INTRODUCTION
The imbalanced bone remodeling in osteoporosis (OP) is associated with an interaction of genetic, hormonal, nutritional, and environmental factors.[1,2] Genetic factors comprise up to 50% to 85% of the risk factors for OP among postmenopausal women.[2] Several genes are assumed to play a role in the pathogenesis of OP, such as vitamin D receptor (VDR) gene, Col1a1, estrogen receptor (ER) gene 1 (ER1 or ERα) and ER gene 2 (ER2 or Erβ). VDR is the first and the most extensively studied candidate gene in the pathogenesis of OP.[3,4] Vitamin D plays a key role in skeletal homeostasis and severe vitamin D deficiency leads to impaired mineralization of bone matrix (i.e., osteomalacia and rickets) while less severe deficiency induces bone loss via secondary hyperparathyroidism.[5] ER1 merits attention as the deficiency of estrogen plays a major role in the pathogenesis of postmenopausal OP.[6] ER1 and ER2 polymorphisms are among the most well-known gene variations associated with OP.[2] ER1 gene has highly polymorphic sites, the most widely studied of which are the PvuII and XbaI.[4] Col1a1 gene is one of the most popular candidate genes associated with OP. Col1a1 encodes the α I chain of type 1 collagen, the major bone matrix protein.[7,8] The Sp1 polymorphism of Col1a1 gene is associated with low bone mineral density (BMD) and a higher risk of fracture. [7,9]
The role of genetic polymorphisms in anti-osteoporotic drug response, i.e., pharmacogenetics of OP, has been a focus of interest in the last decade.[5,10] Bisphosphonates (BPs), being the most commonly used medications, are effective in reducing the risk of vertebral and non-vertebral fractures by increasing or preserving the bone mass.[11] Stable or improving BMD is assumed to be a sign of reduction in the loss of bone mass and hence the risk of fracture, which both indicate a good treatment response. However, a notable portion, up to 53% of the patients with OP demonstrates a decrease in BMD under treatment, indicating a poor response.[5,12]
Individualized drug therapy based on pharmacogenetic data would be rational since drugs are ineffective in some patients.[5,12] Besides, a long period of time, at least one year is required to evaluate the response to BP treatment. BP may cause adverse effects and complications as well, such as osteonecrosis of the jaw or gastroesophageal reflux in some individuals.[11]
Most studies examine the effect of polymorphisms on response to hormone replacement therapy and raloxifene which are no longer used so frequently in OP.[8] However, there is a small number of studies on pharmacogenetics of BP and the results are controversial.[8,10,13]
The aim of this study is to examine the effects of VDR BsmI, Col1a1 Sp1, ER1 PvuII and XbaI gene polymorphisms on the response to BP in Turkish women with postmenopausal OP.
METHODS
1. Study design
This prospective study was conducted at the Department of Physical Medicine and Rehabilitation, Gazi University School of Medicine in Ankara, Turkey. The study was approved by Gazi University Medical Ethics Committee. The study was funded by the coordinators of scientific research projects of Gazi University Medical School. Written informed consents were obtained from all participants. All the procedures were performed according to the World Medical Association Declaration of Helsinki.
2. Study population
Thirty-three postmenopausal women were recruited from physical medicine and rehabilitation outpatient clinic (Fig. 1). Women who were diagnosed as postmenopausal OP according to World Health Organization diagnostic criteria for OP [14] and who were not on any medication for OP for at least 3 months following their scheduled time of medication (e.g., 15 months after the last dose of zoledronic acid or 4 months after the last dose of monthly ibandronic acid) were included in the study. Exclusion criteria were using drugs that might affect bone metabolism such as parathyroid hormone, steroids or cytostatic drugs, metabolic bone diseases like Paget’s disease, osteomalacia or other diseases affecting bone metabolism such as pituitary gland, thyroid, parathyroid diseases, malignancy, immobilization, anorexia or chronic inflammatory diseases.
Participants who commenced alendronate sodium 70 mg/week or ibandronate 150 mg/month, or zoledronic acid 5 mg/year were included in this study. To assess level of physical activity, they were asked whether they attended to any kind of aerobic exercise including walking for 30 min at least 5 days a week. Daily consumption of calcium-rich foods was evaluated using self-reported questions. All subjects were provided 600 mg calcium and 400 IU vitamin D daily. Vitamin D replacement was administered to thirteen patients with low levels of serum 25-hydroxy-vitamin D. Lifestyle modifications such as a balanced diet rich in calcium and vitamin D, regular exercise, smoking cessation, and adequate exposure to sunlight were advised to all patients.
Treatment adherence and side effects were evaluated during physician visits for repeat prescriptions at every 3 months. Twelve patients were lost to follow-up due to several reasons and twenty-one patients completed the study.
3. Data collection
Dual energy X-ray absorptiometry (Hologic QDR 4500; Hologic Inc., Bedford, MA, USA) was used to measure BMD. The equipment was calibrated daily using a standard spine phantom provided by the manufacturer. Lumbar spine L1-L4 and femoral neck BMD was measured on the same machine at baseline and 12 months after the treatment. BMD was expressed as grams per centimetre square (g/cm2) and as T scores which indicate the standard deviations of individual BMD determinations compared to those of young adults. Women who had a T score equal to or lower than −2.5 in at least one of the two above mentioned regions were diagnosed as OP.[14]
Drug response was assessed based on the change in BMD, at baseline and after 12 months of treatment with BP. Drug response was classified as “inadequate” if there is both an incident fracture and a >2% decrease in BMD, “possibly inadequate” if there is either an incident fracture or a >2% decrease in BMD, and “adequate” if there is no fracture and no decrease in BMD >2%.[15]
Serum samples were drawn after an overnight fasting at baseline and after 12 months of treatment. Serum levels of osteocalcin which is a marker of bone formation reflecting osteoblastic activity [16] were analysed. BP is deemed effective when normal premenopausal levels of bone turnover markers are obtained after therapy. A significant decrease in bone turnover markers within 3 to 6 months is associated with a positive long-term treatment response to BP.[16]
Polymorphisms of the genes encoding ER1, VDR and Col1a1 proteins were studied. DNA was extracted from peripheral white blood cells by salting out method. Polymorphic sites of the relevant genes were amplified by PCR and examined using restriction fragment length polymorphism. VDR gene BsmI polymorphism was defined as BB (absence of a restriction site on both alleles), Bb (heterozygous), and bb (presence of a restriction site on both alleles). ER1 gene PvuII and XbaI polymorphism was defined as PP or XX (absence of a restriction site on both alleles), Pp or Xx (heterozygous), and pp or xx (presence of a restriction site on both alleles), respectively. Col1a1 gene Sp1 polymorphism was defined as SS (absence of a restriction site on both alleles), Ss (heterozygous), and ss (presence of a restriction site on both alleles).
4. Statistical analysis
Data were analysed using SPSS for Windows statistical package, version 20 (SPSS Inc., Chicago, IL, USA). Continuous variables were presented as a median and interquartile range (IQR) while categorical were presented variables as numbers and percentage. A χ2 test was used to confirm that the gene polymorphisms were in Hardy-Weinberg equilibrium. Associations between gene polymorphisms (VDR gene Bsm1, Col1a1 gene Sp1, ER1 gene PvuII and XbaI groups) and age, body mass index (BMI), BMD, and serum osteocalcin levels were analysed by Kruskal-Wallis analysis of variance and Mann-Whitney U test. The patients were categorized in terms of response to BP treatment and the χ2 test was used to compare the genotype frequencies of the responder and non-responder groups. A P value of less than 0.05 was considered statistically significant.
RESULTS
In total 33 postmenopausal women on BP treatment were screened, 12 of them were excluded due to several reasons and 21 patients were included in the study (Fig. 1). Demographic characteristics, osteocalcin levels and BMD of lumbar spine and femoral neck of the 21 women with postmenopausal OP who completed the study are presented in Table 1. Two patients were newly diagnosed. Median (IQR) duration of OP was 7 (9) years. Median duration of previous antiresorptive drug use was 2 (4) years. Ten patients (48%) were unable to achieve the minimum recommended physical activity. Dietary intake of calcium was adequate in all patients. None of the patients had previous long-term use of steroids.
Ten patients received ibandronic acid (47.6%), 6 patients received zoledronic acid (28.6%), and 5 patients received alendronate sodium (23.8%). Previous history of osteoporotic fracture was present in 2 (40%), 3 (30%), and 1 (17%) patient in the alendronate, ibandronic acid and zoledronic acid groups, respectively.
During one year of treatment period, no adverse effect was observed except for one patient who had flu-like symptoms for a few days following zoledronic acid injection. No incident fractures occurred during the study.
When categorized according to gene polymorphisms, age and BMI were similar among different polymorphisms except for Col1a1 gene, in which the median age of Ss subjects was significantly lower than that of SS (median [IQR] ages were 59 (16) years for Ss and 73 (10) years for SS; P= 0.035). Baseline serum levels of osteocalcin and BMD of lumbar spine and femoral neck were similar among subjects with different polymorphisms.
The frequency and distribution of VDR Bsml polymorphism was consistent with the expected frequency by the Hardy-Weinberg equilibrium law (χ2=0.192, P=0.661). The genotype distribution of VDR Bsml polymorphism was as follows: BB, 28%; Bb, 48%; bb, 24%. The increase in BMD and decrease in osteocalcin levels after treatment was higher in the Bb group, however the difference was not statistically significant (Table 2, Fig. 2). When the participants were regrouped as bb+Bb and BB, change in BMD of lumbar spine (0.022 and −0.006 g/cm2; P=0.095) and femoral neck (0.032 and −0.006 g/cm2; P=0.622) were greater in bb+Bb group, however the difference did not reach statistical significance. Similarly, change in serum osteocalcin levels was in favor of bb+Bb group, although not statistically significant (-0.03 and 1.89 ng/mL; P=302).
The genotype frequencies for the Col1a1 gene Sp1 polymorphism were 86 % for SS (N=18) and 14% for Ss (N=3). No patient had ss genotype. The distribution of frequency was consistent with the expected frequency by the Hardy-Weinberg equilibrium law (χ2=0.593, P=0.441). The increase in BMD and decrease in osteocalcin levels after treatment were higher in the SS group but the difference was not statistically significant (Table 2, Fig. 2).
The frequency and distribution of ER1 PvuII polymorphism was consistent with the expected frequency by the Hardy-Weinberg equilibrium law (χ2=0.192, P=0.661). The genotypic distribution of ER1 PvuII polymorphism was as follows: Pp, 47.6%; pp, 28.6%; PP, 23.8%. The increase in BMD and decrease in osteocalcin levels after treatment was higher in the PP group but the difference was not statistically significant (Table 2, Fig. 2).
The frequency and distribution of ER1 XbaI polymorphism was consistent with the expected frequency by the Hardy-Weinberg equilibrium law (χ2=0.445, P=0.505). The genotypic distribution of ER XbaI polymorphism was as follows: xx, 42.85%; Xx, 42.85%; XX, 14.3%. There was no significant difference in BMD or osteocalcin levels among genotype groups after treatment (Table 2, Fig. 2).
Drug response was “adequate” for 16 patients (76%), “possibly inadequate” for 5 patients (24%) in lumbar spine. Drug response for femoral neck was “adequate” for 13 patients (62%) and “possibly inadequate” for 8 patients (38%). There was no significant difference in treatment response in BMD (lumbar or femoral neck) among different gene polymorphisms (Table 3, 4). “Possibly inadequate” treatment response was not observed in any patient in ER1 PvuII pp and ER1 XbaI xx polymorphisms, though the difference was not statistically significant (P=0.246 and P= 0.079 for ER PvuII and XbaI, respectively).
DISCUSSION
In our study, we did not observe any significant effect of VDR BsmI, Col1a1 Sp1, ER1 PvuII, and ER1 XbaI gene polymorphisms on the treatment response to BP in women with postmenopausal OP. There was no significant difference among genetic polymorphisms regarding changes in BMD or osteocalcin levels.
Few studies examined the genetic pathways of OP and few investigated the effect of genotype on the treatment response to BP.[1,10,11] Although a strong hereditary component of response to BP therapy is reported, it needs further replication.[5] Elucidating the role of genetic factors in OP treatment is essential to improve the efficacy of therapy.[11]
Polymorphisms of VDR gene are assumed to affect the skeletal geometry, bone turnover, and calcium utilization in addition to BMD.[4] The distribution of VDR BsmI genotype in our study was similar to those from our country [4,7] and other countries such as Slovenia,[17] Greece,[13] and Italy.[18] In our study, the increase in BMD (both femoral neck and lumbar spine) and decrease in osteocalcin levels after treatment was higher in the Bb group, however the difference was not statistically significant. There was no significant difference in the ratio of the patients with “possibly inadequate” treatment response between groups.
There are conflicting results regarding the effect of VDR BsmI polymorphism on the lumbar spine BMD increase following BP treatment.[13,17,19] Marc et al. [17] who investigated the association of VDR Bsml genotype and response to etidronate therapy in lumbar spine of postmenopausal OP, increase in BMD was significantly higher in BB and Bb groups compared to bb group. In line with Marc et al’s findings,[17] Palomba et al. [20] also observed a greater increase in lumbar BMD in BB subjects. following 12 months of raloxifene therapy in women with postmenopausal OP. Contrary to these two studies, Creatsa et al. [13] found that patients carrying at least one b allele showed a significantly greater increase in lumbar BMD, while patients with the BB genotype showed a decrease after one-year alendronate therapy. On the other hand, different from these three studies, we found that Bb subjects had a greater, but not statistically significant improvement in lumbar BMD after a oneyear BP treatment. The difference among all these studies is difficult to explain; it might be suggested that VDR Bsml genotype can interact with other polymorphisms. Therefore, more than one polymorphism in the equation would better predict the response to anti-resorptive treatments.
Only one study investigated the effect of VDR BsmI polymorphism on femoral neck BMD.[13] They observed that BB allele demonstrated a greater, but not significant increase in femoral BMD, which was completely the reverse of its effect on lumbar spinal BMD.[13] Our findings were consistent for Bb allele demonstrating the best increases both in femoral neck and lumbar spine BMD.
Most prominent reduction in osteocalcin levels varied across three studies; bb allele,[17] BB allele,[20] or no difference among different alleles.[13] Interestingly, allele with the highest response in BMD increase did not yield with the most prominent osteocalcin reduction in two studies.[13,17] However, in our study, the Bb allele demonstrated both the highest increase in BMD and the most prominent reduction in osteocalcin levels.
The genotypic distribution of Col1a1 Sp1 polymorphism in our study was in line with the literature, as SS was the dominant genotype.[7,21] The median age of the Ss group was significantly lower than SS group in our study. This finding might be of interest and may suggest patients with Ss genotype may have a predisposition for OP at a younger age. However, there was no difference in age among Col1a1 Sp1 genotypes in literature [9,22] which indicates further studies with greater populations are needed.
To our knowledge, our study is the second to investigate the effect of Col1a1 Sp1 polymorphism on response to BP. Qureshi et al. [9] studied the effect of this polymorphism on the response to etidronate in premenopausal women with osteopenia. Similar to this study [9] which considers s allele unfavorable, we observed a greater increase in BMD after BP treatment in patients with SS genotype, although our results were not statistically significant. The decrease in osteocalcin levels were also higher in SS patients while the difference was insignificant.
Polymorphisms of ER1 gene are associated with decreased BMD, increased bone turnover, and a higher risk of fractures in both sexes.[23] ER1 gene PvuII p and XbaI x alleles are associated with decreased femoral neck T-score in women with Crohn’s disease.[23] PvuII p allele was associated with an increased risk of hip fracture both in women and men.[24] On the other hand, Erdogan et al. [21] found significantly higher BMD in pp patients compared to PP, but there was no significant difference in terms of XbaI polymorphism. A systematic meta-analysis did not report any association between ER1 XbaI and ER1 PvuII polymorphisms and postmenopausal OP risk in Caucasian or Asian women.[6]
The genotypic distribution of ER PvuII polymorphism in our study was in accordance with those of the Caucasian population, as Pp was the dominant genotype.[4,21,25] Conflicting results were reported about the association of ER1 genotype and treatment response. Mondockova et al. [26] showed that women with both ER1 polymorphisms had a poorer response to hormone therapy and raloxifene. On the other hand, different from hormone or selective ER modulators, response to alendronate was not found to have a significant association with ER genetic profiles and treatment in a recent study,[27] which is similar to our results. The effect of ER polymorphisms might vary on different classes of anti-osteoporotic drugs due to the distinct mechanisms of drug actions and pathways, i.e., mechanism of BPs might be less associated with ERs.
Our sample size was relatively small due to a high rate of drop-out and difficulties in recruitment of patients meeting the selection criteria. Due to this small size, some associations may be missed. Nevertheless, genotype distributions were consistent with the expected frequencies by the Hardy-Weinberg equilibrium law. We included three different BP, all of which are nitrogen-containing BP with the same molecular target; the farnesyl pyrophosphate synthase enzyme.[28] A larger sample size would enable the analysis of the subgroups for the effects of each BP agent individually. A strength of our study is that it is a prospective study with a long period of follow-up. This study is the first to examine the effects of VDR BsmI and Col1a1 polymorphisms on treatment response to BP in Turkish women.
CONCLUSIONS
The results of our study failed to support the hypothesis that the polymorphisms of VDR BsmI, Col1a1 Sp1, ER1 PvuII and XbaI have an effect on the treatment response to BP. However, there are promising findings which may achieve statistical significance with a larger sample size. Randomized prospective studies with larger sample sizes are necessary to reveal the key role of genes in the variation of response to anti-OP treatment, and hence to maximize the efficacy while minimizing the risk of adverse effects.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
