average response to treatment was calculated as the mean urinary concentration in urinary CTX-II/cr, after appropriate logarithmical transformation of the data. Response to treatment in urinary CTX-II/cr. was also analyzed separately for subjects with (n = 565) and without OA (n = 2052). Kolmogorov–Smirnov's analysis was used to test for normality of the data. Mann–Whitney U test was applied to test for significant difference in the distribution of CTX-II/cr. between groups. Outliers (>95% confidence interval of the mean values) were excluded from the analysis (i.e., between 100 and 150 at each visit) used for the graph in Fig. 1. Data were logarithmically transformed if appropriate to obtain normality and homogeneity of the data. Data were primarily done as an intention-to-treat analysis, and secondary analysis was performed on valid completers. Patients were classified as having OA if during
220
P. Alexandersen et al. / Bone 40 (2007) 218–222
Fig. 1. Shows the longitudinal changes over time in urinary CTX-II/creatinine ratio between the two groups. Data shown are mean (±SEM) values in the strontium ranelate group (filled square) and mean values in the placebo group (open circles), adjusted to baseline. Data based on intention-to-treat analysis (outliers excluded). At month 3, the mean response in CTX-II/creatinine to strontium ranelate was −15.1 %. The strontium ranelate-treated group displays a significant decrease in the urinary marker after 3 months as compared to placebo. Below the graph: the absolute values of urinary CTX-II/creatinine during the 36 months of treatment with either strontium ranelate or placebo. Outliers included. *p < 0.0001 compared to placebo.
the interview they revealed a medical history of known OA, classical OA symptoms, or regular use of typical OA medication for joint pain. We stratified the population into women with or without OA for evaluating differences in use of concomitant OA medication and need of join surgery. A p-value of less than 0.05 was considered statistically significant. SPSS statistical software was used for the statistical analyses.
Results Tables 1a and b show some baseline characteristics of the 2617 women randomized to either strontium ranelate or placebo in addition to those that were published in the original paper on the TROPOS study [10]. There were no statistically significant differences between groups for any of the traditional parameters tested, including bone mineral density (BMD) of various skeletal sites [10]. At baseline, the number of subjects complaining about symptoms of OA was 280 in the strontium ranelate group and 285 in the placebo group. Among these patients, 107 received treatments for OA in the strontium ranelate group and 106 in the placebo group (no significant difference) (p = NS for both comparisons). Urinary CTX-II/cr. levels were higher among subjects with OA than in those Table 1a Baseline values of urinary CTX-II/cr. (ng/mmol) a in the two study groups (mean ± SD)
OA not clinically present (n = 2052 b) OA clinically present (n = 565 c) a b c
Strontium ranelate (n = 1310)
Placebo (n = 1307)
327 ± 227 (n = 1021)
316 ± 186 (n = 1016)
359 ± 235 (n = 277)
374 ± 243 (n = 283)
CTX-II/cr., urinary CTX-II/urinary creatinine. OA, osteoarthritis. 15 urine samples missing, leaving n = 2037 samples at baseline. 5 urine samples missing, leaving n = 560 samples at baseline.
without clinical symptoms of OA (p < 0.05), although there were no statistical difference between treatment groups (Table 1a). A good level of compliance to the treatment should be pointed out: 87 ± 14 (mean% ± SD) for strontium ranelate group and 88 ± 13 for the placebo group. The treatment exposure was 1060 ± 76 days (mean ± SD) and 1065 ± 71 days for the strontium ranelate and the placebo group, respectively. Repeated measures ANOVA taking into account multiple measurements in the same individual showed that the response in urinary CTX-II was dependent of time (p < 0.0001), and this time dependency differed statistically significantly between the two treatment groups (time × treatment) (p < 0.0003). In addition, we found a statistically significant difference between treatments (p < 0.0001). Adjustment for multiple comparisons (using the conservative Bonferroni method) did not change this statistically significant difference between treatments (p < 0.0001). Fig. 1 shows the longitudinal change in urinary CTX-II between the two treatment groups, shown as intention-to-treat Table 1b Number of subjects (and %) in the study reporting OA symptoms at baseline and taking medication for OA at baseline
OA a at baseline Taking concomitant medication for OA b at baseline
Strontium ranelate (n = 1310)
Placebo (n = 1307)
280 (21.4%) 107 (8.2%)
285 (21.8%) 106 (8.1%)
a “OA at baseline” were selected as having reported an OA at baseline as an adverse event. b Concomitant treatments reported for OA: non-steroidal anti-inflammatory drugs, Cox-2 inhibitors, painkillers type paracetamol (acetaminophen) or codeine, and codeine-family products, chondroitin sulfate sodium, diacerein, glucosamine sulfate, and piascledine.
P. Alexandersen et al. / Bone 40 (2007) 218–222 Table 2 Responses to treatment (in % of baseline) with strontium ranelate or placebo in urinary CTX-II/creatinine in subjects with (A) or without (B) osteoarthritis (OA) at baseline CTX-II/cr.
Strontium ranelate n
Mean
Placebo Mean
SEM
p value
269 269 268 268 271
5.7 7.9 5.0 4.9 9.1
2.11 2.26 2.37 2.51 2.87
<0.001 <0.001 <0.001 <0.001 <0.01
B. Subjects without OA at baseline (n = 2052) Month 3 957 −15.6 0.88 956 Month 6 950 −15.2 1.02 952 Month 12 953 −15.4 1.07 954 Month 24 957 −15.5 1.19 948 Month 36 951 −6.2 1.23 959
0.0 1.0 2.7 4.2 10.7
0.98 1.16 1.19 1.30 1.58
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001
A. Subjects with OA at baseline Month 3 264 −13.6 Month 6 260 − 11.5 Month 12 256 −12.8 Month 24 252 −9.8 Month 36 251 0.3
SEM
n
(n = 565) 1.68 1.95 2.05 2.35 2.66
analysis. In women receiving strontium ranelate, there was a marked decrease (approximately 20%) in CTX-II/cr. compared to placebo that occurred within the first 3 months of treatment. Thereafter, no further reduction was observed. The observed decrease in CTX-II/cr. in the strontium ranelate group was significantly different from that in the placebo group at any time point after 3 months of treatment (p < 0.0001) (Fig. 1). Using the absolute changes in urinary CTX-II excretion, the differences between groups were also highly statistically significant (p < 0.0001). Analysis of the subgroups with (n = 565) and without OA (n = 2052) at baseline showed similar responses to treatment for the collagen type II marker investigated; however, the relative differences between the strontium ranelate and the placebo group tended to be slightly greater in the OA group compared to the non-OA group within the first 12 months (Table 2). When analyzing the data for valid completers, we found very similar results (data not shown). Subgroup analysis of women with and without OA at baseline further revealed no statistically significant difference between groups (p > 0.05) regarding use of concomitant medication for OA or joint surgery due to OA during the course of the study (Table 3). However, absolute numbers of women undergoing joint surgery were low, even when only considering those with known OA (Table 3). Discussion Development of OA is a complex process involving simultaneous changes in both bone and cartilage. Progression of the disease most likely comprises subchondral bone remodeling, eventually leading to the hallmark of OA osteophytes. We sought to investigate whether the effects of strontium ranelate on bone remodeling, simultaneously increasing bone formation and decreasing bone resorption, would have potential beneficial effect on articular cartilage degradation as measured by a marker of type II collagen degradation (urinary CTX-II).
221
To our knowledge, this clinical study is the first to indicate that strontium ranelate given orally (2 g/day) for 3 years significantly decrease the urinary CTX-II levels by approximately 15%. This could reflect an indirect effect on cartilage, potentially through an inhibition of bone resorption, as in the Spinal Osteoporosis Therapeutic Intervention (SOTI) study, strontium ranelate was demonstrated to reduce significantly the serum CTX-I, a bone resorption marker [9]. The observed increase in the CTX-II excretion all over the study in the placebo group could reflect the global progression of the cartilage degradation, this phenomenon being also observed after 12 months in the strontium ranelate-treated group but to a lesser extent, as at month 36, the CTX-II/cr. value is still lower than that observed at baseline in this group (Fig. 1). This could reflect the efficacy of strontium ranelate in preventing the further progression of the cartilage degradation. In this population, we did not find any substantial differences in responses to treatment with strontium ranelate compared to placebo between subjects with OA and without OA, perhaps in part due to the heterogeneity of OA severity in postmenopausal women as represented in this study. Furthermore, the population was not selected on the basis of existing OA and no specific items related to OA were investigated during the study, leading possibly to under-reported rates. Further investigations in carefully selected populations with various degrees of OA are needed to rule out if there are increasing responses to strontium ranelate in this and other collagen type II markers with increasing severity of OA. Timely identification of high-risk patients and monitoring of the quantitative effect of putative agents for prevention of OA are of immense clinical importance for affected individuals. Today, no efficacious treatment modalities exist for OA, and at least as importantly, no commercially available biochemical marker has yet been developed to establish the diagnosis of OA and to monitor the course of its natural history [12]. CTX-II has been reported to be significantly elevated in patients with OA Table 3 Women taking medication or having joint surgery for osteoarthritis (OA) during the study period Strontium ranelate (n = 1310) Women with OA using concomitant medication for OA a during the study period Total number of women in study Hip arthroplasty Knee arthroplasty Joint prosthesis insertion b Number of women with OA Hip arthroplasty Knee arthroplasty Joint prosthesis insertionb
Placebo (n = 1307)
154 (11.8%)
146 (11.2%)
1310 (100%) 3 (0.23%) 1 (0.08%) 4 (0.31%) 280 (21.4%) 0 (0%) 1 (0.36%) 3 (1.07%)
1307 (100%) 3 (0.23%) 1 (0.08%) 7 (0.54%) 285 (21.8%) 2 (0.70%) 1 (0.35%) 3 (1.05%)
NS for all comparisons between the two groups. a Concomitant treatments reported for OA: non-steroidal anti-inflammatory drugs, Cox-2 inhibitors, painkillers type paracetamol (acetaminophen) or codeine, and codeine-family products, chondroitin sulfate sodium, diacerein, glucosamine sulfate, and piascledine. b No precision on the site available.
222
P. Alexandersen et al. / Bone 40 (2007) 218–222
[5] but not in patients with osteoporosis or metabolic bone diseases [12] compared to healthy controls. This large clinical study clearly suggests that strontium ranelate can significantly reduce the urinary excretion of CTX-II in postmenopausal women, supporting the notion that treatment with strontium ranelate may have a beneficial effect on OA progression. However, there was no statistically significant difference between the number of women using OA medication during course of the study or having joint surgery due to OA, certainly due to the fact that patients were selected on the basis of an existing osteoporosis diagnosed at the femoral neck and not on the basis of existing OA. Therefore, the OA-reported cases at baseline were certainly related to various joint sites and displayed various severity stages, leading to an heterogeneity that cannot allow statistical significance to be reached for these aspects. Interestingly, at least 20% of these patients suffering osteoporosis also have reported an OA status at baseline, somehow in contradiction with the idea generally assumed that patients suffering osteoporosis are not prone to develop OA [13]. In the SOTI study, conducted in parallel to the TROPOS study, measurements of bone resorption by CTX-I decreased by approximately 10% and measurements of bone formation by bALP increased by approximately 8% in strontium ranelatetreated patients compared to placebo (p < 0.001), consequently significant increases in BMD were observed [9]. This effect on BMD could be explained by the simultaneous inhibition of bone resorption and anabolic effect on bone formation [14,15]. Thus, strontium ranelate seems to influence bone metabolism on a cellular level by rebalancing the bone turnover through an action on the two opposite bone processes. Based on the observations of strontium ranelate effects on bone, it could be hypothesized that strontium ranelate is able to positively affect on cartilage by attenuating subchondral bone turnover, which per se has been shown to have an inhibitory impact on cartilage degradation [2]. The mechanism of action of strontium ranelate on cartilage is however at present unknown but based on the present data seems to involve collagen type II metabolism either directly or indirectly. Collectively, these data could at least theoretically help to explain the decrease in urinary CTX-II excretion observed in this study, by inhibition of bone resorption and excessive subchondral remodeling, as it has been shown with anti-resorptive therapies [2,3]. In addition, previous studies have suggested direct effects of strontium ranelate on chondrocytes [11]. However, further studies are required to elucidate whether strontium ranelate also has a direct effect on cartilage by demonstrating increase in type II collagen formation markers. Despite the encouraging results, there are some caveats to consider when interpreting the data of this study. First, the study was not primarily designed to study the effect of strontium on OA, and the reduction in CTX-II excretion by strontium offers only indirect proof for its putative structural effect. Thus, for this study we have no data on direct measures of OA progression, such as the joint space narrowing measured on X-rays at various joint sites (knee, hip, spine, etc.) and no direct correlation can be performed between the levels of urinary CTX-II and a clinical assessment of OA. Finally, no accurate
information on the exact nature and location of the OA could be retrieved from the data collected. A direct relationship between CTX-II excretion and clinical elements of OA disease has to be further investigated in future studies. In conclusion, strontium ranelate (2 g/day) appears to significantly reduce the urinary excretion of collagen type II in postmenopausal women. The putative favorable effect on OA, however, needs to be confirmed in future clinical investigations. Acknowledgments We acknowledge Servier, France, for the helpful corporation and assistance, and we are indebted to Dr Inger Byrjalsen for performing the repeated measures ANOVA. References [1] Creamer P, Hochberg MC. Osteoarthritis. Lancet 1997;350:503–8. [2] Hayami T, Pickarski M, Wesolowski GA, Mclane J, Bone A, Destefano J, et al. The role of subchondral bone remodelling in osteoarthritis. Arthritis Rheum 2004;50:1193–206. [3] Ham KD, Loeser RF, Lindgren BR, Carlson CS. Effects of long-term estrogen replacement therapy on osteoarthritis severity in cynomolgus monkeys. Arthritis Rheum 2002;46:1956–64. [4] Ham KD, Carlson CS. Effects of estrogen replacement therapy on bone turnover in subchondral bone and epiphyseal metaphyseal cancellous bone of ovariectomized cynomolgus monkeys. J Bone Miner Res 2004;19: 823–9. [5] Reijman M, Hazes JM, Bierma-Zeinstra SM, Koes BW, Christgau S, Christiansen C, et al. A new marker of osteoarthritis: cross-sectional and longitudinal approach. Arthritis Rheum 2004;50:2471–8. [6] Garnero P, Conrozier T, et al. Urinary type II collagen C-telopeptide levels are increased in patients with rapidly destructive hip osteoarthritis. Ann Rheum Dis 2003;62:939–43. [7] Sondergaard BC, Henriksen K, Wulf H, Oestergaard S, Schurigt U, Brauer R, et al. Relative contribution of matrix metalloprotease and cysteine protease activities to cytokine-stimulated articular cartilage degradation. Osteoarthr Cartil 2006;14:738-48. [8] Garnero P, Peterfy C, Zaim S, Schoenharting M. Bone marrow abnormalities on magnetic resonance imaging are associated with type II collagen degradation in knee osteoarthritis. Arthritis Rheum 2005;52: 2822–9. [9] Meunier OJ, Roux C, Seeman E, Ortolani S, Badurski JE, Spector TD, et al. The effect of strontium ranelate on the risk of vertebral fracture in women with postmenopausal osteoporosis. N Engl J Med 2004;350: 458–9. [10] Reginster JY, Seeman E, De Vernejoul MC, Adami S, Compston J, Phenekos C, et al. Strontium ranelate reduces the risk of nonvertebral fractures in post-menopausal women with osteoporosis: TROPOS study. J Clin Endocr Metab 2005;90:2816–22. [11] Henrotin Y, Labasse A, Zheng SX, Galais P, Tsouderos Y, Crielaard JM, et al. Strontium ranelate increases cartilage matrix formation. J Bone Miner Res 2001;16:299–308. [12] Garnero P. Osteoarthritis: biological markers for the future? Jt Bone Spine 2002;69:525–30. [13] Dequeker J, Boonen N, Aerssens J, Westhovens R. Inverse relationship osteoarthritis-osteoporosis: what is the evidence? Br J Rheumatol 1996;35:813–20. [14] Marie PJ, Ammann P, Boivin G, Rey C. Mechanisms of action and therapeutic potential of strontium in bone. Calcif Tissue Int 2001;69: 121–9. [15] Canalis E, Hott M, Deloffre P, Tsouderos Y, Marie PJ. The divalent strontium salt S12911 enhances bone cell replication and bone formation in vitro. Bone 1996;18:517–23.