By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Differing effects of PTH 1–34, PTH 1–84, and zoledronic acid on bone microarchitecture and estimated strength in postmenopausal women with osteoporosis: An 18-month open-labeled observational study using HR-pQCT
The 1 to 34 amino-acid N-terminal parathyroid hormone (PTH) fragment teriparatide (PTH 1–34) and the intact 1 to 84 amino-acid PTH peptide (PTH 1–84) are used as bone anabolic agents for the treatment of postmenopausal osteoporosis. By preferentially increasing bone formation over bone resorption, these agents stimulate, in intermittent use, the formation of new bone.1 The mechanisms involved include modeling-based bone formation on quiescent bone surfaces as well as remodeling-based bone formation.2 Bone mineral density (BMD) in the spine as assessed by dual-energy X-ray absorptiometry (DXA) increases 7% to 10% after 18 months of treatment.3, 4 Both PTH 1–34 and PTH 1–84 reduce the risk of new vertebral fractures, whereas efficacy against new nonvertebral fractures was documented with PTH 1–344 but not with PTH 1–84.3 Both agents are used widely, although PTH 1–84 is not approved by the Food and Drug Administration in the United States.
PTH 1–34 as an anabolic agent was proposed by Reeve and colleagues in 1980.5 Although the formation of new cancellous bone was convincing, it was observed that PTH 1–34 may have an unfavorable effect in cortical bone. Since then, this cortical steal phenomenon has been widely studied.6 In primates, it has been shown that intermittent PTH 1–34 or PTH 1–84 exposure causes an increase in cortical porosity.7, 8 This was, however, shown not to decrease integral bone strength, as the pores were preferentially located near the endosteal surface, where their biomechanical impact was limited, and this effect was outweighed as treatment also increased the cortical area.7
With the development of newer bone-imaging devices that contrary to DXA methodology allow three-dimensional bone imaging, assessment of treatment effects on bone geometry and compartmental characteristics has become feasible in patients. With the latest generation high-resolution peripheral quantitative computed tomography (HR-pQCT) devices, image resolution has been markedly refined and now permits detailed bone microarchitecture evaluation.9 Noninvasive assessment of both cortical and trabecular microarchitecture at the distal radius and tibia is now possible in vivo with an isotropic image voxel size of 82 µm. Measurement of selected microarchitectural characteristics hitherto limited to evaluation in bone biopsies has become clinically available with this technique. Bone strength may also be estimated using finite element (FE) analysis based on the HR-pQCT images.10 Such FE strength estimates have been shown to be more closely correlated to radius biomechanical competence than BMD by DXA.10
In a recent longitudinal study using HR-pQCT in 11 postmenopausal women, treatment with 18 months of PTH 1–34 decreased cortical BMD at radius and tibia, whereas a nonsignificant increase was seen in cortical porosity in radius.11 Trabecular thinning and reduced trabecular bone volume fraction were also found in radius, whereas FE estimated bone strength was preserved at both sites. In comparison, quantitative computed tomography (QCT) studies with intermittent PTH 1–84 or PTH 1–34 have shown large increases in vertebral cancellous BMD and estimated vertebral strength.12, 13 Treatment may thus not be equally effective in the peripheral skeleton. To address this issue further, the effects of 18 months of bone anabolic therapy with PTH 1–34 or PTH 1–84 were characterized in detail using HR-pQCT. A group of postmenopausal women with osteoporosis treated with zoledronic acid, ie, a potent antiresorptive agent, was also included for comparison.
Materials and Methods
This 18-month prospective, open-label, nonrandomized study involved women (Table 1) from a single clinical site (Department of Endocrinology, Odense University Hospital, Odense, Denmark), who were prescribed treatment for postmenopausal osteoporosis with PTH (20 µg PTH 1–34 [teriparatide] subcutaneous daily [Eli Lilly, Indianapolis, IN, USA] or 100 µg PTH 1–84 subcutaneous daily [Takeda Nycomed, Zurich, Switzerland]) or zoledronic acid 5 mg infused once yearly (Novartis Europharm, Horsham, UK) from April 2009 to December 2010 as part of routine clinical care. Inclusion criteria were postmenopausal status ( > 1 year from last menstruation) and ability to provide informed consent. Reimbursement of PTH 1–34 and PTH 1–84 treatment in Denmark required two or more vertebral fractures or one vertebral fracture and a lumbar spine (L1 to L4) or total hip T-score ≤ –3.0 SD. Reimbursement of zoledronic acid required a lumbar spine (L1 to L4) or total hip T-score ≤ –2.5 SD and a clinical risk factor for osteoporosis. Exclusion criteria were treatment with glucocorticoids (equivalent to more than 5 mg prednisone daily > 3 months), kidney or liver disease, calcium-metabolic or endocrine diseases affecting bone metabolism, or current exposure to drugs with known effect on bone. At least 1 month passed from cessation of previous bisphosphonate treatment (alendronate, etidronate, or oral ibandronate) until initiation of treatment with PTH (PTH 1–34 median 2 months, range 1 to 13 months; PTH 1–84 median 4 months, range 1 to 60 months) or zoledronic acid (median 3 months, range 1 to 14 months). Based on their diet intake, all participants were recommended supplements with (up to) 1200 mg calcium daily and (up to) 20 µg vitamin D3 daily. Patients prescribed anabolic therapy were presented PTH 1–34 and PTH 1–84 injection pens by a study nurse and were assigned to either PTH treatment based on personal preference. Training was provided regarding self-administration of PTH, and participants had their first injection supervised by a study nurse. Subsequent injections were self-administered. Patients were scheduled for visits at 0.5, 3, 6, and 12 months for supervision of injection technique, serum ionized calcium, and assessment of adverse events. Calcium and vitamin D supplements were discontinued if clinically significant hypercalcemia developed during treatment. PTH dosing frequency was reduced if the hypercalcemia persisted. Women treated with zoledronic acid had kidney status and serum calcium evaluated before their first and second infusion. Historical nonvertebral fractures were assessed using interviews, whereas vertebral fractures ( > 20% reduction in any vertebral height) were assessed on radiographs. Prevalence and duration of prior bisphosphonate treatment were assessed using interviews and chart reviews. Participants were informed orally and in writing before informed consent were obtained. The study was approved by the local Ethics Review Board (file no. 2008–0129).
Table 1. Baseline Anthropometrics, Biochemistry, and DXA BMD in Study Participants With at Least One Follow-up Visit
Zoledronic acid (n = 33)
PTH 1–34 (n = 18)
PTH 1–84 (n = 20)
The p values are derived from one-way analysis of variance.
160 ± 6
159 ± 5
159 ± 7
63 ± 9
64 ± 10
64 ± 11
Age at menopause (years)
47 ± 6
48 ± 5
49 ± 6
Prior bisphosphonate treatment, n (%)
Duration of prior bisphosphonate treatment (months)
18 ± 33
17 ± 39
13 ± 29
Individuals with prior vertebral fracture, n (%)
No. of prior vertebral fractures
Individuals with prior nonvertebral fracture, n (%)
No. of prior nonvertebral fractures
Daily calcium supplements (mg)
762 ± 239
752 ± 239
637 ± 238
94 ± 34
96 ± 35
80 ± 30
Spine BMD (g/cm2)
0.72 ± 0.11
0.71 ± 0.13
0.72 ± 0.13
Total hip BMD (g/cm2)
0.70 ± 0.11
0.62 ± 0.13
0.66 ± 0.12
1/3 Distal forearm BMD (g/cm2)
0.52 ± 0.07
0.49 ± 0.11
0.48 ± 0.08
Ultradistal forearm BMD (g/cm2)
0.26 ± 0.04
0.24 ± 0.07
0.26 ± 0.06
Images of the nondominant distal radius and tibia (or dominant side in case of previous fracture at the desired site) were obtained using high-resolution peripheral quantitative computed tomography (XtremeCT, Scanco Medical, Brüttisellen, Switzerland) at baseline, 6 months, and 18 months. Images were evaluated to obtain measures of 1) bone geometry, 2) cortical morphology, 3) trabecular morphology, and 4) overall biomechanical competence. The standard patient protocol for image acquisition was applied as described previously.9 Using an offset from the endplate of 9.5 mm and 22.5 mm in radius and tibia, respectively, 9.02 mm axial bone representations comprising 110 slices were obtained providing images with an isotropic voxel size of 82 µm. Image quality was graded after reconstruction by one author (SH) using a five-step scale as suggested by the manufacturer (1 = best, 5 = worst), and images with grade < 3 were disregarded.14 An autosegmentation approach using periosteal and endosteal thresholds was applied for image segmentation.15 All contours were visually checked on screen and corrected manually if necessary by one author (SH). Total and cortical bone densities were derived from the respective volumes, calibrated using a scan phantom, and expressed in mg/cm3. Cortical parameters included cortical thickness (Ct.Th), measured directly as the endosteal-periosteal distance using a distance transformation method, and cortical porosity measured as void cortical volume divided by total cortical volume.16, 17 Trabecular area was derived from trabecular volume; scan height and voxel size using an annular approach. Metric trabecular parameters were extracted using the manufacturer's default method after fixed threshold segmentation. Bone volume per tissue volume (BV/TV) was derived from trabecular density, whereas trabecular number (Tb.N) was directly measured using a distance transformation method. Trabecular thickness (Tb.Th) and trabecular spacing (Tb.Sp) were derived from Tb.N and BV/TV in analogy to standard histomorphometry.9 An automatic common region matching procedure based on variation in bone cross-sectional area was applied to ensure that only volumes common to scans obtained at baseline, 6 months, and 18 months were used for extraction of the above parameters.18 Finally, radius and tibia failure load was estimated using unmatched images with a micro-finite element analysis solver provided by the manufacturer (Finite Element Analysis Software version 1.15, Scanco Medical), where all bone materials were given a Young's Modulus of 10 GPa and a Poisson's ratio of 0.3. From the models, an estimate of failure load in compression is calculated based on the assumption that bone failure occurs if > 2% of the elements are strained beyond 0.7% strain.10
Quality control was assessed by daily and weekly scan of phantoms (QRM, Möhrendorf, Germany). Based on repeated measurement of 13 individuals after repositioning (mean age 64 years, range 33 to 79 years), the root mean square coefficient of variation (RMSCV) for radius and tibia parameters in our unit ranged from 0.4% to 0.8% for densities, 3.5% to 5.0% for trabecular microarchitecture parameters, 1.0% to 7.2% for extended cortical measures, and 1.2% to 1.7% for FE estimated failure load. This is in agreement with the reproducibility reported by others.9, 16
Areal BMD was measured at the spine (L1 to L4), total hip, and forearm (1/3 distal and ultradistal regions), using DXA (Hologic, Waltham, MA, USA) at baseline, 6 months, and 18 months. In our unit, the RMSCV was 1.5% at both the hip and spine.
Blood samples were drawn in fasting state and stored at –80°C until analysis of bone turnover markers. Serum 25 (OH) vitamin D levels were measured using EZChrom Elite chromatography (Agilent Technologies, Santa Clara, CA, USA). Amino-terminal propeptide of type 1 collagen (PINP) and c-telopeptide of type 1 collagen (CTX-1) were measured using an electrochemiluminescent method (Roche Diagnostics, Mannheim, Germany).
No reports regarding the effects of anabolic or antiresorptive agents on HR-pQCT parameters were available at the time of study planning. Formal hypothesis testing was therefore not preplanned. We did, however, estimate sample sizes required to show significant between-group differences in tibia BV/TV based on the assumption that the effect with either PTH therapy would resemble that of PTH 1–34 in the iliac crest ( + 14% in BV/TV after 18 months of treatment),19 whereas the effect with zoledronic acid would resemble that of risedronate (–9.5% in BV/TV after 3 years of treatment with 5 mg risedronate daily).20 Based on these assumptions and an estimate of tibia BV/TV in osteoporotic women of 9.7% ± 2.5%9 and a similar SD post-treatment, 19 participants in each therapy group would suffice to demonstrate significant between-group treatment differences with alpha = 0.05 and beta = 0.80.
Data from participants with at least one follow-up visit were analyzed. Data are presented as mean ± SD or median (range) as appropriate. Between-group differences at baseline were quantified using one-way analysis of variance. The percentage change from baseline in DXA and HR-pQCT parameters for each participant was determined and the three treatments were compared in a longitudinal analysis by using a linear mixed-effect model with fixed effects for visit, treatment group, and visit by treatment-group interaction. From the models, the fitted mean percentage changes from baseline within each treatment group at months 6 and 18 were determined, and the differences in treatment response between groups at month 18 were quantified. Changes in biochemical markers of bone turnover had a non-normal distribution. The change within groups was therefore quantified using Wilcoxon's match-pairs signed rank test. Between-group differences were quantified using a Kruskal-Wallis test, and if significant, pairwise comparisons were specified using a Mann-Whitney two-sample test. Because the objective of the study was hypothesis-generating rather than to confirm between-treatment differences, we did not apply correction for multiple testing. The p values are two-sided and thus considered statistically significant if below 0.05.
Study recruitment and completion are outlined in Fig. 1. Of those prescribed anabolic therapy (n = 78) or zoledronic acid (n = 68) during study recruitment, 35 met one or more exclusion criteria (PTH, n = 18; zoledronic acid, n = 17), whereas 32 refused to participate (PTH, n = 16; zoledronic acid, n = 16). Of the 79 women included, 71 (90%) completed the month 6 visit and 66 (84%) completed the month 18 visit. Main reasons for discontinuation were no wish to continue and adverse events. Because of hypercalcemia, reduction of PTH dosing frequency was necessary in two patients (PTH 1–34 = 1, PTH 1–84 = 1). More participants prescribed anabolic therapy had prior vertebral fractures compared with those prescribed zoledronic acid (p = 0.04), whereas no differences between groups were observed regarding age, height, weight, years from menopause, number of nonvertebral fractures, calcium supplements, 25(OH)D3, BMD by DXA (Table 1), or HR-pQCT parameters in radius (Table 2). In tibia, HR-pQCT parameters did not differ between groups except that those prescribed PTH 1–34 had smaller CSA compared with those prescribed PTH 1–84 and zoledronic acid (both p < 0.05) and lower total density compared with those prescribed zoledronic acid (p < 0.05). A similar fraction in each treatment group had received prior bisphosphonate treatment (zoledronic acid 16/33, PTH 1–34 = 5/18, PTH 1–84 = 9/20, p = 0.66). All participants were white.
Table 2. Radius and Tibia HR-pQCT Parameters at Baseline, Month 6 (M6), and Month 18 (M18) as Well as Mean Percentage Changes ± SD at M6 and M18 in Patients Treated With Zoledronic Acid, PTH 1–34, or PTH 1–84
Note: Parameters printed bold indicate significant change from baseline within treatment group with p < 0.05.
▵ = change from baseline; CSA = cross-sectional area; BV/TV = trabecular bone volume per tissue volume; FE = finite element.
Outcomes obtained after autosegmentation algorithm.
p < 0.1 compared with baseline within treatment group.
p < 0.05.
p < 0.01 zoledronic acid versus PTH 1–34 at baseline.
p < 0.05 PTH 1–34 versus PTH 1–84 at baseline.
p < 0.01.
p < 0.001 zoledronic acid versus PTH 1–34 at follow-up.
p < 0.05.
p < 0.01.
p < 0.001 PTH 1–34 versus PTH 1–84 at follow-up.
p < 0.05.
p < 0.01.
p < 0.001 PTH 1–84 versus zoledronic acid at follow-up.
The p values are derived from one-way analysis of variance for baseline comparisons and mixed-effects models for longitudinal changes.
Zoledronic acid increased BMD in the spine, total hip, 1/3 distal, and ultradistal forearm after 18 months of treatment (Fig. 2). Similarly, PTH 1–34 increased BMD in the spine and total hip but decreased BMD at the 1/3 distal forearm site. No change at the ultradistal site was observed at month 18. PTH 1–84 increased BMD in the spine, did not change BMD at the total hip, whereas declines were seen both at the 1/3 distal and ultradistal forearm sites at month 18.
Effects of treatment on biochemical markers of bone turnover
Zoledronic acid decreased markers of bone formation and resorption at both months 6 and 18 (Fig. 2). Conversely, markers of formation and resorption increased with PTH 1–84 at months 6 and 18, whereas PTH 1–34 increased P1NP at months 6 and 18 and increased CTX1 at month 6 and nonsignificantly so at month 18 (p = 0.08).
Effects of treatment on bone morphology by HR-pQCT
HR-pQCT data are shown in Table 2. One image at month 6 (PTH 1–84, n = 1) and three images at month 18 (PTH 1–84 = 1, zoledronic acid = 2) were disregarded because of excessive motion artifacts. The average common region in radius was 85%. In radius, zoledronic acid increased BV/TV and trabecular number at month 18, whereas CSA, cortical thickness, cortical porosity, and total and cortical densities were unchanged (Fig. 3). PTH 1–34 increased CSA at month 18 along with cortical thickness and cortical porosity at months 6 and 18. Cortical density was decreased at month 18. Trabecular number was increased at month 6 but was unchanged from baseline at month 18. PTH 1–84 increased cortical porosity at month 6 and month 18, whereas CSA and cortical thickness were unchanged. Cortical density was decreased at month 18. No significant changes in the cancellous compartment were observed at month 18. The response in several parameters differed between treatment groups (Table 2). Most notably, the increase in cortical porosity with PTH 1–34 and PTH 1–84 differed significantly from the response to zoledronic acid (p < 0.01 and p < 0.001).
In tibia, all images had adequate quality, and the average common region was 91%. In tibia, zoledronic acid increased cortical thickness, total and cortical densities, trabecular BV/TV, and trabecular number at month 18, whereas CSA was unchanged. PTH 1–34 increased cortical thickness and cortical porosity at months 6 and 18, whereas CSA was unchanged. Total density increased at month 18, despite a decrease in cortical density. In the cancellous compartment, BV/TV and trabecular number increased with no change in trabecular thickness. PTH 1–84 decreased cortical thickness and increased cortical porosity at month 18 with no change in CSA. Total and cortical densities decreased. In the cancellous compartment, trabecular number increased, whereas trabecular thickness decreased at month 18. The increase in cortical porosity was unique to PTH 1–34 and PTH 1–84 treatments (p < 0.01 and p < 0.001), whereas the decrease in cortical thickness and BV/TV observed with PTH 1–84 at month 18 was significant compared with the responses seen with zoledronic acid (p < 0.01 and p < 0.001) and PTH 1–34 (both p < 0.001). Trabecular thinning was solely observed with PTH 1–84 (p < 0.05 compared with PTH 1–34 or zoledronic acid).
Effects on FE estimated strength
In radius, FE estimated strength was preserved with zoledronic acid and PTH 1–34, whereas it decreased from baseline with PTH 1–84 (Fig. 4). This decline was significantly different from the response with zoledronic acid (p < 0.01) but not PTH 1–34 (p = 0.16). In tibia, estimated failure load was maintained with zoledronic acid at month 6 with a trend toward an increase at month 18. Estimated strength was preserved with PTH 1–34, whereas it declined from baseline with PTH 1–84 at month 18. This decline was significant compared with the response with both zoledronic acid (p < 0.001) and PTH 1–34 (p < 0.001).
In this longitudinal HR-pQCT study, we found that 18 months of treatment with PTH 1–34 or PTH 1–84 was associated with an increase in cortical porosity and a decline in cortical density in both radius and tibia. Also, trabecular number in tibia increased with both anabolic agents as did cortical thickness in both radius and tibia with PTH 1–34. Conversely, with 18 months of treatment with zoledronic acid, radius and tibia cortical porosity was unchanged, whereas trabecular bone volume fraction increased. Bone strength estimated by FE analysis was preserved with PTH 1–34 and zoledronic acid, whereas it decreased with PTH 1–84 at both radius and tibia.
Anabolic therapy increased cortical porosity with a corresponding decrease in cortical density in both radius and tibia. The decline in cortical density is in accordance with findings from a smaller 18-month study on the effects of PTH 1–34 using HR-pQCT (n = 11), which also reported (nonsignificant) increases in cortical porosity.11 The cellular mechanisms underlying these findings cannot be addressed using HR-pQCT. Consistent with the mode of action of anabolic agents documented in animals,7, 8, 21 PTH may, however, accelerate intracortical remodeling as well as endosteal remodeling with a net positive balance between bone formation and resorption on the endosteal surface.22 We also found a slight but significant increase in radius CSA with PTH 1–34. As the change in CSA along the extremity length is used for common region matching in HR-pQCT methodology,18 it is uncertain if this increase relates to the applied image handling or represents apposition of new bone on the periosteal surface. An increase in radius CSA with 18 months of PTH 1–34 (20 µg) treatment compared with placebo-treated control subjects was similarly found using peripheral QCT.23 However, the extent to which intermittent PTH exposure increases bone diameter in humans remains unsettled.
No changes in cancellous microarchitecture parameters with either anabolic agent were observed in radius. In tibia, both anabolic agents caused a significant increase in trabecular number. PTH 1–34 also increased BV/TV, whereas significant trabecular thinning was observed with PTH 1–84. Trabecular thinning was also reported by MacDonald and colleagues11 during PTH 1–34 treatment using HR-pQCT. The underlying cellular mechanism cannot be addressed using HR-pQCT but may include intratrabecular tunneling. This phenomenon has been observed at multiple skeletal sites in animals treated with intermittent PTH 1–8424 and also in the human iliac crest during treatment with PTH 1–34.25
Treatment with zoledronic acid increased cortical density and cortical thickness in tibia with similar nonsignificant trends in radius. No changes in cortical porosity were observed. This is in accordance with previous findings on the effects of the bisphosphonates alendronate and ibandronate using HR-pQCT.26–29 The response in the cancellous compartment with zoledronic acid in our study, with an increase in trabecular number in tibia, is in accordance with findings with alendronate in some29 but not all studies.27 Ibandronate did not exert significant effects on cancellous architecture compared with placebo after 2 years of treatment.28 Whereas the observed increase in cortical thickness may be caused by closure of the endosteal remodeling space, the mechanism underlying the observed increase in trabecular number is not in accordance with the antiresorptive mode of action of bisphosphonates.30 Trabecular tunneling has not been described with bisphosphonate treatment, and with age the number of trabeculae in radius and tibia decrease rather than increase.31 The observed increase in trabecular number in our and other studies with bisphosphonates may thus relate to image segmentation and structure extraction on the endosteal surface.
Neither anabolic nor antiresorptive therapy resulted in an increase in FE estimated bone strength compared with baseline. For zoledronic acid, this is in accordance with a HR-pQCT study assessing the effects of 2 years of treatment with alendronate on FE estimated bone strength in radius and tibia.27 An important limitation to the FE analysis applied in our and in the above-mentioned study is that it assumes fixed, homogeneous material properties. The observed changes in estimated strength are therefore solely attributable to changes in geometry and microarchitecture. This may have underestimated the biomechanical response with zoledronic acid because part of the fracture-reducing efficacy with bisphosphonates is thought to be conveyed through an increase in bone mineralization.32
Increases in FE estimated vertebral strength up to 30% have been found after 1 year of PTH 1–34 treatment.12 In the total hip, FE estimated strength was found to be preserved with 18 months of PTH 1–34 treatment,33 whereas a 2% increase was found after 12 months of PTH 1–84.34 In our study, FE estimated bone strength was preserved with PTH 1–34, whereas it decreased from baseline with PTH 1–84 in radius and tibia. In postmenopausal women, cortical pores have been estimated to cause a 0.6% and 2.6% decrement in FE estimated failure load in radius and tibia, respectively.35 The observed increases in cortical porosity in our study may therefore partly underlie the decreases in estimated strength found with PTH 1–84. In the case of PTH 1–34, the increase in cortical porosity was to a larger extent counterbalanced by beneficial effects on cortical geometry and cancellous mass and architecture whereby overall bone strength was preserved. It should, however, be emphasized that the FE models used in this study were unscaled. The effects on bone strength of new, undermineralized bone tissue formed during anabolic therapy may therefore not have been appropriately accounted for. The finding that PTH 1–34 and PTH 1–84 appear to have different profiles was, however, surprising and warrants confirmation. The extent to which differences in pharmacokinetic profiles between the two agents36, 37 or actions through a separate receptor for the C-terminal end of the PTH peptide, proposed to be expressed in cells of osteoblast lineage,38, 39 cause diverse skeletal effects is unknown. Larger, long-term cohort studies directly comparing fracture rates between patients treated with the two anabolic agents are needed to clarify the clinical significance of these observations.
This study has some important strengths. To our knowledge, this is the first study to directly compare anabolic treatment with PTH 1–34 and PTH 1–84 in postmenopausal osteoporotic women in a head-to-head observational study. Also, a group of participants prescribed bisphosphonate therapy was included, allowing comparisons of treatment effects of both anabolic and antiresorptive therapy within the same study.
However, we acknowledge that our study has limitations. First, a fraction of participants in each treatment group had received prior bisphosphonate treatment, which has been observed to blunt the response to anabolic therapy in the spine and hip.40, 41 There was no consistent pattern to suggest a blunting of the response in BTM, DXA, or HR-pQCT parameters in the case of previous bisphosphonate use in those treated with PTH 1–34 and PTH 1–84 (data not shown). Because the fraction of previous bisphosphonate users in those treated with PTH 1–34 (28%) was (nonsignificantly) lower compared with PTH 1–84 (45%), we cannot, however, exclude that this difference has impacted findings. Second, this was an open-label, nonrandomized study. Because of differences in reimbursement criteria for anabolic therapy and zoledronic acid, those prescribed anabolic therapy had more severe skeletal disease illustrated by a larger fraction with vertebral fractures. Also, total density in tibia was lower in those prescribed PTH 1–34. Treatment groups did not, however, differ significantly in terms of BMD by DXA or microarchitecture parameters in either radius or tibia. Third, the compliance with anabolic therapy and supplements was not ascertained through drug accountability but was addressed through a number of clinical visits throughout the study and subsequent chart review.
To conclude, this 18-month, open-label, observational HR-pQCT study found divergent treatment-specific effects in cortical and trabecular bone with anabolic and zoledronic acid therapy. Anabolic therapy increased cortical porosity and decreased cortical density, which for PTH 1–34 was accompanied by an increase in cortical thickness at both sites. With both anabolic agents and with zoledronic acid, trabecular number increased in tibia. FE estimates of bone strength was maintained with zoledronic acid and PTH 1–34 but decreased with PTH 1–84. The decline in FE failure load with PTH 1–84 was significantly inferior to the response with zoledronic acid in both radius and tibia and to PTH 1–34 in tibia but not in radius. The different responses with PTH 1–34 and PTH 1–84 was surprising and warrants confirmation.
All authors state that they have no conflicts of interest.
We are grateful to Mrs. Steffanie Anthony Christiansen for study coordination and Claire Gudex, MD, for language support.
Authors' roles: Study design: SH and KB. Study conduct: SH. Data collection: SH. Data analysis: SH. Data interpretation: SH, KB, EH, and JEBJ. Drafting manuscript: SH. Revising manuscript: KB, JEBJ, and EH. Approving final version of manuscript: SH, KB, EH, and JEBJ. SH takes responsibility for the integrity of the data analysis.