Comparative Resistance to Teriparatide-Induced Bone Resorption With Denosumab or Alendronate

Abstract

Context:

In postmenopausal osteoporotic women, denosumab fully inhibits teriparatide-induced bone resorption at approved doses. This property of denosumab is distinct from that of alendronate and likely contributes to the efficacy of combination denosumab and teriparatide therapy. Whether denosumab fully inhibits bone resorption when challenged by a higher dose of teriparatide is unknown.

Objective:

We aimed to define the comparative ability of denosumab and alendronate to block the acute proresorptive effects of high-dose teriparatide.

Design, Setting, and Participants:

In this randomized controlled trial, bone resorption (serum C-telopeptide [CTX]) was measured in 25 postmenopausal women prior to and 4 hours after a single 40-μg sc teriparatide injection. Subjects then received either a single injection of denosumab 60 mg or oral alendronate 70 mg weekly for 8 weeks. After 8 weeks, serum CTX was again measured before and 4 hours after a teriparatide a 40-μg injection.

Outcomes:

The primary outcome was the between-group difference in the teriparatide-induced change in CTX from baseline to week 8.

Results:

At baseline, 40 μg of teriparatide induced similar 4-hour increases in mean CTX in both groups (alendronate 47% ± 14%, denosumab 46% ± 16%). After 8 weeks, teriparatide was still able to stimulate bone resorption in women treated with alendronate (mean CTX increase of 43% ± 29%) but not in women treated with denosumab (−7% ± 11%; P < .001 for between group comparison).

Conclusions:

Denosumab, but not alendronate, fully inhibits the ability of high-dose teriparatide to increase bone resorption acutely. These results suggest that combining denosumab with a more potent anabolic stimulus may result in greater separation between bone resorption and formation and hence greater increases in bone mass.

Currently available osteoporosis medications reduce fracture risk but do not cure osteoporosis in most patients with severe disease. Based on mechanism of action, osteoporosis therapy is divided into two categories: antiresorptive or anabolic. The current mainstay of antiresorptive treatment is bisphosphonates, which act by suppressing bone resorption by inhibiting the activity of mature osteoclasts, although they suppress new bone formation as well (1). The only US Food and Drug Administration (FDA)-approved anabolic agent is teriparatide, which acts by stimulating osteoblast activity but also stimulates bone resorption (2). The combination of the antiresorptive denosumab, a receptor activator of nuclear factor-κB (RANK) ligand inhibitor, and teriparatide was recently shown to increase bone mineral density (BMD) more than either drug alone (3). The mechanism of this combination's efficacy may relate to the ability of denosumab, unlike bisphosphonates, to fully block teriparatide-induced resorption while not fully inhibiting bone formation (4–6). In previous studies, the anabolic effect of teriparatide was demonstrated to be dose dependent in that both biochemical markers of bone formation and BMD increase more in women treated with 40 μg teriparatide daily compared with those treated with 20 μg daily (the FDA approved dose) (2, 7).

Based on these findings, we hypothesize that using a more potent anabolic agent in combination with denosumab might allow for the greater separation of bone formation and resorption and hence larger increases in BMD. Defining the properties of teriparatide in the setting of denosumab therapy is essential if we are to develop improved approaches in combination osteoporosis treatments. To test the hypothesis that denosumab retains its capacity to inhibit bone resorption in the setting of high-dose teriparatide, we assessed the acute changes in bone resorption markers in response to 40 μg of teriparatide in postmenopausal women before and after 8 weeks of treatment with either alendronate or denosumab.

Subjects and Methods

Study subjects

Women were recruited to this open-label randomized, controlled trial from March 2013 to April 2014 through targeted mailings, advertisements, and referrals to Massachusetts General Hospital (Boston, Massachusetts). Inclusion criteria were age over 45 years with at least 36 months since the last menses (plus FSH > 25 IU/L if age < 55 years or a history of hysterectomy) plus a high risk of fracture. We defined high fracture risk according to the following criteria: T score −2.5 or less at the spine, hip, or femoral neck; or personal history of low-trauma spine or hip fracture; or calculated 10-year risk of major osteoporotic or hip fracture of greater than 20% or greater than 3%, respectively, based on the World Health Organization FRAX database (http://www.shef.ac.uk/FRAX/).

Women were excluded if they had any of the following: abnormal blood calcium; abnormal blood PTH level; serum 25-hydroxyvitamin D concentration lower than 20 ng/mL or greater than 65 ng/mL; bone disease other than osteoporosis; anemia; history of malignant disease or radiation therapy; severe cardiopulmonary, liver, renal, or major psychiatric disease; or excessive alcohol intake.

Women were also excluded if they had received the following: 1) 5 years or more of cumulative oral bisphosphonate within 5 years of enrollment; 2) 2–5 years of oral bisphosphonates within 3 years of enrollment; 3) up to 2 years of oral bisphosphonate within 1 year of enrollment; 4) glucocorticoids within 6 months of enrollment; 5) denosumab within 6 months of enrollment; 6) estrogen, selective estrogen receptor modulators, or calcitonin within 3 months of enrollment; or 7) any exposure to iv bisphosphonates or strontium ranelate.

Study design

Women were stratified by age (younger than 65 years vs 65 years or older) and previous bisphosphonate use. Subjects were then randomized on a 1:1 basis in random blocks of 2, 4, or 6 to one of two treatment groups (Figure 1): 1) denosumab 60 mg sc (Prolia; Amgen, Inc); or 2) alendronate 70 mg orally once weekly for 8 weeks.

Figure 1. Study schema.

ALN, alendronate; DMAB, denosumab; TPTD, teriparatide.

Prior to any antiresorptive therapy, whole-blood calcium and albumin; serum β-c-terminal telopeptide of type 1 collagen (CTX); osteocalcin (OC); and aminoterminal propeptide of type 1 procollagen (P1NP) were measured both immediately prior to and 4 hours after a single 40-μg sc teriparatide injection. Syringes containing 40 μg of teriparatide (Forteo; Eli Lilly, Inc) were prepared by the research pharmacy. Subjects were fasting during all blood samplings. After the 4-hour collections, denosumab 60 mg was administered by a study physician to subjects assigned to the denosumab group, and subjects assigned to the alendronate group were provided oral alendronate. Adherence to alendronate was assessed by diary. After 8 weeks, fasting serum calcium, albumin, CTX, P1NP, and OC were again measured in all subjects immediately before and 4 hours after a teriparatide 40-μg injection. All subjects were dosed with teriparatide at 8:00 am at both weeks 0 and 8. Adjusted calcium was calculated using the following formula: adjusted calcium (milligrams per deciliter) = calcium (milligrams per deciliter) + 0.8 × (4.0 − albumin [grams per deciliter]).

At the baseline visit, women were asked to complete a food-frequency questionnaire developed by the study investigators to assess calcium intake. Calcium and vitamin D were prescribed to achieve total daily intakes of 1200 mg calcium and 400 IU vitamin D. The study was approved by the Partners Healthcare Institutional Review Board, and all subjects were provided written informed consent. This trial was registered with clinicaltrials.gov (number NCT01750086).

Biochemical measurements

Serum samples were stored at −80°C until the end of the study when biomarkers of bone turnover (OC and P1NP for bone formation and CTX for bone resorption) were assessed. All samples for a particular subject were measured in the same assay, and each assay contained samples from both treatment groups. Serum OC was measured via an electrochemiluminescent immunoassay (Meso Scale Discovery) with inter- and intraassay coefficients of variation (CVs) of 10% and 8%, respectively. Serum P1NP was measured via a RIA (Orion Diagnostica) with inter- and intraassay CVs of 6–10% and 7%–10%, respectively. Serum CTX was measured via a fully automated electrochemiluminescent immunoassay (Roche Diagnostics), with an interassay CV of less than 5%. The limit of detection for serum CTX was 0.01 ng/mL and the reportable range was 0.01–5.99 ng/mL.

Statistical analysis

The primary study end point is serum CTX. Using an analysis of covariance, the primary analysis assessed the between-group difference in the 4-hour teriparatide-induced change in serum CTX at 0 week vs 8 weeks. The 4-hour change in serum CTX at 0 week was included as a covariate to account for the observed 0-week variability. The secondary end points OC, P1NP, and calcium were assessed in the same fashion. Two-sided values of P ≤ .05 were considered significant.

Results

Of the 27 women enrolled, 25 women (93%) completed the study and are included in this analysis (Figure 2). Two women discontinued for medical reasons. One subject experienced lightheadedness 3 hours after the first teriparatide injection, and the other subject experienced lightheadedness and syncope 3 hours after the first teriparatide injection. The subject with syncope was evaluated in the emergency department, and her symptoms were attributed to the potential interaction of teriparatide and extended fasting. Baseline characteristics were similar between the two groups (Table 1). All women in the alendronate group reported taking 100% of their protocol-specified alendronate doses.

Figure 2. Subject disposition.

Table 1.

Baseline Characteristics

	Alendronate (n = 13)	Denosumab (n = 12)	P Value
Age, y	66 ± 8	66 ± 7	.959
Non-Hispanic white	10 (77%)	12 (100%)	.076
Body mass index, kg/m2	25.1 ± 3.7	23.4 ± 2.7	.222
Prior bisphosphonate use	4 (31%)	3 (25%)	.748
Creatinine, mg/dL	0.7 ± 0.1	0.7 ± 0.1	.668
Adjusted calcium, mg/dL	9.9 ± 0.6	9.8 ± 0.4	.638
25-Hydroxyvitamin D, ng/mL	35 ± 18	32 ± 11	.630
Serum CTX, ng/mL	0.51 ± 0.19	0.47 ± 0.28	.577
Serum OC, ng/mL	53.7 ± 19.3	54.7 ± 25.8	.914
Serum P1NP, μg/L	52.5 ± 19.9	51.7 ± 16.4	.915
Posterior-anterior lumbar spine T-score	−2.1 ± 1.1	−1.3 ± 1.1	.086
Total hip T-score	−1.6 ± 0.8	−1.5 ± 0.8	.677
Femoral neck T-score	−1.9 ± 0.5	−2.0 ± 0.8	.921

Data are mean ± SD or number (percentage).

At week 0, teriparatide induced a significant increase in mean serum CTX in the combined cohort (47% ± 15%) (0 vs 4 hours after teriparatide, P < .001), with no significant difference between designated treatment groups (alendronate 47% ± 14%, denosumab 46% ± 16%) (Figure 3). After 8 weeks of antiresorptive therapy, baseline serum CTX before teriparatide injection decreased by 60% ± 19% (from 0.51 ± 0.19 to 0.20 ± 0.11 ng/mL) in the alendronate group and by 90% ± 5% (from 0.47 ± 0.28 to 0.04 ± 0.01 ng/mL) in the denosumab group (P < .001 alendronate vs denosumab). At 8 weeks, teriparatide was still able to stimulate bone resorption in the alendronate group (serum CTX increased 43% ± 29%), but teriparatide did not increase serum CTX in the denosumab group (−7% ± 11%; P < .001 for between group comparison).

Figure 3. Serum CTX levels (mean ± SD) before and 4 hours after teriparatide injection in the alendronate-treated group and denosumab-treated group.

*, P < .05 for the 0- to 4-hour increase. ALN, alendronate; DMAB, denosumab.

At 0 weeks, teriparatide induced an increase in mean adjusted blood calcium in the entire cohort (2.5% ± 5.3%, P = .025). At the week 8 visit, blood calcium showed a trend toward an increase in response to teriparatide in both groups (alendronate, P = .067; denosumab, P = .079) (Figure 4). Only one subject had a calcium level greater than 10.8 mg/dL after teriparatide; the maximum level was 11.1 mg/dL and the subject was asymptomatic.

Figure 4. Adjusted calcium levels (mean ± SD) before and 4 hours after teriparatide injection in the alendronate-treated group and denosumab-treated group.

*, P < .05 for the 0- to 4-hour change. ALN, alendronate; DMAB, denosumab.

At 0 weeks, serum OC decreased from 54 ± 22 ng/mL to 52 ± 20 ng/mL 4 hours after teriparatide administration, but this decrease was not significant (P = .180) (Figure 5). At the week 8 visit, OC decreased significantly in both groups 4 hours after teriparatide administration (alendronate, −10% ± 14%, P = .021; denosumab, −13% ± 14%, P = .005).

Figure 5. Serum OC and P1NP levels (mean ± SD) before and 4 hours after teriparatide injection in the alendronate-treated group and denosumab-treated group.

*, P < .05 for the 0- to 4-hour change. ALN, alendronate; DMAB, denosumab.

At the 0-week visit, serum P1NP decreased from 52 ± 18 μg/L to 50 ± 18 μg/L (−3% ± 8%, P = .043) in response to teriparatide in the full cohort (Figure 5), whereas at the week 8 visit, mean serum P1NP decreased significantly in the denosumab group only (denosumab, −14% ± 13%, P = .001). There were no significant between-group differences in the changes in OC or P1NP.

Discussion

In this study, we demonstrate that denosumab fully inhibits high-dose teriparatide's acute stimulation of bone resorption (as measured by serum CTX 4 h after teriparatide administration), whereas alendronate attenuates but does not fully inhibit the increase in serum CTX.

The observed teriparatide-induced increase in serum CTX within 4 hours at the 0-week visit is consistent with previous studies assessing the acute effect of PTH (via infusion or injection) on bone resorption in humans. In healthy adult men, PTH infusion induced a linear increase in serum CTX and serum cross-linked N-telopeptide of type 1 collagen within 6 hours, which was sustained over 18 hours (8). This acute effect of PTH infusion on resorption was observed seen in hypogonadal men and premenopausal women (9, 10). In a separate study, a single 20-μg sc teriparatide injection or an endogenous increase in PTH induced by acute hypocalcemia increased serum CTX within 3 hours (11). An acute increase in serum cross-linked N-telopeptide of type 1 collagen and urinary CTX was also observed after an initial injection of either 28.2 or 56.5 μg of teriparatide in Japanese postmenopausal osteoporotic women (12).

These acute teriparatide-induced increases in bone resorption contrast with more delayed increases observed when bone resorption is assessed 24 hours after teriparatide administration. When measured after 24 hours, increases in bone resorption are generally detectable only after several months of repeated daily teriparatide administration (7). The clinical significance of these transient increases in bone resorption is unknown but refute the conventional wisdom that once-daily sc injections of teriparatide initially stimulate osteoblast activity without activation of osteoclastic bone resorption.

The observed differential effect on teriparatide-induced bone resorption with denosumab or alendronate is consistent with a previous study assessing the acute effect of PTH combined with antiresorptives. In a study in which blood endogenous PTH was transiently increased by inducing hypocalcemia, alendronate-treated subjects demonstrated less increase in indices of bone resorption than subjects treated with risedronate, raloxifene, or placebo (13).

The binding of RANK ligand to RANK is required for osteoclast differentiation, proliferation, and activation, and denosumab acts as a potent antiresorptive by blocking this binding (14). Its antiresorptive properties are highly potent as demonstrated by the complete absence of osteoclasts in more than 50% of iliac crest biopsy specimens taken from women treated for 24–36 months (15). Its ability to fully block the antiresorptive effects of teriparatide in this study likely stems from this unique mechanism of action as compared with bisphosphonates, which induce apoptosis only in mature osteoclasts actively resorbing bone matrix (1).

Interestingly, biochemical markers of bone formation do not acutely increase (and in some cases decrease) in response to teriparatide administration. This pattern is in contrast to the increases in OC when measured 24 hours after teriparatide administration after as few as 2 days of daily therapy (16). Similar acute decreases in bone formation markers have been observed previously with both PTH infusion and injection (8–10, 12, 17). The mechanism by which teriparatide transiently induces this suppression is unknown but may relate to the immediate effects of teriparatide favoring proliferation as opposed to activity (bone formation). The clinical observation that BMD increases within 6 months of beginning teriparatide therapy (18) shows that the eventual sustained and dose-dependent increase in osteoblast activity is physiologically dominant. Moreover, the ability of denosumab to fully inhibit the acute increase in bone resorption in response to high-dose teriparatide suggests that the eventual stimulation in bone formation may provide for even greater increases in BMD than those observed when denosumab is combined with standard, FDA-approved-dose teriparatide.

The dose of teriparatide used in this study is higher than that currently approved by the US FDA. The degree of hypercalcemia observed in this study was minimal. In the large teriparatide registration trial, the incidence of mild hypercalcemia (defined as calcium > 10.6 mg/dL) 4–6 hours after the injection was 2% in the placebo group, 11% in the 20-μg PTH group, and 28% in the 40-μg PTH group. Treatment was withdrawn for persistent hypercalcemia after a reduction in calcium intake in one woman in the placebo group (total n = 544), one in the 20-μg group (n = 541), and nine in the 40-μg group (n = 552) (2). Whether the incidence of hypercalcemia would be lower in the setting of antiresorptive medication is unknown.

There are several limitations of the current study. First, we provided only an 8-week period of alendronate and denosumab therapy before the second teriparatide injection. It is notable, however, that 8 weeks of both alendronate and denosumab have been shown to be sufficient to ensure that the maximal antiresorptive effect of these agents (19). For example, the percentage change from baseline in serum CTX in patients treated with denosumab at 1, 3, and 6 months is −88%, −86%, and −70%, respectively, and the percentage change in patients treated with alendronate at 1, 3, and 6 months is −62%, −65%, and −65%, respectively. Additionally, we measured serum CTX only at baseline and again 4 hours after teriparatide administration. We based this design on data from our prior studies in which 30 μg of sc teriparatide maximally increased serum CTX at 4 hours (data not shown). Also, bone turnover markers are known to have an inherent diurnal variation, and this could potentially influence our results (20). Because all samples were measured fasting and with consistent timing and serum CTX has not been reported to increase significantly between 8 am and noon (20), we believe any effect of diurnal variation would be negligible. Finally, predicting the long-term effects of teriparatide combined with denosumab on bone formation markers is limited because teriparatide's acute effects on bone formation markers are counter to their long-term effects with repeated daily dosing.

In summary, this study demonstrates that in the presence of denosumab-induced RANK ligand inhibition, teriparatide has no acute effect on bone resorption. The difference in the ability of denosumab and alendronate to inhibit the proresorptive properties of teriparatide may, in part, explain the observed distinct BMD response among the various teriparatide combination therapy studies (3–6). Furthermore, these results suggest that combining denosumab with higher doses of teriparatide or with even more potent anabolic agents may be an even more effective osteoporosis treatment strategy in patients at high risk of fracture.

Funding Statement

This work was supported by Amgen Inc and by Grants 1UL1TR001102-01 and 8 UL1 TR000170-05 to the Harvard Clinical and Translational Science Center from the National Center for Advancing Translational Science.