The Role of Bone Turnover Markers in the Follow up of the Treatment of Osteoporosis

Although bone mineral density (BMD) measurement is the common method for the diagnosis of postmenopausal osteoporosis (PMO, bone turnover markers have been proposed as good indicators for monitoring the efficiency of antiresorptive treatment. We aim to evaluate course of bone turnover markers in the follow-up of treatment osteoporosis and compare the results of these markers with bone mineral densities. 47 postmenopausal patients were included in this study. The age average of the patients was 59.1  7.6 (48-81). Patients were divided into 5 groups depending on the type of the treatment: Calcium and vitamin D (8 patients); calcium, vitamin D and risendronate (10 patients); calcium, vitamin D and alendronate (10 patients); calcium, vitamin D and hormone replacement therapy (tibolone) (9 patients) and calcium, vitamin D and calcitonine (10 patients). Samples were taken from each group, prior to treatment and in the third and sixth months of the treatment. Bone mineral density measurements were carried out before the treatment and in the sixth month. The basal concentrations of bone turnover markers  serum Type I collagen C-telopeptide (CTX), N-mid-Osteocalcin (OC) and urinary deoxypridinoline (Dpd)  and Interleukin-1  (IL-1  ) were compared to the values in the third and sixth month of the therapy and also to BMD.OC, CTX and Dpd are good markers for the evaluation the effectiveness antiresorptive therapy. The changes in CTX and OC seems to be reflect the changes in bone mass in the early period. However, the changes in Dpd levels come out later. HRT (tibolone) and biphosphonates especially risendronate are thought to be effective therapeutic approaches, because a significant decrease in bone turnover markers and an improvement in BMD – especially in lombar vertebra - were observed in the follow up of treatment. A good correlation found between the changes in levels of CTX and Dpd and changes in BMD during treatment, which suggest that these two markers would be useful for


Introduction
PMO is a disease characterised by low bone mineral denisty and microarchitecural impairment -as result of reduction in circulating estrogen causing an accelerated bone turnover and resorption-which lead to bone fragility and an increase risk of fractures. It is a very serious health problem leading to significant morbidity, disability, mortality and decreased quality of life (1,2). Bone mineral density (BMD) measurement is the most common method used for the diagnosis and the follow-up of the treament of osteoporosis (3). However, changes in BMD measured by densitometry are not detected until the bone loss is great, therefore it would be useful to have dynamic tests that could reflect the changes earlier (4). In recent years, for the evaluation of efficiency of the treatment of osteoporosis, bone formation and resorption markers are being used. With these markers, earlier information can be obtained when compared to bone mineral density measurement, so that they may be used to detect patients who do not respond to therapy (5,6). The magnitude and the timing of reduction (within weeks) in the levels of bone turnover markers are thought to be their advantage when compared to BMD measurements by which slower changes and smaller magnitude are obtained (7).
Antiresorptive agents (biphoshonates and calcitonin), hormone replacement therapy (ERT-tibolone) and calcitriol (1,25 dihydroxyvitamin D) with calcium supplementation are the most widely used treatments in PMO. Biphosphonates directly reduce the bone turnover by inbiting the activity of osteoclasts (3,8) . Randomized placebo-controlled studies with alendronate and risendronate show their effect on inducing the gain of BMD, reducing the increased rates of bone turnover and decreasing the risk of fractures (2,9,10). In previous studies, it was shown that calcitriol increases bone mass not only at lumbar spine but also at femoral neck (11,12). Another common therapeutic approach, calcitonin stimulates the increase in bone mass and also decreases in bone resorption markers and OC levels (13,14). Tibolone, synthetic estrogenic, prosgestagenic and androgenic steroid, reduces bone resorption by decelerating bone turnover (15). Recent studies have revealed that tibolone is effective in incerasing the bone mass and that was proved by the increase in BMD measurement and decrease in bone turnover markers (16,17).
In this study, we aim to show that the bone turnover markers are as reliable as BMD measurement in the follow-up of the treatment of PMO and reflect the changes in bone mass earlier and also, to select the earliest marker that can reflect the changes in BMD.

2.1.Subjects
Forty-seven postmenopausal women aged 48-81 (59.17.6) who were admitted to Fatih University Faculty of Medicine Department of Obstetrics and Gynecology were taken in this study. Osteoporosis was confirmed with BMD measurement.
The patients were in menopause for 5 years and diagnosis of osteoporosis was made according to the criteria declared by World Health Organisation. None of the patients had used medication neither for osteoporosis nor effecting bone mass and none had a disease that effects bone metabolism or cause secondary osteoporosis.
All patients were given Ca 1500 mg, D vitamini 400 IÜ daily. They were divided into 5 groups depending on the type of the treatment: only calcium and vitamin D (8 patients); calcium, vitamin D and risendronate (10 patients); calcium, vitamin D and alendronate (10 patients); calcium, vitamin D and hormone replacement therapy (tibolon) (9 patients) and calcium, vitamin D and calcitonine (10 patients). Blood and urine samples were taken from each group, prior to treatment and in the third and sixth months of the treatment. Bone mineral density measurements were carried out before the treatment and in the sixth month.

Assays
The blood and urine specimens were immediately santrifuged, seperated and stored at -20 o C till analysis. Serum calcium (Ca), phosphorus (P), creatinine (Crea), alkaline phosphatase (ALP) levels and urinary calcium and creatinine levels were measured by Hitachi 912 (Roche Diagnostics Co., Mannheim, Germany) autoanalyzer. The creatinine clearance is calculated according to the formula. Serum IL-1 and urinary DPD (Pyrilinks-D) levels were measured by IMMULITE One Analyzer (DPC Diagnostic Products Co., CA, USA) using chemiluminescent immunoassay. Serum CTX (-crossLabs) ve N-mid OC levels were measured by Roche Elecsys 2010 Analyzer (Roche Diagnostics Co., Mannheim,Germany) using electrochemiluminescent immunossay. All the analysis were made in single run after and calibration of the assay was controlled by spesific control seras. The inter-assay and intra-assay coefficients of variation (%CV) are shown in the Table 1.

2.3.Bone Mineral Density
BMD of lumbar vertebra (L2-L4), femur neck and Ward's triangle were measured by dual energy X-ray absorptiometry (Lunar Radiation Corporation) at the beginning and in the sixth month of the treatment.

2.4.Statistics
The sattistical analysis was performed with SPSS for Windows 9.0 software programme. One-way analysis of variance (ANOVA) and Banferroni test were used to compare the changes between groups for normally distributed variables and Kruskall-Wallis analysis of variance for the non-normally distributed variables. To compare the changes due to time with in the groups, ANOVA for the normally distributed variables and Paired t ve Wilcoxon test were used. Pearson correlation analysis method was used for the correlation analysis. Results were given as mean  SD. Statistical significance was assumed with p<0.05.

Results
The groups were well matched in age, the age of menopause and menarche ??, BMI (Body Mass Index), The increase in lombar vertebra, femur neck ve Ward's triangle scores in the sixth month were significantly different for the group I and III when compared to baseline values (p<0.05). For the groups II, IV and V significant difference was seen only in lomber vertebra scores (p<0.05). During the siz month follow up, the changes in bone turnover parameters in time and in and between the groups are shown in Figure1 and Table 4.  borderline difference due to time and treatment groups 2 : significant difference due to time and treatment groups (p<0.01) Baseline OC levels were found similiar between the groups when we compare the five treatment groups. However, in group I the elevation in the 3rd and 6th month found different when compared to other groups (p<0.05). Dpd levels were found significantly lower in 3rd month in group I (p<0.05) and significantly higher in group III (p<0.05). For serum Ca levels, baseline values were significantly higher in group I and lower values in group II and 3rd month values were found significantly higher in group I and lower in group II and IV (p<0.05). Serum baseline P levels were found only different in group I (p<0.05). In the 3rd month, the elevation in group I and the fall in group IV and in the sixth month the elevation in group I and fall in group II and V were found significant (p<0.05). CTX, IL-1, ALP and urinary Ca levels show no difference in groups.
When we compare the parameters in group I due to time, OC levels decrease in the 6th month when compared to baseline and 3rd month (p<0.01). Only significant decrease in CTX levels was seen in between 3rd and 6th month (p<0.01). Dpd levels were found to decrease in 3rd month (p<0.01) and IL-1 levels were found significantly different between the baseline and sixth month (p<0.05). Significant decrease was seen in 6th month when compared to 3rd in serum ca levels (p<0.05). The decrease in all ALP and urinary Ca levels were significant (p<0.05), but no difference was observed in P levels.
For the parameters in group II, OC levels in the 3rd month compared to baseline and in the sixth month compared to 3rd were found significantly lower (p<0.01,p<0.05). CTX levels were lower in 3rd and 6th month compared to baseline (p<0.05). Dpd levels in the 6th month were lower than the baseline and the 3rd month(p<0.05). The decrease in IL- and ALP levels were found significant although they were in normal range (p<0.05). No significant difference were noted for urinary Ca and serum P levels.
OC and CTX were foun lower only in the 6th month compared to baseline(p<0.05) in the group III. The difference in Dpd and IL-1 levels were detected in the 6th month when compared to baseline and 3rd month(p<0.01, p<0.05). No significant difference were noted for serum and urinary Ca, serum P and ALP levels.
We found significant decrease in OC, CTX and Dpd levels in the 3rd and 6th month in group IV when compared to baseline (p<0.05), however the difference between the 3rd and the 6th month was insignificant. The significant decrease were seen between the 3rd and the 6th month for IL-1 levels (p<0.01), in the 3rd month for ALP levels (p<0.05) and in the 6th month for P levels (p<0.05).
In group V, Dpd levels were significantly lower in 3rd month than baseline and in 6th month than 3rd month (p<0.05) and IL-1 levels decrese in the 6th month compared to 3rd month (p<0.05).
When we evaluate the results depending on the comparison either between the groups or due to time, we found that the difference for OC (p= 0.68) and Dpd (p=0.71) were at/in? borderline, for CTX it is significant (p<0.01), and for IL-1, serum Ca, ALP, P ve urinary Ca it is insignificant.
BMD scores seems to correlate with OC, IL-1, Dpd and P levels at the beginning of the treatment and with OC, CTX and Ddp levels in the 6th month. A correlation was observed between OC and CTX levels not only at the beginning but also in the 6th month.

Discussion
Osteoporosis is a common health problem for the postmenopausal women. In recent years, researchers aim to find new bone spesific parameters that will facilitate the early diagnosis and the follow-up of osteoporosis.
Osteoporosis is diagnosed by BMD measurement. It was shown that increased bone turnover is correlated with bone loss. Bone loss can be evaluated by bone turnover markers because, the bone mass is determined by bone turnover (18,19). These markers are used to give information about the BMD in an early period than BMD measurement (20,21) Osteoclastic bone markers Tartarate resistant acid fosphatase (TRAP), Hyroxyproline (Hyp), Pridinoline (Pyd), Deoxypiridinoline (Dpd), Type I collagen N-telopeptide (NTX), Type I collagen C-telopeptide (CTX) reflect the bone resorption and formation markers derived from osteoblasts Osteocalcin (OC), alkaline phosphatase (ALP), type I collagen propeptides) inform us about the future bone loss (20,22,23). In 60-70% of women with early postmenopausal osteoporosis , TRAP,OC, urinary Ca, Hyp and Dpd show bone loss (18,19). Chaki et al. noted that bone spesific ALP, CTX, NTX, OC, Pyd, Dpd can be used to predict future BMD in postmenopausal women (24).
Calcitonin is commonly used in the osteoporosis treatment (25,26). In short-term and long-term placebo controlled studies calcitonin was denoted to be an effective treatment (26,27).
In two previous studies, treatment with tibolone increased BMD and decreased OC and CTX levels and this deceleration in markers are found to be inversely proportional to BMD (28,29). It was also revealed that serum CTX levels reflect the bone loss and future fracture risk (30). Rymer et al. showed that tibolone increased vertebra and femoral neck BMD and cause a decrease in OC and ALP levels in three different long term studies (26,32,32). Also, some other studies support these results and verify the efficiency of tibolone on BMD (17,(33)(34)(35). Parviainen et al.,Christagau et al. and Bjarnason et al noted that tibolone significantly decreases Dpd,.
In their placebo controlled long-term study Greenspan et al. concluded that CTX and NTX predict longterm changes in vertebral BMD in women receiving alendronate (39). Combined treatment with biphosphonates and HRT had a favorable effect on BMD and also decrease bone spesific ALP, NTX, CTX, Dpd levels (40-42). Also, OC levels well correlate with the improvement in BMD made by combined treatment with calcitonin and HRT and can be used in the follow-up of treatment (13,14,43,44). And also by calcitonin therapy while the BMD increases, CTX levels decreases (45). However, there are adverse results of studies about calcitonin in which no change in OC levels were observed (46,47).
In this study, a bone formation marker -OC, two resorption markers -CTX and Dpd-and a cytokine -IL-1-which thought to have effect on bone turnover were analysed and the suitable marker/markers that can reflect future KMD and that can be used to evaluate the efficiency of treatment are aim to be investigated. Just about all the parameters analysed start to decrease in the first three months. However, in risendronate and tibolone groups the decrease were significant. The decrease in OC, CTX and Dpd levels in the 6 th month correlate with the KMD scores.
It is known that the antiresorptive treatment given inhibits bone resorption and indirectly effects on bone formation. In our study, CTX and OC decreases in four groups except the calcitonin group. Also, in tibolone and risendronate group the decreased revealed earlier (in the 3rd month). OC and CTX levels were in correlation during the six months period. The decrease in Dpd levels found insignificant only in the group which was given only Ca and vitamin D. Other groups had shown signifiant decrease in the 6th month, also, this descending began in the 3rd month in tibolone group. No statistical difference was found between CTX and OC levels in evaluating the effectiveness of treatment or reflecting the future bone status.
Estrogen defficiency has the major role in postmenopausal osteoporosis. Recent studies direct us to a fact that, estrogen stimulates the secretion of local factors that effect bone remodeling and in the absence of estrogen this mechanism is impaired (48). IL-1 is one of these factors that stimulates the differentiation and activation of osteoclasts. IL-1 is inreased in some cases with estrogen deficiency except menopause and this supports the modulator effect of estrogen on cytokines (49-52). IL-1 levels were significantly higher in women with potmenopausal osteoporosis compared to healthy ones and also found inversely proportional to the BMD (53,54). Also, studies with different results exist. In these studies no difference was observed in IL-1 levels in postmenopausal period and no correlation was detected with BMD (55,56).
In our study, the IL-1 levels were decreased in risendronate in the 3rd month, but significant difference was detected in other groups in the 6th month. No correlation between IL-1 levels and BMD were observed. This prevensts IL-1 to be the marker.
Only the change in CTX levels between groups and due to time is significant. The significance for OC and Dpd is in/at?? borderline. The changes in IL-1, serum Ca, ALP, P and urinary Ca levels are insignificant.
Lombar vertebra BMD scores were significantly elevated in all treatment groups. The increase in femur neck and Ward's triangle scores are increased in Ca and vitamin D group and in risendronate group. The correlation between the BMD and OC and CTX were significant in the 6th month.
We concluded that, CTX and OC are the suitable markers for the follow-up of the treatment of postmenopausal osteoporosis. The changes in their levels also give information about the future-period bone mineral density. Considering the decrease in these parameters and the increase in the bone mineral density, the most appropriate treatments were determined as tibolone and risendronate and with these treatments, improvement in bone mineral density became most evident in lomber vertebra.