Osteoporosis International

, Volume 19, Issue 3, pp 339–348

Bone turnover and bone collagen maturation in osteoporosis: effects of antiresorptive therapies

Authors

    • Nordic Bioscience A/S
  • D. J. Leeming
    • Nordic Bioscience A/S
  • P. Qvist
    • Nordic Bioscience A/S
  • C. Christiansen
    • Nordic Bioscience A/S
  • M. A. Karsdal
    • Nordic Bioscience A/S
Original Article

DOI: 10.1007/s00198-007-0462-5

Cite this article as:
Byrjalsen, I., Leeming, D.J., Qvist, P. et al. Osteoporos Int (2008) 19: 339. doi:10.1007/s00198-007-0462-5

Abstract

Summary

Bone collagen maturation may be important for anti-fracture efficacy as the reduction in risk is only partly explained by a concomitant increase in BMD during anti-resorptive therapy. Different treatments caused diverse profiles in bone collagen degradation products, which may have implications for bone quality.

Introduction

The aim of the present study was to evaluate the effect of different anti-resorptive treatments on bone collagen maturation measured as the ratio between the degradation products of newly synthesized and mature isomerized C-telopeptides of type I collagen.

Methods

Participants were from cohorts of healthy postmenopausal women participating in double blind, placebo-controlled 2-year studies of alendronate, ibandronate, intranasal hormone replacement therapy (HRT), oral HRT, transdermal HRT, or raloxifene (n = 427). The non-isomerized ααCTX and isomerized ββCTX were measured in urine samples obtained at baseline, and after 6, 12, and 24 months of therapy.

Results

Bone collagen maturation measured as the ratio between ααCTX and ββCTX showed that bisphosphonate treatment induced a collagen profile consistent with an older matrix with a 52% (alendronate) and 38% (ibandronate) reduction in the ratio between the two CTX isoforms vs. 3% and 15% with HRT or raloxifene, respectively.

Conclusions

Anti-resorptive treatments had different effects on the endogenous profile of bone collagen maturation. Whether that effect on bone collagen has an impact on bone strength independent on the treatment-dependent effect on BMD should be investigated.

Keywords

Bone qualityBone turnoverCollagenFracture riskMatrix proteins

Introduction

Bone is a dynamic tissue being continuously remodeled throughout life not only to maintain the calcium homeostasis but also to maintain mechanical strength by replacing fatigued bone by new, mechanically sound bone [1].

Bone formation and resorption are two highly coupled metabolic processes [26], although the balance between resorption and formation differs in different phases of life. The loss of ovarian sex steroids at menopause results in accelerated bone turnover with a predominance of bone resorption over bone formation [7]. The related negative calcium balance promotes bone loss, increases bone fragility, and thereby the risk of future fractures [8]. A rational approach to counter this altered balance is the inhibition of bone resorption, although also bone formation is inhibited due to coupling between these cellular events [911]. Accordingly, pronounced inhibition of osteoclast function can be expected to impair the dynamic renewal of skeletal tissue leading to alterations in the biochemical composition of bone and changed mechanical properties [12]. The inorganic phase of the bone provides the stiffness, i.e., the ability to resist compression, whereas the organic phase, mainly constituted of type I collagen, provides bone its flexibility, i.e., the ability to absorb energy and undergo deformation. How biochemical changes in the bone collagen are associated with fracture and bone quality remains to be investigated and understood.

The bone collagen is continuously renewed and post-translationally modified, including enzymatic cross-linking (pyridinoline and deoxypyridinoline), non-enzymatic glycation (pentosidine), and β-isomerization occurring in the DG motif of the CTX epitope (1207EKAHDGGR1214). It has been shown that the degree of isomerization reflects the skeletal age [13, 14]. Studies on in vitro aging of fetal bovine cortical bone recently demonstrated that the bone mechanical properties changed as a function of time, as did the amount of cross-linking and pentosidine, as well as the degree of β-isomerization [15]. The biochemical changes of collagen were associated with a 30% decrease in bending and compressive yield stress and a 2.5-fold increase in compressive post-yield energy absorption. In another recent study of human lumbar vertebrae the amount of pentosidine and degree of β-isomerization was related to the biomechanical properties of bone after adjustment for BMD [16].

BMD is considered the most important determinant of fracture risk with the World Health Organization (WHO) criteria of osteoporosis defined as BMD-measurements falling 2.5 standard deviation below the average of young adults (T-score ≤−2.5). However, recent studies have shown that up to one-half of patients with incident fractures have BMD above the WHO diagnostic threshold criteria [1719], and it has also been reported that the risk of an osteoporotic fracture is approximately tenfold higher in old compared with young individuals at the same BMD level [20]. Additionally, recent data demonstrate that during anti-resorptive therapy the reduction in fracture risk is only vaguely explained by the concomitant increase in BMD [21]. Even though drugs have become more and more effective in terms of increasing BMD, achievements in terms of anti-fracture efficacy have not followed correspondingly, e.g., the mean increase in spinal BMD to alendronate is sevenfold higher compared to that of calcitonin, yet the reductions in vertebral fracture risk are fairly comparable being 44% and 36%, respectively [21, 22] .

The discrepancies in BMD response and fracture risk protection may be due to the different mode of action of different anti-resorptive treatments. To look into this aspect of anti-resorptive therapy, we investigated the effect on bone collagen age, measured as the ratio of ααCTX to ββCTX in urine samples during different types of anti-resorptive treatment regimens of bisphosphonates, hormone replacement therapy (HRT), and raloxifene (RLX).

Methods

Subjects and study design

All studies were conducted in accordance with Helsinki Declaration II, and approved by local ethical committees. Written informed consent was obtained for all participants.

Bisphosphonate - Alendronate

The participants were part of a double blind, placebo-controlled, randomized 3-year study on the effects of oral alendronate in the prevention of postmenopausal osteoporosis [10]. Study participants who completed the first 2-year study period and who received daily 10 mg of alendronate (n = 14), 20 mg of alendronate (n = 13), or placebo (n = 13) were included.

Bisphosphonate – Ibandronate

The participants were part of a double blind, placebo-controlled, randomized 2-year study on the effects of oral ibandronate in the prevention of postmenopausal osteoporosis [23]. A randomly selected subgroup of study participants receiving daily continuous oral 2.5 mg of ibandronate daily (n = 36), intermittent oral 20 mg of ibandronate every 2nd day for 24 days every 3 months (n = 36), or placebo (n = 26) was included. After the first year of therapy, the placebo group was crossed-over to receive active treatment.

HRT - intranasal estradiol

The participants were part of a double blind, placebo-controlled, randomized 2-year study on the effects of daily intranasal 17β-estradiol for prevention of bone loss in early postmenopausal women [24]. A randomly selected subgroup of study participants receiving daily 300 μg of 17β-estradiol (n = 50), or placebo (n = 25) was included.

HRT - oral estradiol

The participants were part of a double blind, placebo-controlled, randomized 2-year study on the effects of 17β-estradiol continuously combined with drospirenone on the safety and efficacy for prevention of postmenopausal osteoporosis [25]. A randomly selected subgroup of study participants receiving daily 1 mg of 17β-estradiol continuously combined with 1 mg or 2 mg of drospirenone (n = 49), or placebo (n = 33) was included.

HRT - transdermal estradiol

The participants were part of a double blind, placebo-controlled, randomized 2-year study on the effects of daily transdermal 17β-estradiol continuously combined with transdermal levonorgestrel on the safety and efficacy for prevention of postmenopausal osteoporosis [26]. A randomly selected subgroup of study participants receiving daily 45 μg of 17β-estradiol continuously combined with 40 μg of levonorgestrel (n = 35), or placebo (n = 20) was included.

Raloxifene (RLX)

The participants were part of a double blind, placebo-controlled, randomized 2-year study on the effects of raloxifene on bone mineral density in postmenopausal women [27]. A randomly selected subgroup of study participants receiving daily 60 mg of raloxifene (n = 30), or placebo (n = 47) was included.

From all studies second void fasting urine samples were obtained at baseline, after 6 months, 12 months, and 24 months of therapy.

Biochemical markers and bone mineral density

Urine samples were stored at −20°C until analysis. The urinary excretion of ααCTX was measured using the ALPHA CrossLaps ELISA [14], and the urinary excretion of ββCTX was measured using the Serum CrossLaps One Step ELISA (urine samples were pre-dilution 1:10 in assay buffer before measurement)(Nordic Bioscience Diagnostics, Herlev, Denmark) [14, 28]. Urinary creatinine was measured by routine chemistry method and used for calculation of creatinine-corrected marker levels. The urine samples were measured at same time in all subgroups in present study.

Bone mineral density (BMD) was measured by dual-energy X-ray absorptiometry at the spine (BMDspine), and hip (BMDhip) with Hologic QDR-1000, or QD-R2000 densitometers (Hologic, Waltham, MA, USA).

Statistical analysis

To assess longitudinal changes, the values were calculated for each person and expressed as the percentage of the initial baseline value. The data of the creatinine-corrected values of ααCTX and ββCTX and the relative changes were logarithmically transformed to obtain normality and symmetry of variances. Analysis of variance (ANOVA) was used for comparison of baseline data between study groups with Tukey–Kramer adjusted significance levels in the multiple comparisons. The two-tailed Student’s t-test was applied for pair-wise comparisons of data from the active treatment group and placebo within each study and for comparison of selected subgroup with non-selected subgroup within each study. Regression analysis was used to assess relation of baseline characteristics with age, and to calculate the changes in BMD during treatment.

For all tests p < 0.05 was considered significant. All statistical calculations were performed using the SAS software package (release 9.1, SAS Institute Inc., Cary, NC, USA).

Results

Demographic data

Table 1 shows the baseline characteristics of initial study population of each study and the probability of difference of selected subgroups. In the ibandronate study, the selected subgroup was of a statistically significant higher age (68.7 years vs. 67.2), and in the nasal HRT group the selected subgroup was of lower height (163.5 cm vs. 164.7 cm). In all other aspects of age, height, weight, BMI, BMDhip, and BMDspine, the selected subgroups did not differ from the non-selected subjects. Table 2 shows the baseline demographic data of the participants within each study subgroup. The participants in the two bisphosphonate groups represented the youngest and oldest postmenopausal women with a mean age of 52.6 years in the alendronate and 68.7 years in the ibandronate group. The mean age of the three groups receiving HRT were in-between the bisphosphonate groups and ranged from 53.4 years in the nasal HRT to 59.0 years in the oral HRT group. The mean age of the RLX group was 55.7 years comparable to the mean age of 55.4 years in the transdermal HRT group. Minor but clinically non-significant differences in height, weight, and BMI were observed between the groups. The age differences in the groups were reflected in BMDhip and BMDspine with the ibandronate group having the lowest BMD at both sites. The baseline level of urinary ααCTX was comparable in the six groups. Minor differences within a range of ±20% were observed in urinary ββCTX that was related to the individual study. The ratio between ααCTX and ββCTX at baseline differed within ±20% but this difference can be explained by differences in age among the study subjects. The ratio increased with advancing age, i.e., the lowest ratio was found in the study groups of alendronate and nasal HRT aged about 53 years, and the highest ratio was found in the study group of ibandronate aged 69 years. Overall we found an increase in the ratio of 1.9% per year (p < 0.001) in the complete study of 427 postmenopausal women. The increase in αα/ββCTX ratio was age- and not study-related, an increase of similar magnitude was found when analysing the age-dependency in individual study groups.
Table 1

Characteristics of initial study populations at baseline and subgroup comparison

 

n

Age (yrs)

Height (cm)

Weight (kg)

BMI (kg/m2)

BMDhip (g/cm2)

BMDspine (g/cm2)

Alendronate

 Total

79

52.6 (2.2)

163.7 (5.2)

61.1 (6.3)

22.8 (2.3)

0.85 (0.08)

0.92 (0.10)

 Subgroup

40

0.95

0.29

0.11

0.27

0.13

0.86

Ibandronate

 Total

240

67.2 (4.9)

160.3 (5.8)

62.9 (9.3)

24.5 (3.6)

0.70 (0.09)

0.73 (0.07)

 Subgroup

98

<0.001

0.77

0.82

0.94

0.52

0.37

Nasal HRT

 Total

386

53.3 (1.8)

164.7 (5.9)

68.2 (10.8)

25.1 (3.9)

0.87 (0.09)

0.97 (0.09)

 Subgroup

75

0.64

0.05

0.90

0.28

0.89

0.06

Oral HRT

 Total

240

58.5 (3.9)

164.6 (5.6)

72.8 (11.9)

26.9 (4.4)

0.89 (0.09)

0.97 (0.12)

 Subgroup

82

0.20

0.18

0.85

0.43

0.80

0.82

Transdermal HRT

 Total

214

55.2 (3.0)

165.0 (6.2)

67.1 (9.7)

24.7 (3.4)

0.89 (0.09)

1.01 (0.08)

 Subgroup

55

0.63

0.76

0.14

0.08

0.27

0.75

Raloxifene

 Total

248

55.3 (3.1)

164.0 (5.9)

70.2 (10.9)

26.1 (4.1)

0.86 (0.09)

0.95 (0.11)

 Subgroup

77

0.22

0.60

0.28

0.20

0.31

0.64

For each study the upper row gives the mean values (SD) and the lower row gives the p-value of the pairwise comparisons between included and non-included subjects

Table 2

Characteristics of subgroups at baseline

 

Bisphosphonate

HRT

RLX

Alendronate

Ibandronate

Nasal

Oral

Transdermal

Raloxifene

n = 40

n = 98

n = 75

n = 82

n = 55

n = 77

Age

52.6a

68.7

53.4a

59.0

55.4b

55.7b

 (yrs)

(±2.1)

(±4.1)

(±1.8)

(±3.8)

(±2.9)

(±3.3)

Height

164.3a

160.4

163.5a

163.9a

165.2a

164.3a

 (cm)

(±5.6)

(±5.2)

(±5.4)

(±5.2)

(±5.5)

(±6.7)

Weight

62.2a

63.1a

68.3c

73.0b

65.4ac

69.1bc

 (kg)

(±6.1)

(±8.6)

(±9.9)

(±12.8)

(±9.4)

(±9.9)

BMI

23.1a

24.5ac

25.6bc

27.2b

24.0ac

25.6bc

 (kg/m2)

(±2.5)

(±3.3)

(±3.8)

(±4.9)

(±3.3)

(±3.7)

BMDhip

0.87ab

0.70

0.87ab

0.89ab

0.91a

0.85b

 (g/cm2)

(±0.08)

(±0.09)

(±0.08)

(±0.08)

(±0.09)

(±0.10)

BMDspine

0.92b

0.74

0.95b

0.97ab

1.01a

0.94b

 (g/cm2)

(±0.09)

(±0.06)

(±0.08)

(±0.12)

(±0.08)

(±0.11)

ααCTX

0.58a

0.71a

0.62a

0.69a

0.76a

0.57a

 (μg/mmol)

(0.22–1.49)

(0.35–1.44)

(0.36–1.08)

(0.38–1.28)

(0.40–1.45)

(0.30–1.08)

ββCTX

2.93abc

2.55bc

3.47a

2.60bc

3.04ab

2.25c

 (μg/mmol)

(1.10–7.79)

(1.53–4.24)

(2.22–5.43)

(1.65–4.09)

(1.83–5.05)

(1.39–3.65)

Ratio

0.21ab

0.30c

0.19a

0.28c

0.26bc

0.27c

 ααCTX/ββCTX

(±0.08)

(±0.10)

(±0.06)

(±0.08)

(±0.07)

(±0.09)

Values shown are mean (± SD), or geometric mean (± 1SD range)

Superscripts: Mean values that have no superscript in common are significantly different from each other at the 5% level (Tukey–Kramer adjusted significance levels)

ααCTX changes

Figure 1 shows relative level of urinary ααCTX in percentage of baseline values in the postmenopausal women during the six anti-resorptive therapies for 24 months. All treatments were highly efficient in reducing bone resorption with a fully expressed treatment effect after six months of treatment. The most pronounced decrease was associated with the two bisphosphonate regimens with a time-averaged decrease of 91% in the alendronate group and 80% in the ibandronate group. The decrease associated with the HRT treatments ranged between 63% to 75%, and the lowest decrease of 36% was found in the raloxifene group. The decrease in the active treatment group as compared with the study-specific placebo group was statistically highly significant for all six regimens (p < 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-007-0462-5/MediaObjects/198_2007_462_Fig1_HTML.gif
Fig. 1

Urinary ααCTX in percentage of baseline values in postmenopausal women during 24 months of anti-resorptive therapy. Upper left shows the placebo-corrected changes, and upper right the placebo-corrected time-averaged mean during the treatment period. The lower panels show values relative to baseline in each study in the placebo (○) and active treatment groups (●). AL (alendronate); IB (ibandronate); HRT-N (intranasal HRT); HRT-O (oral HRT); HRT-T (transdermal HRT); RLX (raloxifene). Values shown are geometric mean±1SEM. The level of significance denotes difference from the placebo group: *p < 0.05; ***p < 0.001

ββCTX changes

The relative changes of urinary ββCTX during treatment mirrored largely those of ααCTX for all treatment regimens, although subtle differences with slightly less suppressed levels of ββCTX were observed (Fig. 2). Alendronate treatment was associated with the highest reduction in ββCTX with a time-averaged decrease of 82%. A decrease of 69% was found in ibandronate group, and the HRT treatments ranged between 62% to 71%. The lowest decrease of 26% was found in the RLX group. The decrease in the active treatment group as compared with the study-specific placebo group was statistically highly significant for all six regimens (p < 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-007-0462-5/MediaObjects/198_2007_462_Fig2_HTML.gif
Fig. 2

Urinary ββCTX in percentage of baseline values in postmenopausal women during 24 months of anti-resorptive therapy. Upper left shows the placebo-corrected changes, and upper right the placebo-corrected time-averaged mean during the treatment period. The lower panels show values relative to baseline in each study in the placebo (○) and active treatment groups (●). AL (alendronate); IB (ibandronate); HRT-N (intranasal HRT); HRT-O (oral HRT); HRT-T (transdermal HRT); RLX (raloxifene). Values shown are geometric mean±1SEM. The level of significance denotes difference from the placebo group: **p < 0.01; ***p < 0.001

Ratio of ααCTX to ββ CTX changes

The ratio between ααCTX and ββCTX, reflecting the maturation of the resorbed bone collagen, changed most markedly in the two bisphosphonate groups (Fig. 3). After 6 months of treatment the ratio in the alendronate group was decreased by 38%, and progressively decreased with 50% at 12 months, and 61% at 24 months of treatment. In the ibandronate group, the ratio was decreased with 40% at 6 months of treatment, and remained decreased with 36% at 12 months of treatment. The placebo-corrected changes in the ratio was highly statistically significant with a time-averaged reduction in the alendronate group of 52% during the 2-year treatment period (p < 0.001), and a 38% reduction in the ibandronate group during 12 months of treatment (p < 0.001). In comparison the reductions associated with HRT treatment were between 3% and 15%, all being statistically non-significant. Treatment with RLX resulted in a borderline significant decrease of 14% (p = 0.05), comparable to that of HRT. The absolute levels of ratio at baseline and after 24 months of treatment are given in Table 3. In absolute levels the lowest ratios were seen in the two bisphosphonate treatment groups.
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-007-0462-5/MediaObjects/198_2007_462_Fig3_HTML.gif
Fig. 3

Urinary ααCTX/ββCTX ratio in percentage of baseline values in postmenopausal women during 24 months of anti-resorptive therapy. Upper left shows the placebo-corrected changes, and upper right the placebo-corrected time-averaged mean during the treatment period. The lower panels show values relative to baseline in each study in the placebo (○) and active treatment groups (●). AL (alendronate); IB (ibandronate); HRT-N (intranasal HRT); HRT-O (oral HRT); HRT-T (transdermal HRT); RLX (raloxifene). Values shown are geometric mean ±1SEM. The level of significance denotes difference from the placebo group: *p < 0.05; **p < 0.01; ***p < 0.001

Table 3

ααCTX/ββCTX Ratio (mean and 1 SEM) at baseline and at 24 months

 

Placebo

Treatment

n

Baseline

24 Mths

n

Baseline

24 Mths

Alendronate

13

0.21 (0.01)

0.25 (0.03)

27

0.21 (0.02)

0.12 (0.02)

Ibandronate

26

0.29 (0.02)

72

0.30 (0.01)

0.15 (0.01)

Nasal HRT

25

0.18 (0.01)

0.22 (0.01)

50

0.19 (0.01)

0.22 (0.01)

Oral HRT

33

0.29 (0.01)

0.29 (0.02)

49

0.27 (0.01)

0.23 (0.02)

Transdermal HRT

20

0.27 (0.01)

0.28 (0.01)

35

0.26 (0.01)

0.24 (0.01)

Raloxifene

30

0.27 (0.02)

0.28 (0.02)

47

0.27 (0.01)

0.24 (0.01)

BMD changes

The anti-resorptive treatments of the two bisphosphonates and the three HRT regimens increased BMD in a comparable manner (Fig. 4). The placebo-corrected yearly increases ranged from 1.6% to 2.8% for BMDhip, and between 2.7% to 4.1% for BMDspine. These treatment effects were statistically highly significant (p < 0.001). In comparison, treatment with raloxifene was associated with lower increases of 0.8% and 0.9% per year in BMDhip (p < 0.01) and BMDspine (p < 0.001).
https://static-content.springer.com/image/art%3A10.1007%2Fs00198-007-0462-5/MediaObjects/198_2007_462_Fig4_HTML.gif
Fig. 4

Changes in percentage per year in BMDhip (a) and BMDspine (b) in postmenopausal women during anti-resorptive therapy. AL (alendronate); IB (ibandronate); HRT-N (intranasal HRT); HRT-O (oral HRT); HRT-T (transdermal HRT); RLX (raloxifene). In the individual studies the values were adjusted for the corresponding placebo group. Values shown are mean ±1SEM. The level of significance denotes difference from the placebo group: **p < 0.01; ***p < 0.001

Discussion

We have investigated if different anti-resorptive treatments cause diverse age profiles in excreted bone collagen degradation products in face of similar effects on BMD. We measured urine degradation products of collagen type I (both the non-isomerized younger ααCTX isoform and the older mature ββCTX isoform), which are released during osteoclastic resorption of bone.

The main findings were 1) treatment with bisphosphonate, HRT, or RLX was associated with a statistically highly significant suppression in bone resorption of both newly synthesized (ααCTX) and mature (ββCTX) collagen type I degradation with the most pronounced suppression observed in the bisphosphonate groups, and 2) treatment with bisphosphonate induced a lower ratio between ααCTX and ββCTX compared to that of HRT and RLX.

Bisphosphonates are potent agents that target the hydroxyapatite bone mineral surfaces in vivo at sites of active bone remodeling, inhibiting osteoclast activity [29]. Other agents for prevention of osteoporosis include HRT and RLX with a different mode of action. HRT co-administrated with progesterone is a rational approach to counter the consequences of estrogen deficiency in postmenopausal women, whereas RLX acts as an estrogen agonist on skeletal tissue. The new generation of drugs are Zoledronic acid, a bisphosphonate, and Denosumab, which is a high affinity, fully human monoclonal antibody binding to RANKL and blocking the interaction between RANKL and RANK, mimicking the endogenous effects of osteoprotegerin [30, 31].

The bone collagen maturation should preferably be measured in the bone matrix, but the degree of isomerization of bone collagen can non-invasively be measured as the ratio between ααCTX and ββCTX in urine samples. Using the surrogate marker in urine samples, one can assess the dynamic changes in the bone collagen occurring during treatment. In the present study we found that all six anti-resorptive treatment regimens were highly efficient in reducing the level of bone resorption as measured by urinary ααCTX and ββCTX although to different levels. The most suppressed levels were found in the bisphosphonate groups of alendronate and ibandronate. This higher suppression seems to be related to the drugs rather than caused by differences in the study cohorts as the bisphosphonate treatment groups represented both the youngest and the eldest subjects among the studies. The decreases in levels of ααCTX and ββCTX mirrored each other, although ααCTX was slightly more suppressed than ββCTX. This higher reduction in ααCTX is likely the result of attenuated bone turnover caused by the anti-resorptives (2–6). This is in line with the increase in the ratio between ααCTX and ββCTX of 1.9% per year since menopause in untreated postmenopausal women reflecting the general increase in bone turnover caused by the menopause leading to a decrease in skeletal age and to the paradigm that older postmenopausal women may have on average a less mature skeleton than younger postmenopausal women. However, we have not specifically investigated the difference between pre- and postmenopausal women. In particular the bisphosphonate treatments induced a pronouncedly lower ratio between ααCTX and ββCTX with a reduction of 52% in the alendronate group and 38% in the ibandronate group compared to 3 to 15% reduction in the HRT and raloxifene groups. The treatment-dependent effect on the ratio could not be explained by the study-specific placebo-adjustments used in the calculations as the time-averaged mean used for adjustments were 116% in the alendronate, 77% for ibandronate, 117% for nasal HRT, 91% for oral HRT, 101% for transdermal HRT and raloxifene.

Several lines of evidence suggest that BMD gains only to a certain extent explain fracture reduction. In a study on the relationship between change in BMD and vertebral fracture risk in women treated with risedronate, BMD increases of 2–4% were correlated to a decrease in fracture risk, whereas an increase above this threshold presented no additional benefit [32]. Accordingly, others have demonstrated that improvement of BMD explained only a small part of the risk reduction in RLX, bisphosphonate, and calcitonin-treated patients [33, 34]. These discrepancies may partly be accounted for by the fact that treatment effect on bone matrix quality is important to anti-fracture efficacy and a certain level of remodeling is needed to sustain an optimal quality of the skeleton [35].

We recognize the limitations of our study. The analysis is based on randomly selected subgroups from six individual double-blind, placebo-controlled studies. The subgroups differed in size both within the studies and across studies, but the selection is assumed not to affect the reliability of the data based on the randomness in the selection process. Furthermore the urine samples were analyzed without knowledge of the treatment group. All studies included healthy postmenopausal female volunteers, but inclusion criteria differed between studies. However, even though the two bisphosphonate groups represented the youngest and the oldest cohorts in mean age, the two bisphosphonate treatments were associated with a comparable and most pronounced decrease in the ααCTX to ββCTX ratio. The observed differences in treatment effects are, therefore, most probably a consequence of the treatments and not caused by differences between populations. Furthermore, one may speculate whether storage of the urine samples for prolonged time at −20°C affects the results. To our knowledge we have not been able to detect any changes in the measured parameters caused by storage time at this temperature. Also, the data presented are placebo-adjusted values (adjusted within the individual study), if any change occurred it has been corrected by the placebo-adjustment. Finally one may speculate whether differences in the baseline bone turnover rate could explain the observed treatment differences. However, the data of the bone resorption markers of ααCTX and ββCTX, measured at exactly the same time in all the subgroups, were at comparable levels e.g., in the alendronate and the nasal HRT group, yet the treatment response differed between these groups.

In conclusion, we have for the first time demonstrated that different anti-resorptive therapies induce differences in the maturation profile of bone collagen measured as the ratio between ααCTX and ββCTX.

Funding

Conflict of Interest

Inger Byrjalsen, Diana J Leeming, and Per Qvist are employees of Nordic Bioscience A/S, and Per Qvist , Claus Christiansen, and Morten A Karsdal are stock owners of Nordic Bioscience A/S.

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2007