THERAPEUTICS AND MEDICAL MANAGEMENT (E SHANE AND RA ADLER, SECTION EDITORS)

Anabolic and Antiresorptive Therapy for Osteoporosis:Combination and Sequential Approaches

Felicia Cosman

Published online: 24 October 2014# Springer Science+Business Media New York 2014

Abstract In the recent BoneKey Reports review, it was notedthat combinations of anabolic and antiresorptive agents havepotential to improve bone density and bone strength morethan either agent as monotherapy. Small clinical trialshave been performed evaluating combinations of PTH1-34(TPTD) or PTH1-84 (PTH) with a variety of antiresorptivesincluding hormone/estrogen therapy, raloxifene, alendronate,risedronate, ibandronate, zoledronic acid, and denosumab.Most of the studies evaluate dual-energy X-ray absorptiome-try outcomes, and a few trials report volumetric mineral den-sity (BMD) by quantitative computed tomography, followedby finite element modeling to calculate bone strength. None ofthe studies has been powered to assess differences in fractureincidence between combination therapy and monotherapy.BMD outcomes vary based on the timing of introduction ofthe anabolic agent (before, during, or after antiresorptivetreatment), as well as the specific anabolic and antiresorptiveused. Furthermore, effects of combination therapies aresite-dependent. The most consistent effect of combiningantiresorptive agents with PTH or TPTD is a superior hipBMD outcome compared with TPTD/PTH alone. This is mostevident when TPTD/PTH is combined with a bisphosphonateor denosumab. In contrast to findings in the hip, in the major-ity of studies, there is no benefit to spine BMD with combi-nation therapy vs monotherapy. The 2 exceptions to this arewhen TPTD is combined with denosumab and when TPTD isgiven as monotherapy first for 9 months, followed by the addi-tion of alendronate (with continuation administration of TPTD).

Based on what we now know, in patients previously treated withbisphosphonates who suffer hip fractures or who have very lowor declining hip BMD, strong consideration should be given tostarting TPTD and continuing a potent antiresorptive agent(possibly switching to zoledronic acid or denosumab) to improvehip BMD and strength quickly. Furthermore, in treatment naïveindividuals with very severe osteoporosis, such as those withspine and hip fractures, combination therapy with TPTD anddenosumab or TPTD followed by combination treatment with apotent bisphosphonate or denosumab should be considered tomaximize early increases in BMD throughout the skeleton(Cosman BoneKEy Rep 3, 2014)[1].

Keywords Monotherapy . Combination therapy . Anabolic .

Antiresorptive . Sequential therapy . Severe osteoporosis

Introduction

Various anabolic and antiresorptive therapies have been com-bined in an attempt to attain superior bone mass and strengtheffects compared with monotherapy. Although teriparatide(PTH1-34; TPTD) or PTH1-84 (PTH) combined withantiresorptive agents could theoretically produce additive orperhaps even synergistic skeletal effects, combination therapystudies have produced different outcomes based on skeletalsite (spine vs hip vs radius), assessment tool [dual-energyX-ray absorptiometry (DXA) vs quantitative computed to-mography (QCT)], specific antiresorptive agent utilized, andwhether patients are previously treatment naïve or treatmentexperienced [1]. Treatment naïve and treatment-experiencedpopulations differ in the magnitude of active bone surface. Inaddition, acute administration of antiresorptive agents affectparathyroid dynamics (increasing endogenous PTH produc-tion), and acute withdrawal of prior antiresorptive treatmentsaffect osteoclast activity. There may also be unique acute

F. CosmanColumbia College of Physicians and Surgeons, Columbia University,New York, NY 10025, USA

F. Cosman (*)Helen Hayes Hospital, Clinical Research Center, Route 9 W, WestHaverstraw, New York, NY 10993, USAe-mail: cosmanf@helenhayeshosp.org

Curr Osteoporos Rep (2014) 12:385–395DOI 10.1007/s11914-014-0237-9

effects of certain antiresorptive agents on osteoblast function.All of these factors might play a role in producing distinctresponses to combination therapy in previously treated vstreatment naïve individuals. Therefore, in trying to interpretthe many different combination therapy studies, it is importantto take note of these distinctions. In clinical practice, and insome of the most recent trials, the distinction between treat-ment naïve and treatment experienced is starting to blur, as agreater proportion of patients with osteoporosis have beenexposed to prior bisphosphonates. How much time needs toelapse before these patients can be considered ‘treatmentnaïve’ is unclear.

This review summarizes the key combination trials andevolving concepts regarding combination treatment, firstoverviewing studies in treatment naïve women, then in previ-ously treated women and lastly in men.

Treatment Naïve Postmenopausal Women

Teriparatide and Hormone Therapy or Raloxifene

Ste Marie et al. [2] studied treatment naïve women (n=125)and women previously treated with hormone therapy (HT;discussed below, n=122). Within each cohort, women wererandomized to receive HT alone or HT with TPTD 40 mcgdaily for 1 year. In the treatment naïve cohort, there wereBMD increases because of the newly administered HT itself,and much larger increases in the group receiving HT withTPTD (16 % in the spine and 6 % in the hip). The increasesfrom TPTD appeared additive to those of HT, certainly withno obvious blunting of TPTD effect by HT, especially whenviewed in light of findings from another study utilizing TPTD40 mcg monotherapy [3].

Deal et al. randomized 137 postmenopausal treatment-naive women to receive TPTD 20 mcg or TPTD plus ralox-ifene (Ral) for 6 months. Mean serum PINP (aminoterminalpropeptide of Type I Procollagen) level rose similarly in the 2groups, while mean serum CTX (cross-linked C-terminaltelopeptide) level increased to a lesser degree with combina-tion TPTD/Ral vs TPTD monotherapy. Spine BMD incre-ments were similar in the 2 groups, while hip BMD incre-ments were significantly greater in the TPTD/Ral combinationgroup [4]. A larger increment in hip BMD with combinationtherapy has been a consistent finding in almost all of thecombination treatment studies, across treatment naïve, andtreatment experienced individuals.

Teriparatide or PTH and Bisphosphonates

Finkelstein et al. randomized 93 treatment naïve women toreceive alendronate (Aln) for six months prior to the additionof TPTD 40 mcg vs either agent alone [3]. A large proportion

of women (40 %) discontinued the study early or never tookstudy drugs and a substantial proportion of women in bothTPTD groups required dose reduction (by 25 %–50 %) due tohypercalcemia or side effects. BMD gains in both spine andhip by DXA were lower in those given TPTD after the 6-month pretreatment with Aln, compared with those givenTPTD monotherapy; however, the difference in spine BMDgain was not significant if groups were restricted to those whodid not discontinue study medication prematurely [5]. RadiusBMD declined more in women who received TPTD mono-therapy compared with combination treatment and the incre-ment in total body BMD did not differ between groups. Incontrast to the prior studies and the subsequent studiesdiscussed, total hip, and femoral neck BMD increased morewith TPTD monotherapy than with combination therapy (intotal hip 8.1 % for TPTD monotherapy vs 2.9 % for combi-nation; in femoral neck 10.8 % vs 3.1 %, respectively). This isone of the only studies to evaluate effects of TPTD over a full24 month period and hip BMD continued to increase duringthe latter 6 months. The treatment duration is unlikely to be theonly explanation for the difference in hip effect in this study vsall other studies, because the increment in femoral neck BMDwas already significantly higher at 18 months with TPTDmonotherapy vs combination therapy. The TPTD dose usedhere was double the approved 20 mcg dose, and becauseTPTD/PTH effects on BMD are clearly dose dependent, theapplicability to other combination treatment studies is unclear.Bone turnover markers of both formation and resorption in-creased in the combination group above the suppressed base-line (after 6 months of Aln alone) [1].

Black et al. randomized 238 treatment naïve women toPTH with Aln vs PTH and Aln monotherapies [6]. DXAassessed spine BMD increased similarly with PTH monother-apy and PTH/Aln combination treatment (6.3 % and 6.1 %,respectively). However, total hip BMD increased significantly(1.9 %) with the PTH/Aln combination, but not with PTHmonotherapy (0.3 %). Radial BMD declined more with PTHmonotherapy (−3.4 %) than with combination therapy(−1.1 %). QCT-measured increases in integral spine and totalhip were similar between the PTH monotherapy and combi-nation groups, but trabecular spine BMD increased more withPTH monotherapy (25.5 %) than with combination treatment(12.6 %). In contrast, QCT assessed cortical bone densitydeclined in the hip (−1.7 %) with PTH alone, but was un-changed in the combination group. In summary, the DXAresults demonstrated that for the spine, there was no evidenceof additive effect with combination vs PTH monotherapy;however, in the hip, BMD increments were superior with thecombination treatment [6]. Evidence of a blunting effect withcombination PTH/Aln treatment was seen only when evalu-ating trabecular bone by QCT. Because single energy QCT-based BMD increments induced by PTH may be artifactuallyelevated by reductions in bone marrow fat [7, 8], it is unclear

386 Curr Osteoporos Rep (2014) 12:385–395

how to interpret the QCT findings vs DXA results. As allpatients in this trial were on potent osteoporosis medications,there were few clinical fractures, with no group differences.Incident morphometric vertebral fractures were not reported.

Cosman et al. randomized 412 treatment naïve postmeno-pausal women to receive TPTD monotherapy, intravenouszoledronic acid (Zol) monotherapy, or a combination ofTPTD and Zol [9••] for 1 year. With combination therapy,serum CTX level declined similarly to that seen with Zolmonotherapy, whereas serum PINP level declined only mod-estly compared with that seen with Zol monotherapy. At12 months, the spine BMD increase was similar withTPTD/Zol (7.5 %) to that seen with TPTD monotherapy(7.0 %), whereas the hip BMD increase was larger (2.3 %with TPTD/Zol vs 1.1 % TPTD alone) as was the femoralneck BMD increase (2.2 % combination vs 0.1 % TPTDalone). Therefore, there was clearly an additive effect ofTPTD/Zol in the hip region compared with TPTD monother-apy. In the combination group, peak BMD increments werereached the fastest at both spine and hip sites, compared witheither agent alone (Fig. 1), with the most rapid increases seenin the first 3 months, corresponding to the most prominentreduction in level of CTX. Between 6 and 12 months, as theantiresorptive effect of Zol was lost (as Zol was remodeled outof the skeleton), hip BMD plateaued with combination treat-ment and the BMD increase due to Zol monotherapy reachedthe same level as that seen for combination treatment. Thisstudy illustrated quite clearly that the largest BMD gain withcombination treatment occurs when the antiresorptive effect ismost marked. Clinical fractures occurred in 9.5 % of patientson Zol monotherapy, 5.8 % on TPTD monotherapy and 2.9 %on combination treatment (P<0.05 vs Zol alone).

Schafer et al. studied 2 different regimens of PTH andmonthly oral ibandronate (Ibn) [5]. 44 women with low bonemass were randomized to receive 6 months of PTH concurrentwith Ibn followed by Ibn monotherapy for 18 months (total of24 months Ibn); or 2 cycles of sequential regimens involvingPTH for 3 months followed by Ibn for 9 months (total of18 months Ibn). Both regimens improved BMD of the spine,total hip and femoral neck similarly. Increments were similarto those seen with PTH alone over 18 months [6], despite thefact that PTH was only given for 6 months total duration ineither regimen. Both regimens resulted in increases in markersof bone formation during the initial months of PTH treatment,with increases that were somewhat greater in the sequentialarm where PTH was given as monotherapy. Formationmarkers declined during the latter 3 months in the concurrentregimen. During the initial months of PTH treatment,CTX did not increase with either regimen. In the se-quential regimen, however, with the second cycle ofPTH treatment (given after 9 months of Ibn alone),when women were switched back to PTH, CTX levelsincreased significantly from the suppressed baseline (onIbn). This finding of more exuberant bone resorption whenswitching from a bisphosphonate to PTH (vs the bone resorp-tion response seen with de novo treatment with PTH orTPTD) is consistent with other studies discussed below, whenwomen are switched from alendronate or risedronate to PTHor TPTD [10, 11]

Teriparatide and Denosumab

Tsai et al. randomized 94 women (primarily treatment naïve)to receive TPTD monotherapy, denosumab (Dmab)

Fig. 1 LSM percentage change in BMD from baseline at lumbar spine(a), total hip (b), and femoral neck (c) according to treatment. a=p<0.001vs teriparatide (TPTD) alone and vs zoledronic acid (ZOL) alone; b=p<.001vs ZOL alone; c=p<.05 vs TPTD alone; d=p<.05 vs ZOL alone. Data arefrom intent-to-treat population excluding missing values. Bars show

standard error. Republished with permission from John Wiley & Sons:permission conveyed through Copyright Clearance Center, Inc. CosmanF, Eriksen EF, Recknor C, et al. Effects of intravenous zoledronic acidplus subcutaneous teriparatide [rhPTH(1–34)] in postmenopausal osteo-porosis. J Bone Miner Res. 2011;26:503–1. [9••]

Curr Osteoporos Rep (2014) 12:385–395 387

monotherapy or combination TPTD/Dmab [12••]. Previousoral bisphosphonate use was noted in about 35 % of patients,but all use had finished at least 6 months prior to recruitmentinto the trial and mean time since discontinuation of the priorbisphosphonate was between 27 and 42 months for the threetreatment groups. Therefore, this is largely a treatment naïvepopulation. Spine, total hip, and femoral neck BMD (Fig. 2)increased significantly more with TPTD/Dmab combinationtreatment (9.1 %, 4.9 %, 4.2 %, respectively) than with TPTDmonotherapy (6.2 %, 0.7 %, 0.8 %), or Dmab monotherapy(5.5 %, 2.5 %, 2.1 %). In addition, radius BMD increasedmore with combination therapy (2.6 %) vs TPTD monother-apy (−1.8 %). With TPTD/Dmab coadministration, serumCTX levels paralleled the prominent decline seen withDmab monotherapy throughout the first 12 months, whereasserum PINP declined more slowly and did not overlap Dmab

monotherapy values until sometime between 6 and 12months.Mean serum osteocalcin level with TPTD/Dmab coadminis-tration was still above that seen with Dmab monotherapythroughout the first year.

In contrast to the study with TPTD/Zol combination treat-ment, mean levels of the biochemical markers did not exceedbaseline at any point during the trial. Most importantly, incontrast to the serum CTX elevation during the second half ofthe year with the TPTD/Zol coadministration, serum CTXremains completely suppressed with TPTD/Dmab coadmin-istration. The BMD increments with TPTD/Dmab combina-tion treatment were very similar to those seen with the PTH/Zol combination during the first several months, but withTPTD/Dmab coadministration, BMD levels continued to in-crease during the second half of the year, as the antiresorptiveeffect of Dmab persisted.

Fig. 2 (a-d).Mean (SE) percentage changes in bone mineral density (a)Posterior-anterior lumbar spine; (b) Femoral neck; (c) Total hip. (d) Distalone-third of the radial shaft. *P<0.05 vs baseline. †P<0.05 vsdenosumab alone. ‡P<0.001 vs denosumab alone. §P<0.05 vsteriparatide alone. P<0.001 vs teriparatide alone. From Cosman (1).

Reprinted by permission from Macmillan Publishers Ltd: copyright(2014). Cosman F. Combination therapy for osteoporosis: a reappraisal.BoneKEy Rep. 2014;3. [1]. BMD bone-mineral density, DMABdenosumab, TPTD teriparatide

388 Curr Osteoporos Rep (2014) 12:385–395

During the second year continuation of this study, spine,total hip and femoral neck BMD increased in all three groups,however, there were no group differences in the magnitude ofthese increments in the second year with TPTD/Dmab coad-ministration vs monotherapy [13•]. Therefore, there was nodemonstrable advantage of TPTD/Dmab coadministrationover Dmab or TPTD monotherapy after the first year.Consistently, in the second year, levels of PINP and CTXwere similar with TPTD/Dmab coadministration vs Dmabmonotherapy. Mean osteocalcin level continued to declinefrom 12–24 months with TPTD/Dmab coadministration butthe mean level was still above that seen with Dmab mono-therapy at the end of the trial.

Postmenopausal Women on Established AntiresorptiveTherapy

Prior and Ongoing Hormone Therapy or Raloxifene

In 52 women with osteoporosis on long-term established HT[14, 15], daily TPTD produced rapid increases in markers ofbone formation, and delayed increases in markers of boneresorption [15]. This period of time, where augmentation ofbone formation exceeds bone resorption, has been referred toas the anabolic window, and may represent the most efficientbone building opportunity with TPTD. Another importantobservation from this trial is that bone turnover levelsremained elevated for only 18–24months, after which, markerlevels declined [14]. The mechanism of this apparent resis-tance to ongoing TPTD administration has still not beendetermined, but is consistently seen with biochemical markersin both TPTD monotherapy and combination therapy trials,often with marker levels peaking within the first year and thendeclining toward or to baseline. BMD increased by about14% over 3 years in women receiving TPTD/HTcombinationtreatment, with evidence of the most rapid BMD rise withinthe first 6 months. Total body and hip BMD increased by 4 %in patients on TPTD/HTcombination treatment. After 3 years,vertebral deformity occurrence was significantly reduced inpatients receiving TPTD/HT combination therapy comparedwith HT monotherapy [4].

Another study of similar design performed in women whohad previously been treated with HT showed BMD incre-ments by DXA in the TPTD/HT group of 30 % in the spineand 12% in the femoral neck vs placebo [16]. No fracture datawere presented from this trial, and the data have never beenpublished in a peer-reviewed journal. A third study, utilizingTPTD 40 mcg/day, enrolled 1 subgroup of women on priorHT (n=122), as in the previously discussed two trials [2].B’MD increments were approximately 11 % in the spine and3 % in the total hip in women randomized to TPTD 40 mcg

daily plus ongoing HT (both increments significantly greaterthan HT alone).

Cosman et al. evaluated postmenopausal women previous-ly treated with raloxifene for at least 1 year (n=42) andrandomized them to remain on Ral monotherapy or to receiveTPTD/Ral combination treatment. The TPTD/Ral combina-tion group had an increment of about 10% in the lumbar spineand 3 % in the total hip, whereas those randomized to theongoing Ral monotherapy had no BMD change [17].Increases in both biochemical turnover markers at 3 monthscorrelated with increases in spine BMD at 1 year.

These studies clearly indicate that combination anabolictherapy with ongoing HT or Ral consistently produces agreater increment in BMD than continuing HT or Ralmonotherapy.

Prior Bisphosphonate Therapy

Patients maintained and stabilized on long-term bisphospho-nate treatment are a distinct, but clinically very importantpopulation because many of these patients have fractures ordo not achieve a BMD above osteoporotic range, and thus,might benefit from anabolic therapy. At least 50 % of allTPTD/PTH treatment is initiated in patients who have re-ceived prior antiresorptive agents, most of which is bisphos-phonate. Studies evaluating TPTD treatment in treatment-experienced women have followed 2 basic designs:antiresorptive agents are stopped when TPTD is started [11,18, 19], or antiresorptive agents are continued when TPTD isstarted [17, 20]. Outcomes differ with these distinct studydesigns. In studies where bisphosphonates are discontinued,the spine BMD increment is of consistently lesser magnitudeand perhaps more importantly, hip BMD declines consistentlyover the first year, an effect not seen in protocols where TPTDis added to ongoing bisphosphonate administration.

Studies Where bisphosphonates were Stopped when TPTDor PTH was Started (Monotherapy)

In an observational study where TPTD was given to womenafter cessation of long-term Aln or Ral [19], bone turnovermarkers increased as did spine BMD, but these increases weresomewhat delayed and of lower magnitude in patientspretreated with Aln, compared with those in patientspretreated with Ral. In the cohort on prior Aln, but not in thecohort on prior Ral, a significant reduction in hip BMD wasseen at 6 months, although hip BMD was back to baseline by18 months.

Similarly, in a nonrandomized, prospective study of wom-en previously treated with risedronate (Ris; n=146) or Aln(n=146), Miller et al. [11] found that biochemical markers ofbone resorption were already increased within 1 month oftreatment with PTHmonotherapy in both cohorts, an outcome

Curr Osteoporos Rep (2014) 12:385–395 389

not seen within the first month in treatment-naïve patientstreated with TPTD or PTH monotherapy [21, 22].Furthermore, increases in markers of bone resorption at1 month are not seen in patients on prior antiresorptive therapywith combination therapy, where the antiresorptive agent iscontinued during administration of TPTD [14, 15, 20].Consistent with the Ettinger study above, Miller et al. dem-onstrated that hip BMD declined significantly in both thecohort on prior Ris as well as the cohort on prior Aln for theentire 1-year trial [11]. Finally, average spine BMD in a cohortof women who had been on prior bisphosphonates and thenswitched to teriparatide (Eurofors Trial) increased less than in acohort of treatment naïve women (9.8 %–10.2 % for thebisphosphonate treated and 13.1 % for treatment naïve women[18, 23]). More importantly, women on prior bisphosphonates,whowere switched to TPTDmonotherapy, had a decline in hipBMD over the first year of treatment, an effect which was notseen in the treatment naïve women [23] or in women whereTPTD is given in combination with ongoing bisphosphonatecombination treatment [10, 14, 15, 20].

Studies Where Bisphosphonate Therapy was ContinuedWhen TPTD was Initiated (Combination Therapy)

In Cosman et al. 126 women previously treated with long-term Aln (average duration 3.2 years) were randomized tocontinue Aln and to receive daily TPTD, cyclic TPTD (givenin a 3-month on/3-month off regimen), or Aln monotherapy[20]. Over just 15 months, spine BMD rose 6.1 % in the dailyTPTD/Aln combination group, a higher increment than theaverage changes seen in the studies above when the underly-ing bisphosphonate was discontinued and TPTD or PTH wasgiven as monotherapy. Moreover, mean hip BMD did notdecline at any time point during this study of TPTD/Alncombination treatment.

Comparing Teriparatide Monotherapy vs CombinationTherapy in a Randomized Trial of Women on PriorAlendronate or Raloxifene

In order to formally compare, in a randomized trial, the effectof continuing vs stopping the antiresorptive agent when TPTDis initiated, 198 women on prior Aln (n=102) or Ral (n=96)were randomized within each cohort to continue or stop theirantiresorptive when TPTDwas initiated [10]. This study was adirect randomized comparison of TPTD monotherapy vsTPTD combination therapy in treatment-experienced patients.Although an anabolic response was seen both biochemicallyand densitometrically in all groups, all biochemical turnovermarkers increased more in those randomized to TPTD mono-therapy. Of particular note was the early increase in CTX,which was already significantly elevated at 1 month in thepatients assigned to TPTD monotherapy after Aln treatment,

suggesting a truncation of the anabolic window in patientsfollowing this approach, similar to the findings in the studiesdiscussed above [5, 11, 19]. It appears that withdrawal ofbisphosphonate results in exaggerated bone resorption, partic-ularly in cortical bone where there is less incorporated bis-phosphonate. As a result, BMD declined in the first 6 monthsin the hip (consistent with all TPTD or PTH monotherapystudies in bisphosphonate experienced patients). The in-creases at both 6 and 18 months at both spine and hip weregreater in those patients randomized to TPTD/Aln combina-tion therapy compared with TPTD monotherapy, and at notime point did hip BMD decline in the combination therapygroup (Fig. 3). Differences between TPTD combination andmonotherapy were less marked with Ral pretreatment.

Volumetric BMD of the hip did not change in women whoswitched from Aln to teriparatide monotherapy, but increasedsignificantly at both 6 and 18 months in those who receivedcombination TPTD/Aln therapy [24••]. Volumetric BMD ofthe cortical compartment of the hip declined significantly inthe TPTD monotherapy group. Although strength of thehip, assessed by finite element analysis, did not declinewith TPTD monotherapy, implying that the decline incortical BMD was not in an area critical for strength,hip strength increased significantly only with combinationtherapy [24••] (Fig. 4). Again, differences between TPTDmonotherapy and TPTD/Ral combination therapy were muchless apparent.

The findings from this study have important implicationsfor the clinical use of TPTD in patients who have receivedprior bisphosphonates and are at high risk for fractures of thehip and perhaps other skeletal sites that are rich in corticalbone. Such a patient could include a patient on an oral bis-phosphonate who sustains a hip fracture. It may be that thewithdrawal of the bisphosphonate actually facilitates an exag-gerated bone resorption response to TPTD and offsets theexpected positive bone balance, particularly in the corticalskeleton where there is little incorporated bisphosphonate.Using combination therapy in this type of patient would beexpected to provide additional benefit to the other at risk hipand the rest of the skeleton.

Teriparatide Followed by Addition of Raloxifeneor Alendronate

Muschitz et al. randomized 125 postmenopausal women whohad received TPTD treatment for 9 months to subsequentlystay on TPTDmonotherapy or initiate combination therapy byadding Aln or Ral at the 9-month point, while continuingTPTD for an additional 9 months [25••]. Almost all of thesubjects had had prior osteoporosis treatment, about 75 %with Aln and about 25%with Ris (prior to the initial 9 monthsof TPTD treatment). However, there are few details about theextent and recency of this treatment. Between 9 and

390 Curr Osteoporos Rep (2014) 12:385–395

Fig. 3 Percentage change in DXA BMD from baseline. (a) Lumbarspine; (b) total hip; (c) femoral neck, by treatment group. Left panelshows prior alendronate groups, and right panel shows prior raloxifenegroups. #, P≤0.05; and §, P≤0.01 for difference between groups withinthe alendronate or raloxifene stratum. Values are mean ± SE. Republishedwith permission from the Endocrine Society, permission conveyed

through Copyright Clearance Center, Inc. Cosman F, Wermers RA,Recknor C, Mauck KF, Xie L, Glass EV, et al. Effects of teriparatide inpostmenopausal women with osteoporosis on prior alendronate or ralox-ifene: differences between stopping and continuing the antiresorptiveagent. J Clin Endocrinol Metab. 2009;94:3772–80. [10]

Curr Osteoporos Rep (2014) 12:385–395 391

18 months, the serum CTX level was stable in the TPTDmonotherapy group but declined with the addition of Raland declined more prominently with the addition of Aln.Spine BMD by DXA increased more in the combinationtherapy groups (9.2 % TPTD/Aln group, 10 % TPTD/Rlxgroup) than with TPTD monotherapy (6 %). Total hip BMDincreased significantly more in the TPTD/Aln combinationgroup (7 %) compared with either the TPTD/Rlx combinationgroup (4.2 %) or TPTD monotherapy (4.4 %). There weresimilar differences in the femoral neck. Changes in volumetricBMD (by QCT) in the spine were greater in both combinationtherapy groups than with monotherapy. In the hip, volumetricBMD increments were largest in the TPTD/Aln combinationgroup and significantly higher for integral, trabecular, andcortical BMD in the total hip and for integral and corticalBMD at the femoral neck compared with either TPTD mono-therapy or TPTD/Rlx combination therapy [25••].

In a 12-month extension to this trial, the TPTD monother-apy group received only calcium and vitamin D, and the 2combination therapy groups were continued on their respec-tive antiresorptive agent alone [26•]. BMD continued to in-crease in the spine, total hip, and femoral neck (over 4 % ateach site) in those on Aln, but increased modestly (2 %) onlyat the spine in those on continued Ral. In those on no phar-macologic agent, BMD declined slightly in the total hip, butincreased slightly in the spine (1.6 %).

Novel Cyclic Regimens of TPTD/PTH and Bisphosphonates

Two studies evaluated combination regimens where TPTDwas given in 3-month cycles superimposed on continued or

intermittent bisphosphonate treatment. The first investigationmentioned randomized 126 women established on oral Aln todaily TPTD with ongoing Aln vs 3 month cycles of TPTDwith ongoing Aln vs Aln monotherapy [20]. BMD changeswere similar between the 2 arms of the study over 15 months.Findings of the cyclical PTH followed by Ibn trial arediscussed above [5].

Combination PTH Plus Antiresorptive Therapy in Men

Teriparatide Plus Alendronate

Eighty-three men with osteoporosis were assigned toTPTD at 40 mcg/day, Aln monotherapy, or TPTD after6 months of Aln pretreatment, with ongoing TPTD/Alncoadministration [27]. A substantial proportion of menin both TPTD groups required dose adjustment (by25 %–50 %) due to hypercalcemia or side effects.After a total of 24 months of TPTD administration,spine BMD increased most in the TPTD monotherapygroup (18.1 %), compared with that in the combinationgroup (14.8 %) or Aln monotherapy (7.9 %). Similartrends were seen for the lateral spine and femoral neck,but for the total hip and total body, increases were similarin the 3 treatment groups. In contrast, in the radius, BMDdeclined with TPTDmonotherapy, with slight increases in theother groups. Spine trabecular bone density on QCT increased48 % with TPTD alone, 17 % with the combination, and 3 %with Aln alone.

Fig. 4 Effects of adding vs switching to teriparatide in postmenopausalwomen with osteoporosis pretreated with raloxifene or alendronate, in-cluding (a) hip volumetric BMD at 6 and 18 months, and (b) hip strengthat 6 and 18 months. Note that statistically significant changes frombaseline are indicated by footnote symbols, and P values from statisticalcomparisons between the Add and Switch groups in each stratum areshown above each pair of bars. Republished with permission from John

Wiley & Sons, permission conveyed through Copyright Clearance Cen-ter, Inc. Cosman F, Keaveny TM, Kopperdahl D, Wermers RA, Wan X,Krohn KD, et al. Hip and spine strength effects of adding vs switching toteriparatide in postmenopausal women with osteoporosis treated withprior alendronate or raloxifene. J Bone Miner Res. 2013;28:1328–36.[24••]

392 Curr Osteoporos Rep (2014) 12:385–395

Teriparatide Plus Risedronate

Walker et al. randomized 29 men with low BMD to receiveTPTD coadministered with Ris vs TPTD or Ris monotherapy[28]. Spine BMD increased similarly with TPTD/Ris combi-nation vs TPTDmonotherapy (7 % vs 5.7 %) at 18 months. Inthe total hip and femoral neck, however, BMD increasedsignificantly more with combination treatment (3.9 % at totalhip and 8.5 % at femoral neck) compared with TPTD mono-therapy (0.3 % and 3.9 %, respectively). Biochemical markerlevels for both PINP and CTX increased above baseline forthe entire duration of the trial. With the above mentionedstudies utilizing PTH/Aln [6], TPTD/Dmab [12••], andPTH/Ibn [5], in the combination groups, biochemicalmarkers of bone resorption were below baseline or un-changed for the entire trial. With the TPTD/Zol combination[9••], mean CTX level was below baseline for the first6 months of the study. Perhaps the bone resorption markerchanges in this trial are due to the more modest antiresorptivepotency of Ris at the dose utilized compared with Aln, Zol, orIbn. In the study byDeal et al. [4], where TPTDwas combinedwith Ral (another low potency agent), the CTX level wasabove baseline at 3 and 6 months, though the increase wasof lower magnitude with TPTD/Ral coadministration thanwith TPTD monotherapy.

Commentary and Conclusions

All TPTD/PTH combination studies with antiresorptivetherapies produce unique effects, at least in terms of themagnitude of biochemical marker and BMD changes.However, there are some consistent observations. First,it does appear that both HT and Ral are consistentlypermissive with use of TPTD/PTH both in treatment naiveand treatment-experienced women, allowing additive ef-fects of both agents in both spine and hip sites. Withbisphosphonates, the impact of combination administrationof an agent such as PTH with Aln or TPTD with Zol intreatment naïve women or TPTD with Ris in men does notprovide greater increments in spine BMD compared withTPTD/PTH monotherapy, but does result in consistently su-perior hip BMD effects, compared with PTH or TPTD mono-therapy (in all studies utilizing the approved dose of TPTD).This superior hip BMD outcome from combination treatmentstudies in treatment naïve individuals is also consistent withcombination treatment studies in bisphosphonate treatment-experienced women, where switching to PTH/TPTD mono-therapy results in an early decline in hip BMD compared withan increase in hip BMD with combination treatment. In con-trast to the hip site, however, the only combination treatmentswhich clearly show an additive benefit on spine BMD are the

TPTD/Dmab combination [12••] or TPTD followed by com-bination TPTD/Aln [25••].

There are many limitations to the currently available data.Most importantly, there are no adequately powered fractureoutcome studies. Furthermore, when trying to distinguisheffects of treatment combinations in treatment naïve vstreatment-established individuals, it is not known how muchtime needs to elapse after treatment discontinuation to consid-er a patient ‘treatment-naïve’ again. This will be an increas-ingly difficult distinction as more and more medication holi-days from bisphosphonates are being attempted. Full miner-alization of newly formed bone may not be seen in short-termstudies. DXA cannot measure all properties that might im-prove bone strength with TPTD therapy [1].

It is clearly time to re-evaluate the potential benefits ofcombination therapy. Differences in BMD responses betweencombination treatment and monotherapy might be clinicallymeaningful in some patients. For example, in patients on priorbisphosphonates who sustain a hip fracture or who have a verylow hip BMD or declining hip BMD while on their bisphos-phonate, adding TPTD might provide a substantial benefitover continuing bisphosphonate monotherapy or switchingto TPTD monotherapy. Furthermore, in treatment naïve (andpreviously treated) individuals with multiple prior fractures,particularly those involving spine and nonspine sites, certaincombination treatments might be warranted above monother-apy to achieve optimal short-term benefits on bone mass andpotentially bone strength [1].

Compliance with Ethics Guidelines

Conflict of Interest F. Cosman has received research support fromNIH; personal fees from Lilly, Novartis, Amgen, Merck, Zosano, GSK,Pfizer, Enteris, Radius, Asahi-Kasei; and nonfinancial support from Lilly.

Human andAnimal Rights and InformedConsent All studies by theauthor involving animal and/or human subjects were performed afterapproval by the appropriate institutional review boards. When required,written informed consent was obtained from all participants.

References

Papers of particular interest, published recently, have beenhighlighted as:• Of importance•• Of major importance

1. Cosman F. Combination therapy for osteoporosis: a reappraisal.BoneKEy Rep. 2014;3. doi: 10.1038/bonekey.2014.13.

2. Ste-Marie LG, Schwartz SL, Hossain A, Desaiah D, Gaich GA.Effect of teriparatide [rhPTH(1–34)] on BMD when given to post-menopausal women receiving hormone replacement therapy. JBone Miner Res. 2006;21:283–91.

Curr Osteoporos Rep (2014) 12:385–395 393

3. Finkelstein JS, Wyland JJ, Lee H, Neer RM. Effects of teriparatide,alendronate, or both in womenwith postmenopausal osteoporosis. JClin Endocrinol Metab. 2010;95:1838–45.

4. Deal C, Omizo M, Schwartz EN, Eriksen EF, Cantor P,Wang J, et al. Combination teriparatide and raloxifene ther-apy for postmenopausal osteoporosis: results from a 6-monthdouble-blind placebo-controlled trial. J Bone Miner Res.2005;20:1905–11.

5. Schafer AL, Sellmeyer DE, Palermo L, Hietpas J, Eastell R,Shoback DM, et al. Six months of parathyroid Hormone (1–84)administered concurrently vs sequentially with monthlyibandronate over two years: the PTH and ibandronate combinationstudy (PICS) randomized trial. J Clin Endocrinol Metab. 2012;97:3522–9.

6. Black DM, Greenspan SL, Ensrud KE, Palermo L, McGowan JA,Lang TF, et al. The effects of parathyroid hormone and alendronatealone or in combination in postmenopausal osteoporosis. N Engl JMed. 2003;349:1207–15.

7. Gluer CC, Genant HK. Impact of marrow fat on accuracy ofquantitative CT. J Comput Assist Tomogr. 1989;13:1023–35.

8. Kuiper JW, van Kuijk C, Grashuis JL, Ederveen AG, Schutte HE.Accuracy and the influence of marrow fat on quantitative CT anddual-energy X-ray absorptiometry measurements of the femoralneck in vitro. N Engl J Med. 1996;6:25–30.

9.•• Cosman F, Eriksen EF, Recknor C, et al. Effects of intrave-nous zoledronic acid plus subcutaneous teriparatide [rhPTH(1–34)] in postmenopausal osteoporosis. J Bone Miner Res. 2011;26:503–1. This study randomized 412 treatment naïve women toreceive teriparatide, zoledronic acid, or combination therapy for1 year. The increment in hip BMD was superior with combinationtreatment compared with TPTD monotherapy.

10. Cosman F, Wermers RA, Recknor C, Mauck KF, Xie L, Glass EV,et al. Effects of teriparatide in postmenopausal women withosteoporosis on prior alendronate or raloxifene: differencesbetween stopping and continuing the antiresorptive agent. JClin Endocrinol Metab. 2009;94:3772–80.

11. Miller PD, Delmas PD, Lindsay R, et al. Early responsiveness ofwomen with osteoporosis to teriparatide after therapy withalendronate or risedronate. J Clin Endocrinol Metab. 2008;93:3785–93.

12.•• Tsai JN, Uihlein AV, Lee H, Kumbhani R, Siwila-Sackman E,McKay EA, et al. Teriparatide and denosumab, alone or combined,in women with postmenopausal osteoporosis: the DATA studyrandomised trial. Lancet. 2013;382:50–6. In this trial, 94 women(on no prior treatment for at least 6 months before enrollment) wererandomized to receive teriparatide, denosumab, or combinationtreatment. BMD of the spine and hip improved more with thecombination than with either monotherapy over 1 year.

13.• Leder BZ, Tsai JN, Uihlein AV, Burnett-Bowie SA, Zhu Y, Foley K,et al. Two years of Denosumab and teriparatide administration inpostmenopausal women with osteoporosis (The DATA ExtensionStudy): a randomized controlled trial. J Clin Endocrinol Metab.2014;99:1694–700. This study reports data from the second yearof the study described in Reference 12. Although BMD increased inall groups in the second year, there were no group differences in themagnitude of the increments at any skeletal site, suggesting that theBMD gain with this combination over monotherapy is limited to 1year.

14. Cosman F, Nieves J, Woelfert L, Formica C, Gordon S,Shen V, et al. Parathyroid hormone added to establishedhormone therapy: effects on vertebral fracture and maintenanceof bone mass after parathyroid hormone withdrawal. J Bone MinerRes. 2001;16:925–31.

15. Lindsay R, Nieves J, Formica C, Henneman E, Woelfert L, Shen V,et al. Randomised controlled study of effect of parathyroid hormoneon vertebral-bone mass and fracture incidence among

postmenopausal women on oestrogen with osteoporosis. Lancet.1997;350:550–5.

16. Roe EB, Sanchez SD, del Puerto GA, Pierini E, Bacchetti P, CannCE, et al. Parathyroid hormone 1–34 (hPTH 1–34) and estrogenproduce dramatic bone density increases in postmenopausalosteoporosis- results from a placebo-controlled randomized trial. JBone Miner Res. 1992;12(1):S137 [Abstract].

17. Cosman F, Nieves JW, Zion M, Barbuto N, Lindsay R. Effect ofprior and ongoing raloxifene therapy on response to PTH andmaintenance of BMD after PTH therapy. N Engl J Med. 2008;19:529–35.

18. Boonen S, Marin F, Obermayer-Pietsch B, et al. Effects of previousantiresorptive therapy on the bone mineral density response to twoyears of teriparatide treatment in postmenopausal women withosteoporosis. J Clin Endocrinol Metab. 2008;93:852–60.

19. Ettinger B, San Martin J, Crans G, Pavo I. Differential effects ofteriparatide on BMD after treatment with raloxifene or alendronate.J Bone Miner Res. 2004;19:745–51.

20. Cosman F, Nieves J, Zion M, Woelfert L, Luckey M, Lindsay R.Daily and cyclic parathyroid hormone in women receivingalendronate. N Engl J Med. 2005;353:566–75.

21. McClung MR, San Martin J, Miller PD, Civitelli R, Bandeira F,Omizo M, et al. Opposite bone remodeling effects of teriparatideand alendronate in increasing bone mass. Arch Intern Med.2005;165:1762–8.

22. Neer RM, Arnaud CD, Zanchetta JR, et al. Effect of parathyroidhormone (1–34) on fractures and bone mineral density in postmen-opausal women with osteoporosis. N Engl J Med. 2001;344:1434–41.

23. Obermayer-Pietsch BM, Marin F, McCloskey EV, et al. Effects oftwo years of daily teriparatide treatment on BMD in postmenopaus-al women with severe osteoporosis with and without priorantiresorptive treatment. J Bone Miner Res. 2008;23:1591–600.

24.•• Cosman F, Keaveny TM, Kopperdahl D, Wermers RA, Wan X,Krohn KD, et al. Hip and spine strength effects of adding vsswitching to teriparatide in postmenopausal women with osteopo-rosis treated with prior alendronate or raloxifene. J BoneMiner Res.2013;28:1328–36. This study is a QCT follow-up to Reference 10.Women on long-term prior alendronate were randomized to switchto teriparatide monotherapy or add teriparatide to ongoingalendronate in combination. Over 18 months, volumetric BMDand strength of the hip increased in the combination group, butnot in those switched to teriparatide monotherapy.

25.•• Muschitz C, Kocijan R, Fahrleitner-Pammer A, Lung S, Resch H.Antiresorptives overlapping ongoing teriparatide treatment result inadditional increases in bone mineral density. J Bone Miner Res.2013;28:196–205. This very interesting study evaluated 125 womenwho had had prior bisphosphonate treatment and were then treatedwith teriparatide. At 9 months into the teriparatide treatment,women were randomized to alendronate or raloxifene combinationtreatment with ongoing teriparatide vs continued teriparatidemonotherapy. Spine BMD increased more with both combinationtreatment arms than with monotherapy and hip BMD increasedmore with the teriparatide alendronate combination vs teriparatidemonotherapy or teriparatide raloxifene combination treatment.

26.• Muschitz C, Kocijan R, Fahrleitner-Pammer A, Pavo I, Haschka J,Schima W, et al. Overlapping and continued alendronate or ralox-ifene administration in patients on teriparatide: effects on areal andvolumetric bone mineral density—The CONFORS study. J BoneMiner Res. 2014;29:1777–85. In this 1-year extension to the studydescribed in Reference25, women continued their respectivealendronate, raloxifene, or just calcium and vitamin D after a totalof 18 months of teriparatide. BMD continued to increase in bothspine and hip in those on continued alendronate, but there was noincrease in hip BMD in those continuing on raloxifene. Hip BMDdeclined in those on no pharmacologic agent.

394 Curr Osteoporos Rep (2014) 12:385–395

27. Finkelstein JS, Hayes A, Hunzelman JL, Wyland JJ, Lee H,Neer RM. The effects of parathyroid hormone, alendronate,or both in men with osteoporosis. N Engl J Med. 2003;349:1216–26.

28. Walker MD, Cusano NE, Sliney Jr J, Romano M, Zhang C,McMahon DJ, et al. Combination therapy with risedronateand teriparatide in male osteoporosis. Endocrine. 2013;44:237–46.

Author’s Note

This review is an update to a recent review published inBoneKEy Reports I entitled “Combination therapy for osteoporo-sis: a reappraisal” (doi: 10.1038/bonekey.2014.13. I have addedmore information based on reports published since it’s acceptancein December 2013.

Curr Osteoporos Rep (2014) 12:385–395 395