The Effects of Recombinant TSH on BoneTurnover Markers and Serum Osteoprotegerin

and RANKL Levels

Giuseppe Martini,1 Luigi Gennari,1 Vincenzo De Paola,1 Tania Pilli,2 Stefania Salvadori,1

Daniela Merlotti,1 Fabrizio Valleggi,1 Stella Campagna,1 Beatrice Franci,1 Annalisa Avanzati,1

Ranuccio Nuti,1 and Furio Pacini2

Objective: Recently it was found that thyrotropin (TSH) receptors are present both in osteoclast and osteoblastand that TSH can modulate bone remodeling independent of thyroid hormones. The aim of this study was,firstly, to evaluate the effects of acute administration of TSH on bone remodeling markers both in men and inwomen and, secondly, to evaluate if these effects are mediated by variations in serum osteoprotegerin (OPG)and receptor activator of nuclear factor-KB ligand (RANKL).Design: We studied 30 thyroidectomized patients (10 premenopausal and 10 postmenopausal women, 10 men)affected by thyroid carcinoma on L-thyroxine therapy. Eighty age- and sex-matched subjects were used ascontrols. A blood sample was drawn from each patient at baseline and 3 and 5 days after recombinant humanTSH (rhTSH) administration, in preparation for 131I whole body scan, to assess serum bone markers and serumOPG and RANKL levels.Main outcome: At baseline, postmenopausal women and men had significantly higher values of bone turnovermarkers and serum OPG compared to control subjects. In all thyroidectomized patients serum RANKL waslower than in controls. After rhTSH administration, serum N-terminal propeptide of type-I procollagen (PINP), amarker of bone formation, increased significantly in postmenopausal women, while serum RANKL significantlyincreased after 3 days in postmenopausal patients and men returning to baseline values at day 5. Serum OPGlevels did not change significantly.Conclusions: The low serum TSH observed in thyroidectomized patients on L-thyroxine therapy is associatedwith an increase of bone turnover in postmenopausal women and men that is associated with an increase ofOPG and a decrease of serum RANKL levels. The acute TSH administration results in an increase of PINP, anindex of osteoblastic activity, associated with an increase of serum RANKL. The lack of this response in pre-menopausal women suggests an influence of estrogen status on bone reactivity to TSH.

Introduction

Thyroid hormones exert a key role in normal skeletaldevelopment and in the maintenance of adult bone mass

(1). An increase of bone remodeling has been reported inadult thyrotoxicosis, characterized by an imbalance betweenbone resorption and formation, which results in net boneloss and an increased risk for osteoporotic fractures (2). Re-cently, it has been found that thyrotropin (TSH) receptors(TSHR) are present both in osteoclasts andosteoblasts and thatTSH can modulate bone remodeling independent of thyroid

hormones (3). Indeed, both heterozygotic and homozygoticTSHR null mice are osteopenic with evidence of enhancedosteoclast differentiation (4). The suppression of osteoclastactivation by TSH could be due to an inhibition of tumornecrosis factor a (TNFa) production (4). TNFa is a cytokinethat enhances osteoclast formation directly (5) and by upre-gulating the stromal cells production of receptor activator ofnuclear factor-KB ligand (RANKL) other than the respon-siveness of osteoclast precursors to this factor (6–7). RANKLis a cytokine produced by osteoblasts that stimulates osteo-clast activity and inhibits osteoclast apoptosis via its receptor

1Department of Internal Medicine, Endocrinology & Metabolism and Biochemistry, Section of Internal Medicine, University of Siena, Siena,Italy.

2Department of Internal Medicine, Endocrinology & Metabolism and Biochemistry, Section of Endocrinology, University of Siena, Siena,Italy.

THYROIDVolume 18, Number 4, 2008ª Mary Ann Liebert, Inc.DOI: 10.1089=thy.2007.0166

455

RANK, expressed by many tissues and cells, including os-teoclasts (8). In addition, TNFa and RANKL synergisticallyupregulate RANK expression in osteoclast precursors (9).Osteoprotegerin (OPG) is a decoy receptor expressed bystromal cells and osteoblasts that acts as antagonist ofRANKL (10). Alterations of the RANKL=OPG ratio are cri-tical in the pathogenesis of bone diseases that result fromincreased bone resorption. In a study on thyroidectomizedpostmenopausal women the administration of recombinanthuman TSH (rhTSH) caused an increase of bone formationand a decrease of bone resorption markers without anychanges on serum OPG levels (11).

To better understand the molecular basis of TSH action onbone, we evaluated the effect of rhTSH not only on serumOPG levels, but also on serum RANKL. Moreover, in orderto evaluate the potential influence of hormonal status on theresponse to rhTSH administration, we studied both men andpre- and postmenopausal women. Finally we investigated ifthe changes on serum OPG and RANKL are associated withmodifications of bone formation and resorption markers.

Subjects and Methods

We studied 30 consecutive thyroidectomized patients af-fected by differentiated thyroid carcinoma on L-thyroxinetherapy: 10 premenopausal women (aged 32.9� 7.6 years);10 postmenopausal women (aged 67.2� 7.9 years), and 10men (aged 49.5� 13.9 years). The exclusion criteria werechronic diseases associated with thyroid cancer or treatmentwith drugs known to influence bone metabolism during theprevious 12 months before the enrollment. Seventeen pa-tients with persistent disease or at high risk, received L-thyroxine in suppressive or semisuppressive doses (targetTSH:< 0.5 mIU=mL). Thirteen patients who had previousevidence of complete remission were treated with replace-ment doses of L-thyroxine aimed to maintain normal TSHconcentration.

The number of patients with lower values of TSH wassimilar in each group: six were premenopausal women, sixwere postmenopausal women, and five were men.

All patients had normal parathyroid function. The subjectswith thyroid cancer were recruited from the Section of En-docrinology of the University of Siena, among patients un-dergoing periodic staging of their disease. rhTSH (Thyrogen;Genzyme Therapeutics, Cambridge, MA) was administeredat the dosage of 0.9mg i.m. once daily for the first two daysto perform a diagnostic whole body scan and serum thy-roglobulin measurement.

As control groups we studied 15 premenopausal women(aged 35� 6.1 years), 50 postmenopausal women (aged 65�9.1 years), and 15 men (aged 50.6� 8.5 years) recruited fromthe Metabolic Bone Disease Unit of the University of Siena.All control subjects were healthy and were receiving no med-ications known to affect bone metabolism. Controls did notundergo rhTSH administration.

Before including a patient in the study, an informed con-sent was obtained. The study was approved by the localethics committee.

Laboratory analysis

A blood sample was drawn from each patient before and 3and 5 days after rhTSH administration. Blood samples were

collected in the morning between 8:00 and 9:00 A.M. after anovernight fast, centrifuged, and stored at �208C until mea-surements were performed. Sera were assayed for OPG andRANKL by using a sandwich enzyme immunoassay (Os-teoprotegerin and RANKL, Biomedica, Austria). All sampleswere measured in duplicate in single batches and averaged.The assay for OPG detect monomeric and dimeric forms aswell as OPG bound to its ligand RANKL. The RANKL assaydetects soluble, noncomplexed human RANKL. The detec-tion limit was 0.14 pmol=L and 0.4 pmol=L, respectively.

As markers of bone formation, we assayed serum bonealkaline phosphatase (bALP; Alkphase-B, Metra Biosystem,Mountain View, CA), serum osteocalcin (BGP RIA kit, Dia-Sorin, Stillwater, MN), and serum intact N-terminal pro-peptide of type I procollagen (PINP RIA, Orion Diagnostic,Finland). Serum cross-linked C-telopeptides of type I col-lagen (CTX, serum CrossLaps, Osteometer, Herlev, Den-mark) was used as index of bone resorption. Plasma intactparathyroid hormone was assessed by immunoradiometricmethod (N-tact PTH, IRMA kit, DiaSorin, Stillwater, MN).

The observed intra- and interassay coefficients of variationin our laboratory for each marker were, respectively, as fol-lows: 7.9% and 10.4% for OPG, 4.8% and 8.0% for RANKL,2.0% and 4.1% for bALP, 6.2% and 9.8% for BGP, 5.0% and7.5% for PINP, 6.1% and 5.4% for CTX, and 2.8% and 4.0%for PTH.

Serum calcium, phosphate, and creatinine were deter-mined by standard method.

Serum TSH and free thyroxine (fT4) were measured usingan ultrasensitive commercial immunometric assay method(Immulite 2000, Diagnostic Product Corp., Los Angeles, CA).The peak serum TSH was considered as the maximum levelreached by serum TSH after rhTSH. Serum free triiodothy-ronine (fT3) was measured using an enzyme-linked fluor-escent assay (VIDAS, BioMerieux Italia S.p.A, Rome, Italy).

Statistical analysis

Values are expressed as mean� SD. Group means werecompared by analysis of variance (one-way ANOVA) withBonferroni’s correction for multigroup comparison. To testthe effect of treatment, we used the repeated measureANOVA procedures. The p value< 0.05 were consideredsignificant. Analysis were performed using SPSS software forWindows, version 10.0 (SPSS Inc., Chicago, IL).

Results

Demographic and general features of subjects included inthe study are shown in Table 1.

The data were analyzed separately for men and pre-menopausal and postmenopausal women. Patients and age-matched control groups did not significantly differ regardingweight, height, and serum calcium, phosphorus, and creati-nine (data not shown). As expected in thyroid cancer pa-tients, basal serum TSH and fT4 levels were significantlylower and significantly higher, respectively, than in thecontrol group due to suppressive doses of L-thyroxine thatmost of them were receiving. Bone turnover was higher inpostmenopausal patients than in controls and premeno-pausal patients as showed by the significant increase inbALP, PINP, and CTX levels. Serum BGP did not differ

456 MARTINI ET AL.

between groups. A similar increase in bALP and CTX levelswas also observed in thyroidectomized men as compared tocontrols. The highest values of OPG were found in post-menopausal patients, while the highest serum RANKL levelswere found in men treated with L-thyroxine. All patientgroups showed serum RANKL values lower than controlgroups. In healthy subjects, postmenopausal women had thehighest values of OPG, while the highest values of serumRANKL were observed in men. In all patients a significantinverse correlation was observed between OPG and RANKLlevels (r¼�0.38; p< 0.05), while no correlations were foundbetween these cytokines and bone turnover markers. Whenthe patients were divided according to TSH levels, no dif-ferences were appreciated in bone turnover markers andserum OPG and RANKL levels at baseline and after rTSHadministration.

After rhTSH administration, a wide spectrum of values ofpeak serum TSH (mean, 48.2� 18.1 mU=mL; range, 18.1–75)was observed, with many of our patients (19 of 30 pa-tients) showing a peak serum TSH above 40 mU=mL (mean,59.2� 12.5 mU=mL; range, 40.8–75.5 mU=mL). In most pa-tients, the peak serum TSH was observed at day 3 (24 hoursafter the second rhTSH injection) and was lower in malesthan in females (37.4� 12.9 mU=mL vs. 43.9� 16.2 mU=mL,premenopausal patients, and 63.4� 15.0 mU=mL, postmeno-pausal patients, respectively). No significant changes inserum fT3 and fT4 during rhTSH stimulation were observed.In addition serum calcium, phosphorus, creatinine, and PTHdid not significantly change after rhTSH administration.

Indexes of bone turnover showed a statistically significantincrease of serum PINP at day 3 and 5 (Fig. 1A). A statisti-cally significant increase of serum RANKL was found at day3 while at day 5 the values returned to pretreatment levels(Fig. 1B). Serum OPG did not change significantly.

Table 1. General Characteristics of Patients Included in the Studya

Premenopausal women Postmenopausal women Men

Patients Controls Patients Controls Patients Controls

Subjects (n) 10 15 10 50 10 15Age (yr) 32.9� 7.6 35.1� 8.0 67.2� 7.9 61.0� 10.5 49.5� 13.9 52.1� 5.0Serum fT4 (pg=mL) 15.6� 3.3b 8.5� 1.3 14.8� 5.6b 9.1� 1.2 17.5� 2.7b 8.9� 1.5Serum fT3 (pg=mL) 3.0� 0.5 2.9� 0.2 4.0� 0.8b 3.1� 0.2 3.6� 0.8 3.2� 0.2Serum TSH (mU=mL) 0.9� 1.5b 3.1� 0.4 0.6� 0.5b 3.5� 0.2 0.8� 1.2b 3.4� 0.3bALP (pmol=mL) 9.5� 4.1 10.1� 3.5 22.7� 17bd 14� 2.8e 17.7� 6.4b 12.0� 3.5BGP (ng=mL) 2.5� 2.4 2.7� 1.3 2.7� 2.1 3.1� 1.8 2.0� 1.8 2.2� 0.9PINP (mcg=L) 40.3� 28.8 38� 11 67.1� 27.0bd 53� 15e 43.3� 18.9 44� 15CTX (ng=mL) 0.41� 0.2 0.31� 0.1 0.62� 0.2bd 0.40� 0.19e 0.51� 0.2b 0.37� 0.13PTH (pg=mL) 41.1� 7.8 39.2� 5.7 43.5� 6.3 40.1� 5.5 44.0� 7.1 41.1� 8.2OPG (pmol=L) 3.9� 2.0 3.0� 1.1 8.6� 3.3bc 5.91� 2.16f 5.3� 2.6 4.39� 0.9RANKL (pmol=L) 0.19� 0.1b 0.72� 0.2 0.12� 0.46b 0.92� 0.49 0.32� 0.3b 1.31� 0.6f

aValues are expressed as mean� SD. fT4, free thyroxine; fT3, free triiodothyronine; TSH, thyrotropin; bALP, bone alkaline phosphatase;BGP, osteocalcin; PINP, N-terminal propeptide of type-I procollagen; CTX, cross-linked C-telopeptides of type I collagen; PTH, parathyroidhormone; OPG, osteoprotegerin; RANKL, receptor activator of nuclear factor-KB ligand.

bp< 0.05 vs. controls.cp< 0.5 vs. premenopausal women and men (patients).dp< 0.05 vs. premenopausal (patients).ep< 0.05 vs. premenopausal (controls).fp< 0.05 vs. premenopausal and men (controls).

FIG. 1. Changes on serum N-terminal propeptide of type-Iprocollagen (PINP) (A) and serum receptor activator of nu-clear factor-KB ligand (RANKL) (B) levels after recombinanthuman thyrotropin administration in all patients (^), inpremenopausal women (�), in postmenopausal women (~),and in men (&). Values are expressed as mean� SE.*p< 0.05.

TSH AND BONE REMODELING 457

When the bone markers for men and premenopausal orpostmenopausal women were analyzed separately, we ob-served a significant increase of serum PINP only in postmen-opausal patients (Fig. 1A). The increase of serum RANKLwas found in both men and postmenopausal women, whileit was not significant in premenopausal patients (Fig. 1B).

Discussion

Experimental data confirm that TSH can modulate boneremodeling although the finding that TSHR knock-out micehave focal sclerosis in addition to high remodeling osteo-porosis suggests that in this case osteoblast activity is, at leastin part, dissociated spatially from osteoclast-induced boneresorption (3). The view that osteoporosis associated to hy-perthyroidism is due solely to an increase of serum fT3 andfT4 levels has been challenged. Indeed clinical studies haveshown that patients with Graves’ disease exhibited a per-sistent increase in bone turnover even after their serum fT3

and fT4 levels were normalized by treatment, unless theirserum TSH had been restored to normal values (12). More-over in a 3.7-year follow-up study, women with TSH con-centration of 0.1mU=L or less at baseline were at increasedrisk of hip and vertebral fracture (13). A recent study showsan association between low serum TSH and osteoporosis andosteopenia plus a graded increase in bone mineral densitywith increasing serum TSH across the normal range inhealthy women (14).

To better evaluate the in vivo direct effects of acute TSHadministration on bone turnover in the absence of a parallelincrease in thyroid hormones, we studied thyroidectomizedpatients on L-thyroxine therapy. Our patients showed acondition of subclinical hyperthyroidism that gives an in-crease of bone turnover particularly in postmenopausalwomen, but also in men. In premenopausal patients serumbone markers did not differ from those of control group andwere significantly lower than those of postmenopausal pa-tients. These data could be explained by the different effectsof TSH and thyroid hormones on bone according to hor-monal status. Indeed a recent meta-analysis suggests thatpostmenopausal women with subclinical hyperthyroidismare at risk of osteoporosis, whereas no increased risk wasobserved in men and premenopausal women (15).

The increase of bone turnover in postmenopausal women,other than the effect of ageing, justified the observed increaseof serum OPG, while the highest values of serum RANKL inmen has been already reported (16). The known effect ofsubclinical hyperthyroidism on serum OPG is evident inpostmenopausal patient while serum OPG levels are notdifferent from controls in premenopausal women and men.However data about serum OPG levels could be difficult toexplain since high levels of this cytokine are reported bothin hypothyroidism and hyperthyroidism and usually nor-malizes after appropriate medical therapy (17–19). Sub-clinical hyperthyroid patients also had low serum RANKLlevels. This can be considered an independent risk factor fordevelopment of osteoporosis as demonstrated by Schett et al.(20). However no correlation was found between OPG orRANKL levels and bone turnover markers in our study.

After acute TSH administration, the only serum markerthat changed significantly was serum PINP. This is not sur-

prising, because it is well known that PINP gave the highestsignal to noise ratio of all bone turnover markers and itsshort half-life enables it to be used to evaluate the acutechanges of bone turnover (21). Indeed sBALP is not suitablefor short-term observations for its long half-life and serumBGP still presents some problems of pre-analytic instabilityassociated to heterogenicity of BGP fragments in the serumresulting in considerable limitations in clinical application ofthis a priori highly specific marker (22,23). The increase ofPINP could be explained with osteoblast activation by TSHbringing about an increase of bone formation. This process isparticularly evident in postmenopausal patients for the highbone turnover that can amplify all biological responses.

The lack of the effect of TSH on CTX could be due toa delay of the action of TSH on osteoclasts. These data cansuggest an initial uncoupling of the bone remodeling processwith an increase of osteoblast activation followed by a laterosteoclast inhibition. Indeed, TSHR knock-out mice show anunbalance between bone formation and resorption (24). Ourresults differ from those of Mazziotti et al. (11) who found areversible suppression of bone resorption. An alternativeexplanation is that the small variations of serum CTX levelscompared to its biological variation does not allow to detectany significant change in our small group of patients.

Chronic TSH suppression produces a significant reductionof serum RANKL, justifying the early increase of serumRANKL observed after acute rhTSH administration. Im-portantly, RANKL also represents a trigger for bone forma-tion, by directly activating osteoblasts at concentrations wellbelow those necessary to induce osteoclastogenesis (25).

Although the present findings about TSH actions on boneturnover are interesting, they must be interpreted with cau-tion. In fact, the number of patients studied is small andsome concerns can arise when we try to explain what hap-pens in bone microenvironment evaluating serum levels ofcytokines involved. Moreover, the early response to rhTSHcharacterized by an increase of bone formation marker PINPand serum RANKL although provides evidence for a directTSH action on bone is not consistent with data from TSHRknock-out mice in which TSH has an inhibitory effect both inbone formation and resorption (26). However the anaboliceffect of TSH has been recently demonstrated in OVX rats forthe first time: TSH administration was able to enhance thetrabecular and cortical bone formation in 7 months afterOVX in aged animals that had established osteopenia,characterized by low bone turnover and reduced bone vo-lume (27).

Other authors also suggest that bone loss in hyperthyr-oidism mainly results from T3 excess rather than TSH defi-ciency (28). The future directions on the understanding themechanism of action of TSH on bone will be the evaluationof other cytokines such as TNFa, which is a good indicator ofbone loss in postmenopausal women (29–30). Moreover ithas been demonstrated that TSH directly inhibits TNFaproduction in mice suggesting that this factor is a criticalcytokine mediating the downstream antiresorptive effects ofTSH on the skeleton (3).

Another potential limitation of this study is represented bythe wide biological variability of serum bone turnover mar-kers that can be very useful when the changes are high (i.e.,in monitoring bisphosphonates or teriparatide therapy), but

458 MARTINI ET AL.

can result scarcely significant when bone turnover mod-ifications are small as with TSH administration.

In conclusion our data demonstrated that low levels ofserum TSH as in thyroidectomized patients on L-thyroxinetherapy are associated with an increase of bone turnover inmen and postmenopausal women with a compensatory in-crease of OPG and a decrease of serum RANKL. The increaseof osteoblastic activity following acute TSH administration isassociated with an increase of serum RANKL. The lack ofthis behavior in premenopausal women suggests a potentialinfluence of estrogen status on bone reactivity to TSH.

References

1. Bassett JH, Williams GR 2003 The molecular actions ofthyroid hormone in bone. Trends Endocrinol Metab 14:356–364.

2. Greenspan SL, Greenspan FS 1999 The effect of thyroidhormone on skeletal integrity. Ann Intern Med 130:750–758.

3. Abe E, Marians RC, Yu W, Wu XB, Ando T, Li Y, Iqbal J,Eldeiry L, Rajendren G, Blair HC, Davies TF, Zaidi M 2003TSH is a negative regulator of skeletal remodeling. Cell115:151–162.

4. Hase H, Ando T, Eldeiry L, Brebene A, Peng Y, Liu L,Amano H, Davies TF, Sun L, Zaidi M, Abe E 2006 TNFalphamediates the skeletal effects of thyroid-stimulating hormone.Proc Natl Acad Sci U S A 103:12849–12854.

5. Kim N, Kadono Y, Takami M, Lee J, Lee SH, Okada F, KimJH, Kobayashi T, Odgren PR, Nakano H, Yeh WC, Lee SK,Lorenzo JA, Choi Y 2005 Osteoclast differentiation in-dependent of the TRANCE-RANK-TRAF6 axis. J Exp Med202:589–595.

6. Weitzmann MN, Pacifici R 2006 Estrogen deficiency andbone loss: an inflammatory tale. J Clin Invest 116:1186–1194.

7. Cenci S, Toraldo G, Weitzmann MN, Roggia C, Gao Y, QianWP, Sierra O, Pacifici R 2003 Estrogen deficiency inducesbone loss by increasing T cell proliferation and lifespanthrough IFN-gamma-induced class II transactivator. ProcNatl Acad Sci U S A 100:10405–10410.

8. Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Boyle WJ,Riggs BL 2000 The roles of osteoprotegerin and osteoprote-gerin ligand in the paracrine regulation of bone resorption.J Bone Miner Res 15:2–12.

9. Zhang YH, Heulsmann A, Tondravi MM, Mukherjee A,Abu-Amer Y 2001 Tumor necrosis factor-alpha (TNF) sti-mulates RANKL-induced osteoclastogenesis via coupling ofTNF type 1 receptor and RANK signaling pathways. J BiolChem 276:563–568.

10. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS,Luthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shi-mamoto G, DeRose M, Elliott R, Colombero A, Tan HL, TrailG, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, HughesTM, Hill D, Pattison W, Campbell P, Sander S, Van G,Tarpley J, Derby P, Lee R, Boyle WJ 1997 Osteoprotegerin: anovel secreted protein involved in the regulation of bonedensity. Cell 89:309–319.

11. Mazziotti G, Sorvillo F, Piscopo M, Cioffi M, Pilla P, BiondiB, Iorio S, Giustina A, Amato A, Carella C 2005 Recombinanthuman TSH modulates in vivo c-telopeptides of type-I col-lagen and bone alkaline phosphatase, but not osteoprote-gerin production in postmenopausal women monitored fordifferentiated thyroid carcinoma. J Bone Miner Res 20:480–486.

12. Kumeda Y, Inaba M, Tahara H, Kurioka Y, Ishikawa T,Morii H, Nishizawa Y 2000 Persistent increase in boneturnover in Graves’ patients with subclinical hyperthyroid-ism. J Clin Endocrinol Metab 85:4157–4161.

13. Bauer DC, Ettinger B, Nevitt MC, Stone KL; Study of Os-teoporotic Fractures Research Group 2001 Risk for fracturein women with low serum levels of thyroid-stimulatinghormone. Ann Intern Med 134:561–568.

14. Morris MS 2007 The association between serum thyroid-stimulating hormone in its reference range and bone statusin postmenopausal women Bone 40:1128–1134.

15. Heemstra KA, Hamdy NA, Romijn JA, Smit JW 2006 Theeffects of thyrotropin-suppressive therapy on bone metabo-lism in patients with well-differentiated thyroid carcinoma.Thyroid 16:583–591.

16. Martini G, Gennari L, Merlotti D, Salvadori S, Franci MB,Campagna S, Avanzati A, De Paola V, Valleggi F, Nuti R2007 Serum OPG and RANKL levels before and after in-travenous bisphosphonate treatment in Paget’s disease ofbone. Bone 40:457–463.

17. Amato G, Mazziotti G, Sorvillo F, Piscopo M, Lalli E, BiondiB, Iorio S, Molinari A, Giustina A, Carella C 2004 Highserum osteoprotegerin levels in patients with hyperthyr-oidism: effect of medical treatment. Bone 35:785–791.

18. Guang-da X, Hui-ling S, Zhi-song C, Lin-shuang Z 2005Changes in plasma concentrations of osteoprotegerin beforeand after levothyroxine replacement therapy in hypothyroidpatients. J Clin Endocrinol Metab 90:5765–5768.

19. Mikosch P, Igerc I, Kudlacek S, Woloszczuk W, GallowitschHJ, Kresnik E, Stettner H, Grimm G, Lind P, Pietschmann P2006 Receptor activator of nuclear factor kappaB ligand andosteoprotegerin in men with thyroid cancer. Eur J Clin In-vest 36:566–573.

20. Schett G, Kiechl S, Redlich K, Oberhollenzer F, Weger S,Egger G, Mayr A, Jocher J, Xu Q, Pietschmann P, TeitelbaumS, Smolen J, Willeit J 2004 Soluble RANKL and risk ofnontraumatic fracture. JAMA 291:1108–1113.

21. Gallagher JC, Rosen CJ, Chen P, Misurski DA, Marcus R 2006Response rate of bone mineral density to teriparatide in post-menopausal women with osteoporosis. Bone 39:1268–1275.

22. Seibel MJ 2000 Molecular markers of bone turnover: bio-chemical, technical and analytical aspects. Osteoporos Int11(Suppl 6):S18–S29.

23. Seibel MJ 2005 Biochemical markers of bone turnover. Part I:biochemistry and variability. Clin Biochem Rev 26:97–122.

24. Sun L, Davies TF, Blair HC, Abe E, Zaidi M 2006 TSH andbone loss. Ann N Y Acad Sci 1068:309–318.

25. Lam J, Ross FP, Teitelbaum SL 2001 RANK ligand stimulatesanabolic bone formation. J Bone Miner Res 16(S1):1053.

26. Galliford TM, Murphy E, Williams AJ, Bassett JH, WilliamsGR 2005 Effects of thyroid status on bone metabolism: aprimary role for thyroid stimulating hormone or thyroidhormone? Minerva Endocrinol 30:237–245.

27. Sampath TK, Simic P, Sendak R, Draca N, Bowe AE, O’BrienS, Schiavi SC, McPherson JM, Vukicevic S 2007 Thyroid-stimulating hormone restores bone volume microarchitec-ture, and strength in aged ovariectomized rats. J Bone MinerRes 22:849–859.

28. Bassett JH, O’Shea PJ, Sriskantharajah S, Rabier B, Boyde A,Howell PG, Weiss RE, Roux JP, Malaval L, Clement-LacroixP, Samarut J, Chassande O, Williams GR 2007 Thyroidhormone excess rather than TSH deficiency induces osteo-porosis in hyperthyroidism. Mol Endocrinol 21:1095–1107.

TSH AND BONE REMODELING 459

29. Pacifici R, Brown C, Puscheck E, Friedrich E, Slatopolsky E,Maggio D, McCracken R, Avioli LV 1991 Effect of surgicalmenopause and estrogen replacement on cytokine releasefrom human blood mononuclear cells. Proc Natl Acad SciU S A 88:5134–5138.

30. Ralston SH, Russel RGG, Gowen M 1990 Estrogen inhibitsrelease of tumor necrosis factor from peripheral bloodmononuclear cells in postmenopausal women. J Bone MinerRes 5:983–988.

Address reprint requests to:Giuseppe Martini

Department of Internal MedicineUniversity of Siena, Policlinico ‘‘S.Maria alle Scotte’’

Viale Bracci53100 Siena

Italy

E-mail: martinig@unisi.it

460 MARTINI ET AL.