ASBMR 29th Annual Meeting

W161

Identification of cEts1 and STAT1 as Potential Primary Targets in the Secondary Gene Regulation of the Human Fibroblast Growth Factor-23 Gene by 1,25-dihydroxyvitamin D3. A. Hsieh*¹, M. Gurevich*¹, D. Mathern*¹, M. J. Kaczmarska*¹, C. A. Haussler², G. K. Whitfield², M. R. Haussler². ¹Biochemistry & Molecular Biophysics, University of Arizona College of Medicine, Tucson, AZ, USA, ²Basic Medical Sciences, University of Arizona College of Medicine, Phoenix, AZ, USA.

The mammalian fibroblast growth factor-23 (FGF23) gene encodes a protein hormone, produced in osteoblasts, that serves as a potent inhibitor of kidney phosphate reabsorption. The hormonal form of vitamin D, 1,25-dihydroxyvitamin D₃ (1,25D), has a reciprocal role to that of FGF23, promoting both kidney reabsorption and intestinal absorption of phosphate. Recent findings have revealed that FGF23 and 1,25D form an endocrine loop, with 1,25D upregulating FGF23 and the resulting FGF23 inhibiting production of 1,25D to close the loop. The stimulatory effect of 1,25D was previously shown in rat osteoblast-like UMR-106 cells to involve a substantial increase in FGF23 mRNA. Transactivation by 1,25D is typically mediated by the vitamin D receptor (VDR), which binds to vitamin D response elements (VDREs) in the vicinity of regulated genes. However, conserved VDREs of the DR3 or ER6 type have not been found near the FGF23 locus; also, previous cycloheximide experiments indicated that 1,25D regulates FGF23 likely via a secondary mechanism. Utilizing a bioinformatics approach to seek potential secondary regulators, we first located segments of sequence conservation within 130 kb upstream and 20 kb downstream of the FGF23 gene using the UC Santa Cruz genomic browser, which compares sequences among 14 vertebrate genomes. Within these conserved sequences, we then sought consensus binding sites for known transcriptional regulators. This search yielded several candidates, notably members of the STAT and GATA families, as well as cEtsl. As an initial screen, regulation of the relevant mRNAs by 1,25D was examined in rat UMR-106 cells. Two nuclear factors, cEtsl and STAT1, were found to be both well expressed in UMR-106 cells and also reproducibly (∼ 2-fold) upregulated by 1,25D. Examination of the cEtsl promoter revealed two conserved VDREs of the DR3 type located at −24 and −26 kb upstream of the rat cEtsl promoter. These two VDREs were shown to bind VDR in a gel mobility shift and are currently being evaluated for their ability to confer 1,25D responsiveness onto a heterologous luciferase reporter gene. We will also evaluate if VDR binding to these sequences actually takes place in vivo via a chromatin immunoprecipitation assay. Elucidation of the molecular details whereby 1,25D regulates FGF23 will significantly enhance our overall understanding of bone mineral homeostasis in mammals.

Disclosures: G.K. Whitfield, None.

W162

Co-stimulation of 1,25-(OH)2D3 and BMP-2 Enhances the Expression of VDR and RANKL mRNA in Osteoblasts. Y. Yoshikawa, A. Kamada, E. Domae*, S. Goda*, I. Tamura*, A. Kawamoto*, J. Okazaki*, Y. Komasa*, T. Ikeo.. Osaka Dental University, Hirakata, Japan.

1alpha, 25-dihydroxyvitamin D3 (1,25-(OH)2D3) appears to indirectly induce osteoclast formation by up-regulating RANKL expression and down-regulating OPG expression in marrow stromal cells. A recent study has shown that BMP-2 enhances the induction of osteoclast formation by 1,25-(OH)2D3., but little is known about the mechanism of the promoting effect of 1,25-(OH)2D3 and BMP-2 co-stimulation on osteoblast-osteoclast interaction. In this study, we investigated the time-dependent expression of RANKL, OPG and vitamin D receptor (VDR) mRNA in osteoblasts co-stimulated with 1,25-(OH)2D3 and BMP-2.

Osteoblastic MC3T3-E1 cells were grown in culture medium with 1,25-(OH)2D3 and BMP-2 for 3, 6, 12 and 24 hours. Total RNA was extracted from cells and first-strand cDNA was synthesized from total RNA using reverse transcriptase. The mRNA expression level was assessed using real-time PCR.

The mRNA expression of VDR in osteoblasts increased significantly 3 hours after co-stimulation with 1,25-(OH)2D3 and BMP-2 compared with that stimulated with 1,25-(OH)2D3 only (p<0.01). RANKL mRNA expression also increased significantly 6 hours after co-stimulation (p<0.01), while OPG mRNA expression showed no significant difference at any time course.

These results suggested that BMP-2 may accelerate the 1,25-(OH)2D3 -induced osteoclast formation by through enhance the expression of VDR and following RANKL in osteoblast.

Disclosures: Y. Yoshikawa, None.

W163

Plasminogen Activator Inhibitor (PAI-1) Deficiency Enhances Fracture Callus Size but Reduces Cartilage Remodeling During Fracture Repair. C. H. Rundle, X. Wang*, R. M. Porte*, J. E. Wergedal, S. Mohan, K. H. W. Lau.. Research (151), J.L. Pettis VAMC, Loma Linda, CA, USA.

The plasmin system, which functions in extracellular matrix remodeling, plays a critical role in tissue healing, including bone repair. Plasminogen activators (PAs), key activators of the plasmin system, are regulated by plasminogen activator inhibitors (PAIs). Consistent with a regulatory role of PAI-1 in bone, we have found that femur size parameters of adult male PAI-1 knockout (KO) mice [C57BL/6 (B6) background] were reduced relative to adult male B6 mice (periosteal circumference: 14% less, 4.51 ± 0.13 vs 5.23 ± 0.14 mm, p<0.0001; endosteal circumference: 19% less, 3.10 ± 0.12 vs 3.82 ± 0.15 mm, p<0.0001; femur length: 4% less, 15.3 ± 0.31 vs 16.01 ± 0.47 mm, p<0.02). Because we have also found PAI-1 gene expression to be reduced during fracture healing, we evaluated fracture healing in PAI-1 KO mice to determine its functional role in bone repair. Adult male PAI-1 KO mice and B6 mice (12 weeks, n = 6 per group) underwent femoral fracture, and fracture healing was followed by pQCT and histomorphometry at 7, 10, 14, 21 and 28 days post-fracture. When compared with B6 mice, PAI-1 KO fractures displayed increased woven bone area at 10 days (1.66 ± 0.11 vs 1.30 ± 0.36 mm², p<0.05) and 14 days healing (4.77 ± 1.34 vs 2.48 ± 0.86 mm², p<0.002) and increased bone mineral content (BMC) at 10 days (0.43 ± 0.03 vs 0.28 ± 0.13 mg, p<0.03), 14 days (1.27 ± 0.37 vs 0.54 ± 0.29 mg, p<0.001) and 28 days healing (0.57 ± 0.18 vs 0.35 ± 0.08 mg, p<0.04). These results are consistent with increased callus size and bone. Histomorphometry (n = 3 to 5 mice per group) at 7, 14 and 21 days healing showed that the relative cartilage area:total callus area at 14 days healing was greater in PAI-1 KO mice than in B6 mice (38 ± 8% vs 11 ± 4%, respectively, p<0.0003). The augmented callus cartilage and bone were even greater in PAI-1 KO mice than in B6 mice when normalized for the smaller femur size in PAI-1 KO mice. At 28 days, the callus bone remained elevated in PAI-1 KO mice, but was remodeled to cortical bone in B6 mice, suggesting that callus remodeling in PAI-1 KO mice was less effective than in B6 mice. Histology revealed that PAI-1 deficiency not only failed to accelerate bony union of the fracture, but also yielded an abnormal pattern of callus remodeling by 14 days healing in which the surface of the fracture callus cartilage ossified prior to its interior. Conclusions: 1) PAI-1 is an important regulator of bone, since PAI-1 deficiency reduced bone size and marrow space and reduced bone length in KO mice, 2) PAI-1 deficiency may impede fracture cartilage remodeling, resulting in an enlarged callus, and 3) because PAI-1 deficiency yielded an abnormal pattern of bony bridging, the PAI/PA system may have a regulatory role in bony union of the fracture.

Disclosures: C.H. Rundle, None.

W164

Modeling Bone Morphogenetic Protein and Bisphosphonate Combination Therapy in Wild Type and Nf1 Haploinsufficient Mice. A. Schindeler*, M. Ramachandran*, C. Godfrey*, A. Morse*, M. M. McDonald*, K. Mikulec*, D. G. Little*., Orthopaedic Research & Biotechnology, The Children's Hospital at Westmead, Westmead, Australia.

Recombinant bone morphogenetic proteins (BMPs) show promise in treating the orthopaedic complications associated with neurofibromatosis type 1 (NF1), such as congenital pseudarthrosis of the tibia. Minimal scientific data is available regarding the effects of BMPs in the context of NF1. We hypothesized that as abnormalities in both bone formation and resorption have been documented in Nf1-deficient mice, co-treatment with the bisphosphonate zoledronic acid (ZA) could be used to augment the effects of BMP-induced bone formation.

First, primary osteoblasts isolated from wild type (Nf1^+/+) and Nf1^+/−-deficient mice were cultured in the presence and absence of BMP-2. While Nf1^+/+ cells exhibited significantly less alkaline phosphatase (ALP) expression than Nf1^+/+ cells (P<0.01), addition of 100ng/ml BMP-2 for 1 week increased ALP by 2-fold in cells of both genotypes (P<0.01). Analagous results were obtained for matrix mineralization at 3 weeks. To model this response in vivo, 20μg BMP-2 was implanted intramuscularly into the quadriceps of 12 week old mice to induce heterotopic bone. When examined radiographically, less heterotopic bone was present in Nf1^+/− mice than in Nf1^+/− controls at both 3 weeks (P<0.03) and 6 weeks (P<0.01). Mice of an Nf1^+/− genotype also showed an increase in osteoclast number relative to bone surface (Oc.N/BS) suggesting a greater potential for the resorption of BMP-2 induced bone.

To test the effect of an anti-resorptive agent, BMP-2 implanted mice were co-treated twice weekly for 3 weeks with 0.02mg/kg ZA or with saline. ZA treatment led to a synergistic increase in the amount of heterotopic bone in both Nf1^+/− and Nf1^+/+ mice compared with saline controls, as measured by dual-energy X-ray absorptiometry (P<0.03) and histomorphometry (BV/TV, P<0.01). However, significantly less net bone was still observed in Nf1^+/+ mice (P<0.03) indicating that despite suppressing bone catabolism with ZA, bone anabolism remained impaired.

Thus, the anabolic deficiency noted in Nf1^+/+ mice is amenable to stimulation by BMP-2, but mineralized tissue formation remains below that of Nf1^+/+ controls in vitro and in vivo. Bisphosphonate combination therapy is superior to BMP therapy alone in terms of net bone production in vivo in both wild type and Nf1^+/+-deficient mice.

Disclosures: A. Schindeler, None.

This study received funding from: The Australian Orthopaedic Association, The Neurofibromatosis Association of Australia.

W165

Identification of Novel Dentin Matrix Protein-1 (DMP1) Mutations in Two Unrelated Kindreds with Autosomal Recessive Hypophosphatemia. S. Turan¹, M. Bastepe¹, A. Bereket*², T. Akcay*², T. Güran*², O. Mäkitie*³, H. Jüppner¹. ¹Endocrine Unit, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA, ²Pediatric Endocrinology, Marmara University, Istanbul, Turkey, ³Hospital for Children and Adolescents, University of Helsinki, Helsinki, Finland.

DMP1, a non-collagenous bone matrix protein, is thought to play an important role in bone mineralization. Homozygous DMP1 mutations have recently been identified as a cause of an autosomal recessive form of hypophosphatemia (ARHP). Here, we report two new unrelated ARHP kindreds in whom the affected individuals carry novel homozygous DMP1 mutations. The index case in the first kindred was 3 years old when he presented with short stature and bowing of the legs that had become apparent during late infancy after he started to walk. Laboratory investigations revealed hypophosphatemia (2.8 mg/dl) due to renal phosphate-wasting, elevated serum alkaline phosphatase activity (964 U/L) and normal serum PTH levels (47.6 pg/ml). Similar clinical and biochemical findings were observed for his 16 month-old brother and his 12.5 year-old female cousin; the parents of all three affected children are first-degree cousins. In the second family, the index case and his sister are now 78 and 66 years old respectively. Both patients presented in early childhood with bone pain and bowing of their legs, and later in life, developed severe joint pain and contractures, calcifications of paraspinal ligaments leading to complete immobilization of the cervical, thoracic and lumbar spine, and significant cranial hyperostosis. Laboratory studies revealed hypophosphatemia due to renal phosphate-wasting as well as increased markers of bone turnover. Nucleotide sequence analysis of the PCR-amplified exons encoding DMP1 and the flanking intronic regions lead, in the first kindred, to the identification of a novel homozygous frame-shift mutation (485Tdel) in exon 6 resulting in a premature stop codon. The parents and available unaffected siblings were heterozygous for 485Tdel. In the second kindred, a novel homozygous missense mutation was identified that changes a conserved serine located in exon 6 to cysteine (S69C). The identified genetic mutations underscore the importance of DMP1 in the pathogenesis of ARHP and for the normal regulation of phosphate homeostasis.

Disclosures: S. Turan, None.

This study received funding from: National Institutes of Health (R21 DK075856-01).

W166

Bone Formation Rate Falls After Puberty in Brtl OI Mouse Due to Cellular Uncoupling. T. E. Uveges*¹, W. A. Cabral*¹, C. Bergwitz*², F. Legard*³, P. Collin-Osdoby*⁴, A. Forlino*⁵, P. Osdoby*⁴, G. A. Gronowitz*³, J. C. Marini¹. ¹BEMB, NIH/NICHD, Bethesda, MD, USA, ²Endo Unit, Mass Gen/Harvard Med, Boston, MA, USA, ³U Conn Health Ctr, Farmington, CT, USA, ⁴Dept Biology/Bone&Min Metab, Washington U, St. Louis, MO, USA, ⁵Dept Biochem, U of Pavia, Pavia, Italy.

The Brtl mouse is a knock-in model for moderately severe (type IV) osteogenesis imperfecta (OI); it carries a mutation causing a classical glycine substitution on one collal allele (Gly349Cys) and models the molecular, biochemical and phenotypic changes (small size, reduced BMD) of OI. We have previously shown that 2-month Brtl femurs have reduced cortical thickness and cross-sectional area and fracture at a lower maximum load. Geometric parameters remain reduced while load to fracture normalizes after puberty, implying that Brtl pubertal adaptation increases material strength and elastic modulus.

We have now examined Brtl long bone histomorphometry, cellular function and bone formation and resorption at 2 and 6 months of age to further understand the pathophysiology of OI. Both cortical and trabecular bone are reduced in Brtl before and after puberty; Brtl BV/TV is reduced 40–45%. An uncoupling of cellular number develops after puberty, favoring osteoclast number and function. Brtl ObS/BS is comparable to that of wt littermates at both ages. ObS decreases about 40% in both genotypes with age, although Brtl marrow can replenish osteoblasts as efficiently as wt marrow, as demonstrated by normal numbers of CFU at both ages. In contrast, OcS/BS is increased in Brtl compared to wt at both ages (35–45%), as are the number of TRAP positive cells (50–35%). Interestingly, OcS falls only modestly (∼ 1/5) after puberty, Noc is stable and TRAP positive cells decline substantially (∼ 4/5). Brtl MAR and BFR are comparable to wt at 2 months of age, declining to half of wt values only at 6 months. These data suggest that both (1) uncoupling of cellular numbers, with a greater decline in ObS than OcS at 6 months, and (2) a decline in production of bone matrix by Brtl osteoblasts occur as Brtl mice age. Normal levels of urinary DPD crosslinks were detected in Brtl, as in many OI patients with structural defects of collagen; the lack of elevation of bone resorption in spite of increased osteoclasts may reflect the matrix insufficiency of Brtl or resistance of abnormal matrix to resorption. Finally, although Brtl osteoblasts in culture deposit normal amounts of mineral, as detected by alizarin red staining, mineralized Brtl bone has reduced levels and organization of birefringence under polarized light, suggesting that mineral deposited on the abnormal Brtl matrix is improperly aligned. Changes in the composition of bone matrix or mineral may underlie the improved mechanical properties of post-pubertal Brtl bone.

Disclosures: T.E. Uveges, None.

W167

Double KO of the Huntingtin Interacting Protein-1 (HIP1) Family Members, HIP1 and HIP1r, Lead to an Osteopenic Phenotype with Severe Spinal Defects. E. I. Waldorff*¹, T. S. Hyun*², K. I. Oravecz-Wilson*², T. S. Ross*², S. A. Goldstein¹. ¹Orthopaedic Surgery, University of Michigan, Ann Arbor, MI, USA, ²Internal Medicine, University of Michigan, Ann Arbor, MI, USA.

Endocytic protein HIP1 has been shown to be over-expressed in a variety of cancers and both HIP1 and HIP1-related (HIP1r) prolong the half-life of multiple growth factor receptors. Previously generated double HIP1/HIP1r KO (DKO) mice surprisingly demonstrated a dramatic skeletal phenotype distinguished by dwarfism and severe kypholordosis. Single HIP1 KO mice have less severe kypholordosis while HIP1r KO mice have no spinal defects.

To investigate the effect of the HIP1 family on skeletal development; closed, internally-stabilized transverse fractures were created in left tibiae of 3 WT, 7 Het (HIP1r -/-;HIP1 +/ -), and 3 DKO female 3.5 month old mice. Fractures were examined 4 wks post-fracture using microCT and histology. The spine and right tibiae were examined using microCT, while marrow from both femora was used for protein analysis using Western blots.

Compared to Het and WT, DKO mice have smaller spine angle; larger vertebral bodies, smaller vertebral foramen and smaller posterior spinous processes; a trend for decreased TMD of the vertebral body; increased vasculature in the thoracic vertebrae; shorter tibiae, lower tibial cortical and trabecular TMD; smaller cortical thickness; a trend toward decreased BVF of tibial trabecular bone and thinner trabeculae, while trabecular number were similar.

RANK expression of the femoral marrow was similar between groups, but the transcription factor, NFATc1, which acts downstream of RANK, was more highly expressed for DKO mice. Previous studies have shown no change in osteoclast numbers or differentiation, and normal TRAP5b serum levels for DKO mice.

DKO fracture callus was smaller than the other groups, with a trend toward higher BMD within the callus. Histological examinations showed that DKO mice had higher cartilage to callus ratio, and higher ratio of TRAP positive surface to total surface within the callus. Red blood cell numbers in the callus were increased for DKO mice, indicating a possible increase in vasculature.

The interpretation of the data suggests several possible underlying mechanisms that might explain the skeletal phenotype of the DKO mouse. Deletion of the HIP1 family may affect primary bone formation in the long bones through the influence on cartilage matrix formation, or direct bone matrix formation. Results from the fracture study in combination with the vertebral defects, suggest that progenitor cell availability, tissue maintenance and remodeling might also be compromised in this phenotype. Further longitudinal studies will examine the fracture repair over time.

Disclosures: E.I. Waldorff, None.

W168

Aromatase Deficiency in an Adult Chinese Male Caused by Heterozygote for Two New Point Mutations in the CYP19 Gene. O. Wang*, X. Xing*, X. Meng*, M. Nie*, Z. Chen*, M. Li*, W. Xia*, Y. Jiang*., Endocrinology, Peking Union Medical College Hospital, Beijing, China.

The aromatase enzyme encoded by CYP19 gene catalyzes the conversion of androgens to estrogens. Deficiency of aromatase causes estrogen deficiency which may affect bone growth and metabolism. The purpose of this study was to report the first Chinese male with aromatase deficiency and the genetic mutations in his CYP19 gene.

The 24-year-old male was admitted in Peking Union Medical College Hospital for continuing linear growth for 24 years and malformation in his legs for 2 years. He had no growth spurt. Sexual function was normal. When he was 22-year-old he got genua valga which deteriorated gradually. His parents and brother were phenotypically normal. His mother was 165cm, father 172cm and brother 175cm in height. The patient was 182.5cm and the ratio of the upper segment to the lower segment was 0.92. Physical examinations showed genua valga, acanthosis nigricans and enlarged liver. The secondary male characteristics and testicular size were normal. The biochemical markers and BMD were measured. Genomic DNA was extracted from WBC of peripheral blood from the patient and his family numbers. Ten exons of CYP19 gene were amplified by PCR and sequenced directly.

His serum testosterone levels were 14.6 to 20.5nmol/l while estrogen level was undetectable. FSH and LH levels were elevated. Bone age was 16 years old. Epiphyses were unfused in lower extremities, phalange and wings of iliac bones. BMD was reduced at lumbar spine (Z score = −2.9) and femoral neck (Z score = −0.9). Semen analysis was normal. OGTT suggested type 2 diabetes mellitus. Serum cholesterol level was elevated with increased ALT levels and severe liver steatohepatitis identified by ultrasound. Genetic analysis showed two heterozygous point mutations in CYP19 gene, one was Tyr81Cys in exon 3 (from his mother), the other was Leu451Pro in exon 10 (from his father). This new case of aromatase deficiency confirmed previous studies that estrogen plays an important role on bone growth and gain of peak bone mass in males. Estrogen may also affect glucose and lipid metabolism in males. In this patient, two new point mutations in CYP 19 gene are likely to be the molecular basis of his estrogen deficiency.

Disclosures: O. Wang, None.

W169

Osteoprotegerin Deficiency (Juvenile Paget's Disease): Responses to Oral and IV Bisphosphonates in 3 Children. D. Wenkert¹, M. Natter*², M. Otsuka*³, N. Fish*², J. Lopez-Benitez*², M. Markowitz*⁴, W. McAlister*⁵, S. Mumm⁵, M. Whyte⁵. ¹Shriners Hospt Children, St Louis, MO, USA, ²Tufts-New Eng Med Ctr, Boston, MA, USA, ³N. Navajo Med Cntr, Shiprock, NM, USA, ⁴Montefiore Hospt, Bronx, NY, USA, ⁵Wash U Sch Med, St Louis, MO, USA.

Osteoprotegerin (OPG) deficiency, the principal variant of juvenile Paget's disease, is a rare, autosomal recessive osteopathy featuring markedly accelerated bone turnover. Here, we assess responses of 3 affected, unrelated children to bisphosphonate (BP) therapy (Table).

Pt #1: A 4 yr old boy of Cypriot descent suffered leg swelling, periosteal elevation, and ESR = 74 from “Caffey disease” diagnosed at age 1 mo. Alkaline phosphatase (ALP) reached 2470 U/l. Hearing (normal at birth) was lost by 21 mo (cochleas absent on CT). Three fxs and warm, bony expansions deformed all limbs. Growth, except head size (OFC), was stunted. On referral, he had weakness, torticollis, chest and limb deformity, could not sit, and cried when moving. X-rays showed markedly expanded bones, “bowtie” vertebrae, thickened diploic space, and marked coxa vara.

One wk after 1/3 mg/kg pamidronate (PAM), he sat; 3 weeks after the first 3-dose cycle (total ∼2 mg/kg), he crawled. After 3 cycles, he stood. Stamina, appetite, mood, torticollis, and mobility remarkably improved.

Pt #2: A 7 yr old girl of Puerto Rican descent seemed well until an atraumatic femoral fx at age 2. X-rays showed limb deformities from “Engelmann disease” (ALP 1963 U/l). By referral, she had 4 fxs (25% ht, 20% wt, 5% arm span, <5% sitting ht, and > 95% OFC), mild deafness, and arm pain; leg pain limited walking. Femurs and tibias were bowed. X-rays showed periosteal elevation, osteopenia, and expanded bones. At age 10, after 1.5 yrs of alendronate, knee pain was minimal.

Pt#3: An 11 yr old, dwarfed, deaf, Navajo boy (NEJM 2002: 347:175) had macrocephaly, and chest and limb deformity. He became poorly compliant for calcitonin, and declined risedronate. At age 7, PAM was begun (ALP max 2716) with symptom and x-ray improvement.

Pt 1 and 2 had different, homozygous, loss-of-function defects in exon 2 of TNFRSF11B encoding OPG Pt #1 had a frame-shift, single base deletion (c.278delT), likely forming no functional protein, consistent with severe disease. Pt #2 had a missense mutation (c.T349C, p.Phe 117Leu) causing a less severe phenotype. Pt#3 is deleted for OPG

Children with OPG deficiency show varying responsivity to BPs, reflecting the underlying OPG mutation. 

Table  . Biochemical and DXA response to Biophosphonates in 3 children with OPG deficiency

Disclosures: D. Wenkert, None.

This study received funding from: Shriners Hospitals for Children, The Clark and Mildred Cox Inherited Metabolic Bone Disease Research Fund, The Barnes-Jewish Hospital Foundaiton, and The Hypophostasia Research Fund.

W170

Three Novel Mutations of the Phex in Three Chinese Families with X-linked Hypophosphatemic Rickets. W. Xia*¹, X. Meng¹, Y. Jiang*¹, X. Xing*¹, M. Li*¹, O. Wang*¹, Y. Pei*², L. Yu*¹, L. Pang*¹, Y. Sun*¹, Y. Hu*¹, X. Zhou*¹. ¹Endocrinology, Peking Union Medical Colleg Hospotal, Beijing, China, ²Endocrinology, Second Artillery Forces General Hospital, Beijing, China.

X-linked hypophosphatemia (XLH), the most prevalent form of inherited rickets in humans, is a dominant disorder of phosphate homeostasis characterized by growth retardation, rachitic and osteomalacic bone disease, hypophosphatemia, and renal phosphate wasting. The gene responsible for XLH was identified by positional cloning and designated PHEX (formerly PEX) to depict a phosphate regulating gene with homology to endopeptidases on the × chromosome. Recently, extensive mutation analysis of the PHEX gene has revealed a wide variety of gene defects in XLH. The ethnic distribution of the mutations is very broad. Only a few mutations in Chinese were reported previously. To analyze the molecular basis of three unrelated Chinese families with XLH, we determined the nucleotide sequence of the PHEX gene and fibroblast growth factor 23 (FGF-23) gene of affected members. The serum FGF-23 concentrations of these patients with XLH were also measured. Three different types of novel mutations were observed in three families respectively from our study: one deletion mutation c.264delG causing p.W88 X; one missense mutation c.1673C>G causing p.P558A; one non-sense mutation c.1809G>A causing p.W603 X. Serum concentration of FGF-23 in XLH patients of these three families was significantly higher than normal range. These results suggest that PHEX gene mutations were responsible for XLH in these Chinese patients. And these mutations may contribute to higher serum FGF-23 level.

Disclosures: W. Xia, None.

W171

Identification of Sex-specific QTL for Femoral Neck Density and Strength in Inbred COP and DA Rats. I. Alam¹, O. Sun¹, L. Liu*², D. L. Koller², L. G. Carr*³, M. J. Econs³, T. Foroud², C. H. Turner¹. ¹Biomedical Engineering, Indiana University Purdue University Indianapolis, Indianapolis, IN, USA, ²Medical and Molecular Genetics, Indiana University Purdue University Indianapolis, Indianapolis, IN, USA, ³Medicine, Indiana University Purdue University Indianapolis, Indianapolis, IN, USA.

Hip fracture is one of the most common osteoporotic fractures with significant morbidity and mortality. Several studies in humans identified potential chromosomal regions linked to hip size and bone mass. Animal models, particularly the inbred rat, serve as a potential complementary model for studying the genetic influences on hip fragility. Previously we demonstrated that, despite similar body size, Copenhagen 2331 (COP) rats had greater femoral neck biomechanical properties compared with Dark Agouti (DA) rats. To identify genes regulating fracture risk at the femoral neck, we mapped quantitative trait loci (QTL) for femoral neck density, structure and strength phenotypes using 828 (405 males and 423 females) F2 progeny derived from the inbred COP and DA rats. Skeletal phenotypes were measured by peripheral quantitative computed tomography (pQCT) and biomechanical testing. Femoral neck phenotypes included volumetric bone mineral density (vBMD), cross-sectional area, polar moment of inertia (Ip), neck width, ultimate force, and energy to break. We performed a whole-genome screen using 93 microsatellite markers with an average intermarker distance of 20 cM. Genetic marker maps were generated using MAPMAKER/EXP from the COP x DA F2 data and compared with published Rat Genome Database (RGD) maps. These recombination based maps were then employed for genome-wide linkage analyses to detect sex-specific and sex-independent QTLs. Permutation testing was employed to obtain the thresholds for genome-wide significance (p<0.01). Strong evidence of linkage for femoral neck vBMD QTLs was observed on chromosomes (Chrs) 1, 2, 10, 17 and 18 when both males and females F2 were included in the analyses. In addition, QTLs affecting femoral neck structure were found on Chrs 7, 10 and 17 and QTLs for biomechanical properties were detected on Chrs 1, 4, and 15. Sex-specific QTL analyses revealed several QTLs that have male-specific effects: total vBMD on Chrs 2 and 18, cortical vBMD on Chr 2, trabecular vBMD on Chr 17 and biomechanics on Chr 15. Female-specific QTLs were found on Chr 10 for total and cortical vBMD, and on Chr 1 for trabecular vBMD and biomechanics. Several QTL regions identified in this study are homologous to human chromosomal regions previously linked to QTLs contributing to femoral neck and related phenotypes. In summary, we detected evidence that sex-specific QTL contribute to hip fragility in the rat.

Disclosures: I. Alam, None.

W172

Heritability of the Principal Components of a Statistical Shape Model Describing Baboon Midshaft Femur Geometry. T. L. Bredbenner¹, L. M. Havill², M. C. Mahaney², D. P. Nicolella¹. ¹Materials Engineering, Southwest Research Institute, San Antonio, TX, USA, ²Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX, USA.

Cross-sectional geometry is an important contributor to bone strength. Studies in humans and inbred rodents suggest that bone shape is under partial genetic control. It is less clear, however, that genes contribute substantially to normal population-level variation in bone geometry. Using the pedigreed baboon, an established nonhuman primate model for the genetics of bone strength-related traits, we previously established that geometric properties of the midshaft femur are significantly heritable. In the present study, we developed a statistical shape model to efficiently describe the complex geometry of a set of midshaft transverse sections from 110 pedigreed baboons (30 male, 80 female; 5–33 yrs) and investigated the heritability of principal components of the statistical shape model. Midshaft sections were digitized, the images were segmented, and cubic splines were fit to inner and outer cortical boundaries. Parameter vectors were defined for each section using spline point coordinates and a statistical shape model (SSM) representation of all cross-sections was developed using the minimum description length (MDL) method. A Principal Component Analysis of the statistical shape model showed that 90% of the variability in bone cross-section geometry can be described using 6 independent components (Table 1). Shape component scores were determined to reconstruct individual cross-sectional boundaries from the statistical shape model. Maximum likelihood-based variance decomposition methods were used to estimate the mean effects of age and sex covariates and the additive effects of genes on the variability of the first 6 shape factors (Table 1). Covariate effects accounted for ∼10–55% of the variability in shape factor values (p_age, p_sex, and/or p_age*sex < 0.005). The additive effects of genes significantly influence (p < 0.1) the normal variation in the 2 dominant shape factors that described 77% of the variability in baboon midshaft femur geometry. The statistical shape model retains significant dependence on the additive effects of genes and potentially can be used to efficiently quantify the contribution of bone geometry to an individual's risk of fracture. 

Table  . Results of Principal Component Analysis and Variance Decomposition

Disclosures: T.L. Bredbenner, None.

This study received funding from: Southwest Research Institute Advisory Committee for Research.

W173

Single Chromosome Substitutions Alter Functional Interactions Among Femoral Morphological Traits and Mineral Density in Inbred Mice. H. C. Courtland¹, B. Hu*¹, A. Hill*², J. B. Singer*³, E. S. Lander*³, J. H. Nadeau*², K. J. Jepsen¹. ¹Orthopaedics, Mount Sinai Scool of Medicine, New York, NY, USA. ²Case Western Reserve University, Cleveland, OH, USA. ³The Broad Institute of MIT and Harvard University, Cambridge, MA, USA.

Prior work established a hierarchical relationship between heritable whole bone traits (morphology, ash content) and mechanical properties that define fracture susceptibility in inbred mice. Functional interactions among these traits were then determined using Path Analysis on a panel of adult AXB/BXA Recombinant Inbred (RI) strains (Fig. 1)[1]. To test if genetic manipulation can perturb these functional interactions, we examined whole chromosome substitutions of 21 B6-Chr^A strains (C57BL/6J background with one chromosome replaced by the corresponding A/J chromosome), many of which demonstrate a significant alteration of whole bone traits. 16 week B6-Chr^A femoral length (Le), total area (TtAr), cortical thickness (CtTh) and tissue mineral density (TMDn) values were measured. The expected TMDn value for each B6-Chr^A strain was calculated using the AXB/BXA RI structural equation for TMDn. Significant differences between the expected TMDn and actual TMDn were calculated using genome-wide corrected t-tests (p≤0.003). Results showed that four female B6-Chr^A (2, 6, 16, and X) and male B6-Chr^A (2, 3, 4, and 18) strains had observed TMDn values significantly different than expected. For both female and male B6-Chr² mice and male B6-Chr² mice, TMDn was significantly lower than expected based on values for Le, CtTh, and TtAr which were similar to B6. Female B6-Chr^6,16,x mice had Le, CtTh, and TtAr values significantly larger than B6 and yet their observed TMDn values were not different than B6. The observed decrease in TMDn in male B6-Chr⁴ mice was not expected based on an increase in CtTh and in male B6-Chr¹⁸ mice a large increase in TMDn was not expected with a small increase in femoral length. Thus, eight chromsomes alter the functional relationship between femoral morphology and TMDn in inbred mice, suggesting that one or more genes present have a role in maintaining the biological control mechanisms that establish mechanically functional combinations of femoral TMDn, Le, CtTh, and TtAr. Future studies are needed to determine if these chromosomes disrupt functional interactions during growth.

[1] Jepsen et al. Mamm. Gen. 2007 In Press 

Figure Figure 1.. Path diagrams detailing functional interactions among femoral bone traits in female and male RI mice. Straight arrows and path coefficients indicate functionally related bone traits with directionality (causality). Structural equations are listed below each Path diagram.

Disclosures: H.C. Courtland, None.

W174

Genome-wide Scan for QTL Influencing Bone Mineral Density: The Southwest Ohio Family Study. S. A. Czerwinski*¹, M. Lee¹, A. C. Choh*¹, E. W. Demerath*¹, B. Towne*¹, R. J. Sherwood*¹, D. L. Duren*¹, J. Blangero*², S. A. Cole*², R. M. Siervogel¹. ¹Community Health, Wright State University Boonshoft School of Medicine, Kettering, OH, USA. ²Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX, USA.

In order to identify QTL influencing susceptibility to osteoporosis, we performed genome-wide quantitative trait linkage analysis on bone mineral density (BMD) phenotypes collected in a sample of individuals from five large extended families participating in the Southwest Ohio Family Study. A total of 581 (265 males, 316 females) individuals ranging in age from 8-86 years had at least one BMD phenotype and genetic marker data. These individuals were genotyped for 395 autosomal markers spaced approximately every 10 cM along the genome. Our analyses focused on three BMD phenotypes- total body, femoral neck and lumbar spine collected using the Hologic QDR 4500 bone densitometer (Hologic, Inc., Waltham, MA). Using a variance-components based maximum likelihood method (SOLAR) for pedigree data, we calculated initial heritability estimates (h²) and identified quantitative trait loci (QTL) influencing variation in the three BMD phenotypes. In this sample, each of the BMD traits measured was highly heritable ranging from h²=0.31 to 0.56 (p<0.000001) after adjusting for the effects of age, sex, age², age-by-sex, age²-by-sex, ethnicity, and height. Preliminary results of the quantitative trait linkage screen revealed significant evidence for linkage on chromosome 3 (82cM) near marker D3S1300 with a maximum LOD of 3.6. Evidence of suggestive linkage was also found on chromosome 6 (165 cM) near marker D6S441 with a maximum LOD of 2.8, and on chromosome 2 (51 cM) near marker D2S165 with a maximum LOD of 2.0. These results provide preliminary evidence of replication of linkage signals between our study and other previous studies.

Disclosures: S.A. Czerwinski, None.

W175

Positional Cloning Identified ESR2, ESRRB, BMP4 and LTBP2 as the Candidate Genes for the Quantitative Trait Locus in Chromosome 14 for BMD Variation. Z. J. Dai¹, C. L. Cheung¹, P. C. Sham*², V. Chan*¹, K. P. K. Luk*³, A. Paterson*⁴, A. W. C. Kung¹. ¹Department of Medicine, The University of Hong Kong, Hong Kong, China. ²Genome Research Centre, The University of Hong Kong, Hong Kong, China. ³Department of Orthopaedics and Traumatology, The University of Hong Kong, Hong Kong, China. ⁴Hospital for Sick Children, University of Toronto, Toronto, ON, Canada.

Chromosome 14 has previously been linked to BMD variation in several genome-wide linkage studies. To replicate and fine-map the quantitative trait locus (QTL) influencing BMD in southern Chinese, a high-density linkage and association screen of chromosome 14q13-32 was performed using high-density microsatellite markers in 306 extended southern Chinese families having a proband with BMD Z score <−1.28 (equivalent to lowest 10^th percentile of the population) at either hip or spine. Candidate genes were identified within the QTL with gene prioritization using the Endeavour programme. Genotyping of the candidate genes was performed with Tag SNPs in another cohort of 600 pairs of unrelated high (Z score > +1) and low BMD (Z score <−1.28) subjects. Maximum LOD score of 1.89 was detected at 69.2cM (14q23) for total hip BMD. LOD scores greater than 1.2 were detected at 14q21 for femoral neck BMD, 14q12-13 and 14q23 for trochanter BMD. Five candidate genes were identified within these regions, namely estrogen receptor β (ESR2), bone morphometric protein 4 (BMP4), estrogen-related receptor β (ESRRB), latent transforming growth factor beta binding protein 2 (LTBP2), and transforming growth factor beta 3 (TGFB3). Significant association was detected with the following SNPs: T1213C, A110732G of ESR2, rs762642 and rs2071047 of BMP4, rs2980896, rs4903413 and rs12436385 of ERR2, rs2359141, rs3825709, rs2286411, rs2043948 and rs7569 of LTBP2, rs3917201 of TGFB3. Epistasis revealed significant gene-gene interaction between these genes which may be the underlying mechanism explaining the linkage peak. Our results identified multiple genes within the QTL in chromosome 14 that contribute to BMD variation in the general population. These results help in understanding the pathophysiology of low bone mass and osteoporosis.

Disclosures: Z.J. Dai, None.

This study received funding from: Grants from Hong Kong Research Grant Council, Osteoporosis Research Fund, Matching Grant, University Research Committee/Committee of Research and Conference Grants of the University of Hong Kong.

W176

Heritability of Lumbar Trabecular Bone Mechanical Properties in Baboons. L. M. Havill¹, M. R. Allen², T. L. Bredbenner³, D. P. Nicolella³, D. B. Burr², C. H. Turner⁴, M. C. Mahaney¹. ¹Southwest Foundation for Biomedical Research, San Antonio, TX, USA. ²Indiana University School of Medicine, Indianapolis, IN, USA. ³Southwest Research Institute, San Antonio, TX, USA. ⁴Indiana University Purdue University, Indianapolis, IN, USA.

Mechanical properties of bone provide a direct and inclusive measure of bone's fracture resistance. Genetic effects on these properties have been demonstrated in rodents but not in primates, and the degree to which these genes contribute to population level normal variation in mechanical properties is unknown. In this study we 1) characterize normal variation, including age and sex effects, on vertebral trabecular bone mechanical properties in a non-human primate model, and 2) detect and quantify the proportion of variation in these properties that is due to the additive effects of genes (heritability (h²)). Cranio-caudally oriented trabecular bone cores (6 mm × 8 mm) were obtained from the L3 vertebral body of 156 baboons (110 females, 46 males; 6-32 years) from a single extended pedigree. Cross-sectional area and bone volume/tissue volume for each core were obtained via microCT (uCT-20; SCANCO USA, Inc.). Each core was then tested to failure in monotonic compression (0.5 mm/min) using a servohydraulic testing machine (Model 858 Mini Bionix II, MTS Corp.). Ultimate stress, modulus, and toughness were determined from load vs displacement data (10Hz). Age, sex, and additive genetic effects were assessed using maximum likelihood-based variance components methods. After accounting for relatively small but significant age and sex effects, a substantial amount of the residual variation in ultimate stress (0.58 ± 0.20 (p=0.0005)), modulus (0.29 ± 0.24 (p=0.0855)) and toughness (0.64 ± 0.22 (p=0.0003)) is attributable to genes. Subsequent analyses in which bone volume (h²=0.55 ± 0.20 (p=0.0009)) is accounted for yield nonsignificant heritability estimates for ultimate stress (0.15 ± 0.14 (p=0.1042)) and modulus (0.02 ± 0.13 (p=0.4413)), suggesting that much of the genetic effect detected in the first analyses acts through bone volume. Interestingly, the results for toughness differ. When bone volume is accounted for, a significant genetic effect on toughness (0.29 ± 0.23 (p=0.0389)) remains. This indicates that variation in toughness is due, in part, to genes that affect bone volume, but also to genetic effects that are independent of bone volume. These results clearly demonstrate that mechanical properties of vertebral trabecular bone are strongly heritable in the baboon and that these genetic effects are largely, but not entirely, due to genetic effects on bone volume.

Disclosures: L.M. Havill, None.

W177

A Genome-Wide Scan of Metacarpal Bone Geometric Strength Indices in the Framingham Study. N. Shimabuku¹, S. Menn², S. Demissie*³, L. A. Cupples*³, D. P. Kiel², D. Karasik². ¹Creighton U Med Sch, Omaha, NE, USA. ²IFAR, Hebrew SeniorLife, Boston, MA, USA. ³Biostatistics, BU Sch Public Health, Boston, MA, USA.

Bone geometry, which has a substantial genetic component, contributes to a bone's ability to resist fracture. Metacarpal bones have the same characteristic geometry of other long bones in the body and have been routinely measured in previous research studies to predict fractures and quantify bone loss. Thus, to identify chromosomal regions governing bone geometry and strength, we performed linkage analysis of metacarpal bone geometry indices.

In this study, a genome-wide scan (with a set of 615 markers with an average spacing of 5.7 cM) was performed on 1,702 individuals from 330 extended families of the Framingham study. Metacarpal strength indices, namely Metacarpal Cortical Thickness (MCT), Cortical Index (MCI), and Section Modulus (MZ), were obtained at the midshaft, from the digitized x-rays of 1,412 men and women (mean age, 56.6 yrs parents (n=638), 47.3 yrs offspring (n=774)). Data for metacarpals 2, 3 and 4 were averaged. All indices were adjusted for age and age² in each sex. Two-point and multipoint linkage analysis was performed using SOLAR, to determine regions of excess chromosomal sharing.

Heritability (h²) was significant for all indices, ranging from 0.53 and 0.62 (Table). Both two-point and multipoint linkage identified several chromosomal regions with the maximum log10 of the odds ratio (LOD) scores > 2.0. Thus, MCI and MCT were linked to chr. 7 (149 cM) with multipoint LOD>2.0. MZ was linked with chr. 9 (49 cM) (multipoint LOD=2.4). Highest LOD scores (both two-and multipoint) were obtained at chr. 11 (between 101 and 106 cM) with MZ. Additional adjustment for height and weight did not change h² but generally reduced the LOD scores, especially multipoint.

In conclusion, in this preliminary study, we have identified excess allelic sharing at the genomic markers DIIS2000 and DIIS1986 on chr. 11q21-q22, as demonstrated by the section modulus, which reflects bending resistance. Linkage to this and other regions may be in some part due to effects on body size. Identification and subsequent characterization of loci for bone strength of the appendicular skeleton can further elucidate the genetic contributions to bone geometry and resistance to stress. 

Table  .  

* two-point (multipoint)

Disclosures: D. Karasik, None.

This study received funding from: NIAMS/NIA, NHLBI.

W178

Linkage Screen for Bone Mineral Density Phenotypes in Male and Female COP and DA Rat Strains. D. L. Koller¹, I. Alam², O. Sun², L. Liu*¹, M. J. Econs³, T. Foroud*¹, C. H. Turner⁴. ¹Medical and Molecular Genetics, Indiana University School of Medicine, Indianapolis, IN, USA. ²Biomedical Engineering, Indiana University School of Medicine, Indianapolis, IN, USA. ³Medicine, Indiana University School of Medicine, Indianapolis, IN, USA. ⁴Orthopaedic Surgery, Indiana University School of Medicine, Indianapolis, IN, USA.

Studies attempting to localize and identify genes contributing to variation in BMD in both humans and model organisms have demonstrated that multiple genes of modest effect may be involved. Since inbred strains of experimental animals may be informative for only a subset of these genes, we undertook a genome screen to identify BMD QTLs in F2 progeny derived from the inbred Copenhagen 2331 (COP) and dark agouti (DA) strains of rats to complement our previous study in inbred Fisher 344 and Lewis rats. Phenotyping of areal BMD (aBMD) and volumetric BMD (vBMD) at the lumbar spine and femur was completed in 828 animals (405 male and 423 female). Microsatellite marker typing was performed using 93 markers at an average density of 20 cM. Skeletal phenotypes were measured by peripheral quantitative computed tomography (pQCT) and DEXA. Genetic marker maps were generated using Mapmaker/EXP from the genotype data and compared with published maps (RGD). These genotype, phenotype, and mapping data were then employed in the R/qtl software package to perform genome-wide linkage analyses to detect QTLs acting in both sexes as well as those that have sex-specific effects. Permutation tests were conducted to determine genome-wide significance thresholds. A major QTL was detected in both sexes on chromosome 18 for vBMD at midfemur (genome-wide p<0.01). On distal chromosome 1, a QTL was found in both sexes affecting femur and lumbar aBMD as well as vBMD at distal femur, and this QTL appears distinct from the proximal chromosome 1 QTL impacting BMD in our previous Fisher 344xLewis F2 cross. Additional aBMD and vBMD QTLs on chromosomes 2, 6, and 10 were also detected in both sexes of the COPxDA F2. Several sex-specific QTL were also detected, notably for the distal femur vBMD phenotype. These included a male-specific QTL (p<0.01) on chromosome 8 and female-specific QTL on chromosomes 7 and 14 (p<0.01). Interestingly, few of the QTL identified demonstrated overlap with the significant QTLs from the Fisher/Lewis cross. These results confirm the presence of QTL which act in both sexes, as well as some which are gender specific. In addition, unique QTL appear to influence different measures of BMD. Thus, the genetic architecture of BMD in the rat model is quite complex and would appear to be influenced by a number of different genes.

Disclosures: D.L. Koller, None.

W179

Evidence for QTL Underlying Normal Variation in Calcaneal Quantitative Ultrasound Measures: The Southwest Ohio Family Study. M. Lee¹, A. C. Choh*¹, E. W. Demerath*¹, B. Towne*¹, R. J. Sherwood*¹, D. L. Duren*¹, J. Blangero*², R. M. Siervogel¹, S. A. Cole*², S. A. Czerwinski*¹. ¹Community Health, Wright State University Boonshoft School of Medicine, Kettering, OH, USA. ²Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX, USA.

Research in the the genetics of osteoporosis has steadily increased in recent years, but the identification of specific genes that influence osteoporosis risk factors such as calcaneal quantitative ultrasound (QUS) measures is still not well established. Quantitative ultrasound traits are correlated with bone mineral density (BMD), but are useful independently from BMD, as they predict risk for future fracture. The purpose of this study is to conduct a whole genome scan to search for quantitative trait loci influencing normal variation in QUS traits. QUS measures were collected from a subset of 555 individuals (255 males and 300 females) from the Southwest Ohio Family Study who were genotyped and had at least one QUS measurement. Participants ranged in age from 8 to 86 years and were distributed across 5 large extended families. Using the Sahara ® bone sonometer (Hologic, Inc., Waltham, MA), speed of sound (SOS), broadband ultrasound attenuation (BUA), and quantitative ultrasound index (QUI) were collected in each heel (right and left). Variance components based linkage analysis was performed on all six traits using 395 polymorphic genetic markers spaced approximately 10 cM apart across the whole genome to identify quantitative trait loci (QTL) influencing the QUS traits. Age, sex, race and weight effects were simultaneously estimated in all models. Initial heritability estimates (h²) for the 6 QUS traits ranged from 0.54 to 0.68. Suggestive evidence for a QTL influencing SOS was found on chromosome 5p15 (LOD = 2.00) near marker D5S1953, while a suggestive QTL on chromosome 5q32 (LOD = 2.23) near marker D5S436 was found for QUI. An additional suggestive linkage was observed on chromosome 15q21 for both SOS (LOD = 2.27) and QUI (LOD=2.23) near marker D15S978. Additionally, QTL with LOD scores > 1.5 were observed on chromosomes 1, 3, and 16.

Disclosures: M. Lee, None.

W180

A Quantitative Trait Locus Mapping to Chromosome 2p Influences Variation in Bending Resistance of Bone: The Framingham Osteoporosis Study. M. C. Mahaneu¹, L. M. Havill¹, S. Demissie*², L. A. Cupples*², D. P. Kiel³, D. Karasik³. ¹Genetics, Southwest Foundation for Biomedical Research, San Antonio, TX, USA. ²School of Public Health, Boston University, Boston, MA, USA. ³Hebrew SeniorLife, Institute for Aging Research, Boston, MA, USA.

Genes influencing geometric bone indices at the hip have not been clearly identified in previous studies. Thus we used maximum likelihood based variance decomposition methods to detect, characterize, and localize to specific chromosomal regions the effects of genes influencing variation in average buckling ratio (ABR), cross sectional moment of inertia (CSMI), and section modulus (Z) at the femoral neck (FN) and proximal femoral shaft (FS) measured in more than 1350 Framingham Osteoporosis Study participants from 323 nuclear and small extended families. Initial analyses of age- and sex-adjusted residuals following inverse-Gaussian normalization detected significant (P≤1.72ξ10⁻¹⁵) additive genetic effects; with heritability estimates (h²) ranging from h²=0.34 (FN-ABR) to h²=0.61 (FS-Z). Univariate whole genome linkage screens detected genome-wide significant evidence (at α=0.05, LOD=3.01) for a quantitative trait locus (QTL) for two bending resistance traits (FS-CSMI and FS-Z) at the same genomic location: 80 cM from 2pter. Bivariate statistical genetic analyses revealed that nearly all (94%; ρ_G=0.97, P<5.23ξ10⁻³⁷) additive genetic variance in these two measures was due to the shared effects of the same gene(s). Follow-up bivariate linkage analyses focusing on chromosome 2 supported a pleiotropic QTL at this same location (QTL-specific genetic correlation, ρ_Q=1.0, P<0.0001). Although univariate evidence for an FS-ABR QTL on 2p was only suggestive, a subsequent trivariate linkage analysis that included this measure in addition to FS-CSMI and FS-Z found significant (P<0.0001), complete (ρ_Q²=1), and negative (i.e, ρ_Q=1) QTL-specific pleiotropy between it and both FS-CSMI and FS-Z. These observations are consistent with known differences in the relationships of the variables to fracture resistance: i.e., both CSMI and Z are positively correlated with bone strength while increased ABR is associated with higher fracture risk. Additionally, bivariate whole genome linkage screens localized pleiotropic QTLs influencing CSMI and Z at both the FS and FN sites (respectively, maximum bivariate LOD=3.19 and LOD=2.81) at this same location on 2p. Peak evidence places all the QTLs detected in this study just qter of the midpoint of a 16.4 mb region bounded by D2S1352 and D2S1772 and within which no bone related QTLs or obvious candidate genes have been mapped previously. We conclude that this QTL contributes to variation in bending resistance-related traits at both proximal femur sites.

Disclosures: M.C. Mahaney, None.

W181

QTL on Chr 11 Regulates Bone Shape in SAMP6 Mice. B. Otsuki*¹, M. Shimizu*², M. Mori*³, S. Okudaira*², R. Nakanishi*², K. Higuchi*³, T. Tsuboyama*², T. Nakamura². ¹Orthopaedic Surgery, Kishiwada City Hospital, Kishiwada, Japan. ²Orthopaedic Surgery, Kyoto University, Kyoto, Japan. ³Aging Angiology, Shinsyu University, Matsumoto, Japan.

In our previous study using SAM strains, we reported a QTL (Pbd1) on Chr1l with significant linkage to cortical thickness index (CTI, Fig. 1A). The aim of this study was to clarify the effects of Pbd1 and narrow the QTL region. All mice were handled according to the Guideline for Animal Experiments of Kyoto University.

We first generated a congenic strain named P6.P2-Pbd1b that carried a 39 cM SAMP2-derived Chrl1 interval onto a SAMP6 background, and the CTI of P6.P2-Pbd1b was significantly higher than that of SAMP6 (Fig. 2). To narrow the interval, sixteen sub-congenic strains that have smaller overlapping intervals were generated (Fig. 2). From the CTI or BA/TA values, the Pbd1 locus should be seated on the region between D11Mit10 and D11Mit184.

In the morphological analyses of femora using μCT at 16 weeks (Fig. 1B and 1C), the cross-sectional shape of S5 at midshaft was more compressed than that of SAMP6 in the anteroposterior direction (Fig. 1D). We then examined the time-course change of OBL between SAMP6 and S5 (Fig. 1E). OBL was not different at 5 days of age. However, the difference in OBL became remarkable as they grew, indicating that the difference was formed in the process of bone modeling. The shape of epiphyseal growth plate of SAMP6 and S5 was not different; however, dynamic histomorphometrical analysis at 8 weeks revealed that rates of endocortical and periosteal bone formation were different between SAMP6 and S5 (Table 1).

In conclusion, the Pbd1 locus that was seated between D11Mit10 and D11Mitl84 alters the cross-sectional shape of the femur. 

Table  .  

 ,  

Figure Fig 1A.. Lateral view of femur and CTI. B and C, Summary of the morphological parameters of the femur measured using microfocus computer tomography (μCT). D. μCT images of femurs of SAMP6 and S5 mice E. the time-course change of OBL.

Figure Fig 2.. Characteristics of initial seven subcongenic strains and second-generation nine-subcongenic strains. We used CTI and BA/TA for the index of relative bone mass because these values are quite well correlated (R = 0.929, p < 0.0001).

Disclosures: B. Otsuki, None.

W182

QTLs for Femoral BMD Identified in Two Separate Chicken Intercrosses Are Syntenic to Loci Controlling Human BMD. C. Rubin*¹, D. Wright*², A. Sahlqvist*¹, S. Kerie*¹, P. Jensen*³, C. Ohlsson⁴, H. Mallmin⁵, S. Larsson*⁵, O. Ekvall*¹, O. Kämpe*¹, K. Johnsson⁶, L. Andersson*², A. Kindmark¹. ¹Dept Med. Sciences, Uppsala University, Uppsala, Sweden. ²Dept. of Medical Biochemistry and Microbiology, Uppsala University, Uppsala, Sweden. ³IFM Biology, Linköping University, Linköping, Sweden. ⁴Dept. of Internal Medicine, The Sahlgrenska University Hospital, Gothenburg, Sweden. ⁵Department of Surgical Sciences, Uppsala University Hospital, Uppsala, Sweden. ⁶Department of Surgical Sciences, Uppsala University, Uppsala, Sweden.

We have previously conducted QTL-analysis for femoral bone traits in an intercross between the Red Junglefowl (RJ), the wild ancestor of domestic chicken, and White Leghorn (WL) strain L13 (L13), a breed heavily selected for egg-production. We have now performed QTL-analyses in a replicate intercross where RJ was intercrossed to another strain of White Leghorn (the OS-strain). The two WL-strains have been reproductively separated for approximately 50 years, during which genetic recombination has taken place within each strain. QTL-analyses in the OS/RJ-intercross may therefore considerably confine confidence intervals for shared QTLs.

From the RJ/L13-intercross, 337 F2-individuals were genotyped for 548 informative genetic markers and pQCT and DXA were used to measure femoral BMD and bone structure. From the RJ/OS-intercross, 700 F2-individuals were genotyped for 384 informative SNP-markers and were subjected to femoral bone phenotyping by pQCT. QTLs were mapped in the software QTL-express using forward selection for loci with significant marginal effects with bodyweight included as a covariate.

In the RJ/L13 intercross eight QTLs were identified as significant on the 5% genome-wide level. Six QTLs were identified in the RJ/OS intercross, three of whose confidence intervals overlap with RJ/L13 QTLs. The three QTLs which overlap are situated on chicken chromosomes 1, 2 and 3 and affect total and cortical BMD (chr. 1), BMD (chr. 2) and bone size as well as BMD (chr 3). The three overlapping QTL-regions are syntenic to human Xp22 (chr. 1), 18p11 and 18q11-q12 (chr. 2), and 6q21-q26 (chr. 3). A QTL for BMD on human Xp22 has previously been reported, as have murine QTLs in the region syntenic to Xp22. Similarly, a human QTL for BMD of lumbar spine has previously been identified on 18p11. A QTL affecting BMD and biomechanical strength in mice is located in a region syntenic to human 6p22-23. The three chicken loci will be fine-mapped, mapped further for interspecies synteny to mouse and man and will be searched for presence of selective sweeps, the latter of which could help confine QTL-intervals greatly. Furthermore, 500 femoral bone RNA-samples collected from RJ/OS F2-individuals will be used for upcoming eQTL-studies.

Disclosures: C. Rubin, None.

W183

Mouse Chromosome 4 Homologous with Human 1p36: Fine Mapping Femoral vBMD and Trabecular Microstructure. K. L. Shultz, L. G. Horton, H. F. Coombs*, V. E. DeMambro, C. J. Rosen, W. G. Beamer. The Jackson Laboratory, Bar Harbor, ME, USA.

We previously reported (28th ASBMR Mtng) that B6.C3H-4-7 congenic mice carried 15 Mb of C3H alleles from Chr 4 that regulate femoral volumetric (v)BMD and trabecular BV/TV. Ten sublines were developed from B6.C3H-4-7 mice to fine map this bone genetic regulation. Isolated femurs from groups (4 mo old) of female and male mice (B6 controls + new sublines) were phenotyped by pQCT for total femoral vBMD and by Micro CT40 for distal trabecular microstructure (BV/TV, trabecular no., thickness). A haplotype map with 11 chromosomal markers made from genotyping the 10 sublines revealed C3H Chr 4 segments of 0.8 to 12 Mb in size. This genetic map was correlated with bone phenotypes to find patterns of segregation indicative of C3H alleles with bone regulation.

vBMD increased in females from 6 sublines that share a region from 135.00-136.42 Mb, while 4 sublines were B6-like. Males from these 6 sublines also increased vBMD, indicating the presence of a gene(s) that regulates vBMD in both sexes. Males of two sublines with C3H alleles in the 126.86-132.69 Mb region also increased vBMD in male but not female femurs, indicating a second region with male-specific regulation.

Data for female distal trabecular microstructure did not reveal a clear segregation pattern indicative of BV/TV, trabecular number, or thickness regulation. However, B6.C3H-4-19 subline females showed a consistent negative effect for all three phenotypes when compared to B6 controls, suggesting C3H alleles in this subline represent a third regulatory locus. Males carrying C3H alleles in the 126.86-132.69 Mb region showed increased BV/TV and trabecular thickness, while males carrying C3H alleles in the 135.00-136.42 Mb region had increased BV/TV and trabecular number.

The Chr 4 region from 135.00-136.42 Mb carries ∼28 known or predicted genes. Examination of the Novartis gene expression database for tissue specific expression showed strong expression of 3 genes (Cnr2, E2f2, Zbtb40) and weak expression for 2 genes (Luxpl, Aof2) in bone/bone marrow. Cnr2 −/− mice have reduced osteoclast activity and low bone mass (Ofek et al 2006). Akp2 (Klein et al, 27th ASBMR Mtng) reported as a candidate for B6.DBA/2 congenic is distal to this region identified by our B6.C3H-4 congenic sublines. The larger proximal region is 5.85 Mbp and contains ∼100 known or predicted genes.

In summary, 1) congenic sublines partitioned a major QTL into components affecting bone density and microstructure; 2) Chr 4 has at least two distinct bone QTL: A, 1.41 Mb, females & males, vBMD & trabeculae; B, 5.85 Mb, males, vBMD & trabeculae; and 3) gene expression databases suggest several candidate genes in region A; while the male-specific region, B, requires further mapping.

Disclosures: K.L. Shultz, None.

This study received funding from: NIAMS 45433.

W184

Heritability of Serum 25-hydroxyvitamin D Suggests Genetic Factors Contribute to Endogenous Vitamin D Production. E. A. Streeten, K. Hairston*, K. A. Ryan*, B. D. Mitchell, A. R. Shuldiner*. Endocrinology, Diabetes and Nutrition, University of Maryland School of Medicine, Baltimore, MD, USA.

Vitamin D deficiency is an important contributor to poor bone health and fractures. The main source of Vitamin D is skin production responding to UVB sunlight, converting 7-dehydrocholesterol precursors to Vitamin D3. With similar occupational sun exposure in the summer, individual serum 25(OH)D levels have been shown to vary widely. Possible explanations for this variability include genetic factors. We hypothesize that genetic factors play a role in endogenous production of Vitamin D and its metabolites.

In 1151 participants of the Amish Family Osteoporosis Study (AFOS), 25(OH)D was measured by Diasorin RIA, intact PTH (iPTH) was measured by RIA and serum calcium (Ca), phosphorus (P), alkaline phosphatase (AP) were measured by multichannel analysis. BMD was measured at the hip, spine, forearm and whole body by DXA (Hologic). Clinical information collected included over-the-counter supplements. All participants had normal values of serum calcium (8.5-10 mg/dl) and creatinine (<1.3 mg/dl) and were not on chronic glucocorticoids. The heritability (h2) of these measures was estimated using a pedigree-based likelihood approach. Associations of BMD (corrected for age, sex and BMI) with 25(OH)D and iPTH (both corrected for season) was estimated by regression analysis.

The mean age (+SD) of AFOS participants was 50.5 + 15.6 (range 18-91); 481 men and 680 women. The mean 25(OH)D was 22.4 + 7.3 ng/ml (range <5 to 51.2); higher in summer (June-Aug) 25.5 + 6.9, than in winter (Dec-Feb), 18.9 + 6.3 (p<0.0001). PTH was higher in winter (58.7+18.1pg/ml) than in summer (53.9+18.5) (p=0.01). 25(OH)D levels were in the deficiency range (<15 ng/ml) in 17%, insufficiency range (16-29 ng/ml) in 69% and sufficiency range (>30 ng/ml) in 14%. Of those with 25(OH)D<30 ng/ml, 24% had secondary hyperparathyroidism (iPTH > 65 pg/ml). Heritability (h2) was 0.35 + 0.06 for 25(OH)D, 0.32 + 0.06 for iPTH, 0.34 + 0.06 for Ca, 0.25 + 0.06 for P, 0.35 + 0.06 for AP and 0.22 + 0.06 for ln PTH/25(OH)D. BMD (corrected for age, sex, BMI) at total hip (but not other sites) was correlated with 25(OH)D (corrected for season), but only in those with 25(OH)D > 30 ng/ml (p=0.03). BMD (corrected for age, BMI) was correlated with iPTH (corrected for season) for spine and hip in women (p=0.04, 0.02) but not in men (p=0.78, 0.48). Vitamin D supplements were taken in 41% (136/329); daily dose mean 398 IU, median 400 IU).

Vitamin D insufficiency and deficiency are common in the Amish and the serum level of 25(OH)D is heritable. The moderate heritability of 25(OH)D and other factors involved in Vitamin D homeostasis (Ca, P, iPTH) suggest that genetic factors may be important to endogenous Vitamin D production.

Disclosures: E.A. Streeten, None.

W185

Genetic Loci for Bone Structure and Strength Identified in Inbred COP and DA Rats. Q. Sun¹, I. Alam¹, L. Liu*², D. L. Koller², L. G. Carr*³, M. J. Econs³, T. M. Foroud², C. H. Turner¹. ¹Biomedical Engineering, Indiana University, Indianapolis, IN, USA. ²Medical and Molecular Genetics, Indiana University, Indianapolis, IN, USA. ³Medicine, Indiana University, Indianapolis, IN, USA.

Bone strength and structure are major determinants of osteoporotic fracture. These two phenotypes are under substantial genetic regulation and the identification of the genes contributing to their observed phenotypic variation may provide important insight regarding the genetics of osteoporosis and fracture risk. Previous studies have shown that the Copenhagen 2331 (COP) and Dark Agouti (DA) rats have significant differences in bone strength despite their similar body mass. Thus, these inbred rat strains may provide a unique resource to identify the genetics underlying the phenotypic variation in bone strength and structure. A sample of 828 (405 males and 423 females) COP x DA F2 progeny had extensive phenotyping for bone structure measures including cortical bone area and polar moment of inertia at the femur midshaft and total, cortical and trabecular bone areas at the distal femur, and lumbar vertebra 5 (LV5). Bone strength phenotypes included ultimate force of femur and LV5. These skeletal phenotypes were measured using peripheral quantitative computed tomography (pQCT) and mechanical testing. A whole-genome 20 cM screen was conducted in the F2 rats. Genetic marker maps were generated using MARMAKER/EXP from the F2 data and compared with maps from the Rat Genome Database (RGD). The recombination-based maps were then used for genome-wide linkage analyses to detect linkage to the bone strcture and strength phenotypes. Permutation testing was employed to obtain the thresholds for genome-wide significance (p<0.01). A very significant QTL for femur midshaft cortical structure and strength was identified on the distal part of chromosome (Chr) 1 with a LOD score of 33; evidence of linkage was found with both the male and female F2. In addition, Chrs 6, 7, 10, 13, and 15 were linked to femur midshaft structure. Linkage for femur trabecular structure was detected on the proximal part of Chr 1 and Chr 14. Additional QTL for femur strength were identified on Chrs 6 and 15. For LV5, Chrs 2, 16, and 18 harbored QTL for cortical structure. Trabecular structure for LV5 was linked to Chrs 1, 7, 12, and 18. Our study demonstrates genetic linkage for bone structure and strength on several rat chromosomes. The strongest linkage was discovered for several femoral phenotypes on chromosome 1 harboring two distinct QTLs, one at each end.

Disclosures: Q. Sun, None.

W186

Oral Bisphosphonate Treatment Prevents the Changes of Bone Mineral Density and Bone Biomarkers in Postmenopausal Japanese Women with Early Breast Cancer Treated with Aromatase Inhibitor. O. Chaki, K. Kurasawa*, T. Ishikawa*, Y. Nomura*, H. Yoshikata*, R. Kikuchi, E. Hirahara*. Yokohama City University, School of Medicine, Yokohama, Japan.

Many clinical trials have demonstrated the efficacy and safety of long-term adjuvant treatment with aromatase inhibitors (AIs) in women with breast cancer. Whereas aromatase inhibitors are generally well tolerated and, compared with tamoxifen, offer superior efficacy and safety, in certain other respects they are associated with the loss of bone mineral density and an increased risk of fracture. Bisphosphonates are effective inhibitors of osteoclast activity and bone resorption, and are standard treatments for osteoporosis. Bisphosphonates are now being incorporated into breast cancer treatment regimens in order to combat osteoporosis caused by ovarian suppression, chemotherapy treatment, aromatase inhibitors and the postmenopausal state itself.

The aim of our study was to investigate changes in bone loss and bone turnover associated with aromatase inhibitor treatment and the effectiveness of oral bisphosphonate treatment. We conducted an open prospective study in 34 postmenopausal women treated with anastrozole for an early breast cancer. Fourteen patents were treated with alendronate (5mg/day). Bone mineral density (BMD) was measured using DXA at baseline and one year later. The bone turnover markers bone alkaline phosphatase (BAP) and urinary NTX were measured at baseline and 1,3 and 6 months later. In the aromatase inhibitor group, there was a marked bone loss at the lumbar spine by −2.8% (p< 0.001) and hip by −1.3% (p< 0.05), but there was no significant increase in bone turnover (+16.4% for BAP and +5.3% for uNTX). BMD of bisphosphonate treatment group was increased at the lumbar spine by 2.3% and hip by 7.8%.

In conclusion, anastrozole treatment of breast cancer induces an accelerated bone loss but doesn't induce an increased bone turnover in Japanese postmenopausal women with breast cancer. This bone loss is suppressed by oral bisphosphonate treatment completely.

Disclosures: O. Chaki, None.

W187

Dynamic Genomic and Proteomic Patterns of Osteosarcoma Metastasis to Lung: Implication of Macrophage Migration Inhibitory Factor and Tumor Necrosis Factor for Lung Metastasis. B. Whited*¹, C. Wilburn*², G. Holt*¹, H. Schwartz*¹, R. Caprioli*³, L. Gerstenfeld⁴, R. Caldwell¹. ¹Department of Orthopaedics and Rehabilitation, Vanderbilt University School of Medicine, Nashville, TN, USA. ²Department of Medicine/Clinical Pharmacology, Vanderbilt University School of Medicine, Nashville, TN, USA. ³Department of Biochemistry, Mass Spectrometry Research Center, Vanderbilt University School of Medicine, Nashville, TN, USA. ⁴Department of Orthopaedic Surgery, Boston University Medical Center, Boston, MA, USA.

Despite successful treatment of osteosarcoma (OSA), death from lung metastasis occurs in > 30% of patients. Novel molecules permissive for OSA metastasis to lung may improve treatment outcome by existing as prognostic markers or therapeutic targets. The rationale of this study was to examine differences in common osteoblastic gene expression between high and low metastatic OSA and mine the proteome for differences in protein expression. We focused on transcript and protein profiles from an established OSA model using northern blotting analyses, 2-dimensional difference gel electrophoresis (2D-DIGE) and MALDI-TOF mass spectrometry. This balb/c murine model of OSA is characterized by orthotopic primary tumor growth, a period of minimal residual disease, spontaneous pulmonary metastasis, and clonally related variants (K7M2 and K12). K7M2 cells are aggressive and highly metastatic, whereas K12 cells are less aggressive with infrequent pulmonary metastases. Analysis of gene expression during mineralization assays revealed dynamic differences in mRNA levels of COLA1, APase, and OC between the K7M2 and K12 OSA cells in both the proliferative and extracellular matrix deposition stages. 2D-DIGE and MALDI protein profiling demonstrated a comprehensive proteomic evaluation between the K12 and K7M2 OSA cells, revealing several differentially expressed proteins such as macrophage migration inhibitory factor (MIF), ubiquitin, and thymosin beta-4. Computer modeling algorithms further implicated VEGF, TNF, and IL-1B as key mediators in OSA metastasis to lung. Western blotting and IHC validated these findings as we observed increased TNF and MIF expression in the metastatic cells. Taken together, decreased levels of osteoblast markers in K7M2 cells demonstrate dedifferentiation, while up-regulation of MIF and TNF in highly invasive K7M2 cells represent molecular mechanisms to compromise host immunosurveillance during metastasis. These studies are among the first to implicate MIF and TNF as potential biomarkers and therapeutic targets for OSA lung metastasis. These data also represent a rational, comprehensive approach to evaluate the cellular proteome to identify proteins that may be causally associated with metastasis.

Disclosures: R. Caldwell, None.

This study received funding from: Brooks Fund.

W188

Melanoma Bone Metastasis: Osteoclasts Cannot Be Excluded. L. C. Dumitrescu, J. R. Edwards, J. A. Sterling, G. R. Mundy. Center for Bone Biology, Vanderbilt University Medical Center, Nashville, TN, USA.

Recently, there has been much controversy on the role of osteoclasts in bone metastasis in melanoma. Jones et al (2006) have suggested that melanoma bone metastases are not mediated by osteoclasts. Since we have been unable to find convincing evidence that bone osteolysis in cancer ever occurs independent of osteoclasts, we employed the melanoma cell lines B16F10 and Mel A along with primary bone marrow-derived cells to investigate and characterize the effect of this aggressive tumor on osteoclast formation and function, using bone resorption assays, enzymatic and immunohistochemical analysis and RT-PCR for melanoma-derived osteolytic factors. Treatment of primary bone marrow-derived osteoclast-precursors with M-CSF (30ng/ml), RANKL (50ng/ml) and media conditioned by B16F10 or Mel A cells showed a significant dose-dependent increase in the formation of TRAP+ multinucleated cells. These cells were also vitronectin receptor and calcitonin receptor positive. Addition of only 5% B16F10 conditioned media to bone marrow cultures stimulated the formation of 47±1 TRAP+ cells/well, compared to 5.5 ±2.5 in control treated cultures (p<0.05). A similar significant response was observed with media conditioned by Mel A cells, with 49.5±10.5 TRAP+ cells/well seen in treated cultures compared to 8.5±0.5 in control cultures (p<0.01). The quantity of TRAP+ cells formed correlated with a dose-dependent increase in bone resorption and lacunae pit formation on dentine discs. RT-PCR analysis showed clear expression of PTHrP on B16F10 and Mel A cells, although RANKL was not expressed in either cell type. These studies demonstrate a clear interaction between melanoma cells and osteoclast precursors through the release of potent soluble factors, and indicate that melanoma cells are capable of stimulating osteoclast activity. These findings are consistent with a central role for osteoclasts in the extensive osteolysis observed in melanoma bone metastases.

Disclosures: J.R. Edwards, None.

W189

A Phase II Pilot Study of Safety and Effects on Bone Resorption of AZD0530 in Prostate or Breast Cancer Patients with Bone Metastases. R. D. Finkelman¹, F. M. Torti*², A. Lipton*³, R. A. Hannon⁴, A. Hussain*⁵, C. Evans*⁶, M. Akilu*², R. Iacona*¹, A. Swaisland*⁷, M. Stuart*⁷. ¹AstraZeneca, Wilmington, DE, USA. ²Wake Forest University Comprehensive Cancer Center, Winston Salem, NC, USA. ³Hershey Medical Center, Hershey, PA, USA. ⁴Academic Unit of Bone Metabolism, University of Sheffield, Sheffield, United Kingdom. ⁵Greenebaum Cancer Center, University of Maryland School of Medicine, Baltimore, MD, USA. ⁶University of California, Davis, Sacramento, CA, USA. ⁷AstraZeneca, Alderley Park, Macclesfield, United Kingdom.

We have reported previously that once daily dosing for 14 days with AZD0530, a highly selective and orally available Src kinase inhibitor with Bcr-Abl activity, resulted in a dose-dependent reduction of bone resorption markers βCTx and NTx in healthy male volunteers (Hannon et al, ASBMR 2005). A similar outcome was evident in preliminary, unvalidated data from cancer patients with advanced disease. The present study is the first to investigate AZD0530 in cancer patients specifically with metastatic bone disease, and is due to start enrollment in September 2007. Patients with prostate or breast cancer and at least one radiographically confirmed metastatic bone lesion will be randomized to one of two treatment arms: (1) AZD0530, 175 mg once daily; or (2) zoledronic acid, 4 mg iv infusion, every 4 weeks. Patients must have been on stable therapy for their cancer for at least 8 weeks with no anticipated change of therapy for 4 weeks after randomization. Patients will continue on their present therapy during the study. Patients must have had no previous exposure to bisphosphonates. Concomitant chemotherapy with cytotoxic or novel molecularly targeted agents is not allowed on study or within 4 weeks prior to study entry. Concomitant therapy with bisphosphonates is not allowed on study for patients randomized to the AZD0530 treatment arm. Fasted serum and urine samples for assessment of bone turnover biomarkers βCTx, NTx, bone ALP, PINP, ICTP, and TRAP 5b will be collected at baseline and at 1, 2, and 4 weeks. PTH, PTHrP, calcium, phosphate, and total ALP will also be assessed in serum at the same timepoints. Serum PSA will be assessed in patients with prostate cancer at baseline and at 4 weeks. Changes from baseline will be compared between treatment arms. Data from this study will be used to determine the future development of AZD0530 in the treatment of metastatic bone disease.

Disclosures: R.D. Finkelman, AstraZeneca 3.

This study received funding from: AstraZeneca.

W190

Zoledronic Acid Decreases Serum Low-density Lipoprotein Cholesteol in Multiple Myeloma Patients. G. Martini¹, A. Gozzetti*², L. Gennari¹, S. Salvadori*¹, D. Merlotti¹, A. Avanzati*¹, B. Franci*¹, M. Campagna*¹, F. Lauria*², R. Nuti¹. ¹Metabolic Disease Unit, University of Siena, Siena, Italy. ²Hematology and Transplants Division, University of Siena, Siena, Italy.

Nitrogen-containing bisphosphonates (N-BPs) have been designed to inhibit osteoclast-mediated bone resorption and therefore they are the most widely used and effective anti-resorptive agent for the treatment of diseases in which there is an icrease of osteoclastic resorption as postmenopausal osteoporosis, Paget's disease and tumor-associated osteolysis. The mechanism of action of these drugs has not been completely clarified but it has been observed that N-BPs may inhibit squalene synthase, as ibandronate, or farnesyl pyrophosphate synthase, as alendronate or pamidronate, inducing also a reduction of cholesterol biosynthesis. Zoledronic acid (ZA) represents a novel N-BPs which has also antitumor activity. “In vitro” studies suggest that ZA can inhibit cholesterol production, but 90% of the biosynthesis was rescued after 48 hours.

To explore the effects of ZA on serum lipids “in vivo”, we studied 26 patients with smoldering myeloma at diagnosis. Sixteen patients were treated with ZA (4 mg ev) at baseline and at months 1, 2, 4 and 6. The remaining 10 served as controls. At baseline and after 1, 3 and 6 months, a blood sample was drawn from each patient to evaluate total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C) and C-terminal telopeptide of Type I collagen (CTX). In the control group no significant changes were observed throughout the study period for any of the biochemical variables. As expected, CTX decreased significantly by 40-50% after ZA administration. In treated patients, we observed a progressive and significant reduction of TC with a maximum decrease of 13% at 6 months; moreover LDL-C decreased by 21% at 6 months while no significant differences were appreciated as regards HDL-C and TG. Also the indexes of cardiovascular risk improved after ZA administration: TC/HDL-C ratio progressively decreased by 18% and HDL-C/LDL-C ratio increased by 35% showing an effect that appears to be cumulative.

In conclusion we demonstrated that ZA, a third class of N-BPs, is able to alter lipid profile more than older N-BPs as pamidronate or neridronate. This effect can be due to the potency of ZA to inhibit FPP synthase that is about 70-fold than pamidronate.

Disclosures: L. Gennari, None.

W191

Characterization of Leupaxin as a Focal Adhesion-associated Protein in Prostate Cancer Cells. A. Gupta.. Biomedical Sciences, Dental School, University of Maryland, Baltimore, Baltimore, MD, USA.

We have identified the presence of leupaxin (LPXN), which belongs to the paxillin extended family of focal adhesion-associated adaptor proteins, in prostate cancer (PC) cells. Previous studies have demonstrated that LPXN is a component of the podosomal signaling complex found in osteoclasts, where LPXN was found to associate with the protein tyrosine kinases Pyk2 and c-Src, and the protein-tyrosine phosphatase-PEST (PTP-PEST). In the current study, LPXN was detectable as a 50 kDa protein in PC-3 cells, a bone-derived metastatic androgen-independent prostate cancer cell line. Second, we demonstrate that the subcellular localization of LPXN is both cytoplasmic and nuclear, suggesting the possibility of nucleo-cytoplasmic shuttling in PC-3 cells. Third, in PC-3 cells, LPXN was also found to associate with Pyk2, c-Src, and PTP-PEST. Fourth, a siRNA-mediated inhibition of LPXN expression resulted in decreased in vitro PC-3 cell migration. Fifth, a recombinant adenoviral-mediated overexpression of LPXN resulted in an increased association of Pyk2 with LPXN, whereas a similar adenoviral-mediated overexpression of PTP-PEST resulted in decreased association of Pyk2 and c-Src with LPXN. Sixth, the overexpression of LPXN in PC-3 cells resulted in increased migration, as assessed by in vitro Transwell migration assays. On the contrary, the overexpression of PTP-PEST in PC-3 cells resulted in decreased migration. Finally, overexpression of LPXN resulted in increased activity of Rho GTPase, which was decreased in PTP-PEST-overexpressing cells. Our data demonstrates that LPXN forms a signaling complex that may play a role in PC-3 cell function.

Disclosures: A. Gupta, None.

This study received funding from: AR-47942.

W192

Effects on Bone Turnover of the Potent, Once-Daily, Oral Src Inhibitor AZD0530 in Patients with Advanced Solid Malignancies. R. A. Hannon¹, R. D. Finkelman², G. Clack³, R. B. Iacona*², P. Baker*³, M. Rimmer*³, F. Gossiel*¹, I. C. Smith*³, A. Robinson*³, R. Eastell*¹. ¹Academic Unit of Bone Metabolism, University of Sheffield, Sheffield, United Kingdom. ²AstraZeneca, Wilmington, DE, USA. ³AstraZeneca, Alderley Park, Macclesfield, United Kingdom.

AZD0530 is a potent, orally active, once-daily Src inhibitor with Bcr-Abl activity that modulates multiple key signaling pathways in cancer and has potential activity in many different tumor types. We have previously demonstrated significant AZD0530 dose-dependent reduction in bone resorption in healthy males. In the present Phase I study, we assessed the effect of AZD0530 on bone resorption in adult patients with advanced solid malignancies resistant to standard therapy.

Patients were randomized into three parallel groups to receive AZD0530 50, 125, or 175 mg/day. After a single dose on day 1 followed by 6 days washout, patients received once-daily doses for 21 consecutive days. Fasted serum and second-morning void urine were collected on day 1 predose and on days 2, 3, 17, and 28. Serum cross-linked C-terminal telopeptide of type I collagen (sCTX) and urinary cross-linked N-terminal telopeptide expressed as a ratio to urinary creatinine (uNTX) were assessed as markers of bone resorption. An analysis of covariance model was fitted to the log-transformed baseline-scaled ratio data at day 28, with treatment as a fixed-effect factor and log-transformed baseline (predose day 1) as a covariate.

There were significant reductions in sCTX in all treatment groups and in uNTX at 125 and 175 mg. Percent changes from baseline at day 28, derived from the adjusted geometric mean baseline-scaled ratios, were 36.3% (95% CI 57.6%, 4.4%; P=0.030; n=12), 61.7% (95% CI 73.6%, 44.5%; P<0.001; n=14), and 74.5% (95% CI 83.3%, 61.3%; P<0.001; n=11) for sCTX in the 50, 125, and 175 mg treatment groups, respectively; and 12.7% (95% CI 32.6%, 13.1%; P=0.293; n=12), 48.3% (95% CI 59.2.%, 34.4%; P<0.001; n=14), and 50.1% (95% CI 61.8%, 34.7%; P<0.001; n=11) for uNTX, respectively. There was a dose-response trend for reductions in biomarkers.

The significant reductions in the biomarkers are in accordance with inhibition of osteoclast-mediated bone resorption by suppression of Src activity. The reduction (>60%) in sCTX at 125 and 175 mg of AZD0530 is similar to reductions reported with bisphosphonate therapy. These data suggest that AZD0530 may have therapeutic benefit in osteoclast-driven metastatic bone disease in addition to its inhibitory activity on Src in tumor cells. The trend for dose-dependent changes in biomarkers indicates that the maximum tolerated dose should be taken forward in further development.

Disclosures: R.A. Hannon, AstraZeneca 2.

This study received funding from: AstraZeneca.

W193

A New Urine Assay for Sensitive Detection and Assessment of Cancer-induced Skeletal Perturbations. D. J. Hillegonds¹, D. W. Burton*², J. S. Vogel¹, E. Denk*³, T. R. Walczyk³, M. Yang*⁴, S. Vijayakumar*⁵, D. A. Herold*², L. J. Deftos², R. L. Fitzgerald². ¹Center for Accelerator Mass Spectrometry, Lawrence Livermore National Laboratory, Livermore, CA, USA. ²University of California, San Diego, San Diego, CA, USA. ³Laboratory of Human Nutrition, ETH Zurich, Zurich, Switzerland. ⁴Anticancer, Inc., San Diego, CA, USA. ⁵Davis Cancer Center, University of California, Davis, Sacramento, CA, USA.

The need for earlier detection and improved management of metastatic bone disease has driven the search for new bone remodeling biomarkers, but existing techniques are insufficiently sensitive for reliable use in individuals. We report on a novel tracer technology for bone health wherein one calcium-41 (41Ca) dose fully equilibrates with the living skeleton and remains measurable in urine and blood for many years. Urinary 41Ca/Ca is dominated by tracer incorporation and re-release from mineralized bone. In healthy humans, intra-individual urinary variability averaged 3.3+/-0.2% diurnally and 6+/-3% over 200 days. In a preclinical prostate cancer metastasis model, serum 41Ca/Ca directly and significantly correlated with bone disruption, tumor size, serum calcium, osteoprotegerin, and parathyroid hormone-related protein. We posit that longitudinal urinary 41Ca/Ca measurements may sensitively reveal skeletal perturbations, enabling improved clinical management through rapid identification of therapeutic success and non-invasive detection of the earliest stages of cancer growth in bone.

Disclosures: D.J. Hillegonds, None.

W194

Five Years Results of Bisphosphonate Treatment in Long-term Survivors of Highly-Malignant Osteosarcoma. G. Holzer, M. Dominkus*, R. Kotz*. Department of Orthopaedics, Medical University of Vienna, Vienna, Austria.

Purpose: Two thirds of long-term survivors of highly-malignant osteosarcoma treated with chemotherapy protocols including high-dose methotrexate (MTX) have low bone mineral density (BMD) and one third report about fractures after completion of chemotherapy.

Methods: Ten patients with BMD < 2.5 STD (6 male, 4 female; mean age: 33 {plusmn} 1.4 years) participated in this study. Patients received risedronate orally once daily (5 mg) or weekly (35 mg) plus a calcium / vitamin D combination daily. BMD of lumbar spine (LS) and proximal femur of the non-operated side (PF) were measured by dual energy X-ray absorptiometry at twelve months after beginning of the treatment (mean time after chemotherapy at beginning of oral biphosphonate treatment: 13.8 {plusmn} 2.3 years).

Results: After five years of oral bisphosphonate treatment BMD increased at a mean by 2.28% in the LS, whereas in the PF no changes were seen. No fracture occurred during the study time. Conclusions: It was shown that even in long-term survivors of highly-malignant osteosarcoma treated with chemotherapy protocols including MTX increases in BMD values are possible.

Disclosures: G. Holzer, None.

W195

Activity of the Akt Inhibitor Perifosine and Heat Shock Protein (HSP)-90 Inhibitor 17-(Dimethylaminoethylamino)-17-Demethoxygeldanamycin (17-DMAG) in Modulating Osteoclastogenesis and the Bone Marrow Microenvironment (BMM) in Multiple Myeloma (MM). A. Huston¹, X. Leleu*², X. Jia*², J. Anderson*³, S. Vallet*², A. Roccaro*², A. Moreau*², J. Runnels*², H. Ngo*², E. Hatjiharissi*², G. D. Roodman³, Y. Tai*², P. Sportelli*⁴, T. Hideshima*², P. Richardson*², K. Anderson*², I. Ghobrial*². ¹Hematology/Oncology, James P. Wilmot Cancer Center, Rochester, NY, USA. ²Medical Oncology, Dana-Farber Cancer Institute, Boston, MA, USA. ³Medicine-Hematology/Oncology, University of Pittsburgh, Pittsburgh, PA, USA. ⁴Keryx Biopharmaceuticals, New York, NY, USA.

Alterations in the BMM are thought to be critical to MM pathogenesis. The PI3K/Akt and HSP-90 pathways have previously been shown to be dysregulated in MM. We hypothesize that combining two agents that interact at the level of Akt will lead to synergistic cytotoxic activity among MM cells, osteoclastogenesis and the BMM.

Effects of treatment with perifosine (5-10μM; Keryx, NY) and 17-DMAG (50nM; NCI) upon the BMM were evaluated through osteoclastogenesis and osteoclast (OCL) precursor development assays, co-culture of MM cells with bone marrow stromal cell (BMSC) and endothelial cells, in vitro capillary formation, and migration/adhesion assays in the presence/absence of stromal derived factor-1 (SDF-1) and vascular endothelial growth factor (VEGF).

Perifosine and 17-DMAG induced a near complete inhibition of OCL formation. Effects were observed regardless of perifosine treatment timing (100% inhibition first vs. final week), vs. 17-DMAG (100% inhibition first vs. 60% final week). CFC assays for CFU-GM development resulted in significant inhibition after 17-DMAG treatment, as compared to perifosine or control (mean 2 ± 0.5 colonies/well vs. 18 ± 3 vs. 20 ± 3; p=0.01 17-DMAG vs. control). Treatment with perifosine, 17-DMAG and the combination resulted in significant inhibition in MM cell growth following co-culture with BMSCs (p<0.001). Addition of IL-6 or IGF-1 resulted in MM cell proliferation, which was blocked by perifosine, 17-DMAG or the combination (p=0.013). Co-culture of HUVEC endothelial cells with MM cells or addition of VEGF to MM cells resulted in increased proliferation, which was blocked by the combination of agents (p=0.02 in HUVEC co-culture system; p=0.024 with VEGF). Similarly, perifosine, 17-DMAG and the combination significantly inhibited MM.1S cell migration (2.5-fold inhibition) and adhesion in the presence of SDF-1 or VEGF.

The PI3K/Akt and HSP-90 pathways play a critical role in regulating the BMM in MM, including OCL formation, angiogenesis and MM cell migration and adhesion. Targeting these pathways with perifosine and 17-DMAG represents an attractive approach for modulating the BMM in MM.

Disclosures: A. Huston, Amgen Oncology Institute Hematology/Oncology Fellowship Grant (6/2005-6/2006) 2.

W196

Prostaglandin E2 (PGE2) Receptor EP4 Antagonist Attenuates Osteolysis Due to Cancer Metastasis. M. Takita¹, M. Inada¹, M. Hirata¹, T. Maruyama*², C. Miyaura¹. ¹Department of Biotechnology and Life Science, Tokyo University of Agriculture and Technology, Koganei, Tokyo, Japan. ²Discovery Research Laboratory, Minase Research Institute, Ono Pharmaceutical, Co., Ltd., Osaka, Japan.

Bone metastasis of cancer is accompanied by severe bone destruction with increased bone resorption. We recently identified that prostaglandin E2 (PGE2)-induced osteoclast formation was mediated by RANKL expression in osteoblasts, EP4 is the major PGE receptor than in the other EP subtypes (EP1-EP3). Recent studies suggest a possible role of EP4 receptor on cancer growth, but the roles of PGE2 and EP4 on host osteoblast in bone metastasis is still unclear. Here we examine the effects of EP4 antagonist, a selective inhibitor of EP4 signaling in bone destruction due to malignant melanoma. When mice were injected with B16 cells from the left heart ventricle, metastasis was detected as black spots in distal femurs on day 12 after injection. In the B16 injected group, RANKL and cyclooxygenase-2 mRNA expression was elevated in the femur associated with B16 metastasis. PGE2 production was also elevated in B16 metastatic femur. Bone mineral density (BMD) measured by DEXA showed significant decrease of BMD in femur with B16 metastasis. Histmorphometry indicated that increased bone resorption at the site of B16 metastatic region, BV/TV and Tb.Th were significantly reduced, whereas Tb.Sp and ES/BS were increased, that was compared with femur of control mice. When EP4 antagonist was orally administered to mice injected with B16 cells, the osteolysis due to bone metastasis in femurs was suppressed, and the morphometric parameters were restored to the levels of femur of control mice. To know further mechanisms of B16 induced bone resorption, B16 cells were cultured with bone marrow cells and osteoblasts to evaluate the capacities of osteoclast formation. Contact with un-viable B16 cells fixed with formalin induced RANKL expression and PGE2 production in osteoblasts in the absence of bone-resorbing factors, and RANKL-dependent osteoclast formation was seen in the co-culture of bone marrow cells, fixed-B16 and osteoblasts. In the co-culture, EP4 antagonist completely inhibited osteoclast formation. These results suggest that cell-to-cell interaction between B16 and osteoblasts elicits PGE2 production by osteoblasts, and EP4 signaling induces RANKL expression in osteoblasts to form osteoclasts. Therefore, EP4 antagonist may be a possible candidate for the therapy of bone metastasis due to cancers.

Disclosures: M. Inada, None.

W197

Vitamin D Deficiency in Oncologic Patients — an Ignored and Potentially Life Threatening Condition, Possible Role of Increased Osteoblastic Activity in Induction of Hypocalcemia. S. Ish-Shalom¹, E. Segal*¹, S. Felder*², H. Yoffe- Sheinman*², M. Volner*², E. Gez*², B. Raz*³, Z. Shen-Or*³, N. Haim*³. ¹Bone and Mineral Metabolism Unit, Rambam Medical Center, Haifa, Israel. ²Department of Oncology, Rambam Medical Center, Haifa, Israel. ³Endocrine Laboratory, Rambam Medical Center, Haifa, Israel.

Introduction: Monthly administration of bisphosphonates to oncologic patients with skeletal involvement decreases bone turnover and in combination with poor vitamin D status and consequent decrease in calcium absorption may lead to life threatening hypocalcemia.

Routine or sporadic (based on medical history) evaluation of vitamin D status and subsequent vitamin D supplementation is not included in the national or international guidelines for bisphosphonate treatment for metastatic bone diseases.

The aim of this study was to assess vitamin D status and bone turnover in oncologic patients and their impact on the risk of hypocalcemia

Patients / Methods: Following IRB approval, we have assessed serum calcium, phosphate, albumin, 25(OH)D3, alkaline phosphatase and plasma PTH, procollagen type1 nitrogenous propeptide (P1NP), C-terminal telopeptide of collagen (CTX) in 41 consecutive patients with metastatic bone disease, aged 58.2 ± 12.4 years, range 39- 86. The patients were treated with intravenous bisphosphonates, pamidronate or zolendronate, every 3-4 weeks.

Results: Serum albumin corrected calcium was 9.26 mg/dl ± 0.66, range 6.7- 10.72. 25(OH)D3 serum concentration was 12.3 ± 7.04 ng/ml (max. value 27.6); vitamin D deficiency 25(OH)D3 < 10 ng/ml was observed in 17 pt (41.5%); vitamin D insufficiency 10–15 ng/ml in 10 pt (24.5%), vitamin D inadequacy <30 ng/ml in 14 pt (35%). Plasma PINP was 115.686 ± 192.6 ng/ml

Albumin corrected calcium strongly negatively correlated with P1NP (R(- 0.517)P = 0.001) and with serum alkaline phosphatase (R(- 0.562), P< 0.001). Plasma CTX level was 0.265 ± 0.3 (within premenopausal range of 0.14–0.299 ng/ml) and has not correlated with albumin corrected calcium.

Conclusions: Vitamin D deficiency is common in oncologic patients; they are prone to hypocalcemia due to bisphosphonate treatment and absence of supplementation with calcium and vitamin D, especially in presence of increased osteoblastic activity.

Hypocalcemia and vitamin D deficiency are a preventable conditions; evaluation of vitamin D status and adequate calcium and vitamin D supplementation should be part of management of oncologic patients.

Disclosures: S. Ish-Shalom, MSD 2; Novartis 2; Lilly 2.

W198

Microarray Analysis of Craniofacial and Appendicular Osteosarcoma. C. Kong*, M. F. Hansen.. Center For Molecular Medicine, University of Connecticut, Farmington, CT, USA.

Although both primary craniofacial and appendicular osteosarcoma are classified under the rubric of osteosarcoma, their clinical phenotypes are actually quite distinct. Most significantly, chemotherapy treatment leads to vastly different morbidity outcomes; craniofacial patients have a 40% survival rate compared to an 80% survival rate for appendicular osteosarcoma. Given the differences in outcome with conventional chemotherapeutic agents for this disease, there is a clear need to identify differentially expressed genes in these two tumor types as potential new targets for therapy. We have used the Illumina Human Cancer Panel microarray platform to compare craniofacial and appendicular osteosarcoma tumor cell lines to identify quantitative and qualitative alterations in gene expression that distinguish these two diseases. Our comparison of seven osteosarcoma tumor cell lines (1 craniofacial OS tumor cell line, USAC1, and 6 appendicular OS tumor cell lines: SAOS2, U2OS, MG63, HOS, SJSA1, OHS50) revealed distinct differences in gene expression with a bias towards alteration of expression of genes involved in cell DNA damage response and repair and bone signaling. Genes known to be involved in bone signaling such as FGF8, WNT2, WNT1, BMP4, WNT8B, and WNT5A were significantly overexpressed in the craniofacial osteosarcoma cell line when compared to the appendicular osteosarcoma cell lines. In contrast, genes involved in DNA repair such as MSH6, WRN, MSH3, FANCA, RECQL, XPA, PCNA, and BRCA2 were significantly underexpressed in the craniofacial osteosarcoma cell line when compared to the appendicular osteosarcoma cell lines. It is possible that at least some of the differences in phenotype between craniofacial and appendicular osteosarcomas are the result of the developmental pathways by which the bones of craniofacial and appendicular skeleton arise; the craniofacial bones arise from cranial neural crest cells while the appendicular skeleton arises from mesenchymal-derived cells. Our results indicate that craniofacial and appendicular osteosarcoma tumor cell lines show distinct patterns of gene expression. These patterns may provide insight into the molecular basis of these two diseases and possible targets for chemotherapeutic intervention.

Disclosures: C. Kong, None.

W199

A Comparative Study of Anti-resorptive and Anti-cancer Treatments in the Geometrical and Biomechanical Properties of Rat Femurs with Tumor-induced Osteolysis. T. Lee*, X. Wang*, L. Li*, K. Si-hoe*. Division of Bioengineering, National University of Singapore, Singapore, Singapore.

While much has been learned about the mechanisms of metastatic spread of cancer to bone, there has been little headway in establishing guidelines for monitoring the response of metastases to treatment. The objective of this study was to investigate the efficacy of two different treatments (Anti-Resorptive:Ibandronate and Anti-Cancer:Paclitaxel) of tumor induced osteolysis in terms of geometrical and biomechanical parameters.

A rat femur model for tumor-induced osteolysis using W256 cancer cells was adapted from a previous study. Of the 30 rats implanted with cancer cells, 12 were untreated (CANCER), 9 received Ibandronate (IBAN), and 9 received Paclitaxel (PAC). Another 12 underwent a sham operation (CONTROL). Micro-computer tomography (μCT) scans of the femurs extracted post-mortem were used to quantify standard geometrical parameters such as bone volume (BV) and cross-sectional area (CSA). A 3-point bending test was used to assess the femurs' mechanical stiffness. Finally, serum levels of deoxypyridinoline (Dpd) were measured to obtain the degree of bone resorption, which in turn gave an indication of cancer activity.

A summary of the data recorded 30 days post-surgery (20 days for PAC group) are shown in Table 1. CSA data — taken at 25% of the total femur length from the distal end — from the CANCER group had the largest % difference between the left and right femurs (corresponding with a 13.4% increase in Dpd levels). IBAN rats showed the least % difference in CSA (38.1% drop in Dpd levels), followed by PAC rats. This pattern is repeated in the % differences in BV. The anti-cancer effect of PAC can be inferred from the lower Dpd levels of the PAC group (8.86% increase) as compared to the CANCER group. Both IBAN (2.85%) and PAC (-3.86%) maintained the stiffness properties of the operated femurs, with PAC being marginally more effective.

The initial results suggest the viability of this rat model for bone metastasis, and that IBAN was most effective in maintaining skeletal geometry. Interestingly, PAC seemed to be more capable of preserving the biomechanical properties of the femur, despite deterioration in its geometrical properties. A subsequent full scale study, with additional bone mineral content and histomorphological analysis, is in progress to confirm these results. 

Table  .  

Disclosures: T. Lee, None.

This study received funding from: National Medical Research Council, Ministry of Health, Singapore.

W200

Wwox Knockout Mice Reveal Impaired Steroidogenesis, Defects in Bone Metabolism, and Formation of Osteosarcomas. R. I. Aqeilan*¹, M. Q. Hassan*², A. de Bruin*¹, S. Hussain*², S. Lee*², J. P. Hagan*¹, S. Volinia*¹, G. S. Stein², J. B. Lian², C. M. Croce*¹. ¹Department of Molecular Virology, Immunology and Medical Genetics, Comprehensive Cancer Center, Ohio State University, Columbus, OH, USA. ²Department of Cell Biology and Cancer Center, University of Massachusetts Medical School, Worcester, MA, USA.

The WW domain-containing oxidoreductase (WWOX) gene encodes a 46kDa tumor suppressor that is not present in the majority of cancer cell lines. The bona fide in vivo tumor suppressor activity was recently reported (Proc.Natl.Acad.Sci. 104:3949, 2007). Loss of both alleles of Wwox resulted in the formation of frequent juvenile osteosarcomas, while loss of one allele increased the incidence of spontaneous and chemically-induced tumors. In the present study we investigated WWOX function in vivo by examining Wwox deficient (Wwox^−/−) mice for phenotypical abnormalities. Wwox^−/− mice are significantly reduced in size and die between 3 and 4 weeks. Mutant homozygous mice are born with gonadal abnormalities, displayed impaired steroidogenesis and suffer of a metabolic disorder. Wwox^−/− mice exhibit a delay in bone formation and develop metabolic bone disease. LacZ staining of whole embryos revealed Wwox expression in the axial skeleton, craniofacial bones and limbs. An osteopenic phenotype was observed histologically and by μCT, suggesting a high rate of bone resorption, particularly given the serum values of 50% lower calcium and 20% increased phosphate in the KO mouse. Higher acid phosphatase activity was observed in femur bone sections of KO mice compared to the WT, indicating increased osteoclast activity in KO mice. Gene expression in calvaria and femur bone was examined. A 30-40% reduction in Runx2 and alkaline phosphatase (Alkp) suggested a decrease in cells recruited into the osteoblast lineage. The mechanism for these findings was addressed. WWOX protein, via its first WW domain, interacts with several transcriptional activators containing proline-tyrosine motifs and suppresses their function. We find WWOX physically associates with Runx2, the principal transcriptional regulator of osteoblast differentiation, and functionally suppresses Runx2 transactivation ability in osteoblasts. Expression of Wwox in MDA-MB-231 breast cancer cells reduces the mRNA levels of all Runx2 target genes related to metastasis to bone including MMP9, VEGF, and osteocalcin. These results reveal a vital requirement for WWOX in postnatal survival, steroidogenesis and suggest a central role of WWOX in bone metabolism and repression of tumor growth in bone in part coupled to regulation of Runx2 transcriptional activity.

Disclosures: J.B. Lian, None.

W201

Computed Microtomography of Bone Specimens in the Diagnosis of Bone Metastases. H. Libouban*¹, M. Audran², E. Legrand², N. Ifrah*³, M. F. Basle*¹, D. Chappard¹. ¹Faculté de Médecine, INSERM EMI 0335 — LHEA, Angers, France. ²CHU d'Angers, INSERM EMI 0335 — Service de Rhumatologie, Angers, France. ³CHU d'Angers, Service d'Hématologie, Angers, France.

Bone metastases occur in more than half the patients with advanced cancer. Breast and prostate cancers are specially metastazing to bone. When anchored in the bone marrow, malignant cells release soluble factors which interact with bone cell metabolism. Bone metastases from breast or lung cancer usually exhibit a mixed osteolytic and osteosclerotic, with osteolysis usually predominating. Osteosclerosis is a common finding in prostatic cancer although lytic areas can be identified by computed tomography within the sclerotic lesions. Pathological examination of bone biopsies performed in metastatic areas plays a key role in the diagnosis of malignancies. However, histological analysis often needs several days. The recent development of microCT has allowed non-destructive and fast examination of bones.

118 patients who presented at least one bone metastasis (identified by scintigraphy), an overt myeloma or lymphoma or a monoclonal gammopathy of undetermined significance (MGUS) were studied. All patients had a bone biopsy studied by histomorphometry for the diagnosis of the malignant invasion. During the fixation time, the bone cores (7mm in diameter) were analyzed on a Skyscan microCT (resolution 11μm/pixel). On the 3D reconstructed models, signs of metastasis were searched: excess of bone resorption, focal disorganization of microarchitecture, plexiform bone apposition (woven bone), osteosclerosis. Results were compared with data obtained by histomorphometry using kappa test. Excellent agreement was obtained for bone metastases and overt myeloma or lymphoma. MicroCT identified excess bone resorption on trabecular surfaces when ES/BS were > 11%. MicroCT failed to identified patients with smoldering myeloma (with little tumor mass) or some lymphomas with microresorption. MicroCT data are rapidly obtained (< one day) and can confirm the malignant invasion of bone marrow when excess of bone resorption/formation is obtained. However, these signs are not specific and must be confirmed by histopathological analysis.

Disclosures: H. Libouban, None.

W202

Development of a Highly Sensitive High-throughput Mass Spectrometry-based Assay for Rat Procollagen Type I Propeptide (PINP). J. E. Hale*¹, B. Han*¹, L. V. Hale*², M. Sato², J. You*³, M. Copeland*³. ¹Lilly Research Laboratories, Greenfield, IN, USA. ²Lilly Research Laboratories, Indianapolis, IN, USA. ³Indiana Center for Applied Protein Sciences, Indianapolis, IN, USA.

Type I procollagen aminoterminal propeptide (PINP) is a reliable marker of osteoblastic bone formation activity, and is routinely used in clinical trials. PINP has also recently been shown to be a reliable early bone formation marker in the rat. PINP is proteolytically derived from the type I procollagen molecule. Thus, serum level of PINP reflects new collagen matrix synthesis, the majority of which subsequently become newly mineralized bone. Currently the only commercially available assay for PINP is a radioimmunoassay for human PINP, which may not cross react with PINP from other species. It is important to have PINP assay for non-primate species to investigate the mechanism of bone formation and for preclinical development of drugs that can enhance bone formation.

We developed a highly sensitive, high-throughput mass spectrometry-based assay for PINP in rat plasma or serum that does not rely on antibody reagents. Cysteine sulfhydryl groups in plasma or serum proteins were reduced and alkylated then digested with trypsin. The resulting peptides were separated by HPLC and the N-terminal tryptic fragment of the PINP was detected and quantified using tandem mass spectrometry by monitoring selected ion transitions upon high-energy fragmentation. The N-terminal glutamine residue was found to have quantitatively converted to pyroglutamate and an internal Asn residue was found to be deamidated to Asp.

Enrichment of PINP prior to HPLC-MS was necessary due to low levels of PINP in adult rats. This was achieved by combining reduction-alkylation of the serum proteins with a one-step precipitation of the majority of the serum proteins using acetonitrile. By optimizing the sample preparation method and HPLC conditions, up to 1000 samples could be processed in duplicates by a single person in one week. Circulating levels of PINP showed an age-dependent decrease in concentration. While PINP concentration in the 5 week-old rats was 200-250 nM. it dropped to 10-20 nM in the 9 month-old rats. Ovariectomy of adult female rats initially increased PINP, consistent with increased remodeling of the bone, which includes increased formation and resorption activities. Daily treatment of rats with parathyroid hormone (PTH) showed a statistically significant increase in PINP as has been observed in humans.

Disclosures: J.E. Hale, Eli Lilly & Co 3.

W203

C-Reactive Protein, a Marker of Inflammation, in Patients with Osteoporotic Non-Traumatic Fractures. L. A. Holzer*¹, U. Willinger*², G. Holzer¹. ¹Department of Orthopaedics, Medical University of Vienna, Vienna, Austria. ²ENT Clinic, Medical University of Vienna, Vienna, Austria.

Purpose: Inflammation could be partially responsible for the aetiology of osteoporosis. C-reactive protein (CRP) is a sensitive marker of inflammation, tissue damage, and infection. The purpose of this study was to investigate the inflammatory status expressed by levels of high-sensitive CRP in patients with osteoporosis to assess a possible role of inflammation in the development of osteoporosis.

Methods: Eight hundred twenty-four individuals (495 female and 329 male, mean age: 63.2 ± 14.0 years) were examined at our institution. All individuals answered a questionnaire concerning history, had x-ray and densitometry of the lumbar spine, and, if possible, of the proximal femur. Individuals were classified semi-quantitatively due to the spine fracture status and due to the bone mineral density (BMD) (t-score) of the lumbar spine (LS) and the proximal femur (PF). Laboratory examinations, sexual and thyroid hormone levels and, more specifically, bone turn-over markers were done routinely. Serum high-sensitive CRP was measured using an ultrasensitive immuno-nephelometry (N Latex CRP mono, Behringwerke AG, Marburg, Germany) on a BNA Behring nephelometer. Statistical analysis was done using Pearson correlation, t-tests (unpaired), discriminant analysis, and analysis of variance (ONEWAY). For data handling and analyses SPSS for Windows, Version 10.0 was used.

Results: Group differences between individuals without fractures (mean: 0.32 ± 0.38) and such with multiple non-traumatic fractures (mean: 0.56 ± 0.70) in respect to CRP were statistically significant (t=-2.3, dF=100, p=0.026), as were group differences between controls (mean: 0.25 ± 0.18) and patients with osteoporosis (mean: 0.47 ± 0.57) (t=-2.8, dF=53,5, p>0.0008).

Conclusions: The relationship between high-sensitive CRP serum levels and low BMD and multiple non-traumatic fractures in otherwise healthy individuals, is confirming the possible role of inflammation in the development of osteoporosis.

Disclosures: L.A. Holzer, None.

W204

Optimum Testing Strategy for Urinary N-Telopeptide. S. R. Johnson, I. Gioni*, K. Lehane*, J. Ellis*. R&D, Unipath, Bedford, United Kingdom.

Effective therapies exist to treat osteoporosis, such as bisphophonates. However, non-compliance and persistence mean that many women do not receive the full benefit of their treatment. Bone markers, such as urinary N-Telopeptide (uNTx) can be used to track response to therapy. This can help improve compliance and enable clinicians to identify poor or non-responders, so therapies can be optimised. The aim of this study was to ascertain the best testing strategy for uNTx to enable an accurate determination of levels, whilst providing a method that is practicable for both patients and clinicians.

Twenty women were recruited into each of three age-related cohorts, 31-40, 41-50, 51-60 years. The volunteers provided a second morning void urine sample following an overnight fast, for 30 consecutive days. The Osteomark uNTx ELISA was then used to measure NTx, adjusted for creatinine.

The effect of using 1, 2, 3, 4, 5 or 6 days worth of tests to provide a summary level for uNTx was compared to the weekly mean. Using only 1 or 2 tests produced uNTx estimates that could be greater than 20% different from the weekly mean. However, using 3 or more days worth of testing provided a good estimate of uNTx level, with more tests providing increased accuracy. A 3 day strategy was selected for further analysis because it provided sufficient accuracy, with less effort than 4+ day testing.

The effect of testing on consecutive days, testing every other day and random testing over a week was examined to determine the best 3 test regime. Testing strategies beginning at the start of the week produced the lowest variability, with consecutive measurements providing the most accurate result. Although random testing produced standard deviations higher than testing strategies starting on a Monday, it was no worse than consecutive or alternate measurements made later in the week.

As the influence of testing strategies on providing an accurate level was so small, the recommended testing strategy would be to use any 3 samples within a week. With no rigidity on providing a sample on particular days, more emphasis can be placed on ensuring the sample is second void after fasting. This relaxed strategy also enables sampling to fit in with clinic visits and is more achievable by the patients. 

Table  .  

Disclosures: S.R. Johnson, Unipath 3.

This study received funding from: Unipath.

W205

Some but not all Estrogen-Like Molecules Have Both Bone and Cartilage Protective Effects: Evidence for Complete Uncoupling of Bone and Cartilage Effects. M. A. Karsdal, I. Byrjalsen, D. Leeming, P. Ovist, C. Christiansen.. Nordic Bioscience, Herlev, Denmark.

Postmenopausal osteoporosis is associated with increased bone resorption and increased cartilage degradation. Previous investigators have found chondroprotective effects of compounds with estrogenic properties in both preclinical and clinical settings, in which bone and cartilage degradation seemed to be closely coupled.

The objective of the present study was to determine whether Tibolone, a synthetic steroid with estrogenic, androgenic, and progestogenic properties, would have similar dual actions on both bone and cartilage turnover.

Methods: Ninety-one healthy postmenopausal women aged 52-75 yrs entered a 2-yr double blind, randomized, placebo-controlled study of treatment with either 1.25 mg/day (n=36), or 2.5 mg/day Tibolone (n=35), or placebo (n=20). Second void morning urine samples were collected at baseline, and at 3, 6, 12, and 24 months. The biochemical markers CTX-I and CTX-II were measured in Urine CrossLaps™ ELISA and Urine CartiLaps™ ELISA (Nordic Bioscience A/S) as markers of bone resorption and cartilage degradation, respectively.

Results: Bone and cartilage turnover was evaluated as the mean relative change from baseline during the 2-yr study period, corrected for placebo. Tibolone strongly and highly significantly (P<0.001) suppressed bone resorption as measured by CTX-1, Figure 1a. In contrast, an increase rather than a decrease on cartilage degradation was observed, Figure 1B.

Discussion: Previous studies have demonstrated positive effects of estrogens and various SERMs on both cartilage and bone degradation. In the present study we demonstrate a complete uncoupling of the bone and cartilage effects of the synthetic steroid, Tibolone. Bone resorption was highly significantly decreased as expected, whereas cartilage degradation was elevated. These effects are in direct contrast to those observed and published for other SERMs e.g. raloxifene and levormeloxifene decreasing both bone and cartilage degradation. In conclusion, these results may have important implications for the development of improved SERMs with additional effects on cartilage health supplementary to that of the primary endpoint, bone health. Interestingly these data clearly demonstrate that only some estrogen-like substances but not all have dual actions on both bone and cartilage

Disclosures: M.A. Karsdal, None.

W206

Development of Multiplex Immunoassay Panels for Simultaneous Quantification of Bone Metabolism Markers using Luminex xMAP Technology. W. Lie*, H. Hwang*, J. Wang*, J. Mistry* MILLIPORE, St Charles, MO, USA.

Biochemical markers of bone metabolism play an important role in the assessment of bone diseases such as osteoporosis, arthritis, chronic inflammatory disorders, and bone metastasis with cancers. Millipore has developed various multiplexed immunoassay panels for the simultaneous measurement of multiple bone metabolism markers in mouse, rat, and human, using the Luminex xMAP platform. These bead-based sandwich assays are rapid, sensitive, and reproducible and require ≤ 25 μl of serum/plasma sample or tissue culture supematants per well in 96-well plates. The mouse bone panels allow simultaneous quantitation of the following biomarkers in any combinations in the same sample well: Mouse Panel 1 — OPG, Osteocalcin, ACTH, Insulin, Leptin, TNFα, IL-6, and IL-1β and Mouse Panel 2 — RANKL, Osteocalcin, ACTH, Insulin, Leptin, TNFα, IL-6, and IL-1β. The dynamic ranges for the mouse analytes are 10-40,000 pg/mL (OPG, Osteocalcin, RANKL, and Leptin), 24-100,000 pg/mL (Insulin), and 2-10,000 pg/mL (ACTH, TNFα, IL-6, and IL-1β). The rat bone panels allow simultaneous measurement of the following biomarkers: Rat Panel 1 — OPG, ACTH, Insulin, and Leptin, Rat Panel 2 — RANKL, ACTH, Insulin, and Leptin, and Rat Panel 3 — Osteocalcin and PTH. The dynamic ranges for the rat analytes are 10-40,000 pg/mL (OPG, PTH, Leptin), 5-20,000 pg/mL (Osteocalcin), 4-15,000 pg/mL (RANKL), 24-100,000 pg/mL (Insulin), and 2-10,000 pg/mL (ACTH). The human bone metabolism panels allow for simultaneous measurement of the following biomarkers in any combinations: Human Panel A (serum samples) — OPG, Osteocalcin, Osteopontin, and PTH and Human Panel B (tissue culture samples) — OPG, Osteocalcin, Osteopontin, PTH, Leptin, Adiponectin, Insulin, ACTH, TNFα, IL-6, IL-1β, and VEGF. We have also developed a sensitive immunoassay for measuring the serum levels of human RANKL. The dynamic range for this Human RANKL Single-plex assay is 5-20,000 pg/mL. The above bone metabolism panels exhibited acceptable analytical performance characteristics in terms of sensitivity, intra- and inter-assay precision, linearity of dilution, spike recovery, and antibody pair specificity. In conclusion, various multiplexed assay panels were developed and validated for the measurement of multiple bone metabolism markers in mouse, rat, and human samples. The availability of these multiplexed Bone Metabolism Panels and the RANKL, OPG, and Osteocalcin Single-plex assays may provide a powerful tool in studying biological functions of these biomarkers as well as the pathological roles of these molecules.

Disclosures: W. Lie, None.

W207

Increased Bone Resorption in Knee Osteoarthritis with Severe Joint Surface Wear. T. Mashiba, S. Mori, Y. Kaji*, K. Iwata*, S. Komatsubara*, T. Manabe*, T. Yamamoto*. Orthopedic Surgery, Kagawa University Faculty of Medicine, Kagawa, Japan.

It has been known that the patients with knee osteoarthritis (OA) have little higher bone mineral density (BMD) when compared with normal controls. Recently, however, it have been suggested that subchondral bone metabolism is associated with the progression of osteoarthritis of the knee. The objective of this study is to examine the relationship of bone mineral density and bone metabolic markers to radiological progression of knee OA. One hundred sixty-six postmenopausal women (55-89 y.o.) with radiographic knee OA were enrolled in this study. Weight-bearing antero-posterior knee radiograph with the patellae in central position were taken in all women. They were divided into 5 groups based on Koshino's grading system of knee OA (Grade1: sclerosis or osteophyte formation, Grade2: joint space narrowing < 3mm, Grade 3: joint space disappearance, Grade 4: subchondral bony wear < 5mm, Grade 5: subchondral bony wear > 5mm). All women underwent bone mineral density (BMD) measurements of lumbar spine and hip. Serum osteocalcin (OC) and urinary deoxypyridinoline (DPD) were measured. There were no significant differences among groups in body mass index or BMD at any site. DPD was significantly higher in grade 5 group than in others, and Spearmann' rank correlation analysis showed significant association between elevated OA grade and increased DPD value. However, no significant differences were found in OC among all groups. In this study, bone resorption expressed by DPD was elevated especially in women knee OA with the severest joint surface wear. Our results indicates that increased bone resorption was associated with the progression of subchondral bony wear in knee OA although it is not clear whether increased bone resorption is cause or results of joint surface wear.

Disclosures: T. Mashiba, None.

W208

Fluctuation of the Normal Values of the Biochemical Markers of Bone Turnover in Greek Premenopausal Women., E. Metania*¹, P. Dimou*², P. Katsimbri*¹, T. Kaplanoglou*¹, I. Koulouris*¹, P. N. Soucacos*¹, G. Skarantavos¹. ¹1st Department of Orthopaedic Surgery, Metabolic Bone Diseases Unit, General University Hospital ‘ATTIKON’., Athens, Greece. ²Biochemical Department., Asclipieion Hospital, Voula., Athens, Greece.

Aim: In clinical practise, normal values of the biochemical markers of bone turnover are concidered the values defined by the manufactory companies of the various reagent kits used universally. The fact that the abnormal rate of bone turnover in Greek women is based on normal values of the biochemical bone markers of women of a different nationality, has led to the need to create a data base based in normal values in Greek premenopausal women.

Material-Methods: A detailed medical history was recorded and blood samples were collected from 100 healthy, Greek premenopausal woman, between 30 to 54 years, with normal physical activity. Women in pregnancy, or women who had diseases, or taking pharmacological agents affecting bone metabolism were excluded. In all participants detailed information was given regarding the purpose of the study, and written consent was obtained. Blood samples were collected and, after centrifuging, the values of 3 bone formation markers [osteocalcin, bone alkaline phosphatase, PINP], and 2 bone resorption markers [serum NTX and CTX] were determined. Osteocalcin, PINP (Roche Diagnostics GmbH Mannheim) and serum CTX (Roche Diagnostics GmbH Mannheim) was measured with an electrochemiluminescence immunoassay. Bone alkaline phosphatase (Quidel Corporation, San Diego, CA USA) and serum NTX (Osteomark NTx Serum, Wampole Laboratories, NJ USA) was measured with an ELISA. The participants were categorized in three groups according to their age and statistical analysis of the results was performed.

Results: According to our study the normal values for premenopausal Greek women are: 1) Osteocalcin: 3,56 to 33.8 ng/ml (Mean: 18,68, SD:7,56. 2) Bone Alkaline Phosphatase: 4,52 to 27,08 U/L (Mean: 15,80, SD:5,64). 3) PINP: 12,01 to 58,49 μg/1 (Mean: 35,25, SD:11,62). 4) NTX: 5,18 to 21,98 ng/ml (Mean: 13,58, SD:4,20). 5) CTX: < 0,531ng/ml, (Mean: 0.265, SD:0,133). The reproducibility of our measurements was 5-8%.

Conclusion: 1. The present study reflects the reference values for the biochemical bone markers that determine the normal rate of bone turnover in the Greek women. 2. Thereby, it is feasible to more accurately determine high or low bone turnover in postmenopausal Greek women. 3. In all bone markers that were examined, we observed difference in the values of the Greek population compared with the reference values, especially for osteocalcin and bone alkaline phosphatase. 4. Our results suggest that, during the premenopausal period, the biochemical bone markers show no age variability.

Disclosures: E. Metania, None.

W209

Comparison of Serum Tartrate-resistant Acid Phosphatase Type 5b Assays and Other Bone Resorption Markers for Monitoring Raloxifene Therapy. Y. Mochizuki*¹, A. Oishi*¹, Y. Igarashi*², N. Inaba*¹. ¹Department of Obstetrics and Gynecology, Dokkyo Medical University, Tochigi, Japan. ²Department of Basic Medicine, Dokkyo Medical University, Tochigi, Japan.

Objectives: Tartrate-resistant acid phosphatase type 5b (TRACP 5b) is a reliable bone marker derived specifically from osteoclasts and its serum levels are not affected by feeding, renal or hepatic function. The purpose of this study was to compare the clinical performance of two different TRACP 5b assays and other markers of bone resorption for monitoring raloxifene therapy. Methods: Twenty five postmenopausal women with osteoporosis/osteopenia were treated raloxifene 60mg daily. Serum TRACP 5b, serum and urinary N-terminal telopeptides of type I collagen (S-NTX and U-NTX), urinary deoxypridinoline (DPD) were measured at baseline and 1, 3 and 6 months after treatment. Lumber bone mineral density(LBMD) was determined at baseline and six months.The TRACP 5b assays and other markers were as follows: A novel assay system for TRACP 5b called fragments absorbed immunocapture enzymatic assay (FAICEA) (Nitto Boseki Co., Ltd., Fukushima, Japan), A pH-selective immunoassay measuring TRACP 5b activity (BoneTRAP® Assay, IDS Ltd., NE, UK), S-NTX and U-NTX (Mochida Pharmaceutical Co., Ltd., Osaka, Japan) and DPD (DS Pharma Biomedical Co., Ltd., Osaka, Japan). Results: Serum TRACP 5b values were reduced significantly by raloxifene on 1 month using 2 assay kits. The reduction rates on 1, 3 and 6 months were 22.7%, 26.9% and 28.3% (FAICEA), 13.7%, 18.6% and 22.7% (BoneTRAP®), respectively. On the other hand, the reduction rates of other markers on 1, 3 and 6 months were 16.9%, 14.3% and 17.2% (S-NTx), 18.7%, 31.0% and 25.8% (U-NTX), 15.6%, 22.6% and 18.9% (DPD), respectively. Serum TRACP 5b change exceeded the least significant change (LSC) on 6 month in 66.7% (FAICEA; LSC 31.6%) and 33.3%(Bone TRAP®; LSC 29.5%) of the subjects. The changes of LBMD at 6 months only correlated significantly with the changes of TRACP 5b with FAICEA (r²=0.190, p<0.05) at 3 months. Conclusions: These results suggest that serum TRACP 5b with FAICEA is an excellent bone resorption marker for monitoring raloxifene treatment.

Disclosures: Y. Mochizuki, None.

W210

Serum PINP Is a Useful Marker of Bone Formation in Rat Ovariectomy Model. J. Morko¹, J. P. Rissanen¹, M. I. Suominen¹, Z. Peng¹, S. Rasi*², J. Risteli*³, J. M. Halleen¹. ¹Pharmatest Services Ltd, Turku, Finland. ²SBA-Sciences, Oulu, Finland. ³Department of Clinical Chemistry, University of Oulu, Oulu, Finland.

Serum procollagen I N-terminal propeptide (PINP) is a useful marker of bone formation in humans. We report here the development and characterization of a non-radioactive immunoassay for rat PINP, and the use of PINP as a bone formation marker in rat ovariectomy (OVX) model. A polyclonal PINP antiserum was developed in rabbits using a synthetic peptide QEDIPEVS as antigen, corresponding to the first amino acids of the N-terminal propeptide of rat type I procollagen α1 chain. Intra- and inter-assay coefficients of variation (CVs) were calculated from a panel of rat serum samples in three different PINP levels (n=10 in each level). Two OVX studies were performed, both with 3-month old rats. A pilot 14-day study contained a sham-operated control group and an OVX group, and included measurement of serum PINP and other bone turnover markers before the operation and at days 2, 4, 7, 10 and 14. An extensive 8-week study included a sham-operated control group, an OVX group receiving vehicle, and an OVX group receiving 17β-estradiol (E2, 10 μg/kg/day s.c.). PINP and other bone turnover markets, including serum osteocalcin, C-tenniual cross-linked telopeptides of type I collagen (CTX) and tartrate-resistant acid phosphatase 5b (TRACP 5b), were measured before the operations and at 2 and 8 weeks. Trabecular bone parameters were determined from tibial metaphysis using pQCT analysis and histomorphometry at 8 weeks. The PINP immunoassay had an intra-assay CV of 2.8%, inter-assay CV of 7.5%, dilution linearity of 95% and recovery of 107%. PINP increased during the first 2 weeks after OVX, and returned to sham-level at 8 weeks. Osteocalcin and CTX were increased at 2 weeks, but they were also increased at 8 weeks. TRACP 5b was decreased at 2 and 8 weeks, and the values at 8 weeks correlated strongly with the histomorphometrically determined absolute number of osteoclasts (N.Oc/T.Ar). E2 prevented the increase of PINP, as well as the changes of the other bone turnover markers caused by OVX. Changes in PINP showed a strong correlation with changes in other bone turnover markers at 2 weeks, and PINP values at 2 weeks correlated significantly with trabecular bone parameters at 8 weeks. These results demonstrate that rat serum PINP is a useful marker of bone formation, and short-term changes in PINP predict long-term changes in trabecular bone parameters in rat OVX model. Osteocalcin behaved similarly than CTX and differently than PINP at 8 weeks, suggesting that a substantial amount of rat serum osteocalcin may be derived from bone resorption rather than bone formation.

Disclosures: J. Morko, None.

W211

Automated Measurement of 1,25 Dihydroxyvitamin D on the LIAISON® Analyzer. D. M. Heldman*, D. L. Ersfeld*, C. L. Ross*, P. J. Krohn*, G. T. Olson, J. A. Schmidt. Research & Development, DiaSorin Inc, Stillwater, MN, USA.

1,25-dihydroxyvitamin D (1,25-Vit-D) is the metabolically active, most potent form of vitamin D. Its production in the kidneys is tightly regulated by serum calcium, phosphorus, and PTH concentrations. However, numerous disorders exist that can disrupt this carefully controlled system. These include renal failure, type I and II vitamin D dependent rickets, as well as extra-renal disorders such as sarcoidosis and rheumatoid arthritis. Because of these and similar conditions it is becoming increasingly important to have an accurate measurement of 1,25-Vit-D. Therefore, we have developed a sensitive and precise automated assay for the measurement of 1,25-Vit-D on the LIAISON® Analyzer.

Using a competitive inhibition, chemiluminecent immunoassay and a solvent based C-18 column extraction, we were able to demonstrate the following performance parameters over a range of 4-200 pg/mL. Analytical and functional sensitivity were <4 pg/mL and <15 pg/mL respectively. Recovery ranged from 96-112%, with an average of 103%. The coefficient of variance (CV) for intra-assay precision was <13% and inter-assay precision was <15%. Linear regression for dilution linearity yielded a slope of 0.95 and r-value of 0.99. Linear regression for the method comparison of 51 patient samples compared to the DiaSorin 1,25 Dihydroxyvitamin D RIA kit was LIAISON® = 0.91(RIA) + 2.1, r = 0.94. Results obtained during development demonstrate that the LIAISON® 1,25 Dihydroxyvitamin D Assay is an accurate and precise means of measuring serum 1,25-Vit-D levels. This product significantly enhances the expanding DiaSorin LIAISON® bone and mineral product line.

Product availability subject to required regulatory approvals.

Disclosures: G. T. Olson, DiaSorin Inc. 3.

This study received funding from: Diasorin, Inc.

W212

Analytical and Clinical Performance of an Automated Assay for the Measurement of Osteocalcin on the LIAISON® Analyzer. A. L. Podgorski*¹, D. L. Vaught*¹, J. J. Body*², F. Vertongen*³, G. T. Olson¹, F. A. Blocki¹, J. A. Schmidt¹. ¹Research and Development, DiaSorin Inc., Stillwater, MN, USA. ²Department of Medicine, Institut J. Bordet, Universite Libre de Bruxelles, Brussels, Belgium. ³CHU Saint Pierre, Universite Libre de Bruxelles, Brussels, Belgium.

Osteocalcin (OC) is a small bone matrix protein (Mr 5800) produced by osteoblasts (OB). Three of OC's 49 amino acids are calcium binding gamma-carboxyglutamic acids. OC comprises nearly 1% of bone's organic matrix: smaller amounts are synthesized in dentine by odontoblasts. Expression of OC is vitamin K-dependent and under the control of 1,25-dihydroxyvitamin D. OC is a marker of bone formation reflecting increased or decreased bone turnover in patients with altered bone metabolism. Although OC serum levels are broadly accepted as reflecting increased OB activity, intact OC and OC fragments from the resorption of bone matrix have recently been reported as contributing to serum OC levels.

The LIAISON® Osteocalcin Assay utilizes a 25 μL serum sample, 20 μL magnetic particles coated with a monoclonal anti-OC (directed to residues 20-43) and 175 μL tracer, a second Mab anti-OC (directed to residues 1-19) labeled with an isoluminol derivative. Following a 10-min incubation, wash, and chemical trigger addition, light is measured that is proportionate to the amount of OC in the sample. Results are available in 20 minutes with a throughput of 180 tests/hour. Performance was established following standard CLSI protocols. Analytical and functional sensitivities are ≤0.3 ng/mL and ≤3 ng/mL, respectively. Intra- and inter-assay imprecision with samples throughout the range is ≤ 9%, following a 20 day precision protocol. Mean recovery is 93% and dilution linearity is (observed) = 1.05(expected) + 0.12, R = 1.00. A correlation using 249 serum samples across the range is: LIAISON = 0.99(commercial assay) − 1.01, R = 0.98.

In a clinical laboratory samples from patients with osteoporosis or metastatic breast cancer were tested using the LIAISON® Osteocalcin Assay to demonstrate the clinical utility of this assay for treatment evaluation. The results of the LIAISON® Osteocalcin assay corresponded well with another approved automated assay.

The accurate, precise LIAISON® Osteocalcin Assay enhances the expanding LIAISON® bone and mineral product line.

Product availability is subject to required regulatory approvals.

Disclosures: A.L. Podgorski, DiaSorin Inc. 3.

This study received funding from: Diasorin Inc.

W213

A Comparison Between Gene Expression in Skeletal Muscle and Bone of Postmenopausal Osteoporotic Females and Their Controls. S. Reppe*¹, O. K. Olstad*², V. T. Gautvik*¹, S. Nygård*³, L. S. Nissen-Meyer*¹, P. I. Høvring*¹, R. Jemtland*⁴, K. M. Gautvik*¹. ¹Medical Biochemistry, University of Oslo, Oslo, Norway. ²Dept. of Clinical Chemistry, Ullevaal University Hospital, Oslo, Norway. ³Mathematics, University of Oslo, Oslo, Norway. ⁴Section of Endocrinology, Rikshospitalet University Hospital, Oslo, Norway.

It is well known that patients with osteoporosis also have reduced muscle strength and show increased tendency to falls, thereby increasing the fracture risk. There is a question whether the entire musculo-skeletal system is affected in postmenopausal osteoporosis or only the skeleton. Maximum bone density and muscle mass are both 60%–80% genetically determined. In this study we have assessed the global gene expression patterns in trans-iliac bone and in muscle biopsies from the same individuals from osteoporotic females (n=12,) and healthy controls (n=12), using total RNA (pooled two by two) and Affymetrix HG U133 Plus 2.0 arrays. The osteoporotic group had T-scores < −2.5 in vertebrae and hips and low impact fractures. The MAS 5.0 software was used to identify and filtrate out absent genes, while signal values were obtained using PLIER. Students t-test were used for significant differences (p<0.05). 310 mRNAs (genes) showed significantly changed expression between healthy and osteoporotic women in the bone biopsies while the corresponding number in muscle biopsies were 130 at the level of significance of p < 0.05. Using Ingenuity Pathway Analysis we found that the top canonical pathways involving more than 5 genes associated with the altered gene expression in bone were: calcium signaling (p<0.000), Cardiac β-adrenergic signaling (p<0.000) and actin cytoskeletal signaling (p < 0.001), represented by 12, 9 and 9 genes, respectively. In muscle no pathway with more than 3 genes was found to be affected in osteoporotic women as compared to healthy controls. The major physiological systems and functions affected in bone were “skeletal and muscular system development and function” and “tissue morphology” involving 62 and 42 genes, respectively, while the corresponding top systems in muscle were “hematological system development and function” and “Immune response” both with 4 genes. Thus, the osteoporotic phenotype is reflected by altered gene expression in bone. The present study also indicate abnormal gene expression in skeletal muscle, in support of the hypothesis of musculo-skeletal disorder in female osteoporosis.

The EU project LSHM-CT-2003-50294 is acknowledged for financial support.

Disclosures: S. Reppe, None.

W214

A Non-radioactive Immunoassay for Mouse PINP. J. P. Rissanen¹, M. I. Suominen¹, J. Morko¹, S. Rasi*², J. Risteli*³, J. M. Halleen¹. ¹Pharmatest Services Ltd, Turku, Finland. ²SBA-Sciences, Oulu, Finland. ³Department of Clinical Chemistry, University of Oulu, Oulu, Finland.

Serum procollagen I N-terminal propeptide (PINP) is a useful marker of bone formation in humans. We have reported the development and characterization of a non-radioactive immunoassay for rat PINP. Here we demonstrate that the assay can also be used for measuring PINP from mouse serum samples. We used the PINP immunoassay to study the use of PINP as a bone formation marker in a mouse model of glucocorticoid-induced osteoporosis. PINP was measured from mouse serum samples with a polyclonal PINP antiserum developed in rabbits using a synthetic peptide QEDIPEVS as antigen, corresponding to the first amino acids of the N-terminal propeptide of rat type I procollagen α1 chain. Twelve weeks old Balb/c male mice (n=8/group) were given either vehicle (control group) or 3, 10 or 30 mg/kg/d of prednisolone orally for 4 weeks. Serum osteocalcin and PINP were measured from fasting blood samples before the start of treatment, and at 2 and 4 weeks. For dynamic histomorphometry, tetracycline labelling was performed at day 10 and calcein labelling at day 2 before termination. Parameters of dynamic histomorphometry were determined from trabecular bone and cortical bone at the end of the study. Statistical analysis was performed with ANCOVA for follow-up measurements and either ANOVA or Kruskal-Wallis test for end-point measurements after checking the assumptions for normality and homogeneity of variances. Linear contrasts of means or Mann-Whitney test were utilized for pairwise comparisons. Parameters of dynamic histomorphometry were dramatically decreased by prednisolone. The decrease was dose-dependent in trabecular bone parameters, whereas maximal reduction was observed already with the lowest dose of prednisolone in cortical bone parameters. PINP values were dose-dependently decreased by prednisolone treatment at both 2 and 4 weeks. The values were decreased by more than 50% already with the lowest dose of prednisolone at 2 weeks. Similar, although not as dramatic changes were observed for serum osteocalcin values. These results demonstrate that PINP is a useful marker of bone formation in this mouse model of glucocorticoid-induced osteoporosis. Prednisolone changes PINP values in a similar pattern than it changes parameters of dynamic histomorphometry, particularly in trabecular bone.

Disclosures: J.P. Rissanen, None.

W215

Reproducibility of Biochemical Markers of Bone Turnover in Clinical Practice. A. L. Schafer*¹, E. Vittinghoff*², R. Ramachandran*³, R. Mahmoudi*⁴, D. C. Bauer¹. ¹Medicine, University of California, San Francisco, San Francisco, CA, USA. ²Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA, USA. ³Laboratory Medicine, University of California, San Francisco, San Francisco, CA, USA. ⁴Hospital Medicine, Kaiser Permanente, Oakland, CA, USA.

Recent investigation has shown that biochemical markers of bone turnover may be useful to predict fracture risk and to assess response to osteoporosis therapy. Although several bone turnover markers are FDA-approved, there is little information about reproducibility of marker measurements in clinical practice. This study aimed to determine the laboratory reproducibility of two markers: urine N-telopeptide (NTX) and serum bone-specific alkaline phosphatase (BSAP).

Fasting morning serum and urine were collected from 5 postmenopausal women (mean +/-SD age 65 +/- 6.3 years) not taking osteoporosis therapy. The serum and the urine were pooled, divided into identical aliquots, and stored at −80C. An identical specimen was sent to each of 6 high-volume US commercial labs on 5 dates over an 8-month period. On the fifth date, 5 identical specimens were sent to each lab. Fictional identifiers were used, and specimens were sent by the authors' institutional lab, so that the commercial labs were unaware of the investigation. Means, SDs, and coefficients of variation (CVs) with 95% confidence intervals (CIs) were calculated.

Longitudinal reproducibility was evaluated as one specimen was sent to each lab on each of 5 dates. For urine NTX (Table), CVs varied from 5.4% to 37.6%. Longitudinal variation was significantly higher for labs using the Osteomark assay (CV 40.2%, CI 26.6-86.9) than for those using the Vitros ECi assay (CV 7.0%, CI 5.3-10.4). For BSAP, CVs ranged from 3.1% (CI 1.9-9.1) for Esoterix to 23.6% (CI 13.9-77.2) for LabCorp. 

Table  .  

Within-run reproducibility was evaluated as each lab was sent 5 identical specimens on one date. For urine NTX, see Table. For BSAP, Esoterix produced 5 identical measurements, and CVs for the other labs ranged from 2.2% (CI 1.3-6.3) for Quest to 15.5% (CI 9.2-47.1) for LabCorp.

We conclude that the reproducibility of urine NTX and serum BSAP is highly variable at US commercial labs. Poor reproducibility is a barrier to the use of biochemical markers of bone turnover in clinical practice, particularly if clinicians do not consistently use the same assay and laboratory.

Disclosures: A.L. Schafer, None.

This study received funding from: Procter & Gamble Pharmaceuticals.

W216

Measurement of 25-OH Vitamin D₂ by the LIAISON® 25 OH Vitamin D TOTAL Assay. J. A. Schmidt, E. M. Frenzel*, M. A. Friedberg*, A. L. Podgorski*, J. S. Fenske*. Research & Development, DiaSorin Inc, Stillwater, MN, USA.

The role of vitamin D in bone and mineral metabolism was recognized from its first identification as a factor that could cure rickets. However, vitamin D is now recognized as a prohormone with multiple roles in maintaining optimal health. Vitamin D₃ (cholecalciferol) is the form of vitamin D that occurs naturally in humans, but vitamin D deficiency is commonly treated in the United States with vitamin D₂ (ergocalciferol). Fortified foods or supplements may contain either form. Thus, it is important that an assay for use world-wide measure both D₂ and D₃ forms of 25 OH vitamin D.

We compared the results obtained for 25 OH vitamin D in 110 serum samples by a validated liquid chromatography-tandem mass spectrometry (LC-TMS) method and by LIAISON® 25 OH Vitamin D TOTAL, a chemiluminescent immunoassay. LC-TMS gives separate results for 25OH D₂ and 25OH D₃, but there is no clinical utility in separate values; the LIAISON® assay gives a single result. Of 110 samples tested, 46 were found to contain detectable amounts of 25OH D₂. Actual values were 4-45 ng/mL 25OH D₂ and 6-53 ng/mL total 25OH D. Regression equations were similar:

LIAISON® = 0.83(LC-TMS) + 2.7; R = 0.87 for all samples; and

LIAISON® = 0.80(LC-TMS) + 3.6; R = 0.78 for samples containing 25OH D₂.

Furthermore, there was no trend to higher or lower bias with increasing 25OH D₂. These results demonstrate that the LIAISON® 25OH Vitamin D TOTAL assay measures both the D₃ and D₂ forms of the vitamin equally.

Product availability subject to required regulatory approvals.

Disclosures: J. A. Schmidt, DiaSorin Inc. 3.

This study received funding from: Diasorin. Inc.

W217

Development of a Bone Specific Alkaline Phosphatase Assay for the LIAISON® Analyzer. K. E. Paulsen, C. M. Klatt*, F. A. Blocki, P. J. Krohn*, G. T. Olson, M. A. Friedberg*, J. A. Schmidt. R&D, DiaSorin Inc, Stillwater, MN, USA.

Bone Specific Alkaline Phosphatase (BAP) is a serum marker for osteoblastic bone formation. Measurement of BAP is useful in diagnosing Paget's disease and osteoporosis and in monitoring the response to antiresorptive therapy in these patients, and for assessing bone turnover in patients with chronic kidney disease — mineral and bone disorders (CKD-MBD). BAP is also an emerging marker of bone formation within the context of reversing adynamic bone disease. We report here the development of a chemiluminescent immunoassay for BAP on the LIAISON® Analyzer.

Approximately 95% of total alkaline phosphatase activity in normal human serum is due to the presence of the liver and bone isoforms present in approximately equal quantities. BAP and Liver Alkaline Phosphatase (LAP), alternatively known as TNALP, tissue non-specific alkaline phosphatase, share a common primary sequence and differ structurally solely due to post-translational glycosylation. Antibody with specificity toward the carbohydrate portion of the BAP enzyme is needed in order to minimize cross reactivity to the LAP isoform in the assay.

The BAP assay uses 50 μL of serum sample mixed with paramagnetic particles coated with monoclonal antibody to BAP in a 20 minute incubation. A tracer of a second monoclonal antibody labeled with an isoluminol derivative is then added followed by a 15 minute incubation. Particles are washed and a chemiluminescent reaction produces a signal proportionate to the quantity of BAP in the sample.

The LIAISON® Bone Specific Alkaline Phosphatase assay measures in the range of 0.1 to 120 ng/mL with analytical sensitivity < 0.1 ng/mL, functional sensitivity < 0.4 ng/mL, and inter-assay imprecision ≤ 8%. The assay demonstrates good correlation to a predicate device, LIAISON® = 0.94x + 1.78 ng/mL, R = 0.90. Dilution Linearity yields a regression equation of y = 1.04 (expected) − 4.0, R = 0.98. Cross reactivity to the LAP isoform in this two-site mass based assay is comparable to or better than other methods which use phosphatase activity for detection.

The LIAISON® Bone Specific Alkaline Phosphatase assay is an accurate and precise assay, with minimal cross reactivity toward liver alkaline phosphatase, significantly enhancing DiaSorin's expanding bone and mineral product line. Product availability subject to required regulatory approvals.

Disclosures: J.A. Schmidt, DiaSorin Inc 3.

This study received funding from: Diasorin, Inc.

W218

Age-Related Changes in Bone Turnover in Men: Evidence for Impairments in Both Osteoblast and Osteoclast Function. B. Srinivasan, B. L. Riggs, E. J. Atkinson*, L. J. Melton III, S. Khosla. Mayo Clinic, Rochester, MN, USA.

While bone turnover generally increases after menopause and with aging in women, age-related changes in bone remodeling in men have been less clearly defined. Most studies have found decreases in biochemical markers of bone formation with age in men, consistent with an age-related defect in bone formation. Results with bone resorption markers have been more variable; some studies have found an increase and others report no clear changes. Serum CTx is a well established marker of bone resorption, reflecting the net effects of osteoclast number and activity on bone. By contrast, data from in vitro, animal, and human studies indicate that measurement of serum tartrate-resistant acid phosphatase (TRAP) 5b (which is specific for TRAP derived from osteoclasts) principally reflects osteoclast numbers. To dissociate possible age-related changes in osteoclast numbers versus activity, we measured serum CTx and TRAP 5b levels using highly sensitive and specific immunoassays in a large, population-based sample of men (n = 313) age 22 to 9] yrs. We also calculated the ratio of CTx:TRAP 5b as an index of resorptive activity per osteoclast. As a measure of bone formation, we assessed serum PINP levels. Panel A shows the age-related changes (mean ± SE, expressed as T-scores relative to men age 20–29 yrs) in serum PINP levels, demonstrating a 45% decrease in bone formation between age 20 to 90 yrs (P < 0.0001). Panel B shows the changes in serum CTx and TRAP 5b; as is evident, while serum CTx decreased over life by 40% (P < 0.001), serum TRAP 5b levels increased by 32% (P < 0.001), suggesting that despite an apparent agerelated increase in osteoclast numbers, net bone resorption decreased with age in men. Panel C shows the ratio of CTx: TRAP 5b, representing an index of activity per osteoclast, which decreased in men over life by 36% (P < 0.001).

These findings thus demonstrate that bone formation clearly decreases with age in men. While TRAP 5b (an index of osteoclast numbers) increases, CTx (an index of net bone resorption) decreases with age in men, consistent with an age-related defect in osteoclast function. Since there is increasing evidence that osteoclasts directly regulate osteoblastic activity, the observed deficit in osteoclast function with aging in men may contribute to the age-related impairment in bone formation in men.

Disclosures: B. Srinivasan, None.

This study received funding from: NIH/NIA — AR27065.

W219

Utilize of s-RANKL, s-FREE RANKL and Osteoprotegerin Serum Levels for Prediction of Bone Density. D. Stejskal*¹, M. Karpisek*², P. Solichova*¹, R. Ochmanova*³. ¹Department of Laboratory Medicine, Sternberk hospital, Sternberk, Czech Republic. ²Department of Toxicology, University of Veterinary and Pharmacology Sciences, Brno, Brno, Czech Republic. ³Department of Internal Medicine, Sternberk hospital, Sternberk, Czech Republic.

Background: RANKL and its inhibitor OPG are decisive for osteoclast differentiation and osteoresorption function so that they became the aim of an intensive research. Recently, there are some informations about importance of free RANKL for BMD prediction.

Aim: to determine OPG, s-RANKL, s-RANKL free concentrations in persons with various degree of reduced bone density and to find correlations between BMD and listed markers. Methods: We examined 69 patients who under the follow-up for osteopathy. All individuals were examined for OPG (Biovendor), s-RANKL (Biovendor), s-RANKL free (Immunodiagnostic), osteokalcin, b-ALP, DPD/creatinine index. We performed DXA of the skeleton. Patients were divided into the subgroup of persons with normal bone density (T score>−1 “N”) and individuals with reduced bone density (T score< −1, “OP”).

Results: 27 patients had normal bone density and 42 probands had osteopenia or osteoporosis. OPG: individuals “N” had median values 2,9 vs 3,7 pmol/1 in probands “OP” (p<0.05). S-RANKL-free: values didn't differ significantly in “OP” versus “N” patients (0.1 vs 0.09 pmol/1). S-RANKL: individuals “N” had median 311 vs 127 pmol/1 in probands “OP” (p<0.05). OPG strongly correlated with s-RANKL (r= −0.35; p< 0.01) and BMD (r= −0.4; p< 0.01). Index s-RANKL/OPG had sufficient efficacy for reduced BMD. Discrimination analysis revealed that the known s-RANKL/OPG index allows a correct prediction into given groups in 83% of cases.

Conclusions: The negative correlation between OPG and s-RANKL was detected. S-RANKL/OPG index discriminates probands with significant BMD decrease. S-RANKL-free cannot be used for mathematical assessment of bone density.

Disclosures: D. Stejskal, None.

W220

Clinical Significance of Pentsidine in Glucocorticoid-induced Osteoporosis (GIO). I. Tanaka¹, H. Oshima². ¹Department of Laboratory Medicine, Fujita Health University School of Medicine, Toyoake, Japan. ²Department of Internal Medicine, National Hospital Organization Tokyo Medical Center, Tokyo, Japan.

[Objective] In GIO, bone fractures are often seen even at high bone mineral density. This suggests bone quality is an important factor in GIO for bone strength. Recently, it is reported that Pentsidine (PEN) reflects bone quality. In this study, we tried to clarify the clinical usefulness of PEN in predicting the bone quality in GIO. [Subjects & Methods] Fifty nine patients (47 females and 12 males) with collagen diseases under glucocorticoid treatment were enrolled in this prospective study. An incidence of the vertebral compression fracture in two years after the start of glucocorticoid therapy was evaluated in 37 patients among those. The mean of age, daily glucocorticoid dosage (prednisolone equivalent), and total glucocorticoid dosage were 46 years old, 20.3mg/day, and 6.3g, respectively. NTX in the urine (uNTX) and Pentosidine (PEN) were measured by the ELISA methods. Vertebral compression fractures were evaluated with X-ray photography. [Results] 1 ) After the start of glucocorticoid therapy, PEN increased significantly. 2) Doses of glucocorticoids, were significantly correlated with PEN positively. 3) PEN were significantly correlated with uNTX positively. 4) PEN was lower in patients without incident fractures in 2 years than in those with the fractures. 5) Incident fractures in 2 years was high in patients over 180ug/mgCr. [Conclusion] 1) These results suggested that PEN was an important risk factor for incident fracture in patients with glucocorticoid-induced osteoporosis. 2) Fracture risk in patients over 180ug/mgCr of PEN was high.

Disclosures: I. Tanaka, None.

W221

Effect of Food Intake on Bone Resorption in Cynomolgus Monkeys: A Preliminary Study. F. Vlasseros*¹, A. Varella¹, S. Y. Smith¹, D. Henriksen². ¹Charles River Laboratories, Preclinical Services Montreal, Senneville, PQ, Canada. ²Sanos Bioscience A/S, Roedovne, Denmark.

The objective of this preliminary study was to monitor the effect of food intake on biochemical markers of bone turnover in young adult cynomologus monkeys. Food (Rhesus liquid diet) was provided via oral gavage in the morning (approximately 8 to 10 a.m.) following a minimum period of 12 hours fasting and blood samples collected (n=2 to 4 per time-point) prior to feeding and at 10, 15, 20, 30 and 60 minutes, and 4, 6 and 8 hours post feeding. Samples were analyzed for serum N-telopeptide, TRAP5b, osteocalcin and GLP-2. GLP-2 (glucagon-like-peptide-2) is a gut hormone released following food intake, and has been associated with significant decreases in bone resorption markers with no effect on bone formation markers in humans (Henriksen D.B, et al. 2003, JBMR 18 (12):2180–89).

Relative to baseline levels (pre-feeding), NTx showed postprandial decreases of approximately 52% and 25% at 4 and 6 hours, respectively. NTx levels showed no meaningful changes up to 60 minutes following food intake. At 8 hours post feeding NTx was similar to pre-prandial fasting levels. There were no consistent changes in TRAP5b or osteocalcin levels. Relative to baseline levels, GLP-2 plasma levels showed increases 20 to 30 minutes following food ingestion for all animals, although the magnitude of the response was variable and ranged from 20% to 3-fold.

These preliminary data suggest bone resorption activity as measured by serum NTx levels may be decreased 1 to 4 hours following food intake with values rising to fasting levels > 8 hours later. This may be related to transient increases in GLP-2 levels 20 to 30 minutes following food intake. A similar spatial and causal relationship has been reported in humans. Importantly, these data suggest measurement of bone resorption activity in samples collected 1 to 8 hours after feeding may be lower than at any other timepoints. Samples can be collected early morning following overnight fasting and later the same day more than 8 hours after feeding. These preliminary data warrant further investigation into the effects of food intake on biochemical markers of bone turnover and the possible role of gastrointestinal hormones in Cynomolgus monkeys.

Disclosures: F. Vlasseros, None.

W222

Specificity of Mouse and Rat Circulating Serological Osteocalcin Values Using a Novel Immunoassay. C. D. Wisherd*, N. E. Nasser*. Specialty Products Group (SPG), Quidel Corporation, San Diego, CA, USA.

Osteocalcin (OC) or BGP (bone gla protein) is a 5800 molecular weight extrahepatic vitamin K dependent protein produced by osteoblasts found exclusively in bone tissue. It contains three gamma-carboxyglutamic acid residues that are thought to be involved in calcium ion and hydroxyapatite binding and accounts for 10-20% of the noncollagenous protein in bone. While the in vivo function of osteocalcin is unknown, its affinity for bone mineral constituents implies a role in bone turnover.

Many current enzyme immunoassay (EIA) methods used to detect intact human osteocalcin (hOC) show negligible cross reactivity with mouse osteocalcin (mOC) and rat osteocalcin (rOC). This can be explained by current anti-human osteocalcin antibodies' inability to recognize mouse and rat proteins. Development was performed to establish an EIA with the ability to recognize these proteins while maintaining intact hOC detection. Monoclonal antibodies were developed to capture mOC and rOC in addition to hOC. EIA detection is achieved using hOC-coated plates competing with hOC calibrators and neat serum samples to capture a horseradish peroxidase (HRP) conjugated monoclonal antibody. Tetramethylbenzidine (TMB) substrate is subsequently used to detect the enzyme. The reaction is quenched with 2N H₂SO₄ and read at an absorbance wavelength of 450 nm. The total assay time is 45 minutes.

Specificity studies were run using purified mOC and rOC. mOC was prepared via extraction from mouse bone and purification via G75 gel filtration, DEAE 52 and HPLC. rOC was prepared with rat bone utilizing the mOC extraction method. Balb–C, C57BL6 and CD-I mouse sera were spiked with 6.7 ng/mL of mOC and tested in the EIA. Average spike recovery of mouse sera was 116% (n=13, range: 86–135%). Sprague-Dawley and Wistar rat sera were spiked with 34.3 ng/mL of rOC. Average spike recovery of rat sera was 109% (n=7, range: 102–115%).

The results of this study indicate that this novel osteocalcin EIA overcomes shortcomings of the previous methods including: limited species cross reactivity, additional secondary antibody steps, and provides improved sensitivity and short overall assay time.

Disclosures: CD. Wisherd, None.

W223

Comparison of alpha CTX Levels in Healthy Prepuberal Children, Adolescent, Pre And Postmenopausal Women. S. N. Zeni¹, G. G. Pellegrini*¹, M. Linari*², N. Piazza*², J. Somoza*¹, M. E. Rio*³. ¹Sectión Osteopatías Médicas, Hospital de Clinicas, Universidad de Buenos Aires, Buenos Aires, Argentina. ²Public Health Secretary of Vicente López, Epidemiology. Infantil Nutrition., Buenos Aires, Argentina. ³National Council of Scientific and Technical Research, School of Pharmacy and Biochemistry. UBA., Buenos Aires, Argentina.

Collagen type 1 fragments are formed by osteoclasts and same of them can be determined using an specific immunoassays. The CTX epitope EKAHDGGR comprises a DG-motif susceptible to post-translational modifications. In newly synthesised collagen this motif is in the α-CTX form, but during aging of bone it is isomerized to β-CTX form. The α-CTX form can be determined in urine using a new sandwich ELISA. The aim of the present study was to assess the ability of this marker to discriminate the degradation of new bone in different stage of aging. A total of 243 healthy subjects were included. The distribution of the subjects were: 173 prepuberal children (70 girls and 103 boys) and 30 adolescent (12 girls and 18 boys) attending public schools of Vicente Lopez, Buenos Aires and 20 premenopausal and 20 postmenopausal women). The α-CTX (Alpha CTX ELISA; Nordic Bioscience) was assessed in the second void urine. Results of α-CTX/creatinine levels (ug/mMcreatinine) as mean ± SD were:

Prepuberal children: 20±7.2 (6 to <7 years); 15.7±4.5 (7 to <8 years); 14.9±6.7 (8 to <9 years) and 15.3±6.3 (9 to <10 years), adolescents: 4.2±1.9 (11-12 years); premenopausal women: 0.61±0.26 (36.9±7.5) and postmenopausal women: 0.93±0.58 (58.0±5.3 years). There was a significant differences among all groups however the levels of significance between postmenopausal vs. premenopausal women (p<0.05) was the lowest. The present results suggest that α-CTX marker reflects well the increase of bone resorption associated with bone modeling at childhood and with high bone turnover after menopause.

Disclosures: S.N. Zeni, None.

This study received funding from: UBACyT, B703. SU Project.

W224

Hip Geometry and Density Parameters Derived from Volumetric DXA (VXA) Correlate Strongly with 3D QCT. Q. M. Ahmad*¹, K. Ramamurthi*², E. Thrall*³, D. Karasik³, M. Bouxsein³, K. E. Wilson², K. Engelke⁴, R. H. Taylor*¹. ¹CISST, Johns Hopkins University, Baltimore, MD, USA. ²Hologic, Inc., Bedford, MA, USA. ³Orthopaedic Biomechanics Laboratory, BIDMC, Boston, MA, USA. ⁴Institute of Medical Physics, University of Erlangen, Erlangen, Germany.

3D assessment of hip geometry and density using volumetric DXA (VXA) may improve assessment of skeletal fragility. In this study we tested the correlation between structural parameters and volumetric density of the proximal femur derived from VXA and those derived from 3D QCT.

Methods: Male and female statistical models of the proximal femur, capturing both the 3D geometry and density variation, were created using in vivo CT scans from 62 Caucasian men and 57 Caucasian women. Each statistical model comprises a mean-shape and a set of modes of variation. The mean-shape represents the average 3D geometry and density, and consists of a tetrahedral mesh with density functions for each tetrahedron in the mesh. The modes of variation are vectors, derived from principal component analysis on the set of femurs which comprise the statistical models. The modes of variation allow the meanshape to geometrically vary to match a particular subject's femur. A particular instance of the statistical model can be transformed into a voxel-volume and projections can be calculated to simulate DXA scans. To achieve a 3D model from DXA, 4 DXA scans are acquired at different projection angles (chosen to minimize pelvic overlap in vivo), and an iterative process is performed wherein the DXA projections are compared to projections of the statistical model and the modes of variation are varied to minimize the error between these simulated DXA projections of the model and the actual DXA images acquired. To test the fidelity of this approach, we performed 3D QCT (1 mm slice thickness, GE Lightspeed) and VXA in 20 human cadaveric femora (11 F, 9 M, age 62 to 95). Each resulting model was then voxelized and aligned to the QCT. We examined the association between VXA and QCT-derived femoral neck length (FNL), volumetric density of the total proximal femur (vBMD), as well as the area of bone mineral in the cross section (CSA_b), the mass-weighted polar moment of inertia (CSMI_p) computed at a corresponding VXA and QCT slice at the mid-neck, perpendicular to the neck axis.

Results: We found very strong linear correlation between VXA and QCT for all parameters: FNL (r=0.97), vBMD (r=0.96), CSA_b (r=0.94), and CSMI_p (r=0.96). In conclusion, reconstructions that contain both volumetric shape and density information can be made from four DXA views that can be obtained readily in vivo with limited radiation exposure. These VXA reconstructions are highly correlated with structural parameters and volumetric density measured with 3D QCT.

Disclosures: O.M. Ahmad, Hologic Inc. 2.

This study received funding from: Hologic, Inc.

W225

A Technique for Calculating Area Moment of Inertia of Long Bones Using Computed Tomography (CT) Data. A. K. Aiyangar*, H. Ploeg*. Mechanical Engineering, University of Wisconsin, Madison, Madison, WI, USA.

A technique to compute the Area Moment of Inertia (MI) of long bones is presented. In problems related to bending of beams and beam-like structures, there are two main parameters that are indicative of the resistance of the structure to bending: Young's Modulus (E), and MI. While Young's Modulus is a material property, MI is, essentially, a geometrical property of the long bone and depends on a reference axis and the distribution of the material around that axis at every cross-section along the bone. There is not yet a significant appreciation of the effect of ignoring the deviations from standard symmetric beam bending theories as applied to homogeneous materials, while interpreting data from bending tests. As a consequence, there is a strong requirement for a standard method to accurately compute the MI, taking into account the heterogeneities in the bone structure and irregular geometrical shape. Accurate calculation of MI for bone is further complicated as a composite materials model needs to be employed, wherein the different constituent materials are given weighting factors according to their respective elastic moduli. This can be accomplished by using the grayscale variations in computed tomography (CT) images, which is linearly dependent on the density distribution in the bone. The density distribution, in turn, correlates with the elastic modulus. Segmented CT data of a porcine femur bone was used for the study. A porcine femur bone was scanned in a CT scanner (GE Lightspeed. 120kV, 0.352mm pixel size, 0.625mm slice thickness, bone algorithm). The images from the scanning procedure were processed and segmented using Mimics 10.0 software (Materialise, Ann Arbor MI) to obtain the density distribution of the bone. The data for each cross-section was digitized and mapped in the form of pixels (0.125mm²). A custom-written MATLAB function using a composite materials model was used to compute the MI. The equations are shown below.

Where:

, = x and y coordinates of centroid respectively

Ixx, Iyy = MI about the x- and y- axes respectively

R = density ratio for each material

r = constant correlating density to the stiffness

gr = density given as hounsfield units

x_i, y_i = x and y coordinates respectively of pixel “i” in a given cross-sectional slice

The weighted and un-weighted moments of inertia (the difference between a homogeneous versus a composite materials model) differed by 20%. Details of the technique will be presented.

Disclosures: A.K. Aiyangar, None.

This study received funding from: Ali Seireg Memorial Fellowship.

W226

In Vivo Comparison of CT and DXA Methods of Proximal Femur Cross-sectional Geometry Measurement Using HSA. T. J. Beck¹, J. K. Brown², S. Gustafsson*³, K. Zhu³, I. Dick⁴, S. Hentzell*⁵, V. H. Low⁶, K. E. Wilson⁷, R. L. Prince³, ¹Radiology, Johns Hopkins University, Baltimore, MD, USA. ²Mindways Software Inc., Austin, TX, USA. ³Endocrinology and Diabetes, University of Western Australia, Perth, Australia. ⁴Medicine, University of Western Australia, Nedlands, Australia. ⁵Endocrinology and Diabetes, University of Western Australia, Baltimore, Australia. ⁶Radiology, Sir Charles Gairdner Hospital, Nedlands, Australia. ⁷Hologic Inc., Bedford, MA, USA.

The Hip Structure Analysis (HSA) method for measuring femur geometry from 2D DXA scans needs in vivo 3D validation, but methods for comparison with QCT are lacking. In this ongoing study 54 elderly post-menopausal women from the longitudinal CAIFOS/CARES study of ageing, had standard hip DXA (Hologic Discovery A) and CT scans (Phillips 64 slice) of the pelvis/hips. CT axial images were acquired with 0.78 mm pixels and 1 mm slices at 1 mm intervals. Results were analyzed with CTHSA, a collaborative development between Mindways Software Inc. and Johns Hopkins University. The software extracts cross-sections at DXA HSA femur locations: narrow neck (NN), intertrochanter (IT) and femur shaft (FS) and reformatted them to 0.78 mm pixel spacing. The method then measures the geometry for ultimate use in an automated 3D engineering analysis under development. For the present work, we compared CTHSA geometry and bending properties corresponding to frontal plane DXA with those from DXA HSA. In pooled comparisons of all three regions, bone cross sectional area, outer diameter, cross-sectional moment of inertia, section moduli and buckling ratio showed R² values of 0.91, 0.81, 0.95, 0.93 and 0.49, respectively. Within HSA regions, section moduli R² values for FN, IT and FS regions were 0.66, 0.67 and 0.47 respectively. Neck values corresponded to the line of identity but shaft and intertrochanteric values were shifted upward (Fig). Overall, results correlate well between frontal plane geometry measurements measured by DXA and QCT in elderly women and with prior ex vivo studies. These preliminary results show a small dimensional error in CTHSA integrals which appear to explain the shift and should be correctable. Although work remains, results show that CTHSA may provide an in vivo geometry gold standard for DXA as well as a stand-alone method for automated engineering analysis of the hip for use with recent generation multislice CT scan data.

Disclosures: T.J. Beck, Hologic Inc. 7.

This study received funding from: Hologic Inc.

W227

Automated Assessment of Vertebral Shape by Statistical Shape Modelling on Lateral Radiogrpahs. A. D. Brett*¹, J. Krasnow*², C. Miller³, J. Haslam*¹, T. Ozanian*¹, C. van Kuijk*⁴, K. Abrams*⁵, R. Cope*³, M. Beke*⁵, C. Hayes*⁶. ¹Optasia Medical Ltd, Manchester, United Kingdom. ²Roche Pharmaceuticals, Nutley, NJ, USA. ³Bio-Imaging Technologies Inc, Newtown, PA, USA. ⁴VU University Medical Center, Amsterdam, The Netherlands. ⁵Novartis Pharmaceuticals, East Hanover, NJ, USA. ⁶VCU Medical Center, Medical College of Virginia, VA, USA.

We describe the construction methodology, accuracy and reproducibility of a statistical model for the automated annotation and characterisation of vertebral shape from lateral radiographs.

Prevalent vertebral fractures are important predictors of future osteoporotic fractures in the spine and hip. Accurate computerized vertebral fracture detection and classification may benefit from capture of vertebral shape information beyond standard 6-point morphometry, similar to the visual cues that characterize semi-quantitative vertebral assessment. Accurate and reproducible acquisition of vertebral shape may also enhance the prediction of subsequent osteoporotic fractures.

Using a statistical learning technique, a vertebral model that can be altered in its shape and appearance was constructed from lateral radiographs. The model was trained on a set of manually-annotated radiographs (165 subjects from the Canadian Multicentre Osteoporosis Study [CaMos], enriched for prevalent deformities). Each vertebra, from T4-L4, was described using 95 points representing the circumferential vertebral borders, including right/left/central endplate margins, anterior/posterior margins, and osteophytes when present. Radiographs of 100 randomly chosen subjects were used to build the model; the remaining 65 subjects were used for accuracy and reproducibility testing. The resulting model may be used in the annotation of a previously unseen image; vertebral landmarks are searched for automatically after manual initialisation.

Accuracy was assessed by measuring the mean absolute distance between manually placed vertebral contours and automatically fitted points on the 65 test subjects. 79895 points were assessed on 841 individual vertebrae. The mean accuracy calculated over each vertebra in each test image was 1.11 +/- 0.94 mm. Inter-operator reproducibility was simulated by varying the placement of model initialisation points by adding random offsets of up to 4 mm and running 10 searches on each test image. The observed mean standard deviation per vertebra of the resulting annotations is 0.31 +/- 0.24 mm.

These results indicate that statistical modelling can provide a robust tool for the accurate and reproducible automated annotation of vertebral body shape. This method may prove useful as a workflow tool to aid the physician in vertebral fracture assessment.

Disclosures: A.D. Brett, None.

This study received funding from: Novartis Pharmaceuticals.

W228

HR-pQCT Assessment of Cortical Bone: Correlations to FE Derived Mechanical Properties. A. J. Burghardt, K. Davis*, T. M. Link, S. Majumdar. Radiology, University of California, San Francisco, San Francisco, CA, USA.

High-resolution peripheral quantitative computed tomography (HR-pQCT) is a promising new tool for longitudinal evaluation of bone quality. Cortical bone geometry is known to be an important factor in whole bone strength. Various methods, which rely on specific assumptions of cortical structure, exist to quantify cortical thickness. In this study, two common definitions of cortical thickness were evaluated using correlations to micro Finite Element (μFE) derived mechanical properties as a measure of significance. The distal radius and tibia (n=28 and n=24, respectively) from normal volunteers (age 23-76) were imaged using HR-pQCT (Scanco XtremeCT, 82 μm isotropic nominal resolution). The cortical bone compartment was segmented using the standard clinical routines provided by the manufacturer; a strong Gaussian blurring filter (σ=1.5, kemel=7) to wash out the trabecular structure, followed by a fixed global threshold to define the cortical bone phase. Ct.Th was calculated in 2 ways: 1) as the ratio of the cortical area to the perimeter of a semi-automatically defined periosteal contour, and 2) using a direct 3D sphere filling technique. As a standard of reference, a 1% uniaxial compressive strain (in the direction of the long axis of the radius) was simulated in the complete structure using a voxel based finite element technique. The total reaction force and the estimated failure load were determined [1].

Visually, the use of a large blurring filter clearly eliminated thinner cortical components, particularly in the ultra-distal radius. The area-to-perimeter estimate of cortical thickness included these regions as segments with zero thickness. In contrast, the direct 3D measure only considered the cortical volume that survives the segmentation step. No significant difference in tibial cortical thickness was found between methods. In the radius, the direct 3D method yielded thickness values that, on average, were 0.2 mm larger than the area-to-perimeter method (p < 0.0001). Both methods were positively correlated to μFE derived reaction force (R²=0.72 and R²=0.48 respectively) and to the estimated failure load (R²=0.70 and R²=0.46 respectively).

In conclusion, the results suggest that there is significant site dependence in the agreement between different definitions of cortical thickness in limited resolution scenarios. The μFE results indicate that, for HR-pQCT, the area-to-perimeter definition is a better predictor of mechanical competence, likely due to its inherent incorporation of thin segments lost during the segmentation process.

[1] Pistoia, W. Bone 2002

Disclosures: A.J. Burghardt, None.

W229

Topological Analysis and Spatial Distribution of Apparent Trabecular Bone in Radiographs Correlate to Biomechanical Bone Properties: An In Vitro Study of the Proximal Femur. J. Carballido-Gamio*¹, M. B. Huber*¹, K. Fritscher*², R. Schubert*², M. Haenni*³, C. Hengg*⁴, S. Majumdar¹, K. Link¹. ¹University of California, San Francisco, San Francisco, CA, USA. ²University of Health Sciences, Medical Informatics and Technology, Innsbruck, Austria. ³AO Development Institute, Davos Platz, Switzerland. ⁴Medical University Innsbruck, Innsbruck, Austria.

Analysis of the trabecular bone (Tb) pattern observed in radiographs has recently received more attention as significant correlations to different biomechanical bone properties have been found. Most of the analyses however have been based on Fourier, fractal, or morphological approaches trying to characterize texture.

The purpose of this study was therefore to develop an image analysis technique to characterize the apparent (App) Tb architecture observed in radiographs and evaluate it based on biomechanical bone properties.

Radiographs of 13 specimens of the proximal femur were obtained and digitized. Bone mineral density values (BMD) were measured at the femoral neck, and biomechanical testing was performed based on failure load (FL). Five regions of interest (ROI) were automatically positioned in each specimen (head, upper and lower neck, and trochanteric and inter-trochanteric compartment). Regions were normalized for contrast variations and an App Tb map was computed automatically by using soft fuzzy c-means clustering. The skeletons of these maps were computed and their junctions and terminations automatically identified. Voronoi diagrams were constructed using the junctions as nodes, and 1st and 2nd order moments of area of the cell polygons, as well as the distribution of App Tb based on the Voronoi cells were computed and correlated to FL.

Significant correlations were found for the lower neck and inter-trochanteric ROI for different moments of area and App Tb distributions. Correlation values to FL went from 0.577 (p<0.039) for the mean area of the polygons normalized by the mean distance of the nodes of the polygons to their corresponding centroids, up to −0.866 (p<0.0001) for the standard deviation of the distributions of App Tb normalized in a similar manner. When significant correlations were found to FL, no significant correlations were found to BMD. However, the mean elongations of the cell polygons of the upper neck ROI showed significant correlations to BMD (r = −0.568; p<0.042). Similar analysis was performed for the terminations of the skeletons yielding additional significant correlations. The correlation between BMD and FL was 0.695 (p<0.008).

Results suggest that the topology of the App Tb architecture and the distribution of the App Tb observed in radiographs can be quantified and that they are closely related to the biomechanical properties of the Tb in the proximal femur.

Disclosures: J. Carballido-Gamio, None.

This study received funding from: AO Foundation.

W230

Influence of A1330V LRP5 Gene Polymorphism on Volumetric BMD and Structural Parameters of Bone in Men and Women. C. Cepollaro¹, F. Lauretani*², A. Gozzini*¹, L. Masi¹, A. Falchetti¹, F. Del Monte*¹, S. Carbonell Sala*¹, G. Leoncini*¹, A. Tanini*¹, A. Corsi*², S. Bandinelli*³, M. Brandi¹. ¹Department of Internal Medicine, University of Florence, Florence, Italy. ²Tuscany Health Regional Agency, Florence, Italy. ³Azienda Sanitaria di Firenze, Geriatric Rehabilitation, Florence, Italy.

Hereditability studies show that genetic factors may contribute to the variability in bone mineral density (BMD), metabolism and microarchitectural deterioration leading to bone fragility. Several candidate genes, identified since their mutations cause severe Mendelian bone phenotypes, exhibit nucleotide sequence variants with modest effects on bone structure and remodeling. Mutations in the LRP5 gene have been associated with rare inherited syndromes characterized by extremely low (loss of function) or high (gain of function) BMD; but little is known about the contribution of this gene to the development of osteoporosis and determination of BMD in a normal population. A recent large population-based study showed that distinct genetic polymorphisms of LRP5 are associated to BMD and to fracture risk namely in elderly men. No data have been published on the relationships between LRP5 polymorphisms and structural parameters of bone as assessed by pQCT. The aim of the present study was to investigate the possible association of Alal330Val polymorphism in LRP5 gene with volumetric BMD and structural parameters of bone in men and women.

We studied 959 subjects (451 men and 508 women), participating to the InCHIANTI study. In all subjects we performed pQCT (XCT 2000, Stratec, Germany) at the tibia level obtaining the follow parameters: trabecular vBMD (vBMDt, mg/cm³), cortical vBMD (vBMDc, mg/cm³), cortical bone area (tCSA mm²) and cortical thickness (Ct.Th, mm). Ala1330Val genotypes were determined by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP).

In men and women, the LRP5 1330-valine variant was associated with decreased values of all pQCT parameters, reaching the statistical significance (p<0.05) for vBMDt in men and in women and for Ct.Th. and tCSA in women. These results show that Ala1330Val polymorphism may contribute to the determination of BMD and geometric parameters of bone in the male and female populations.

Disclosures: C. Cepollaro, None.

W231

Texture Analysis and Geometry Measurements on Plain Radiographic Femurs: Correlation with Ultimate Load. C. Chappard¹, V. Bousson*², C. Bergot*³, A. Marchadier*¹, T. Moser*³, D. Mitton*⁴, C. Benhamou¹, J. Larédo*³. ¹U 658 Inserm, Inserm, Orleans, France. ²Laboratoire Radiologie Expérimentale, Université Paris VII, Paris, France. ³Laboratoire Radiologie Expérimentale, Université Paris VII, Paris, France. ⁴Laboratoire Biomécanique UMR 8005, ENSAM CNRS, Paris, France.

Osteoporosis diagnosis and fracture risk at the upper femur are usually assessed by Bone Mineral Density (BMD) measurements with Dual X-ray Absorptiometry (DXA). We propose a new method combining texture analysis and geometrical measurements, obtained from plain femur radiographs. Forty pairs of excised femurs were obtained in 18 males and 22 females (mean age: 81.5±12.3 years). Total femur BMD was measured by DXA. Film radiographs (GE Prestilix 1600x) were digitized with a Fonction Transfert Modulation of 166 μm. We selected 3 square Regions of Interest (ROIs) of 3.2*3.2cm in Femoral Neck (FN), Greater Trochanter (GT) and in Inter-Trochanteric regions (IntT) and finally a ROI of 3.6*3.6 cm in the Femoral Head (FH). We calculated texture parameters derived from the cooccurrence matrix such as Correlation (COR), Contrast (CON), Entropy (ENT), Homogeneity (HOM), Dissymmetry (DIS), Inverse of Differential Moment (IDM), Angular Second Moment (ASM) and Maximum (MAX). Geometrical parameters were directly measured on the radiographs: the total length of femoral neck (FNL), from the base of the greater trochanter to the medial limit of the femoral head, and the neck-shaft angle. All femurs were randomly assigned for axial or lateral compression testing to measure ultimate load (N), axial compression mimicking femoral neck fracture and lateral compression pertrochanteric fracture. All textural parameters were significantly correlated with total femur BMD at FH and GT and few of them were marginally significant at FN and intT. After checking Gaussian distribution, we combined textural and geometrical parameters to obtain the best fit with ultimate load. The best results were obtained with 3 parameters, more parameters in the regression calculation did not improve the results. 

Table  .  

In term of prediction of ultimate load, the combination of textural parameters with geometry was close to BMD alone. Except for Femoral head ROI, the addition of FNL measurements in the model substantially improved the prediction of ultimate load.

These results show the potential of measurements on plain radiographs to predict bone strength at the upper femur.

Disclosures: C. Chappard, None.

W232

Comparison of Femur Structure Measurements Derived from DXA and QCT. X. Cheng*¹, H. Barden², J. K. Brown³, O. Zhou⁴. ¹Radiology Department, Beijing University Beijing Jishuitan Hospital, Beijing, China. ²GE Healthcare, Madison, WI, USA. ³Mindways Software Inc., Austin, TX, USA. ⁴GE Healthcare Asia, Shanghai, China.

The strength of the proximal femur is a function of both bone mineral density (BMD) and spatial distribution of bone mass determined from structural measurements such as cross-sectional moment of inertia (CSMI), cross-sectional area (CSA) and hip axis length (HAL). These measurements, which assist in estimating hip fracture risk, are made by both dual-energy X-ray absorptiometry (DXA) and quantitative computed tomography (QCT). Because of the ability to control the projection plane, three-dimensional measurement of femur geometry with QCT can be used for evaluating accuracy of other techniques. We compared CSMI, CSA and HAL measured with DXA (Prodigy, GE Healthcare, software version 10.5) and CT (Toshiba 64-slice, 1.0 mm slices).

We measured the left femur in 14 healthy subjects, aged 35 to 78 years (64.7 ± 3.01 yrs) with total femur DXA BMD from 0.729 to 1.179 g/cm² (0.921 ± 0.126 g/cm²). The DXA automatically measures CSMI (mm⁴) as the minimum CSMI within the neck region of interest (ROI), and CSA of the minimum CSMI section within the neck ROI. These DXA measurements usually occur in the section with the minimal neck diameter. A commercial QCT application (QCT PRO version 4.1, Mindways Software Inc.) was used to generate CSMI, CSA and HAL values from the QCT data. The QCT cross-sectional image at the narrowest portion of the femoral neck that most closely approximated the DXA CSMI measurement site was chosen for comparison purposes. QCT values were adjusted by the ratio of the DXA/QCT reference for average physical density of bone (1.85/1.05 g/cm³). Results showed very high correlations for CSMI (r=0.99), CSA (r=0.98), and HAL (r=0.86), with regression slopes not significantly different from identity. We also found high correlations (r∼0.95) between DXA and QCT BMD values at the hip. We conclude that DXA structural measurements at the proximal femur had very high correlation and agreement with similar QCT measurements. 

Table  .  

Disclosures: X. Cheng, None.

W233

Trends in Femur Structural Strength with Aging in Canadian Adults: Canadian Multicentre Osteoporosis Study. K. S. Davison¹, T. J. Beck², J. D. Adachi³, D. Goltzman⁴, J. P. Brown¹. ¹U. of Laval, Quebec, PQ, Canada. ²Johns Hopkins U., Baltimore, MD, USA. ³McMaster U., Hamilton, ON, Canada. ⁴McGill U., Montreal, PQ, Canada.

This study details changes in hip structural strength parameters during aging in a large sample of randomly-selected community-based individuals from the Canadian Multicentre Osteoporosis Study (CaMOS). Hip Structural Analysis (HSA) was performed on all CaMOS baseline DXA hip scans at sites with HSA calibration data. A total of 5334 scans, 3724 women, and 1610 men were used. At the narrow neck (NN), inter-trochanter (IT), and shaft (FS) regions, BMD, outer diameter (OD), bone cross-sectional area (CSA), section modulus (SM) and estimates of cortical thickness (CT) and buckling ratio (BR) were computed. All eligible CaMOS participants were separated by sex and grouped according to age: <30y, 30-39 y, 40-49y, 50-59y, 60-69y, 70-79y, and 80+y. Percent differences relative to the <30y group were used to approximate the age trend. The mean (SD, range) age was 61.0y (12.44, 25-92) for women and 57.6y (14.30, 25-90) for men. Structural and strength values provided in parentheses below represent the range in the parameter between men and women from the 80+y age group, unless otherwise noted. At the NN structural trends were similar in men and women with small almost linear increases in OD (3-9%) and large decreases in CSA and CW (21-28%). At the NN there were large near-linear decreases in SM (12-18%) and exponential increases in BR (44-59%) with aging in both sexes. The structural changes with aging at the IT region were similar to the NN for both sexes. The IT region showed an exponential increase in BR for both sexes (30-50%) while the SM increased moderately at the 60-69y group (6-12%) and then slowly decreased in the oldest group of women. At the FS, there were small increases in OD (8-9%), moderate decreases in CSA (4-13%), and large decreases in CT (13-24%) with aging in both sexes. The FS strength changes were similar to those of the IT region. The BR for the FS region increased exponentially with aging (31-49%). In general, with age there was a slow transformation toward a wider, thin-walled bone. With regards to strength, there were relatively steady losses of CSMI and SM at the NN with aging after 40-49y; however, at the other regions there were small to moderate increases in CSMI and SM during a large portion of adulthood and generally small losses later in life resulting in a preservation of strength in comparison with the reference group. For all regions and for both sexes there was an exponential increase in BR with age which may suggest this as being one of the mechanisms of fracture in old age.

Disclosures: K.S. Davison, None.

This study received funding from: CIHR.

W234

Association Between DXA-Assessed Muscle-Bone Proportionality and Fractures in Pre- and Post-Menopausal Women. R. Capozza*¹, C. Cure Cure*², G. Cointry*¹, M. Meta³, P. Cure Ramirez*², L. Plantalech⁴, J. Rittweger*⁵, J. L. Ferretti¹. ¹Centro de Estudios de Metabolismo Fosfocálcico, Faculty of Medicine, UNR, Rosario, Argentina. ²Universidad de Barranquilla, Barranquilla, Colombia. ³UCSF, San Francisco, CA, USA. ⁴Hospital Italiano, Buenos Aires, Argentina. ⁵Manchester Metropolitan University, Manchester, United Kingdom.

This study aimed to evaluate the DXA-assessed bone mass (BMC) / muscle mass (lean mass, LM) relationship as an indicator of nonmechanical (i.e. “systemic”) disturbances of the bone-muscle relationships in women in relation with the presence of fractures in different sites. Graphs of the BMC(y)/LM(x) proportionality showing CIs for ± 1, 2, and 3 SDs from the regression line were obtained from the whole body and lower limbs (HB,LL) of 1,035 pre-MP and 1,556 post-MP healthy women as normal references (No-Fx group). Other 614 pre-MP and post-MP women with fractures in “osteoporotic” sites (hip, spine, long-bone metaphyses; Type-II Fx, n=386) or in other sites (Type-I Fx, n=228) were studied the same way. Individual SD-scores of the BMC/LM relationship (BMC/LM SD-scores) were calculated for all pre- and post-MP women as per the determined CIs of the regression curves of the healthy controls.

The BMC/LM SD-scores of all Type-I-Fx women and of the pre-MP women with Type-II Fx were similar to those of their normal references, but lower than that in the post-MP women with Type-II Fx, especially in those with hip fractures. While the BMC-LBM curves were linear and similar in all the other groups, Type-II-Fx post-MP women showed nonlinear relationships, with progressively decreasing BMC and BMC/LBM SD-score values as their LBM decreased. Results were similar in WB and LL. Age and body weight and height were discarded as significant sources of variation for the studied relationships. Results show that the BMC/LM relationship may predispose somehow to Type-II Fx in post-MP women. They also suggest that both LBM and BMC/LBM SD-score determinations, either in the WB or in the LL, can help to differentiate between osteopenias/osteoporoses that ought to receive different treatments. Low BMC/LM SD-scores would define “metabolic” osteopenias which should be treated pharmacologically. Normal BMC/LM SD-scores would indicate a “mechanical” nature of the osteopenia for which only a physical or perhaps no other treatment should be indicated. The SD-scores would allow also to optimize monitoring following biomechanical criteria at low cost

Disclosures: J.L. Ferretti, None.

This study received funding from: FONCyT (Argentina).

W235

Reference Graphs of Age-Related Changes in Bone Mass, Volumetric Density, Design and Strength and of Muscle-Bone Interactions in Normal Men and Women. S. Feldman*, R. F. Capozza*, G. R. Cointry*, S. E. Ferretti*, P. S. Reina*, B. Homse*, M. Zapata*, J. L. Mansur, J. L. Ferretti. Centro de Estudios de Metabolismo Fosfocálcico, Faculty of Medicine, UNR, Rosario, Argentina.

Tomograhic scans of forearms and legs (pQCT, 6 different sites) of normal, unfractured men and pre- and post-MP women (n = 60, 80, 120) aged 25-85 yr were performed to describe the variations and interrelationships of indicators of bone “mass” (BMC, cortical area, trabecular vBMD), bone material “quality” (cortical vBMD, vCtD), and bone design (moments of inertia, MIs), and of muscle strength (muscle cross-sectional area, mCSA).

Three different patterns of variation were shown: 1. Bone “mass” indicators were higher in males than females and decaed slightly after MP. 2. Bone material “quality” (as assessed by vCtD) was higher in pre-MP women than men and decayed dramatically after MP. 3. Bone design was better in men than women but did not decay after MP. Calculated Bone Strength Indices (BSIs = Mis. vCtD) varied similarly to bone mass indicators. Bone mass, design and strength indicators (not so the vCtD) were linearly correlated with mCSA in all 3 groups, with parallel slopes for men and pre-MP women and lower slopes for post-MP women. The SD-scoring of the relationships between bone mass, design or strength indicators and the mCSA for men and pre-MP women allowed calculation of individual SD-scores of every relationship for post-MP women. The SD-scores for the relationships between bone mass and strength (not MIs or vCtD) and mCSA correlated negatively with time since MP (TSMP). Results show that women would tend conveniently to maintain bone design after MP, with an independent variation of bone mass, material quality, design and strenght. From a practical point of view, 1. the SD-scored charts of the correlations between the different bone indicators, and of these with TSMP, provided normal references suitable for evaluation of the two true determinants of “bone quality”, namely, bone material quality and design, and 2. the reference charts of the bone-muscle relationships allowed a non-invasive distinction between “disuse” and “systemic” osteopenias (with normal or reduced bone-muscle SD-scores, respectively) requiring substantially different treatments

Disclosures: J.L. Ferretti, None.

This study received funding from: FONOCyT.

W236

Relationships Between Diaphyseal Geometry and Cortical Mineral Density as Assessed by pQCT in the Human Tibia. R. Capozza*¹, P. Reina*¹, G. Cointry*¹, S. Feldman*¹, S. Castellini*¹, J. Rittweger*², J. L. Ferretti¹. ¹Centro de Estudios de Metabolismo Fosfocálcico, Faculty of Medicine, UNR, Rosario, Argentina. ²Manchester Metropolitan University, Manchester, United Kingdom.

Indicators of tibial cortical bone “mass” (cortical BMC, CtC, g), mineral density (cortical vBMD adjusted for the partial-volume effect, Rho-vCtD, g/cm³) and design (cross-sectional oments of inertia, CSMIs, mm⁴) were determined in calf pQCT scans of normal men, and pre- and post-MP women (n = 60, 80, 120) aged 25-85 yr to evaluate the mechanical efficiency of the distribution of the available compact tissue as a function of its intrinsic properties.

The CSMIs (y) and Rho-vCtD (x) values correlated negatively (“distribution/quality”, d/q hypérboles) showing high-CSMI / low-Rho-vCtD values for men; low-CSMI / high-Rho-vCtD values for pre-MP women, and low-CSMI / low-Rho-vCtD values for post-MP women. Parallel, linear correlations between the CSMIs and the CtCs of the same scans from every group (“distribution/mass” curves) were observed, with a common slope (b = 11,796 mm4/gm.cm-1, p<0.001). Statistical adjustment of all CSMI values (Adj-CSMIs) to a common CtC value (3.5 mg/cm) according to that slope allowed expression of the mechanical ability of the diaphyseal design (CSMIs) achieved as a funciton of the available mineralized mass in every individual. Re-built d/q graphs after substituting Adj-CSMIs for the raw CSMI values showed a substantially reduced but still significant difference between curves for men and pre-MP women, and similar curves for pre- and post-MP women.

Results suggest that cortical density has an effect upon the geometrical design of the tibia diaphysis. As far as the reduced density may reflect a reduced Young's modulus of the tissue, the d/q curves can be regarded as expressing the bone ability to self-distribute the available cortical tissue (as assessed by the CSMIs) as a function of its intrinsic stiffness (as approached by its Rho-vCtD) according to the “Mechanostat” Theory. The gender-related differences between curves can be attributed to sex-hormone influences on that system.

Disclosures: J.L. Ferretti, None.

This study received funding from: FONCyT (Argentina).

W237

Trabecular Bone Mass Evaluation as Related to Cortical Mass in the Human Leg: A pQCT Study. G. R. Cointry*, R. F. Capozza*, S. Feldman*, P. Mortarino*, L. Maffei*, A. Alvarisqueta, J. L. Ferretti. Centro de Estudios de Metabolismo Fosfocálcico, Faculty of Medicine, UNR, Rosario, Argentina.

In uniaxial compression, the relevant factors to bone strength are the amount and material properties rather than the distribution of bone mass, and the nature of bone structure.

The distal 38% of the tibial length undergoes chiefly uniaxial compression. Toward the mid-diaphysis, an increasing proportion of bending and torsion is combined. At 4% of bone length from the heel, tibial structure is mostly trabecular. At 14% distal, almost exclusively cortical, the diameter is minimal and the section is closely rounded, with a fairly homogeneous cortical thickness. At 38% distal the shape departs from circularity. We have determined the total BMC (TC) by pQCT at the 4%, 14% and 38% sites in the leg in normal men and pre- and post-MP women (n = 60,80,120) in order to determine 1. the proportion between total bone mass at the 4%, 14% and 38% sites in the different groups; 2. the functions describing those relationships[J3], and 3. whether gender and estrogen status have any impact on these associations.

The TC at the 4% site (y) correlated linearly with that at the 14% site in all men, pre-MP and post-MP women, either separately or taken together (whole group, y = −0.45 + 1.54 x; r=0.938, R2=0.881, p<0.001, SEE=0.25) and at the 38% site (whole group, y = −0.31 + 1.08 x, r=0.903, R2=0.815, p<0.001, SEE=0.33) with similar intercepts for each group. Results indicate that 1. the mechanical efficiency of the combined bone structure at the 4% site in compression would approximate 66% of that of the cortical structure at the 14% site, and 2. the mechancial solicitations of cortical structure at the 38% site would require the same amount of cortical mass than that of the combined structure at the 4% site, regardless of gender and reproductive status. SD-scored graphs of both relationships studied might provide normal references of the biomechanical proportionality between trabecular and cortical masses. As far as trabecular bone mass is usually more sensitive than cortical mass to metabolic changes, this may provide a diagnosis of metabolic bone diseases as based on data from the same individuals, obviating any comparison with references taken from different populations.

Disclosures: J.L. Ferretti, None.

This study received funding from: FONCyT (Argentina)

W238

Evaluation of Bone Microarchitecture with High Resolution-pQCT in Patients with Thalassemia. L. M. Frieling*¹, R. Grosse*², E. B. Fung*³, R. Fischer*², H. Kruse*¹, G. E. Janka*². ¹Osteoporosis Center, Hamburg, Germany. ²Pediatric Hematology, University Medical Center, Hamburg, Germany. ³Children's Hospital & Research Center, Oakland, CA, USA.

Thalassemia is a genetic disorder of hemoglobin synthesis. Due to improved blood transfusion and chelation therapy, survival has been increased with the consequence of complications like osteoporosis not seen during childhood and adolescence. The obvious shortcomings of conventional BMD methods like dual energy x-ray absorptiometry (DXA), can be overcome by simultaneously assessing the microarchitecture of the bone using high-resolution peripheral quantitative computed tomography (HR-pQCT), which may improve the estimation of the fracture risk in patients with thalassemia. In 17 regularly transfused patients (age: 13–43 y, 9/17 female) with beta-thalassemia major (n = 10), -intermedia (n = 6), and CDA-II (n = 1), the BMD of lumbar spine (LS) and total hip was measured by DXA (Hologic QDR1000+, Bedford, USA). Age, gender and ethnic specific BMD Z-scores were calculated. In addition, we assessed the volumetric BMD and the trabecular architecture of the non-dominant distal radius and tibia by HR-pQCT (XtremeCT®, SCANCO Medical AG, Bassersdorf, Schweiz). Liver iron concentration and endocrinological parameters were also determined. In 15/17 patients low BMD values (LS Z-score range: −1.1 to −3.1) measured by DXA were significantly correlated with total volumetric density (range: 91–388 mg/cm³, p = 0.002) measured by HR-pQCT at the distal radius. In 6/17 patients (> 28 y), all with latent hypogonadism, the spongiosa was porous or nearly dissolved. Patients with hypogonadism (n = 9) were significantly different from normals with respect to radial trabecular inhomogeneity parameter TbSp SD (p = 0.02), but not to LS Z-score. Patients with fractures (n = 5) had lower total densities (p = 0.02) and trabecular TbSp SD (p = 0.02) at the tibia and started blood transfusions at a higher age (p = 0.023). However, Z-scores did not reflect the fracture risk in this patient group (p = 0.11). Liver iron was mainly correlated with tibial TbSp SD (Rs = 0.54, p = 0.025).

In patients with thalassemia BMD Z-scores seem to underestimate fracture risk because a normal cortical thickness and density may conceal a porous trabecular structure. Endocrinological failures, especially hypogonadism, were responsible for the pathological microarchitecture of distal radius and tibia, while bone marrow expansion as in thalassemia intermedia and liver iron concentration seem to play a minor role. These initial results from bone microarchitecture measurements in thalassemia should be confirmed in a larger sample of patients with greater age range.

Disclosures: I.M. Frieling, None.

W239

Trabecular Microarchitecture Assessed by MicroMRI in Postmenopausal Women with Osteoporosis on Therapy. P. Greeley*¹, J. M. Wagner*¹, P. Seaman*², S. Perera*³, B. R. Gomberg*², O. Ganel*², M. Kleerekoper², S. L. Greenspan¹. ¹Medicine, University of Pittsburgh, Pittsburgh, PA, USA. ²MicroMRI Inc., Philadelphia, PA, USA. ³Medicine/ Biostatistics, University of Pittsburgh, Pittsburgh, PA, USA.

Although bone mineral density (BMD) is the gold standard to diagnose patients with osteoporosis prior to fracture, and to follow response to therapy, changes in BMD may only account for 30% of the reduction in fracture risk following therapy. Other factors such as trabecular micro-architecture contribute to fracture risk. A novel, noninvasive technique based on high resolution MRI (MicroMRI), examines trabecular microstructure and provides an index of the trabecular rods and plates. In order to examine the associations between standard measures of BMD by DXA at the wrist, hip and spine and distal radius micro-architecture by MRI in women on treatment, we recruited 20 postmenopausal women (mean age 70 years). Ten women had previously been on a bisphosphonate and 10 women, previously on a bisphosphonate short term, had just initiated treatment with teriparatide. Outcomes included bone mineral density by DXA (Hologic Discovery A) of the right wrist, PA spine, and hip (total, femoral neck,). MicroMRI was examined using a right bird cage wrist coil with a Sigma 1.5 Tesla MRI scanner and analyzed by MicroMRI Inc. (Philadelphia, PA). MicroMRI indices included: BV/TV (bone volume/total volume), Surf (topological surface density), Curv (topological curve density), Surf/Curv (surface to curve ratio, number of platelike to rodlike trabeculae, higher values indicate intact trabeculae); Erosion Index =(ratio of parameters that increase with deterioration). Table Pearson Correlation Coefficients, *p<0.05, ** p<0.01 

Table  .  

Indices of MRI micro-architecture at the radius had the greatest association with BMD at the ultra distal radius but was also associated with BMD at the 1/3 distal radius and femoral neck sites. Correlation at the spine was poor. Older age was associated with fewer plates, lower plate to rod ratio and greater deterioration as assessed by the erosion index.

We conclude that measures of microMRI at the wrist are associated with the wrist and hip BMD and demonstrate a deterioration in architecture in postmenopausal women with osteoporosis on therapy. Additional studies are needed to examine these changes prospectively.

Disclosures: P. Greeley, None.

This study received funding from: NIH/NCRR.

W240

Bone Strength and Toughness Are Reduced by Loss of Architecture — A Non-linear FE Analysis. G. J. Gross*, H. Hong*, B. Borah*, R. J. Phipps, T. E. Dufresne*. Procter & Gamble Pharmaceuticals, Mason, OH, USA.

Reductions in bone volume and deterioration in bone architecture occur in osteoporosis. We performed non-linear computational modeling to understand the effect of architectural changes and bone loss on bone strength and toughness.

MicroCT scanning was performed on iliac crest biopsies collected from a placebo-controlled clinical study in postmenopausal osteoporosis. Finite element models were created directly from these microCT data over a wide range of bone volumes (11–29% BV). The volume of interest was maximized to include the largest number of bone elements possible. A custom program was written to simulate bone loss, which used a 3D Euclidian distance map for each sample to randomly select bone erosion sites. A 26% loss of BV was used to match the 3-yr placebo-treated results from the clinical study. This 26% simulated bone loss resulted in architectural changes comparable to those observed experimentally in clinical biopsies. The plasticity of bone was represented using a nonlinear constitutive model with apparent properties measured by compression loading each FE model.

With a 26% loss of BV, reductions in bone strength and toughness were similar, but depended on initial bone volumes. 

Table  .  

To understand the “dose response” of strength and toughness to bone loss better, a single sample (19% BV) was analyzed at five levels of bone loss (10, 15, 20, 26 and 35%). Both strength and toughness were reduced across all bone loss levels even after only 10% loss. Strength and toughness reductions followed similar trends. 

Table  .  

Use of non-linear tissue properties made these predictions in bone toughness possible and is a valuable technique in assessing bone fragility. Our results show that deterioration of trabecular architecture directly effects both strength and bone toughness. Toughness, the ability to absorb energy and prevent crack propagation, is important in bone fragility and will be further investigated in larger sample groups.

Disclosures: GJ. Gross, Procter & Gamble Pharmaceuticals 3.

W241

Trabecular Bone Microstructure Assessments of the Distal Radius Using a Compact MRI. S. Handa*¹, B. R. Gomberg*², T. Haishi*³, K. Kose*¹. ¹Insitute of Applied Physics, Tsukuba, Japan. ²MicroMRI Inc., Philadelphia, PA, USA. ³MRTechnology Inc., Tsukuba, Japan.

Trabecular bone (TB) microstructure measurements have shown promise for estimation of bone strength and evaluation of drug therapies for osteoporosis. Up to now, several groups have studied this extensively using whole body MRI (wbMRI) system, however the wbMRI systems are expensive and heavily utilized for other clinical indications. A compact peripheral MRI (cpMRI) system dedicated to measure TB microstructure would offer a solution to overcome these disadvantages.

In this study, we have evaluated a cpMRI system (MRTechnology Inc. Tsukuba, Ibraki, Japan) for TB microstructure measurements of the distal radius.

The cpMRI consists of a 1.0 T permanent magnet (100 mm gap, homogeneous volume: 60 mm diameter sphere), gradient coil set, radio frequency probes, and an MRI console[1]. To acquired data, we developed 3D driven equilibrium spin-echo pulse sequence (matrix size: 512 × 384 × 32, voxel size: 150 × 150 × 500 micron cube, data-acquisition time: 17 minutes) for the distal radius imaging of five healthy male volunteers.

Ages ranged from 21 to 25 years (mean 22.4 years) and volunteers were scanned three times independently over two weeks.

Image data processing and analysis was performed semi-automatically using digital topological analysis (DTA — MicroMRI Inc., Philadelphia, PA, USA). Table 1 shows the reproducibility of BV/TV and selected DTA structural parameters in the distal radius. The root mean square CV (RMS-CV) values range from 2.9 % to 14 %. These values are close to those previously reported for the distal radius and tibia using a wbMRI system[2]. In conclusion, the 1.0 T permanent magnet cpMRI system was found to be capable of performing trabecular bone micro-architectural assessment of the distal radius consistent with the results from wbMRI systems, showing great promise of our system for evaluating bone quality in clinical settings.

References

1. Haishi, T., et al., Magn Reson Imaging, 2001. 19(6): p. 875–80.

2. Gomberg, B.R., et al., Bone 2004. 35(1): p266–76. 

Table Table 1. DTA parameter averages, CVs and their RMS-CVs from five volunteers (S1-S5). T_THICK: trabecular bone thickness, SCR : surface-to-curve ratio, TEI : topological erosion Index. All parameters except T_THICK are unitless.

 

Figure Figure 1.. Cross-sectional image and projection image of microscopic TB structure in the distal radius.

Disclosures: S. Handa, None.

W242

Vertebral Fracture Risk: Is Size More Important than BMD. A. Hoiseth*. Sentrum Røntgeninstitute, Oslo, Norway.

Size is a major (bio)mechanical parameter, strongly associated to structural strength; compressive strength thus being associated to radius² and bending strength to radius⁴. BMD, often defined as a gold standard for fracture risk assessment, was constructed to eliminate the size factor, being the best approximation to bone density as measured by a two-dimensional projection. A dimension-less parameter, however, may not adequately reflect structure and thus strength. By DXA, however, also assessment of size as projectional area and bone mass (BMC) are derived. PURPOSE: To report the association for DXA parameters with fractures and with clinical and social consequences of fractures. MATERIAL: DXA parameters were measured in a total of 41.217 women aged > 40 years. Fracture history was obtained in two cohorts of these, respectively 8.584 and 5.138 large. METHOD: After correcting for age, using the total material, the DXA parameters were ranked from high to low values into 10 equally large groups. Identical, but separate fracture analyses were then performed in the two cohorts. Fractures were classfied as a) those having at least one vertebral fracture but without femoral fractures and b) those having at least one femoral fracture. Risk ratio (RR) was calculated for each tenth using the average number of fractures in the cohorts as RR=1. Only consistent results between the two cohorts were considered as “significant”. RESULTS: For the femoral fracture group there was an exponential increas in RR from high to low BMC and BMD in “total femur”. RR for the lower tenths being respectively 2.5 and 2.8. For area there was no difference in RR between the tenths. For the vertebral fracture group there was a linear increase in RR from high to low vertebral BMC and BMD, RR in the lowest tenth being respectively 1.5 and 1.4. For spine area there was an exponetial increase in RR to 1,8 in the lowest tenth. The RR for having fractures with clinical or social consequences also increased exponetially with lower BMD and BMC in the femur, and with spine area. DISCUSSION: The results indicate a need for giving attention to bone as a structure, not merely as a material. Exploring different mechanically relevant parameters may help us to more correct predictions of the risk of vertebral fractures, a fracture that to a large degree is independent to BMD.

Disclosures: A. Hoiseth, None.

W243

Age, Gender and Region Related Changes in Bone Mineral Density of Korean Adult. W. Choi, S. Hone*. Internal Medicine, Hanyang University, College of Medicine, Seoul, Korea, Seoul, Republic of Korea.

Background: Early diagnosis by measurement of bone mineral density(BMD) and treatment can reduce osteoporosis complications. But we had no large population data & Korean specific reference value in diagnosis of osteoporosis.

Method: we performed cross sectional study involving 50208 people(lumbar BMD / M:4,810, F:45,398), and 50026 people (femur neck BMD /M:4745, F:44,700) form Seoul, Gyeonggi, Daegu, jeonbuk. BMD was measured by using DEXA (HOLOGIC, QDR 4500 system) Result: Female peak BMD was obtained at 35-45 years (lumbar: n=2224, 0.985±0.129 and femur: n=6800, 0.864±0.114 mg/cm²). Male peak BMD was obtained at 25-35years (lumbar: n=458,1.010±0.118, femur: n=489 0.960±0.113 mg/cm²). And BMD was decreased with age. But decrease of BMD was more prominent with female and other regions than with male and Seoul. Especially female lumbar and femur wards BMD decrease was accentuated at menopause period. (figure I) The reference value in diagnosis of osteoporosis from our data was lower than the previous used reference data (Japanese data) except female femur. (figure2) Conclusion: Korea is racially homogeneous and not large nation. But there was regional difference in the BMD change. Also the diagnosis of osteoporosis may be overestimated. So further research is needed considering with social factors and fractures.

Disclosures: W. Choi, None.

W244

In Vivo Proteomic Analysis of Circulating Monocytes in Chinese Premenopausal Females with Extremely Discordant Bone Mineral Density. E. Y. Deng*¹, Y. Z. Liu¹, C. Jiang*², L. M. Li*², S. Wu*², Y. Chen*², H. Jiang*², E. Yaw*², P. Xiao*³, S. M. Xiao*², L. J. Tan*², X. Sun*², J. X. Xiong*², X. Z. Zhu*², M. Y. Liu*², S. F. Lei*², X. D. Chen*², J. Y. Xie*², G. G. Xiao*⁴, S. P. Liang*², H. W. Deng¹. ¹Departments of Basic Medical Science and Orthopeadic Surgery, University of Missouri — Kansas City, Kansas City, MO, USA. ²Hunan Normal University, Changsha, China. ³Osteoporosis Research Center, Creighton University, Omaha, NE, USA. ⁴Department of Pediatrics, Harbor-University of California, Los Angeles, Torrance, CA, USA.

Osteoporosis (OP) is a major public health problem and bone mineral density (BMD) is an important determinant of OP. Circulating monocytes (CMCs) may serve as progenitors of osteoclasts and produce a wide variety of factors important to bone metabolism. However, little is known about the specific roles of CMCs in the pathogenesis of OP. Using proteomics techniques of 2-Dimentional Gel Electrophoresis (2-DE) coupled with Matrix Assisted Laser Desorption and Ionization Time-of-Flight/ Time-of-Flight Mass Spectrometry (MALDI-TOF/TOF-MS), we performed a comparative protein expression profiling study of CMCs in Chinese pre-menopausal females with extremely high or low BMD. A total of 38 differentially expressed proteins on 2-DE gels were identified. We confirmed with western blotting five proteins that potentially might have important roles in osteoclastogenesis: up-regulation of ras suppressor protein 1 (RSU1), gelsolin (GSN), and manganese-containing superoxide dismutase (SOD2), and down-regulation of glutathione peroxidase 1 (GPX1) and prolyl 4-hydroxylase beta subunit (P4HB) in low vs. high BMD group. RSU1, GSN, SOD2 and GPX1 might affect monocyte trans-endothelium (via regulating its adhesion, morphological changes, or locomotion), differentiation, and/or downstream osteoclast function (via regulating osteoclast podosome formation, adhesion, motility, or activity). Thus, these proteins may contribute to differential osteoclastogenesis and finally lead to BMD variation. This is the first in vivo proteomics study of OP in humans and the five identified proteins may serve as biomarkers for early prevention and novel targets for clinical intervention of this debilitating disease.

Disclosures: H. W. Deng, None.

This study received funding from: Natural Science Foundation of China, NIH.

W245

The Loss of Follow up After Fragility Hip Fractures in Some Regions in Czech Republic. T. Hala¹, F. Senk*², P. Zivny*³, B. Skyvarova*⁴, M. Carda*⁵, T. Dedek*⁶. ¹SYNARC/CCBR Czech, Pardubice, Czech Republic. ²Osteocentre, Hospital Havlickuv Brod, Czech Republic. ³Osteocentre, Teaching Hospital Hradec Králové, Czech Republic. ⁴Osteocentre, Hospital Jablonec nad Nisou, Czech Republic. ⁵Traumatology, Regional Hospital Pardubice, Czech Republic. ⁶Traumatology, Teaching Hospital Hradec Kralove, Czech Republic.

Fragility fracture is a major risk for osteoporosis and has been identified as the only clinically relevant marker of bone quality. It is important to identify whether patients who experienced fragility fractures are being assessed and treated for osteoporosis in order to reduce the risk of future fracture. The patients with fractures should be managed in accordance with evidence-based clinical guidelines for osteoporosis.

Objective: The main objective of this study was to establish the BMD testing rate in osteoporosis management for people over 50 years of age who had already suffered a hip fracture. This is the first study in Czech Republic conducted for this purpose.

Methods: We conducted a retrospective cohort study using data from 5 fracture clinics and bone disease centres. Study population consisted of all individuals over 50 years of age who sustained hip fracture between January 1, 2004 and December 31, 2005 from regional hospitals where one fracture clinic and one bone disease centre are in house, for accurate data analysis. We used Czech classification system for fractures: hip (S 72.0), pertrochanteric fracture (S 72.1) and subtrochanteric fracture (S 72.2). We included patients with osteoporotic fractures and with DXA measurements of L1-L4 and hip. We excluded those with high-energy trauma fractures.

Results: We analyzed data from 1465 patients, 1054 (71.95%) women and 411 (28.05%) men with proximal femur fractures. We have documented that 99 patients (6.75%) with fractures underwent densitometric evaluation. The number of women and men was 86 (5.87%) and 13 (0.88%), respectively. Out of those 99 patients, 67.3% were over 70 years of age. 

Table  .  

Conclusions: There is evidence of a care gap between the occurence of fragility fractures and the diagnosis of osteoporosis in some regions in Czech Republic. This study provides evidence that many Czechs who experienced fragility fracture are not being diagnosed and treated for prevention of future fractures. The prevalence of bone mass measurements or physician follow-up is enormously low.

Disclosures: T. Hala, None.

W246

Skeletal and Vitamin D Status in Two Indigenous North American Populations. I. V. Haller¹, W. D. Leslie², J. Jaakola*³, H. A. Weiler*⁴, D. Krueger⁵, N. Binkley⁵. ¹Education and Research, SMDC Health System, Duluth, MN, USA. ²Faculty of Medicine, University of Manitoba, Winnipeg, MB, Canada. ³Human Services Division, Fond du Lac Band of Lake Superior Chippewa, Cloquet, MN, USA. ⁴School of Dietetics and Human Nutrition, McGill University, Montreal, PQ, Canada. ⁵Osteoporosis Clinical Research Program, University of Wisconsin, Madison, WI, USA.

Bone health and vitamin D status have received only limited study among North American indigenous populations. To explore potential geographic differences in peripheral BMD and wintertime 250HD levels, we conducted a secondary analysis of data from studies of two indigenous North American populations of women: First Nations (FN) in Canada and American Indian/Alaska Native (AI/AN) in the US. The FN group was comprised of 96 women from the First Nations Bone Health Study (FNBHS); the second group included 78 AI/AN women from the Great Lakes region. All participants had weight, height, and calcaneal BMD measurements (Lunar PIXI) performed. Serum calcium, PTH, and 250HD were evaluated during winter months (January-March). In the FNBHS study, the Diasorin RIA and DPC EIA were used to measure 25OHD and PTH, respectively, while in the AI/AN cohort, HPLC and the DPC IRMA were utilized for these measurements.

In this analysis, the FN women were on average younger than their AI/AN counterparts (42.4±11.3SD vs. 50.3±9.9SD years old). There were no differences between the two groups in weight (82.7±18.9SD vs. 86.4±18.7SD kg) or BMI (30.8±6.3SD vs. 32.2±6.6SD kg/m²). Due to the effects of age and weight on peripheral BMD, we used general linear modeling (GLM) to test the difference in absolute values of calcaneal BMD between the two groups adjusted for weight and age. Adjusted BMD values in FN and AI/AN women did not differ (0.515±0.009SEM vs. 0.542±0.010SEM g/cm², respectively, P=0.052).

In bivariate analysis, the FN group had lower serum calcium (9.0±0.4SD vs. 9.4±0.3SD mg/dl, P<0.001), higher PTH (55.1±22.2SD vs. 42.5±23.6SD pg/ml, P<0.001) and higher 25OHD (21.2±10.0SD vs. 18.1±8.9SD ng/ml, P=0.032). Serum 25OHD was negatively correlated with PTH in the overall sample (r=-0.19, P=0.014); no correlation between 25OHD and BMI was observed. Similar proportions of women in both groups had serum 25OHD concentrations below 30 ng/ml (82.1% in FN vs. 89.7% in AI/AN group). There was a significant effect of group on serum 25OHD concentrations after adjustment for age, BMI and PTH levels using GLM: 22.0±1.0SEM ng/ml in FN group vs. 17.1±1.2SEM ng/ml in AI/AN group (P<0.001). In conclusion, low vitamin D status is common in these two populations of indigenous North American women. Investigation of approaches to optimize vitamin D status in FN and AI/AN women is indicated.

Disclosures: I. V. Haller, None.

W247

Low Bone Mass Is Common in Premenopausal Black Women with Systemic Lupus Erythematosus on Prednisone: High BMI Is Protective. J. K. Jenkins, I. Srivstava-Hadley*, K. Demoruelle*. Division of Rheumatology, University of Mississippi Medical Center, Jackson, MS, USA.

The pupose of this study was to examine a premenopausal black female systemic lupus erythematosus (SLE) population on prednisone to determine the prevalence of low bone density (LBD) and osteoporosis (OP) and identify modifying clinical factors.

We performed a retrospective chart evaluation to examine factors affecting bone mass in premenopausal black women with SLE on steroids. Protected by ethnicity, this population has a higher SLE disease burden that may affect the development of low bone mass (LBM, i.e., LBD and OP). We identified patients from our lupus specialty clinic in 2004-2006 who also had a bone density assessment in the Division of Rheumatology. All patients were on daily prednisone > 2 mg/d. Clinical information was collected from the DXA records and clinic charts. ACR criteria were used for the classification of SLE.

Sixty-one women met our criteria. Nine others were excluded because of incomplete data. Average age at the time of the DXA scan was 32.9 yrs. Average disease duration was 7.1 yr. Prednisone was first used 7.1 yrs prior to DXA. Average prednisone dose was 11.9 mg/d at the time of DXA. Patients took: 0.7 other medications that may be associated with LBM (anticonvulsants, furosemide, anti-ulcer drugs, warfarin); 1.2 steroid sparing drugs (hydroxychloroquine, azathioprine, methotrexate, and mycophenolate); and 0.7 drugs that were potentially protective (calcium, HCTZ, bisphosphonates). Average BMI was 30.3. There were 3 non-vertebral fractures including one hip fracture.

DXA results were (WHO classification): 41% normal, 39% LBD, 20% OP. Regression analysis was performed to identify modifying factors. Duration of steroids, number of steroid sparing drugs, and the number of LBM-associated drugs or -protective drugs used were not associated with LBM. Bone density was strongly positively correlated (p<0.01, r=0.7) with BMI at all sites. Frank obesity (BMI>30, n=32) was protective, and was associated with a higher BMD (p<0.003) at all sites by 12-17%. In contrast to one report in which non-white SLE patients had a higher prevalence of hip OP, we find the usual higher prevalence of vertebral OP (20%, vs. 3 % hip) with steroids in the black SLE population.

This is a large study of black women with SLE on prednisone examining multiple factors affecting the development of low bone mass. Low bone density and osteoporosis appear to be as common in young premenopausal black women with SLE on steroids as in white women in previous reports (23%). This may be due to a greater disease burden and steroid use. Obesity is a significant protective factor against prednisone-associated osteoporosis in SLE in black women.

Disclosures: J.K. Jenkins, None.

W248

What Are the Characteristics of Postmenopausal Women with the T-score of the Total Hip that Is Lower than that of the Lumbar Spine?, M. Kang*, M. Kim*, D. Lim*, S. Lee*, K. Baek*, J. Han*, H. Kim*, K. Lee*. Department of Internal Medicine, The Catholic University of Korea, Colleage of Medicine, Seoul, Republic of Korea.

Background. Generally, T-score of the lumbar spine is lower than that of the total hip in postmenopausal women because of an accelerated loss of trabecular bone after menopause. However, some women had the reversed T- score which meant that the T-score of the total hip was lower than that of the lumbar spine. It is well known that PTH and thyroid hormone preferentially involve cortical bone rather than trabecular bone. Thus, we aimed to investigate the characteristics of postmenopausal women with T-score of the total hip that is lower than that of lumber spine.

Methods. We enrolled 246 postmenopausal women (mean age: 55.87±6.99 years) in our study. Serum levels of TSH, free T4, PTH, FSH, Estradiol, Alkaline phosphatase. Calcium, Phosphorous were measured by the standard methods. BMD at the lumbar spine and femur were measured by dual energy X-ray absorptiometry.

Results. Among 246 postmenopausal women, there were 25 cases in which the T- score of the total hip were lower than that of the lumbar spine (low hip T-score group). Body weight and BMI were significantly lower in the low hip T-score group, when compared to group that had similar lumbar BMD (55.41±6.13 vs. 60.51±7.84, p=0.011, 22.83±2.63 vs. 24.57±3.17, p=0.034). Age, YSM(years since menopause), free T4, TSH, PTH, FSH and Estradiol showed no differences between the two groups.

Conclusions. We have found that postmenopausal women, having the lower T-score of the total hip than that of the lumbar spine, have low body weight and low BMI. But, we could not detect any statistically significant differences in PTH and thyroid hormone levels. Additional studies are required to find the characteristics of postmenopausal women having the lower T-score of the total hip than that of the lumbar spine.

Disclosures: M. Kang, None.

W249

The Effect of Home Parenteral Nutrition on Bone Mineral Density. H. Karakelides*¹, J. L. Burnes*², J. M. Nadeau*², D. G. Kelly*², D. L. Hurley¹, K. A. Kennel¹. ¹Endocrinology, Mayo Clinic College of Medicine, Rochester, MN, USA. ²Gastroenterology, Mayo Clinic College of Medicine, Rochester, MN, USA.

Home parenteral nutrition (HPN) or long-term parenteral nutrition has been associated with poor bone health. Altered nutrition, components of parenteral nutrition formulations (e.g., aluminum) and underlying gastrointestinal (GI) disorders have been suggested as reasons for reduced bone density in this population. In addition to changing nutrition support practices, there have been advances in pharmacotherapies for metabolic bone disorders. However, there is little data on the effects of bone active drugs in patients who are on long-term parenteral nutrition. We conducted a retrospective study of patients on HPN to assess risk factors for bone loss and effects of bone active drugs on bone mineral density (BMD). Subjects studied had been treated by the Mayo Clinic Rochester HPN team since 1994 and had been on HPN for at least 1 year and had at least 2 BMD measurements performed at least 1 year apart. 52 subjects were included with a mean age 59.8 ± 2.0 years, mean duration of HPN 115.1 ± 14.2 months, mean body mass index (BMI) 23.0 ± 0.5 kg/m². There was no association between gender, age, BMI, duration of HPN or underlying GI pathology (inflammatory vs non-inflammatory disorders) and change in BMD during follow-up. The subjects were then divided into those who had been treated (tx, n=31) with bone active drugs versus those who had not been treated (untx, n=21). There was no difference in time on HPN, time between BMDs, underlying GI pathology, BMI or age between the two groups. As may have been expected, subjects in the tx group had a lower baseline BMD of the lumbar spine (LS) (0.92 ± 0.03 in tx vs 1.03 ± 0.04 g/cm² in untx, p<0.05) and femoral neck (FN) (0.76 ± 0.03 in tx vs 0.87 ± 0.03 g/cm² in untx, p=0.01). However, there was no significant impact of treatment with bone active drugs on the change in LS or FN BMD from baseline to final observation. In conclusion, in subjects on HPN, gender, age, BMI, duration on HPN and underlying GI disorder were not good predictors of BMD. Subjects with lower BMD were more likely to be treated with bone active medications.

Disclosures: H. Karakelides, None.

W250

Gonadotropins Associated with Bone Mineral Density in Males. A. Kåss*, H. Gulseth*. Department of Rheumatology, Betanien Hospital, Skien, Norway.

It has been recently suggested that FSH regulates bone mass (1.2) independent of the action of oestrogen. FSH was shown to stimulate osteoclastogenesis and bone resorption by acting on Gi₂d-coupled FSH receptors, which are localised on the surface of mouse and human osteoclasts and their precursors. We wanted to examine the levels of FSH, LH in men and explore their relationship to bone mineral density (BMD). One hundred and eighteen males (age range 49-90 years) with low impact forearm fractures were included. Lumbar spine BMD was measured according to standard procedures with Lunar DXA equipment. Serum FSH, luteinizing hormone (LH) and testosterone were measured using fluooimmunoassay. Multilple linear regression was used to assess the relationship between BMD and these hormones controllling for age and BMI. Of males with established osteoporosis (defined as a T score below 2.5 SD), 28.6% had above normal (> 12.0 IE/L) FSH levels, 21.4% had above normal (> 12.0 IE/L) LH levels and 33.3% had below normal (<8.0IE/L) testosterone levels. FSH and LH were significantly associated (β=-0.775, p=0.021, β=0.751, p=0.013 respectively) with BMD after controlling for age and BMI. To our knowledge this association has only been previously reported in women (3). Further studies are required to assess the oestrogen-independent role of FSH in osteoclastogenesis. It is unknown whether the associations observed between LH and BMD are confounded by FSH or gonadal hormones, or whether LH itself might have oestrogen-independent effects on bone mass; this remains to be studied. References: (1) Sun L., Peng Y, Sharrow AC, Iqbal J, Zhang Z, Papachristou DJ, Zaidi S et al. FSH directly regulates bone mass. Cell 2006; 125: 247–260. (2) Martin TJ, Gaddy D. Bone loss goes beyond estrogens. Nat Med 2006; 12: 612–613. (3) Sowers MR, Greendale GA, Bondarenko I, Finkelstein JS, Cauley JA, Neer RM, Ettinger B. Endogenous hormones and bone turnover markers in pre-and perimenopausal women: SWAN. Osteoporos. Int 2003; 14;:191–197.

Disclosures: A. Kåss, None.

W251

Spine Bone Mineral Density Predicted Incident New Vertebral Fractures in Postmenopausal Korean Women. S. Kim¹, Y. Rhee¹, H. Choi*¹, Y. Kim², E. Kang*¹, B. Cha*¹, E. Lee*¹, H. Lee*¹, S. Lim¹. ¹Internal Medicine, Yonsei University, Seoul, Republic of Korea. ²Internal Medicine, NHIC Ilsan Hospital, Goyang, Republic of Korea.

The identification of individuals at risk for osteoporotic fracture is essential for prevention. Low bone mineral density has been shown to be an important predictor of increased fracture risk. However, it is not known whether bone mineral density (BMD) in the axial skeleton is related to fractures in postmenopausal Korean women. We examined the relationship between BMD measured by dual energy x-ray absorptiometry (DXA) and incident vertebral fractures in postmenopausal Korean women. Baseline data collected from 2000 to 2002 (670 Korean women aged 49-87 years) at Severance Hospital, Seoul, Korea; mean follow-up of 3.2 years. Of the 670 subjects included in the study, 514 had no vertebral fractures and 156 had at least one. The women with vertebral fractures were older, smaller, and lower BMD compared to those without. In a binary logistic regression model with multivariate adjustment, age (p<0.001), spine BMD (p<0.01), and total hip BMD (p=0.001) were independent risk factor for prevalent vertebral fractures. During follow up period, 77 women (34 in 156 women with prevalent vertebral fractures and 43 in 514 women without) developed new incident vertebral fractures. In women without prevalent vertebral fractures, age (OR=1.070; 95% CI, 1.005-1.140) and spine BMD (OR=0.559; 95% CI, 0.332-0.940) were independent risk factor for incident new vertebral fractures in a binary logistic regression model with multivariate adjustment. In models adjusted for age. Spine BMD was associated with increased risk of incident new vertebral fractures (p<0.05) in women with baseline vertebral fractures.

Decreased spine BMD was associated with an increased risk of incident new vertebral fractures in postmenopausal Korean women.

Disclosures: S. Kim, None.

W252

Longitudinal Study of Changes in Bone Mineral Density in Women Undergoing Screening for Osteoporosis. K. Kurasawa*, K. Katayama*, Y. Nomura*, R. Kikuchi*, O. Chaki, F. Hirahara*. Obstetrics and Gynecology, Yokohama City University, Yokohama, Japan.

Objective: Measurements of bone mineral density (BMD) are an important tool for the diagnosis and treatment of osteoporosis. Low BMD is an independent risk factor for fracture. However, longitudinal studies of BMD are few. Longitudinal studies and analyses of the time course of BMD are essential. Methods: The study group comprised 69 women free of diseases potentially affecting bone metabolism who presented at our department from 1993 though 2006 to undergo measurement of BMD. Their mean age was 58.8 years (range, 43 to 71). The subjects were classified into 3 groups according to status: perimenopausal women, early postmenopausal women, and late postmenopausal women. Height, body weight, lumbar-spine bone mineral density, and femoral BMD were measured over time. The percent changes in these values were compared. Lumbar-spine radiography was performed in 1998 and 2006. The presence or absence of new fractures was examined. Results: At the beginning of the study, lumbar-spine bone mineral density (0.954 ± 0.04 mg/cm²), femoral BMD (total, 0.710 ± 0.05 mg/cm²; neck, 0.801 ± 0.06 mg/cm²), radial bone mineral density (0.447 ± 0.04 mg/cm²), and calcaneal bone mineral density (91.54% ± 13.42%) were significantly higher in perimenopausal women than in early and late postmenopausal women (p<0.01). In perimenopausal women, lumbar-spine BMD significantly decreased by 7.71% at 5 years, by 10.29% at 7 years, and by 13.49% at 13 years. BMD at all other sites also significantly decreased during follow-up. In early and late postmenopausal women, lumbar-spine and femoral BMD did not decrease significantly. In contrast, calcaneal BMD significantly decreased in all groups, suggesting that calcaneal BMD may not reflect the estrogen-deficiency-related decrease in BMD after menopause. Height did not significantly decrease in any group, and there were no new fractures. Conclusions: The age-related decrease in BMD seen in women differs depending on the site examined. Early after menopause, BMD decreases rapidly at all sites. Subsequently, longitudinal monitoring confirmed that the percent decreases in BMD gradually converged. When screening for osteoporosis, the BMD of sites other than the calcaneus should be measured mainly during the perimenopausal period. In women with no decrease in height, bone density can be measured at intervals of 1 year or longer.

Disclosures: K. Kurasawa, None.

W253

No-vertebral Fracture Rates Are Lower in Hong Kong Chinese Men than American Caucasians but Bone Mineral Density Predicts Fracture Risk Similarly in the 2 groups-Results from Mr Os International. E. M. C. Lau¹, J. Lapidus², P. C. Leung³, T. Kwok³, D. Bauer⁴, J. Cauley⁵, S. R. Cummings⁶, E. S. Orwoll, for the Mr Os Study group⁷. ¹Hong Kong Orthopaedic and Osteoporosis Center, Hong Kong, China. ²Oregon Health and Science University, Portland, OR, USA. ³Chinese University of Hong Kong, Hong Kong, China. ⁴California Medical Center, San Francisco, CA, USA. ⁵University of Pittsburg, Pittsburg, PA, USA. ⁶San Francisco Co-ordinating Center, California Pacific Medical Center, San Francisco, CA, USA. ⁷Oregon Health and Science University, Portland, OR, USA.

The purpose of this study is to compare fracture rates in Hong Kong (HK) Chinese and American (US) Caucasian men and to study the relationship between BMD and fracture risk in these 2 groups by a cohort method.

The Mr Os cohort consists of 5,995 ambulatory men recruited from 6 communities in the US, plus 2000 men recruited from a single site in Hong Kong. Protocols in the two countries are comparable. BMD was measured with DXA, and standardized to account for equipment differences. Age and body mass index (BMI) were collected at baseline examination. Fractures were ascertained using standardized country-specific protocols, and all were confirmed by physician adjudication. Numbers and rates of fracture were compiled for each country individually. Data from the two countries were pooled, and Cox regression models used to compare associations with fracture in HK vs. US. Regression models were adjusted for age, BMI and clinical site in US.

There were 70 non-spine fractures in 69 Hong Kong participants; 500 fractures in 413 US participants. Rates of all non-spine fractures were significantly lower among HK Chinese (967.1 per 100,000) than among US Caucasians (1818.3 per 100,000) (RR = 0.53,95% CI for RR 0.41–0.68). Rates of hip fractures were also lower (218.5 per 100,000) in Hong Kong than in US (281.3 per 100,000), but not significantly so (RR=0.78,95% CI: 0.45–1.33).

Hong Kong participants, on average, were younger (72.4 vs. 73.8 years), had lower BMI (23.4 vs. 27.4), and lower femoral neck BMD (0.79 vs. 0.84) (all p<0.0001). We adjusted for age and BMI. The relative hazard of non-spine fracture per 0.1 gm/cm² decrease in femoral neck BMD did not significantly differ in Hong Kong, 1.39 (95% CI: 1.10–1.76) and US, 1.48 (95% CI: 1.36–1.62), p=0.63. The relative hazard of hip fracture per 0.1 gm/cm² also did not differ between the countries (Hong Kong RH: 2.23,95% CI 1.30–3.85; US RH: 2.76,95% CI 2.13–3.59), p=.0.49.

In conclusion, the rates of non-spine fracture were lower among HK Chinese than among US Caucasians, but the relationship between femoral neck BMD, and fracture was similar in the two countries.

Disclosures: E.M.C. Lau, None.

This study received funding from: The National Institute on Aging (NIA) and the National Cancer Institute, under the following grant nos:U01 AG 18197.U01 AR45614, U01 AR45632. U01 AR45647,U01 AR45654.U01 AR45583.AG18197.M01 RR000334.

W254

Bone Mineral Density in Hispanic Lupus Patients. I. Moldovan. Loma Linda University, Loma Linda, CA, USA.

The purpose of this study is to examine the bone mineral density (BMD) and its clinical determinants in a cohort of Hispanic women with systemic lupus erytematosus (SLE). The charts of one hundred consecutive Hispanic SLE patients were reviewed and twenty women who had BMD assessment by dual X-ray absorbtiometry (DXA) were identified. Ten patients were premenopausal and ten were postmenopausal. The student t-test was used to assess statistical differences between the two groups. Clinical determinants of BMD were evaluated, including age, height, weight, body mass index (BMI), smoking and alcohol use, duration of SLE, steroid treatment, SLE disease activity index (SLEDAI) at the time of the DXA scan, SLE damage index (SLICC), as well as use of immunosuppressives and osteoporosis medications. The mean age was 34.3 in the premenopausal group and 52.6 in the postmenopausal group. The mean duration of SLE was 6.55 years in the premenopausal women and 19.1 years in the postmenopausal group. The disease activity score trended towards more active disease in the younger patients, however, the average doses of prednisone were comparable. BMI results were also comparable in both groups and were in the range of overweight. Four patients in the younger group and five in the postmenopausal ones had a BMI consistent with obesity. Overall, seventeen patients (85%) were taking glucocorticoids, but only five (25%) were taking medications for osteoporosis, including bisphosphonates, calcium and vitamin D. None of them were drinking alcohol and only one was an active smoker. The results showed that 35% of patients had osteopenia at the spine, 40% at the femoral neck and 10% at the total hip. Osteoporosis was present in 10% of the patients at both femoral neck and spine. One patient had a vertebral compression fracture. When comparing BMD in the pre- and postmenopausal groups, there was a trend towards lower BMD at the spine in the postmenopausal patients, and no statistical significant difference was detected for the total hip and the femoral neck. In conclusion, this study is one of the first ones looking at Hispanic lupus patients in terms of the prevalence of low BMD. In this small retrospective study, the prevalence of low BMD in the studied Hispanic SLE patients was comparable to reports from the literature in other ethnic groups. The small number of patients taking osteoporosis medications, despite widespread use of glucocorticoids underscores the importance of educating both physicians and patients regarding this matter. Somewhat surprising was the fact that there was no statistical difference in BMD, regardless of menopausal status. One explanation can be the small sample of patients analyzed. Another possibility could be that the higher BMI of the patients analyzed confers a protection against bone loss. Further, larger, prospective studies in this ethnic group are needed.

Disclosures: I. Moldovan, None.

W255

Lower Hip BMD Among Ethnically Diverse Postmenopausal Women with Anemia—Results from a Subgroup of the Women's Health Initiative. J. S. Nicholas, T. Bassford*, Z. Chen. University of Arizona, Tucson, AZ, USA.

Previously, we reported a significantly increased risk for hip fracture but not total fracture among anemic women in the Women's Health Initiative (WHI) Observational Study (OS). Additionally, a relationship between anemia and bone density was reported in a population of older community-dwelling anemic Italians, as well as among a variety of groups with specific conditions (thalassemia, hemodialysis, sickle cell anemia). Our aim was to investigate whether this relationship was evident among an ethnically diverse subcohort (n= 5,874) of postmenopausal women, age 50-79 at baseline from three clinical centers that performed bone density assessments in the WHI observational study. Questionnaires were used to collect demographic, medical, and lifestyle information. Bone mineral density (BMD) was measured by dual-energy x-ray absorptometry. Anemia was defined using the WHO criteria of hemoglobin<12 g/dl in women using baseline measurements. Multivariate regression analysis was used to assess the relationship between whole body and total hip BMD with anemia after adjusting for age, ethnicity, body mass index (BMI), physical activity, total calcium (Ca) intake, and hormone use. Participants were 63.7 (SD=7.4) years old at baseline, 22.4% non-White minority and had a mean hemoglobin level of 13.5 (SD1.1) g/dl. There were 393 (6.7%) participants with anemia (mean hemoglobin level 11.3 g/dl SD=1.02) of whom 47.8% were non-White minority. No statistically significant (at a p=0.1 level) ethnic interactions with anemia on BMD were found with one exception for the Native American group in which only 6 women were anemic. The ethnic stratified results were similar across all the ethnic groups in the direction of the association, so the combined results for the relationship between anemia and BMD are presented in Table 1. Results from this analysis suggest that anemia is associated with a reduced hip BMD, which may explain the increased hip fracture risk we found previously among anemic women in the entire WHI-OS cohort. The relationship between whole body BMD and anemia is much weaker, which is also inline with our previous findings of no statistical significant association between total fracture and anemia. 

Table Table 1. Regression coefficients (SE) for the association between anemia and BMD

^a adjusted for age & ethnicity ^b adjusted for age, BMI, physical activity, total Ca intake, hormone use

Disclosures: J.S. Nicholas, None.

This study received funding from: NIH.

W256

Impact of Birth Weight and Age at Menarche on Skeletal Growth in Young Women. H. Ohta, T. Kuroda*, S. Orito*, Y. Onoe, Y. Miyabara*, R. Yoshikata*, M. Ohara*, K. Ishitani*, H. Okano. Dept. of Obstetrics and Gynecology, Tokyo Women's Medical University, Tokyo, Japan.

Objectives: Skeletal growth is assumed to occur under the influence of congenital genetic as well as acquired environmental factors. In recent years, fetal programming as developmental origins of health and disease (DOHaD) has drawn attention as a factor affecting lifestyle related disease.

However, it has also been suggested that birth weight and age at menarche, as well as lifestyle factors such as nutritional intake and physical activity, may affect skeletal growth. In this study, therefore, we investigated how these factors might affect skeletal growth in young women.

Subjects and Methods: Following approval of the study protocol by the institutional ethics committee at Tokyo Women's Medical University, birth weight and age at menarche, as well as lifestyle habits were surveyed in a total of 315 healthy young female volunteers aged 19 to 29 years old who were nursing school students at the institution and who gave informed consent to participate in the current study.

Results: The subjects' mean age was 22.1 ± 3.0 (mean ± SD); their mean birth weight, 3,143 ± 45 g; their birth week, 39.3 ± 1.8 weeks; their age at menarche, 12.0 ± 1.2 years of age; their height, 158.8 ± 4.8 cm; their body weight, 53.4 ± 7.3 kg; and their BMI, 21.1 ± 2.6 kg/m².

Birth weight was found to be significantly positively correlated with current BMI and BMD, and current BMI and L_2-4 BMD were found to be significantly higher in those who weighed 3,500 g or greater at birth, while there was no correlation seen between birth week and current BMI or L_2-4 BMD.

In those with normal menarche occurring between 10 to 15 years of age, age at menarche was found to be negatively correlated with current BMI and L_2-4 BMD, and positively correlated with body height.

Conclusions: The study results demonstrated that the greater the birth weight, the greater the current BMI and L_2-4 BMD, thus suggesting a role for birth weight in skeletal growth.

It was also suggested that the earlier the age at menarche, the shorter the body height, likely due to the resulting early closure of the epiphyseal space, and, as a consequence, the greater the BMI and the BMD in these young women.

It was therefore concluded that birth weight and age at menarche are independent factors affecting skeletal growth.

Disclosures: H. Ohta, None.

W257

Association Study of IL-15 Polymorphisms with Bone Mineral Density and Fracture Risk in Postmenopausal Korean Women. J. Koh*¹, B. Oh*², J. Lee*², J. Lee*¹, K. Kimm*³, J. Kim*⁴, J. Lim*⁴, T. Kim*⁵, J. Hong*⁵, G. Kim*¹, S. Kim⁶, E. Park⁴. ¹University of Ulsan College of Medicine, Seoul, Republic of Korea. ²National Genome Research Institute, Seoul, Republic of Korea. ³Eulgi University School of Medicine, Daejeon, Republic of Korea. ⁴School of Dentistry, BK21, Kyungpook National University, Daegu, Republic of Korea. ⁵Skeletal Diseases Genome Research Center, Kyungpook National University Hospital, Daegu, Republic of Korea. ⁶Kyungpook National University Hospital, Daegu, Republic of Korea.

Interleukin-15 (IL-15) has been implicated in both inflammation and bone destruction. In the present study, we investigated the genetic effect of IL-15 polymorphism on bone mineral density (BMD) in postmenopausal women. IL-15 polymorphisms were genotyped in study participants (n = 728). BMD at the lumbar spine and proximal femur was measured using dual-energy X-ray absorptiometry. We found that IL-15+13815A>T polymorphism was significantly associated with BMD of various bone sites. IL-15+13815A>T polymorphism was associated with high BMD of the lumbar spine (p=0.006-0.01), femoral neck (p=0.001-0.02), trochanter (p=0.009-0.05), femoral shaft (p=0.008-0.009), and total femur (p=0.008-0.04). Haplotype (ht) analyses showed that ht1 and ht2 were also significantly associated with high BMD at various sites (p=0.001-0.05). Moreover, IL-15+13815A>T polymorphism, ht1, and ht2 showed an association with risk of fracture (OR (95%CI) = 0.62(0.39-0.99), P = 0.04). These findings indicate that the +13815A>T polymorphism, ht1, and ht2 of IL-15 may be useful genetic markers for bone metabolism and risk of vertebral fracture.

Disclosures: E. Park, None.

W258

Percutaneous Injection of Allantoin Accelerates Fracture Repair. A. A. Selim*¹, K. Mezghani*², A. Abou-Samra³, D. Al-Tamimi*⁴. ¹Biotechnology Center, KFUPM, Dhahran, Saudi Arabia. ²Mechanical Engineering, KFUPM, Dhahran, Saudi Arabia. ³Wayne State University School of Medicine, Detroit, MI, USA. ⁴Pathology Department, College of Medicine, King Faisal University, Al-Khobar, Saudi Arabia.

Allantoin is a botanical extract of the comfrey plant and is used for its healing, soothing, and anti-irritating properties. Allantoin helps to heal wounds and skin irritations and stimulate growth of healthy tissue. Its chemical formula is C4H6N4O3. It is also called 5-ureidohydantoin or glyoxyldiureide. In this study, we examined the effects of allantoin on the proliferation of the osteoblastic cell lines, MC3T3 and ROS17/2.8, in vitro; and on fracture repair in vivo. Cell cultures were treated with allantoin (1, 10 and 100 nM) for seven days (From Day 1 of culture to Day 7). On day 7, MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) and cell counting assays were performed to examine the effect of allantoin on osteoblast proliferation. Allantoin significantly induced osteoblast proliferation in a dose dependent manner as compared to control cultures (35% and 56% increase for 10 and 100 nM, respectively, p<0.05). It is important to mention that Carcinogenic Potency Project (CPP) studies demonstrated that allantoin has no carcinogenic effects on male or female rats and mice. We further examined the effect of allantoin on fracture repair. Experimental fractures were performed on 8-week Sprague-Dawley rats under anesthesia as follow: from an anterior incision we performed a transverse osteotomy at the mid tibia with a bone saw. The animals were then treated with vehicle or allantoin (weekly percutaneous injections; 10 mg/Kg) for up to 4 weeks. Torsional biomechanical testing indicated that the stiffness of the allantoin-treated fractures was 2 fold higher than that of control groups at the two, three, and four-week time-points. The strength of the allantoin-treated fractures was 34%, 60% and 77% greater than that of controls at 2 (p<0.05), 3 (p<0.005) and 4 weeks (p<0.005), respectively. At 4 weeks, the stiffness and strength of the allantoin-treated fractures were equal to those of the intact contralateral tibia, whereas the buffer-treated and untreated fractures were significantly weaker than the intact tibiae. These data demonstrate that percutaneous injections of allantoin accelerate fracture repair in this fracture repair model. Our data on osteoblastic cell lines in-vitro suggest the possibility that the beneficial effects of allantoin on bone may result from increasing osteoblast proliferation and thereby bone formation at the fracture site.

Disclosures: A.A. Selim, None.

This study received funding from: KFUPM and KFU.

W259

Threshold Associations between Aspects of Body Composition and Bone Mineral Mass. T. G. Travison, A. B. Araujo, G. R. Esche*, J. B. McKinlay. New England Research Institutes, Watertown, MA, USA.

Reduced weight is associated with decreased bone mineral content and density (BMC and BMD). It is unclear, however, whether lean mass (LM), fat mass (FM), or the combination of both (expressed through weight or BMI) is most important in stimulating osteogenesis and therefore bone strength. Using data from a racially/ethnically diverse sample of 1,209 men enrolled in the Boston Area Community Health / Bone study, we examined BMC and BMD at the hip, spine and forearm as a function of weight, body mass index (BMI), waist circumference, total body FM, and total body LM.

Data were obtained on 363 non-Hispanic black, 397 Hispanic, and 449 non-Hispanic white randomly-chosen male residents of Boston, MA, USA. Subjects' ages ranged from 30 to 79 y. BMC and BMD were determined by dual x-ray absorptiometry. Analyses indicated that weight, BMI, waist circumference and FM displayed strong positive associations with BMC or BMD up to certain threshold values, beyond which associations were attenuated. LM, by contrast, displayed a strong and consistent association over all values (Figure); for instance, models controlling for fat mass and other covariates indicated a 12.8% and 9.2% increase in femoral neck BMC and BMD, respectively, per 10 kg unit in LM (p<0.001). Once the effect of LM was accounted for, the magnitude and significance of associations between other parameters and BMC or BMD were sharply reduced. Percent increases in BMC or BMD corresponding to increased body size were greater in magnitude at load-bearing sites (hip and spine) than at the wrist; for instance, percent difference in BMD per 10kg difference in LM was significantly (p<.0001) greater at the lumbar spine (6.9%) than the wrist (3.4%). Results were similar for all race/ethnicity groups. These results indicate that decreased body size parameters are strongly associated with decreased BMD, conferring increased risk of fracture at all sites and among all race/ethnicity groups. However, the protective effect of increased weight, BMI, waist circumference or fat mass on BMC and BMD manifests only within certain ranges of body size, beyond which the effect is diminished. On the other hand, it appears that increased LM is positively associated with BMD over the entire range of values, indicating that the protective influence of body composition on BMC or BMD is largely mediated by total lean mass.

Disclosures: T.G. Travison, None.

W260

Correlations Between Bone and Body Composition in Post-pubertal Daughters and Their Premenopausal Mothers. K. Uenish¹, H. Ishida², M. Konishi*³, Q. Zhou³. ¹Kagawa Nutrition University, Laboratory of Phisiological Nutrition, Saitama, Japan. ²Kagawa Nutrition University, Laboratory of Administrative Dietics, Saitama, Japan. ³GE Healthcare Asia, Shanghai, China.

Evidence suggests that both environmental and genetic factors play important roles in the development of osteoporosis. A strong correlation between lean mass (LM) and bone mineral density (BMD) as well as bone mineral content (BMC) has been repeatedly demonstrated in individuals. However, studies of the relationships of bone and body composition values in families are lacking. Dual-energy X-ray absorptiometry (DXA) provides accurate, precise measurements of regional and total body composition. We measured total body BMD, BMC, LM, fat mass (FM), and %fat in 104 post-pubertal healthy Japanese girls (aged 19.6±1.3 years) and their healthy premenopausal mothers (aged 48.0±2.5 years) with DXA (Lunar DPX-IQ, GE-Healthcare). Pearson's correlation coefficients were calculated to examine the relationship between mothers and daughters. Weight, height, BMI, BMD, BMC, FM, LM, and %Fat showed low-to-moderate similarity between the girls and their mothers (r= 0.23-0.47, p<0.01). Significant heritability (h² ± SE) was detected for BMD (0.506 ± 0.03, P < 0.001) and BMC measured with DXA (0.658 ± 0.04, P < 0.001). Among the girls, BMD was moderately correlated with weight, BMI, FM and LM (r= 0.34-0.49, p<0.01). While among their mother, correlations of BMD with weight, BMI and FM were slightly lower (r= 0.26-0.31, p<0.01), and there was no significant relationship between BMD and LM (p>0.05). We conclude that the bone and body composition values of these post-pubertal daughters are significantly related to their premenopausal mothers. The relationship between BMD and LM in these Japanese girls was stronger than in their mothers. 

Table Table 1. Correlations between bone mineral and body composition in daughters

 

Table Table 2. Correlations between bone mineral and body composition in Mothers

Disclosures: K. Uenishi, None.

W261

Racial Differences in Skeletal Health. M. D. Walker¹, R. Babbar*², J. Vargas-Jerez*¹, J. Udesky*¹, C. Gagel*¹, D. J. McMahon¹, R. Lantigua*¹, G. Liu*², J. P. Bilezikian¹. ¹College of Physicians and Surgeons, Columbia University, New York, NY, USA. ²New York Downtown Hospital, New York, NY, USA.

Although there are racial differences in bone mineral density (BMD) and fracture risk, little data are available regarding racial differences in bone mass accrual, peak bone mass (PBMD) achieved and bone loss. We compared changes in BMD between Chinese American, Hispanic and Caucasian women across all decades of adult life. In this cross-sectional study, 1120 Chinese American and Hispanic American women, 80 per decade from age 20 to 90 in each racial group, are being recruited. Along with DXA of the total hip (TH), femoral neck (FN) lumbar spine (LS) and forearm, demographic, familial, medical, nutritional and behavioral information is being obtained. To date, 743 subjects (411 Chinese American and 332 Hispanic women) have been recruited. BMD data for these cohorts were compared to Caucasian referent values. BMD means and standard deviations for each decade were calculated. Decade means were used to approximate longitudinal changes in bone density. These preliminary data suggest that Chinese American women achieve peak bone at the LS later (age 40-49) than Caucasian or Hispanic women (age 30-39). Chinese American women achieved 96.7% of the Caucasian PBMD at the LS (p=0.1), while the Hispanic group reached 99.3% of the Caucasian PBMD (p=0.74). Elderly Caucasian women (age 80-89) lost 28.2% of their bone mass at the LS from peak BMD compared to 22.2% (p<0.01) for Chinese American and 25.8% for Hispanic American women (p=0.2). Chinese and Hispanic American women achieved peak TH BMD later (age 40-49) than Caucasian women (30-39). TH peak BMD for Chinese American women was 96.2% of the Caucasian TH PBMD value (p=0.1) while TH peak BMD for Hispanic women was 104.5% of the Caucasian TH PBMD (p=0.04). Aged Caucasian women (age 80-89) lost 26.6% of their bone mass at the TH from peak BMD compared to 20.7% for Chinese American (p<0.01) and 20.5% for Hispanic American women (p<0.01). Hispanic American women achieve peak FN BMD later (age 40-49) than Caucasian and Chinese American women (age 30-39). Chinese American women reached 93.8% of the Caucasian PBMD at the FN (P=0.01). Hispanic American peak FN BMD was similar to that of Caucasian women (101.8% of Caucasian PBMD; p=0.51). Elderly Caucasian women (age 80-89) lost 32.9% of their bone mass at the FN from peak BMD compared to 25.8% for Chinese American and 27.5% for Hispanic American women (p<0.01). The differences we have shown in the rate of bone mass accrual and decline among these 3 groups may contribute to racial and ethnic disparities in fracture rate.

Disclosures: M.D. Walker, None.

W262

Peak Spine and Hip Bone Mass in Korean Women. Y. Won¹, H. Kim*¹, Y. Shin*². ¹Endocrinology and Metabolism, Kwandong University College of Medicine, Goyang, Republic of Korea. ²Clinical Nutrition Research Center, Kwandong University College of Medicine, Goyang, Republic of Korea.

The achievement of maximal peak bone mass early on in life is one of the most important strategies for prevention of osteoporosis in women. The aim of this study was to clarify the correlation between body morphometry and bone mineral density. Bone mineral density (BMD) of the spine, femoral neck, total hip, and bone mineral content (BMC) of spine, femoral neck, total hip were assessed by dual energy x-ray absorptiometry in 322 healthy women (age 20-84 years).

The peak lumber spine BMC was gained during fifth decades (40s) of life. However, peak BMD in lumber spine was gained during fourth decades (30s) of life. The femur neck and total hip peak BMC were gained during third decades (20s) of life. However, peak BMD of femur neck and total hip were gained during fourth decades (30s). The age was all negative significant correlation of all these bone measurements. The weight and height were all positive significant correlation of all these bone measurements. In multiple stepwise regression analysis, age and weight were independent variable on spine BMD (R² = 0.305), femur neck BMD (R² = 0.294), total hip BMD (R² = 0.283). The weight was a stronger predictor than height for all sites.

We conclude that age of attaining peak bone mass at the hip is younger than at the spine, and age and weight were most important predictor on achievement of peak bone mass in Korean women.

Disclosures: Y. Won, None.

W263

In Vivo Genome-wide Expression Study of Human Blood B Cells Suggests a Novel ESR1 and MAPK3 Network for Postmenopausal Osteoporosis. P. Xiao¹, Y. Chen*¹, H. Jiang*¹, T. L. Yang*¹, F. Pan*¹, Z. H. Tang*¹, Y. Z. Liu², R. R. Recker¹, H. W. Deng². ¹Osteoporosis Research Center, Creighton University, Omaha, NE, USA. ²Department of Orthopedic Surgery and Basic Medical Sciences, University of Missouri-Kansas City, Kansas City, MO, USA.

Osteoporosis is characterized by low BMD resulting from bone resorption (by osteoclasts) exceeding bone formation (by osteoblasts). Estrogen deficiency evokes increased osteoclastic activity and bone loss.

Studies showed that B cells may participate in osteoclastogenesis via expression of osteoclast-related factors, such as NF-kappaB ligand (RANKL), transforming growth factor beta (TGFB), and osteoprotegerin (OPG). However, the role of B cells in bone metabolism and osteoporosis is still largely unknown, particularly at the systematic expression level in vivo.

In this study, we recruited 20 unrelated postmenopausal Caucasian females aged 54–60, 10 with high (spine or hip Z-score > 0.84) and 10 with low BMD (spine or hip Z-score < −0.84). Total RNA of the freshly isolated blood B cells from those subjects were extracted and hybridized individually to Affymetrix HG-UI33A GeneChip^@ arrays to identify genes differentially expressed between low and high BMD subjects. Significance of differential expression was tested by t-test and adjusted with Benjamini and Hochberg (BH) procedure for multiple-testing.

Twenty-nine genes were down-regulated in the low vs. high BMD group. Those genes were further analyzed using Ingenuity Pathways Analysis (Ingenuity® Systems, www.ingenuity.com) and a network involving estrogen receptor 1 (ESR1) and mitogen-activated protein kinase 3 (MAPK3) was identified (Fig. 1). Real-time RT-PCR confirmed the differential expression of 8 genes, including ESR1 (p = 0.044) and MAPK3 (p = 0.002).

This is the first in vivo genome-wide expression study on human B cells for osteoporosis. Our results suggest a novel mechanism for postmenopausal bone loss that downregulation of ESR1 and MAPK3 in B cells further regulates the secretion of factors leading to increased osteoclastogenesis or decreased osteoblastogenesis.

Disclosures: P. Xiao, None.

This study received funding from: State of Nebraska LB595 and NIH.

W264

Sarcopenia in Ageing Men with Osteopenia, Osteoporosis and Normal Bone Mineral Density. C. A. Zerbini, M. G. Pippa*. Rheumatology, Hospital Heliopolis, Sao Paulo, Brazil.

Ageing men is accompanied by a loss of muscle mass. Sarcopenia, the decrease of muscle mass with age has been widely studied because of its relation to impaired gait, disability and falls. Aimed to evaluate sarcopenia, we measured body composition (BC) in 220 men at 50 years and older by dual X-ray absorptiometry (DXA). Sarcopenia was defined as a relative skeletal muscle index (RSMI) (appendicular skeletal muscle mass divided by height) below 7.26 kg/m²,according to Baumgartner classification. Osteoporosis and Osteopenia prevalence were studied as well as its occurrence in sarcopenic subjects. L1-L4 bone mineral density (BMD), femoral (FNBMD) and total body BMD (TBBMD) were also measured by DXA). Total Lean mass (TLM), Total Fat Mass (TFM), percentage of adiposity (% adp), arms lean mass (ALM), legs lean mass (LLM), and body fat distribution (BFD) were also determined. Sample was gathered in different groups according to age:50-60y, 61-70y and 71 or older. The adherence to normal curve was evaluated by Kolmogorov Smirnov test. Cochran-Armitage Trend Test was used to test proportion linear trend. To compare the difference of quantitatives variables the Student t test was used. The effective P value to be considered signifcant was < 0.05. Data were analysed using the statistical computer program Minitab vertion 15.0% and SAS vertion 9.1.3 The average age was 61.8 years (SD =8.15),average weight 72.9 Kg (SD=12.0),average L1-L4 BMD 1.146 g/cm² (SD= 0.17), NFBMD 1.366 g/cm² (SD= and TBBMD was 1.145 g/cm² (SD= 0.10). A hundred ninety five men were not sarcopenic (88.6%) and 25 men presented sarcopenia (11.36%). Sarcopenia prevalence presented a linear decrease according to BMI increase (p < 0.0001). Ageing showed a linear increase in sarcopenia prevalence ( p = 0.0305). According to FNBMD T-score, 38.8% of sarcopenic showed osteoporosis (OP); 10.5% osteopenia and 9.58% normal BMD. Using L1-L4 BMD T — score 31.5% had OP, 12.7% osteopenia and 13,12% had normal BMD. According to TBBMD T — score, 40.9% had OP, 10.5% osteopenia and 12.1 normal BMD. There was a statistic significant difference between sarcopenic group (SG) and non sarcopenic group (NSG) second these variables: age (66.3 ± 9.50 SG), (61.24 ± 7.80 NSG), p value =0.016; BMI (21.7 ± 2.0 SG), ( 27.3 NSG), p value <0.0001; TBBMD ( 1.041 g/cm² SG), (1.158 g/cm² NSG) p < 0.0001; T- score L1-L4 BMD ( −1.5 SG), (-0.5 NSG), p value = 0.001; T-score FNBMD (-1.8 SG), (-1.0 NSG), p= 0.000. Our data confirm higher prevalence of OP in Sarcopenic as well as negative influence of ageing and lower BMI in muscle mass loss.

Disclosures: C.A. Zerbini, None.

W265

Measures of Fitness and Body Composition in Childhood Are associated with Calcaneal Quantitative Ultrasound in Adulthood: A 20 Year Prospective Study. S. J. Foley¹, S. Quinn*¹, T. Dwyer*², A. Venn*¹, G. Jones¹. ¹Menzies Research Institute, Hobart, Australia. ²Murdoch Children's Research Institute, Melbourne, Australia.

Peak bone mass is a major determinant of fracture risk later in life. Weight-bearing exercise during growth is generally accepted as one factor that can influence bone mass, however the long-term effects of childhood exercise and body composition on bone mass remain uncertain as many studies have been retrospective in design. The aim of this inception cohort study was to describe the associations between childhood physical performance measures and body mass index (BMI), and adult bone mass. We measured an Australia wide representative random sample of 1,434 children, aged 7 to 15 yrs, originally included as part of the Australian Schools Health and Fitness Survey in 1985 and approximately 20 yrs later (mean age 31). Measures included 1.6 km run and 50 meter sprint (childhood only) and leg strength, standing long jump and physical work capacity at 170 beats/min (PWC₁₇₀) (in both childhood and adulthood). Body composition was assessed at both time points by BMI. A single Sahara bone ultrasound densitometer was used at follow-up to determine heel bone mass (CV = 1%). In females, there were modest but significant beneficial relationships between physical measures and adult bone mass (1.6 km run: r = +0.08 to +0.10, p=0.02-0.049; 50 m sprint: r = +0.08 to +0.11, p=0.03-0.10; standing long jump: r = +0.06 to +0.08, p=0.20-0.33; PWC₁₇₀: r = +0.12 to +0.21, p=0.004-0.09), of which the last was independent of adult performance (R² = 1-4%). In males, childhood BMI (but no performance measures) was positively associated with all adult QUS parameters after adjustment for adult BMI (r = +0.11 to +0.12, p=0.001-0.003) (R² = 1-2%). An age-PWC₁₇₀ interaction was documented with PWC₁₇₀ at age nine, in both sexes, having a greater influence on adult bone mass (r = +0.25 to +0.32, p=0.002-0.01) (R² = 5-8%) than it did at ages 12 and 15 (r = −0.01 to +0.02, p=0.80-0.97). On further examination, a sex-PWC₁₇₀ interaction was also observed in the 12 yr olds, with childhood PWC₁₇₀ also having a positive effect on female adult bone mass (r = +0.19 to +0.25, p=0.045-0.13), but not male adult bone mass (r = −0.19 to −0.24, p=0.06-0.13). In conclusion, childhood fitness levels, particularly in females and in the prepubertal years, are predictive of adult skeletal status as measured by QUS, explaining up to 8% of the variation in adult bone mass. Furthermore, childhood BMI in males was predictive of all adult QUS parameters. These results strongly suggest that increased skeletal loading in childhood (especially in the prepubertal years) leads to an increase in peak bone mass which is maintained many years later independent of current fitness or body mass index.

Disclosures: S.J. Foley, None.

This study received funding from: National Health and Medical Research Council of Australia, Heart Foundation of Australia, Tasmanian Community Fund.

W266

Dietary Protein and Bone Mineral Density (BMD) in the Presence of Low vs High Calcium in Men and Women of the Framingham Study. C. E. McLennan*¹, K. B. Broe*¹, K. L. Tucker*², R. R. McLean*¹, D. P. Kiel¹, L. A. Cupples³, M. T. Hannan¹. ¹HSL, Boston, MA, USA. ²HNRC-Tufts, Boston, MA, USA. ³BUSPH, Boston, MA, USA.

Short-term metabolic studies show that increasing dietary protein creates negative calcium balance, yet epidemiologic studies support a beneficial effect of protein on bone. However, long-term influences of typical protein intake may rely upon concurrent calcium. We examined the cross-sectional association of dietary protein on BMD in presence of low calcium in Framingham Offspring and Original Cohort members. We hypothesized that protein intake would be positively associated with BMD only in those with low calcium.

3,988 participants (2919 Offspring members with BMD obtained in 1996-2000, and 1069 Original Cohort members with BMD from 1988-89 or 1992-93) had BMD obtained at femoral neck (FN), trochanter (TR) and lumbar spine (LS). Dietary protein and calcium intakes were estimated via 126-item Willett food frequency questionnaire. At each BMD site, by sex and by cohort, we used multivariable linear regression to examine the association of BMD with %protein (% of total energy from protein) adjusted for calcium intake (mg/d), indicator of calcium < 800 mg/d, total energy, age, weight, height, current smoking, vitamin D, supplement use of calcium or vitamin D, and in women menopause and estrogen use. Further models adjusted for physical activity, caffeine and alcohol. We then stratified by low calcium (< 800 mg/d) adjusting for calcium intake within strata.

Mean age of Original Cohort members was 75 (SD 5) and of Offspring cohort was 60 (SD 9). Mean %protein intake was 16% and 17% respectively. Over 60% of both groups had calcium intake < 800 mg/d. No associations were seen between protein and BMD in either men or women of elderly Original Cohort (all p>0.30). In Offspring cohort, no associations were seen in men but there was positive effect between protein and all 3 BMD sites in 1639 women. Adjusting for physical activity, caffeine or alcohol did not change results. Upon stratifying (table), similar though not significant results were seen for 1006 women with low calcium while no association was seen for 633 women with calcium > = 800.

Calcium intake modified the cross-sectional effect of dietary protein on BMD but only for middle-aged Offspring cohort women. Our findings suggest that protein and BMD studies may also need to consider concurrent calcium intake.

Table Table. Framingham Offspring Women, Protein coefficient

Disclosures: M.T. Hannan, None.

This study received funding from: NIH.

W267

Clusters of Dietary Protein and Relation to Bone Mineral Density (BMD) in Men and Women of the Framingham Offspring Study. K. L. Tucker*¹, N. Qiao*¹, C. E. McLennan*², K. E. Broe*², R. R. McLean*², D. P. Kiel², L. A. Cupples*³, M. T. Hannan². ¹HNRCA, Tufts University, JM USDA, Boston, MA, USA. ²HSL Institute for Aging Research, Boston, MA, USA. ³BUSPH, Boston, MA, USA.

Several studies have shown that dietary protein intake may be protective of BMD rather than contributing to calcium loss. We have previously shown a positive relation between BMD and protein in the elderly Framingham Original Cohort. As protein-rich foods are also rich sources of other nutrients, we now further examined this relation by food source of dietary protein in the primarily middle-aged members of the Framingham Offspring cohort.

We evaluated the cross-sectional relation between food sources of protein and BMD in 2919 Framingham Offspring Study participants (mean age 61y) with BMD at femoral neck (FN), trochanter (TR) and lumbar spine (LS) obtained by Lunar DPX-L in 1996-2000. Dietary intakes were estimated using the Willett 126-item food frequency questionnaire and we used cluster analysis to classify individuals into 5 groups determined by major source of protein (table). Multivariable linear regression was used to compare least-squares adjusted mean BMD among protein clusters, adjusting for: total energy, sex, age, weight, height, smoking, total protein intake, calcium, vitamin D, physical activity, caffeine, alcohol and in women, menopause status and estrogen use. We further adjusted for fruit and vegetable servings, an indicator of overall healthy diet also associated with higher BMD.

The dietary cluster for which the primary source of protein was low-fat milk had significantly greater FN and TR BMD than the other clusters (this cluster also tended to have greater LS BMD, p<0.10). Associations remained after further adjustment for fruit and vegetable servings.

In summary, increased protein intake continues to be associated with higher BMD, but a dietary pattern with primary protein sources of low-fat milk and breakfast cereal appears to have the greatest effect, even after adjusting for calcium, vitamin D and fruit and vegetable intakes. These findings show that the beneficial effect of protein on BMD may depend upon the source from which protein is obtained. 

Table  .  

* Significantly different from cluster 1 at p<0.05 † Significantly different from cluster 1 at p<0.10

Disclosures: M.T. Honnan, None.

This study received funding from: NIH.

W268

Withdrawn

W269

Vitamin D status and Bone Health in Adolescents: The Northern Ireland Young Heart's Project. T. R. Hill*¹, J. Wallace*², P. J. Robson*², W. Dubitzky*², L. Murray*³, J. Strain*², A. Flynn*¹, M. Kiely*¹, K. D. Cashman¹. ¹Department of Food and Nutritional Sciences, University College Cork, Cork, Ireland. ²Northern Ireland Centre for Food and Health, University of Ulster, Coleraine, United Kingdom. ³Department of Epidemiology and Public Health, Queens University, Belfast, United Kingdom.

Our objective was to investigate the relationship between vitamin D status {as measured by serum 25-hydroxyvitamin D [25(OH)D]} and adolescent bone health.

The present cross-sectional study used data from the Northern Ireland Young Heart's Project cohort of adolescents, which included 468 girls and 458 boys, aged either 12 or 15 yr who were sampled throughout the year. Bone mineral density (BMD) was measured by DXA at the non-dominant forearm and dominant heel. Stored serums were analysed for 25(OH)D, parathyroid hormone (PTH), osteocalcin (OC) and C-telopeptide of collagen metabolism (CTX) using enzyme-immunoassays. Multivariate models were used to examine the relationship between bone parameters and tertiles of serum 25(OH)D.

Significant age x gender interactions (p<0.05) were evident in the bone parameters, which justified analysing each age-gender group separately. After adjustment for physical, lifestyle and dietary variables, there was no significant association between tertiles of serum 25(OH)D on heel BMD in any of the age-gender groups. In addition, no significant associations were observed between tertiles of serum 25(OH)D and forearm BMD in boys. However, after adjustment for physical, lifestyle and dietary variables, girls in the highest tertile of serum 25(OH)D had significantly (p<0.05) higher forearm BMD than did the girls in the lowest tertile (β = 0.018; 95% CI: 0.003, 0.033; and β − 0.018; 95% CI: 0.001, 0.034, for 12 and 15-y old girls, respectively). Mean serum 25(OH)D levels for the lowest, middle and highest tertile of serum 25(OH)D in girls were 35.1, 58.7 and 94.6 nmol/l. In general, biochemical bone indices (PTH, OC, CTX), after adjustment for covariates, were significantly (p<0.001-0.082) higher in girls (aged 12 and 15-y olds) in the lowest tertile compared to those in the highest tertile of serum 25(OH)D. When limiting the female subjects to only those sampled in winter (December to February), when vitamin D status is at its lowest, girls in the lowest tertile of serum 25(OH)D (mean: 33.5 nmol/l) had significantly (p = 0.05) lower forearm BMD (β = 0.016; 95% CI: −0.00005, 0.033) compared to those in the highest tertile (mean: 81.0 nmol/l). In conclusion, low vitamin D status appeared to be related to lower bone mass and higher bone turnover in adolescent girls, but not in boys. This may have an impact on peak bone mass development in girls.

Disclosures: T.R. Hill, None.

This study received funding from: Higher Education Authority, Ireland.

W270

Higher Dietary Calcium Intake along with Higher Vitamin D Status Could Suppress Bone Turnover and BMD Decrease in Asian Osteoporotic Patients. T. Hirota¹, H. Ikeda*¹, K. Hirota*². ¹Research Laboratory, Tsuji Academy of Nutrition, Osaka, Japan. ²Department of Obstetrics and Gynecology, Nissay Hospital, Osaka, Japan.

Recently some date failed to prove that calcium and vitamin D supplementation reduces the risk of fractures in Caucasian population. The purpose of this study was to evaluate the effects of dietary calcium intake and vitamin D status on bone turnover and BMD in osteoporotic Asian women whose habitual intake of calcium is thought to be much lower and vitamin D is higher than that of Western people.

We evaluated 215 postmenopausal Japanese women (aged 49-87 yrs, median 67 yrs) diagnosed as osteoporotic by L2-4 BMD or vertebral fractures. They had not taken any medicines affecting endocrine systems and bone turnover for 2 months nor bisphosphonates for a year. Subjects could walk and visit the physicians as out patients. Their calcium intakes were evaluated by food frequency questionnaires and vitamin D status was determined from serum 25-hydroxyvitamin D [25(OH)D] measured by RIA. Markers of bone turnover, including serum bone-specific alkaline phosphatase (BAP), osteocalcin (OC), procollagen type I N-propeptide (PINP) for bone formation, and urinary deoxypyridinoline (DPD), type I collagen N-telopeptide (NTX) for bone resorption, along with serum PTH and 1,25(OH)₂D were determined. The BMD of the lumbar spine and the total hip was measured by DXA.

The mean calcium intake in the patients was 814±307 mg/day, which was much higher than non-osteoporotic postmenopausal controls (548±257 mg/day, n=158 aged 45-74 yrs), probably due to an improvement in diet following awareness of a previous calcium shortage. However, their mean serum 25(OH)D was 43±13 nM, which was still low compared with that reported for healthy, elderly, Japanese women. Calcium intakes in patients were negatively associated with all bone turnover markers (BAP, OC, PINP, DPD and NTX). Serum 25(OH)D was negatively associated with OC and DPD, as well as PTH. The association of 25(OH)D with bone resorption markers (DPD, NTX) was clearly observed in younger patients (aged 49-66 yrs). These associations were independently observed after adjustment for 25(OH)D and calcium intake. Higher calcium intake (more than 800 mg/day), together with higher 25(OH)D status (more than 50 nM) showed not only a significant additive decrease in NTX and BAP, but also a significantly higher hip BMD.

We suggested that many osteoporotic patients could improve their diets and succeed in calcium intake but not vitamin D. Higher dietary calcium intake and vitamin D status could suppress bone turnover and BMD decrease in Asian osteoporotic patients. We recommend therapy for osteoporotic patients should include dietary calcium intake greater than 800 mg/day and serum 25(OH)D status of more than 50 nM.

Disclosures: T. Hirota, None.

W271

Differential Accumulation of Lead and Zinc in the Tidemark of Normal and Osteoarthritic Human Articular Cartilage and Subchondral Bone. N. Zoeger*¹, J. G. Hofstaetter², P. Roschger³, C. Jokubonis*¹, G. Pepponi*⁴, G. Falkenberg*⁵, R. Simon*⁶, P. Fratzl*⁷, A. Berzlanovich*⁸, C. Streli*¹, P. Wobrauschek*¹. ¹Vienna University of Technology, Atominstitut, Vienna, Austria. ²Ludwig Boltzmann Institute of Osteology at the Hanusch Hospital of WGKK and AUVA Trauma Centre Meidling and Department of Orthopaedics, Medical University of Vienna, Vienna, Austria. ³Ludwig Boltzmann Institute of Osteology at the Hanusch Hospital of WGKK and AUVA Trauma Centre Meidling, 4th Medical Dept., Hanusch Hospital, Vienna, Austria. ⁴ITC-irst, Centro per la Ricerca Scientifica e Tecnologica, Trento, Italy. ⁵Hamburger Synchrotronstrahlungslabor HASYLAB am Deutschen Elektronen- Synchrotron, Hamburg, Germany. ⁶Institut für Synchrotronstrahlung, Forschungszentrum Karlsruhe, Karlsruhe, Germany. ⁷Max-Planck Institute of Colloids and Interfaces, Dept. of Biomaterials, Potsdam, Germany. ⁸Department of Forensic Medicine, Medical University of Vienna, Vienna, Austria & University of Munich, Munich, Germany.

Little is known about the spatial distribution of the toxic element lead (Pb) and other trace elements in normal and osteoarthritic articular cartilage and subchondral bone. We assessed the elemental distribution of lead (Pb), zinc (Zn), strontium (Sr) and calcium (Ca) using synchrotron radiation induced micro X-ray fluorescence (SR μ-XRF) in the chondral and subchondral bone region of 4 normal, 4 osteoarthritic femoral heads and 3 normal patellae from adults with no history of work-related exposure to Pb. SR μ-XRF line and area scans in confocal geometry were correlated to backscattered electron images visualizing mineralized tissue. We found a highly specific accumulation of Pb in the tidemark, the transition zone between calcified and non-calcified articular cartilage. Pb fluorescence intensities in single tidemarks of normal samples were 13-fold higher when compared to subchondral bone and were strongly correlated with Zn, but were distinctly different from Ca and Sr. However, in the double tidemarks of osteoarthritic samples, Pb fluorescence intensities in the inner tidemark were 3-fold higher than in the outer and were approximately 36.1-fold higher when compared to subchondral bone. Interestingly, Zn-intensities did not correlate with Pb-intensities in double tidemarks. Zn-intensities were approximately 5-fold higher in both, the inner and outer tidemark, when compared to subchondral bone. Our data showed a very specific spatial accumulation of Pb in the tidemark of articular cartilage in normal and osteoarthritic samples. Temporal differences in the accumulation of Zn and Pb in the mineralization front of calcified articular cartilage suggest different mechanisms in Pb and Zn accumulation in the mineralization front.

Disclosures: J.G Hofstaetter, None.

W272

Effect of Four Weeks of Vitamin D3 Supplementation During Late Winter on Markers of the Vitamin D Endocrine System and Bone Turnover. K. Holvik¹, A. A. Madar*¹, H. E. Meyer¹, C. M. Lofthus*², L. C. Stene*³. ¹Institute of General Practice and Community Medicine, University of Oslo, Oslo, Norway. ²The Hormone Laboratory, Center of Endocrinology, Aker University Hospital, Oslo, Norway. ³Division of Epidemiology, Norwegian Institute of Public Health, Oslo, Norway.

A seasonal variation in vitamin D status is commonly observed at northern latitudes with a drop in serum 25(OH)D and a corresponding increase in serum PTH and bone turnover during late winter, which is most evident in non-supplement users. We investigated the effect of four weeks' daily supplementation with 10 μg (400 IU) vitamin D3 on vitamin D metabolites, PTH, and bone turnover in healthy young adults during late winter.

In a trial designed to compare the effect of two common types of vitamin D supplement on vitamin D status ⁽¹⁾, 55 men and women aged 19-48 y recruited in the community received a daily dose of 10 μg vitamin D3 during a 4-week period in late winter 2005. Venous serum samples were drawn at baseline and after 28 days for the analysis of serum concentration of 25-hydroxyvitamin D (s-25(OH)D), 1,25-dihydroxyvitamin D (s-1,25(OH)₂D), intact parathyroid hormone (s-iPTH), and osteoclast-specific tartrate-resistant acid phosphatase (s-TRAP).

As previously reported, mean s-25(OH)D increased from 44 (95% CI 38, 51) nmol/l to 78 (95% CI 72, 85) nmol/l during the period (p<0.001), and the increase did not differ according to type of vitamin D3 supplement.⁽¹⁾ Mean s-iPTH decreased from 5.85 (95% CI 5.14, 6.57) pmol/l to 4.70 (95% CI 4.14, 5.26) pmol/l; p=0.001. Mean s-1,25(OH)₂D increased slightly during the period; from 120 (95% CI 109, 131) pmol/l to 134 (95% CI 124, 144) pmol/l; p=0.057. Mean s-TRAP was 2.65 (95% CI 2.46, 2.84) U/l at baseline, and it increased to 3.03 (95% CI 2.84, 3.21) U/l;p<0.001).

We conclude that four weeks of supplementation with 10 μg vitamin D3 led to a decrease in s-iPTH, a tendency to an increase in s-1,25(OH)₂D, and an increase in s-TRAP concentration. The suggested increase in osteoclast activity in spite of the improvement in vitamin D status and corresponding PTH suppression is surprising.

(1) Holvik K, Madar AA, Meyer HE, Lofthus CM, Stene LC. A randomised comparison of increase in serum 25-hydroxyvitamin D concentration after four weeks of daily oral intake of 10 μg cholecalciferol from multivitamin tablets or fish oil capsules in healthy young adults. Br J Nutr 2007 [in press]

Disclosures: K. Holvik, None.

This study received funding from: Peter Möller (now MöllerCollett AS).

W273

Comparison of iDXA Measurements of Mid-Body Fat Mass and Tape-Measured Circumferences in Overweight Women. P. Liu*, H. Shin*, S. Pounds*, J. Folkert*, A. P. Crombie*, J. Z. Ilich. Nutrition, Food and Exercise Sciences, Florida State University, Tallahassee, FL, USA.

Studies show mid-body fat is more predictive of cardiovascular risk factors than total body fat. New bone densitometers have ability to measure body fat in android (abdominal) and gynoid (hip) regions utilizing total body scans. The data comparing those values obtained by iDXA series (GE Healthcare, Madison, WI) with tape-measured circumferences in overweight women are not available. Our objective was to compare android and gynoid measurements from iDXA with tape-measured circumferences in abdominal and hip regions. Tape measurements were performed according to the USDHHS guidelines (PHS, NHANES III, 1996). Participants were 37 postmenopausal women aged 56.3±3.7 y, BMI 31.7±4.8 kg/m² (mean±SD). Results showed high correlation (Pearson's r ranging from 0.7 to 0.9) between corresponding tape-measured and iDXA-analyzed regions, all p's<0.0001. In multiple regression models with android fat (g) as dependent variable and controlled for weight and age, tape-measured abdominal circumference remained significant predictor, t-ratio=2.55, p<0.05. In models with gynoid fat (g), controlled for weight and age, tape-measured hip circumference remained significant predictor t-ratio=2.23, p<0.05. We conclude that fat mass measured by iDXA in android and gynoid regions corresponds well with tape-measured circumference values and could be utilized during the total body scan and particularly rendered useful if there is no trained personnel to perform tape measurements in overweight women.

Disclosures: J.Z. Ilich, None.

This study received funding from: USDA/CSREES/NRI #2004-05287.

W274

Influence of Overweight/Obesity on BMD of Various Skeletal Sites in Early-Postmenopausal Women. P. Liu*, J. Folkert*, S. Pounds*, H. Shin*, A. Crombie*, J. Z. Ilich. Nutrition, Food and Exercise Sciences, Florida State University, Tallahassee, FL, USA.

We reported earlier that higher body weight was associated with higher BMD in women of wide age range. However, excess weight beyond BMI of 30 kg/m² was not associated with additional increase in BMD. The objective of this study was to examine closer the relationship between weight and BMD in either overweight or obese postmenopausal women. Participants included 37 early (from 2-10y) postmenopausal women aged 56.3±3.7 y who were classified according to their BMI as either overweight (BMI=25-29 kg/m², n=17), or obese (BM1>30 kg/m², n=20). BMD of the whole body, femur (neck, Ward's triangle, shaft, trochanter and total femur), lumbar spine and forearm were measured with iDXA (GE Healthcare, Madison, WI). Non-parametric statistics was used and based on Mann-Whitney tests, obese women did have statistically higher BMD at radius, total forearm, Ward's, trochanter, shaft and total femur (all p<0.05). There was no statistical difference in spine, femoral neck and total body BMD between overweight and obese women. Based on this preliminary study focused on overweight/obese postmenopausal women only, it appears that weight status classified as obesity might lead to higher BMD in some skeletal sites compared to just overweight. More studies with larger sample size are necessary to gain better understanding of the influence of overweight/obesity on BMD.

Disclosures: J.Z. Ilich, None.

This study received funding from: USDA/CSREES/NRI #2004-05287.

W275

Isoflavone and Soya Foods Intakes in Brazilian Population: Data from the BRAZOS Study. N. O. Jacques*¹, J. P. Roque*¹, M. M. Pinheiro², R. M. Ciconelli*³, M. B. Ferraz*³, L. A. Martini¹. ¹Nutrition, Universidade de Sao Paulo, Sao Paulo, Brazil. ²Reumathology, UNIFESP, Sao Paulo, Brazil. ³Paulista Center of Economy in Health, CPES, UNIFESP, Sao Paulo, Brazil.

Epidemiological studies have been demonstrated a positive association between dietary isoflavones and/or soy foods and bone mineral density. The purpose of present study was to evaluate the mean isoflavone and soy foods intake in men and women older than 40 years. This study was part of the Brazilian Osteoporosis Study (BRAZOS). A total of 2392 subjects (1693 women) were enrolled in the nutritional evaluation at the BRAZilian Osteoporosis Study — BRAZOS, undertaken in 120 cities across 5 regions (North, Northeast, Central, Southeast and South) of the country, and included people from five categories of economical level (from A to E). The sampling was based in the census data from IBGE (Brazilian Institute of Statistics and Geographic) 2000 and PNAD (National Research of Home Sampling) 2003 and it was calculated accordingly with to the probabilistic and representative sample of Brazilian population. Dietary intakes were evaluated by a 24h Food Record. For the nutrient analysis the Nutrition Data System software (Minneapolis, MN 2005) was used. The use of supplements was not considered in the present analysis. Statistical analysis was performed by SPSS (v 12.0). Student T test was used for gender comparison and One way ANOVA for comparisons between regions and economical level. Data was log-transformed since there are no normal distribution in isoflavone intakes. Significance was considered as p<0.05. Data is presented as median (minmax). The isoflavone/soy intake in the population evaluated was very low, only 1% of the sample presented an isoflavone intake higher than 1 mg/day. Considering the isoflavone intake in all participants, the mean daidzein was 0.08 mg/d (0.01-69.5) and the mean genistein was 0.03 mg/d (0.01-64.8). The mean genistein intake in men was 0.02 mg/d (0.01-15.78) and in women was 0.13 mg/d (0.01-69.54). For daidzein the mean intake in men was 0.03 mg/d (0.01-11.93) and in women 0.03 mg/d (0.01-69.54). No significant differences were observed between gender and between regions of the country. The major contributors to isoflavone intake were soy milk and soy grains, 52% and 28%, respectively. Besides soy milk and soy-grains, there was others foods that contributed to isoflavones intake, as such as: wholemeal bread, meat-substitute foods containing textured vegetable protein, soy protein isolate and toasted soy beans. Considering the benefits of isoflavones intake for bone health, the nutritional advice for osteoporosis patients, besides calcium and vitamin D, should include an increase in soy products.

Disclosures: N.O. Jacques, Wyeth Health Consumer 2.

This study received funding from: Wyeth Health Consumer.

W276

Low-Level Radiation Can Enhance Trabecular Bone Quantity and Architecture. L. Karim*¹, M. Vazquez*², S. Judex¹. ¹Stony Brook University, Stony Brook, NY, USA. ²NSRL, Upton, NY, USA.

Detrimental effects of high-level radiation exposure to the skeleton range from cell death to cancer and may be strongly catabolic to the skeleton, in particular to trabecular bone. Charged particle radiation such as iron ions and its secondary fragmentation products are of particular concern due to their high charge and energy deposition. However, little is known about the long-term effects of these particles, applied at relatively low-levels, on trabecular and cortical bone morphology. Here, it was hypothesized that even low-level charged particle radiation can deteriorate bone quantity and architecture. Adult male C57BL/6J mice were exposed to full-head FE ion radiation (56-Fe, 1GeV/n) with only 14% exposure for the appendicular skeleton at 120cGy (16.8cGy to the extremities), 60cGy (8.4cGy to the extremities), or 0.0 cGy (age-matched controls) for 11.5 months (n=11 per group). The metaphysis, epiphysis, and mid-diaphysis of the femurs were analyzed for differences in bone quantity and architecture by microCT. Three-point bending tests evaluated cortical bone mechanical properties. In the metaphysis, 8.4cGy radiated femurs had significantly greater (p<0.05) bone volume fraction (+61%) and trabecular number (+17%) than controls. In the epiphysis, the same mice showed a 31% greater bone volume fraction and a 31% smaller structural model index (a smaller structural model index is typically associated with greater bone strength). Bones radiated at 16.8cGy showed a similar trend as 8.4cGy mice but morphological indices were not significantly different from controls. There were no significant differences in cortical bone morphology between any radiated group and control mice. The mechanical tests revealed that bone's material properties did not deteriorate upon radiation at either level. In contrast to our original hypothesis, this data indicated that low-level iron radiation may enhance, rather than deteriorate, trabecular bone in the femoral metaphysis and epiphysis of mice. Mice receiving the smallest dosage showed the greatest beneficial response. The mechanisms by which these anabolic and/or anti-catabolic effects were induced remain to be determined but may be related to processes involving cell differentiation and proliferation, in particular for cells of the more radio-sensitive osteoclast lineage.

Disclosures: S. Judex, None.

W277

Does Pre- or Post-menopausal Soy Consumption Affect Postmenopausal Bone Loss?C. J. Lees, H. Chen*, J. R. Kaplan*. Wake Forest Universiy School of Medicine, Winston-Salem, NC, USA.

Chronic soy consumption has been associated with a decreased incidence of hip fracture in Asian women, an effect often attributed to the estrogen-like compounds (isoflavones) found in soy. However, it is not known whether pre- or post-menopausal soy exposure initiated in fully adult individuals will affect postmenopausal bone loss.

Ninety-six female Macaca fascicularis were randomized into 2 dietary treatments that differed only in major protein source, casein/lactoalbumin or isoflavone-enriched soy. The animals consumed these diets for 30 months and then all were ovariectomized (OVX). Post-OVX, one half of monkeys in each diet treatment switched to the opposing diet while the remainder continued to consume their original diet, resulting in 4 pre and postmenopausal treatment conditions (casein-casein {CC}, casein-soy{CS}, soy-casein{SC}, soy-soy{SS}). Bone mass was determined throughout the pre-menopausal phase and six months after OVX. Percent change was calculated from 30 month premenopausal and 6 postmenopausal measurements.

All groups had a significant decrease in whole body (WB, p< 0.04) and lumbar vertebral bone mineral content (LVBMC, p < 0.03). 

Table  .  

In the immediate postmenopausal period, consumption of soy significantly slowed the rate of bone loss, but only in animals that had consumed casein diet in the pre-menopausal period. Pre-menopausal consumption of soy did not affect the rate of bone loss during the first 6 months after OVX.

Disclosures: C.J. Lees, None.

This study received funding from: NIH.

W278

Use of a Web-Based Fracture Risk Assessment Tool. B. Ettinger¹, C. Gross¹, K. Cody*², R. Walker*³. ¹Medicine, University of California, San Francisco, CA, USA. ²Foundation for Osteoporosis Research and Education, Oakland, CA, USA. ³Walker Associates, Long Beach, CA, USA.

Current methods for interpretation of bone densitometry results are inadequate. Reporting of T-score has been the standard for many years, but experts throughout the world have called for a more accurate and clinically useful report format. We modified an existing fracture risk model that was developed at Kaiser Permanente's Division of Research* and made it available to providers through the Foundation for Osteoporosis Research and Education (FORE) website. The model calculates 10-year fracture risk and portrays patient risk graphically, categorized into 3 zones: “low-green “ 20%. The interactive model will be available at the poster site for ASBMR attendees and for general use at http://riskcalculator.fore.org.

Physicians who had referred at least one patient to FORE's densitometry service during 2004-2006 were mailed invitations to use the fracture assessment tool on a special FORE-sponsored password-protected website. We assessed the percentages of invitees who logged on to the site and who performed a patient calculation using the tool.

Of 241 providers who received the mailed invitation 11 (4.9%) logged on to the fracture risk assessment website within 8 weeks. Of those logging on, 10 (91%) completed the necessary quantitative data entry fields (age, height, weight, total hip Z-score, spine Z-score), answered 7 yes/no clinical questions (prior osteoporotic fracture, mother/sister with hip fracture, current smoking, alcohol intake > 2 drinks/day, exposure to glucocorticoids, rheumatoid arthritis), and submitted these data for risk calculation. The median usage of the risk calculator among those with any usage was 3 times. When invitees were categorized into 4 quartiles based on number of prior DXA referrals, there was a trend for greater likelihood of using the tool with greater prior referral: 1 referral(n=60) 3.3% use; 2-5 (n=69) 1.4%; 6-20 (n=56) 5.4%; > 20 (n=56) 7.1%.

We conclude that, based on a mailed outreach and web-based practicum, only a small percentage of providers referring patients to a well-recognized center of bone densitometry excellence appear to be interested in learning about fracture risk assessment. In the near future, fracture risk assessment is expected to become a key element of densitometry reporting and would likely become the standard method for osteoporosis counseling/decision-making. More efforts will be necessary to reach providers and teach them how to interpret and use fracture risk assessment.

* Ettinger B, Hillier TA, Pressman AR, Che M, Hanley DA. Computer model for calculating fracture risk. J Women's Health 2005; 14:158-172.

Disclosures: B. Ettinger, None.

This study received funding from: Procter & Gamble.

W279

Progression of Vertebral Deformities but no Change in Bone Mineral Density in Patients with Sarcoidosis: A 4 Year Follow-up Study. A. Heijckmann¹, M. Drent*², A. Nieuwenhuijzen Kruseman*³, P. Geusens³, M. Huijberts*³. ¹Internal Medicine, Hospital Bernhoven, Veghel/Oss, The Netherlands. ²Sarcoidosis Management Center, University Hospital, Maastricht, The Netherlands. ³Internal Medicine, University Hospital, Maastricht, The Netherlands.

Introduction: Several disease related factors may influence bone mineral density (BMD) and bone quality in sarcoidosis. We have previously demonstrated in a cross-sectional study that BMD is unaffected in a large cohort of individuals with sarcoidosis. Nevertheless, increased bone turnover and a high prevalence of vertebral deformities (20%) were found in these patients. This may imply that bone strength rather than BMD is decreased in sarcoidosis. The aim of this study was to determine whether this may have consequences for the incidence of new or progressive vertebral deformities during follow up.

Methods: BMD of the hip (DXA) and vertebral fracture assessment (VFA) with a lateral single energy densitometry was performed at baseline and follow-up (45 months, range 35-49) in 66 patients with sarcoidosis. In addition, clinical risk factors and glucocorticoid use were assessed. Potential predictors of new or progressive vertebral deformities were assessed using logistic regression analysis.

Results: The BMD of the total group was unchanged after follow up, even in the groups with current or previous glucocorticoid use. The prevalence of vertebral deformities increased from 20 to 32% of all subjects, and in 17 subjects one or more new or progressive vertebral deformities were diagnosed (26%). 84% of the subjects with a new or progressive deformity had a BMD in the low normal range (T-score ≤ −0.5). Logistic regression analysis revealed that a lower T-score of the femoral neck at baseline (OR=2.5 (CI: 1.0-5.9), p=0.04) and a mother with a hip fracture (OR=14.1 (CI:1.4-142.6), p=0.02) were determinants for new or progressive vertebral deformities.

Conclusions: Our study shows that in subjects with sarcoidosis the number of vertebral deformities increases in the course of this disease, despite unchanged BMD. The combination of a low normal bone mass in combination with a family history of fragility fractures contributes to an increased risk of the incidence of these deformities.

Disclosures: A. Heijckmann, None.

W280

To What Extent do Vertebral Fractures, Disc Height Loss, Low Bone Density, and Poor Muscle Strength Contribute to Hyperkyphosis in Older Women?, D. M. Kado¹, K. Prenovost*¹, L. Palermo², K. Stone³, T. A. Hillier⁴, S. R. Cummings³. ¹Medicine, UCLA, Los Angeles, CA, USA. ²UCSF, San Francisco, CA, USA. ³California Pacific Medical Center Research Institute, San Francisco, CA, USA. ⁴Center for Health Research, Kaiser Permanente, Portland, OR, USA.

Thoracic hyperkyphosis has been traditionally attributed to underlying fractures of the thoracic spine, but only 36-38% of those with the worst kyphosis have underlying fractures. Since thoracic hyperkyphosis is associated with adverse health outcomes including poor pulmonary and physical function, and increased mortality, we sought to determine other important, and possibly modifiable causes. Using baseline data from the Study of Osteoporotic Fractures, we randomly sampled 1196 women, aged 65 or older who had baseline radiographs available to measure Cobb angle kyphosis (T4-T12), disc heights (T4-T12), and prevalent vertebral deformities (3SD definition, from T4-L5). Participants also underwent baseline calcaneal bone mineral density and hip abduction strength testing, and answered questions regarding their medical history and health behaviors. Because vertebral fractures represent a continuous rather than discrete break in bone, we examined both a dichotomous vertebral fracture index and anterior to posterior vertebral height ratios from T4 to T12. We hypothesized that even if not classifiable as a vertebral fracture, vertebral wedging may substantially contribute to degree of kyphosis. In multivariable regression, the vertebral height ratio (averaged over T4-T12) independently explained 60% (consistent with previous literature), average disc height ratio explained 22%, and a paternal history of stooped posture explained 0.2% of Cobb angle variance. All reported values are squared semi-partial correlations that factor out inter-correlations among predictors, that in this case also included prevalent vertebral fractures, calcaneal bone mineral density, and hip abduction strength (each explaining less than 0.1% of Cobb angle variance). The overall total adjusted R² for this model was 74.2%. Whether osteoporosis treatments can decrease vertebral wedging and improve kyphosis is unknown.

Disclosures: D.M. Kado, None.

This study received funding from: NIH/NIAMS and NIA AG24246.

W281

5 Year Incidence of Fractures in Patients with Severe Osteoporosis: The Canadian Multicentre Osteoporosis Study (CaMos). S. M. Kaiser¹, C. Ioannidis², J. Lorraine³, W. P. Olszynski⁴, J. C. Prior⁵, D. A. Hanley⁶, J. P. Brown⁷, A. Papaioannou², R. G. Josse⁸, T. Anastassiades⁹, K. Siminoski¹⁰, S. Kirkland¹, C. Joyce*¹¹, D. Goltzman¹², J. D. Adachi². ¹Dalhousie University, Halifax, NS, Canada. ²McMaster University, Hamilton, ON, Canada. ³Eli Lilly and Company, Toronto, ON, Canada. ⁴University of Saskatchewan, Saskatoon, SK, Canada. ⁵University of British Columbia, Vancouver, BC, Canada. ⁶University of Calgary, Calgary, AB, Canada. ⁷Laval University, Sainte-Foy, PQ, Canada. ⁸University of Toronto, Toronto, ON, Canada. ⁹Queen's University, Kingston, ON, Canada. ¹⁰University of Alberta, Edmonton, AB, Canada. ¹¹Memorial University, St. John's, NF, Canada. ¹²McGill University, Montreal, PQ, Canada.

WHO defines patients with severe osteoporosis as those with a bone mineral density (BMD) T-score of < −2.5 and one or more fragility fractures. Patients with multiple fragility fractures also have severe bone disease. To assess future fracture risk, we classified CaMos participants 50 years and older, into two “severe” groups, and two comparator groups, according to their prevalent fracture status and BMD T-scores: (1) Severe Group 1: patients defined by the WHO classification (SG1, n=372, 91.4% women); (2) Severe Group 2: patients with 2 or more prevalent fragility fractures irrespective of their BMD T-score (SG2, n=390, 78.5% women); (3) Low BMD group (LG): patients with BMD T-scores ≤ −2.5 and no prevalent fragility fractures (n=3498, 67.7% women); and (4) Normal BMD group (NG): patients with T-scores > −1 and no prevalent fragility fractures (n=1437, 58.7% women). Multivariable logistic regression analyses were conducted to examine differences between the “severe” groups, and the two comparator groups, in the likelihood of developing a new fracture. Subclinical vertebral fractures were not examined, and fractures of the fingers, toes and face were excluded. Odds ratios (OR) and 95% confidence intervals (CI) were calculated. Over the 5 year period, 25.7% (80/331), 24.2% (80/315), 7.1% (221/3096), and 5.2% (66/1272) of the SG1, SG2, LG and NG developed new fractures. The likelihoods of developing a new fracture are summarized in the table below. 

Table  .  

Subjects with severe osteoporosis, either as defined by WHO or as defined by having 2 or more fragility fractures, are more likely to develop a new fracture as compared with participants with no prevalent fragility fractures and normal or osteopenic BMD T-scores.

Disclosures: S.M. Kaiser, Alliance for Better Bone Health 5; Servier 5; Novartis 5; Eli Lilly and Company 5.

This study received funding from: Government of Canada, Pharmaceutical Industries.

W282

An Epidemiological Study of Hip Fractures in Korean Women: Estimated from a National Health Insurance Claims Database. H. Kang*¹, K. Yang*², S. Moon³, S. Kwon*⁴, J. Park*⁵, D. Kang*⁶, S. Park*⁷. ¹School of Public Health, Yonsei University, Seoul, Republic of Korea. ²Orthopaedics, Yonsei University, Seoul, Republic of Korea. ³Orthopedics, Yonsei University, Seoul, Republic of Korea. ⁴Orthopaedics, Catholic University, Seoul, Republic of Korea. ⁵Orthopedics, Konkuk University, Seoul, Republic of Korea. ⁶Severance Hospital Clinical Trial Center, Yonsei University, Seoul, Republic of Korea. ⁷Health Insurance Review Agency, Seoul, Republic of Korea.

Purpose: Previous epidemiologic studies in Korea were limited because the incidence rate was derived from the population in a single region and they require extrapolation from small populations to the national level. Thus, this study was conducted to provide a national estimate of population incidence rates of hip fracture in Korean women over the age of 50 using the Korean National Health Insurance (NHI) claims database. Methods: All claims records of visits or admissions of female patients aged 50 or older, containing a diagnosis of hip fracture (fracture of femur (ICD-10 code: S72), fracture of neck of femur (S72.0, S72.00), pertrochanteric fracture (S72.1, S72.10), fracture of pubis (S32.5, S32.50), osteoporosis with pathological fracture (M80.05, M80.55, M80.85, M80.95)) from 2002 to 2004 were identified from the NHI claims database. The first 6-month period (January — June, 2002) was set to be a ‘window period,’ such that patients were defined as incident cases only if their first records of fracture visit or admission was observed after June 30, 2002. Among the incidence cases identified, we excluded traumatic fractures and those with multi-fracture. Results: Average annual incidence was 17,826 fractures. Age-standardized annual incidence rates adjusted to Korean population of 2000 and to US non-Hispanic population of 1990 were 305.0 and 459.2 per 100,000 women, respectively. The incidence rate of hip fracture increased with age, ranging from 22.0 per 100,000 for 50-54 years old to 1,988.5 for 85 or older. Mortality rate within 1 year following hip fracture was 177.3 per 1,000 patients, with an increase in an exponential fashion after age 65. Conclusion: Our results support a high incidence rate of hip fracture for women living in Korea. The high incidence rate is a sentinel signal of the impact of hip fracture in Korea, in terms of high mortality, morbidity, and health care costs to treat hip fracture. It is hoped that this national epidemiological study contributes to raising the awareness of hip fractures among elderly population and to motivating the adoption of effective treatment options for osteoporosis needed to prevent the increase in hip fracture.

Disclosures: H. Kang, Sanofi-Aventis 2.

This study received funding from: Sanofi-Aventis.

W283

Risk Factors for Osteoporotic Fractures in Brazilian Men and Women — The BRAZilian Osteoporosis Study (BRAZOS). M. M. Pinheiro¹, R. M. Ciconelli*¹, L. A. Martini², M. B. Ferraz*¹. ¹Rheumatology, Unifesp, Escola Paulista de Medicina, Sao Paulo, Brazil. ²Nutrition, Faculdade de Saude Publica, Universidade de Sao Paulo, Sao Paulo, Brazil.

Background/ Aims: Osteoporotic fractures are an important public health problem in worldwide. Several European, North American and Asian studies have demonstrated that the identification of risk factors for osteoporosis has a relevant role to early diagnosis and for treating people at high risk. The BRAZOS is the first epidemiological study performed in a representative sampling of the Brazilian population. The purpose of the present study was to evaluate the risk factors for osteoporotic fractures in Brazilian men and women older than 40 years. Patients and Methods: A total of 2420 subjects (1695 women) were enrolled in the BRAZilian Osteoporosis Study — BRAZOS, undertaken in 120 cities across 5 regions (North, Northeast, Central, Southeast and South) of the country, and included people from five categories of economical level (from A to E). This study evaluated aspects on lifestyle, fractures, dietary intake, physical activity, falls, quality of life, willingness to pay and knowledge on osteoporosis by quantitative and personal research that was applied in a face-to-face way. The sampling was based in the census data from IBGE (Brazilian Institute of Statistics and Geographic) 2000 and PNAD (National Research of Home Sampling) 2003 and it was calculated accordingly with to the probabilistic and representative sample of Brazilian population. The coefficient of variation is 2.2% with 95% confidence interval. After that, the data were slanted to total Brazilian population. Results: Subjects were classified as White (50%), Mixed Race (35%), Black (13%), Asian (1%) and Indian (1%) race. Mean age, height and weight for men and women were 58.4±12.8 and 60.1±13.7 years, 1.67±0.08 and 1.56±0.07 m and 73.3±14.7 and 64.7±13.7 kg, respectively. Men (56%) and women (51%) had BMI > 25.1 kg/ m². Menopause's mean age was 47 years and 39% of them were at postmenopausal and 15% was currently taking hormone therapy. Osteoporotic fracture was reported by 5.29% of the subjects and the main sketetal sites were forearm, femur, ribs, spine and humerus. Osteoporosis was observed in 6% of this population. Conclusions: Our results have demonstrated that main risk factors for osteoporotic fractures in Brazilian men and women were age, sedentary, familial history of hip fracture, smoking and low intake of calcium and vitamin D. This study suggests that a better comprehension of clinical risk factors in our population could be an important tool to identify men and women with higher risk for osteoporotic fractures.

Disclosures: M.M. Pinheiro, Grants from Wyeth Health Consumer 2.

This study received funding from: Wyeth Health Consumer.

W284

National Differences in the Risk for Falls and Fractures. L. Dukas*¹, M. Runge*², E. Schacht³. ¹Acute Geriatric University Clinic, Kantonsspital Basel, Basel, Switzerland. ²Center for Muscle and Bone Research, Aerpah Klinikum, Esslingen, Germany. ³Zurich Osteoporosis Research Group ZORG, Zurich, Switzerland.

A Creatinine clearance (CrCl) of <65ml/min. was described in multiple national studies from Germany, Switzerland and the USA as a significant risk factor for falls and fractures. In a multinational study in 7 different European countries including Russia we investigated this association. We also investigated if there are national differences in the risk for falls and fractures. For this cross-sectional study data we included 1190 women and 127 men aged 60year and older with a mean BMI of 27.2kg/m2 and a mean CrCl of 58.6ml/min., newly diagnosed and therefore untreated osteoporosis. Patients with a CrCl of <65ml/min.had in multivariate controlled analyses a 72% significantly increased risk for falls (RR 1.72, 95%CI 1.30-2.29,p<.001) and a 47% significantly increased risk for fractures (RR 1.47, 95%CI 1.05-2.07, p=.02) compared to patients with a CrCl of > 65ml/min. Interestingly we found a multivariate-controlled significant difference in fall incidence according to the country of origin: Within the past 12 month we found in patients from Russia, Lithuania, Hungary, Romania and Slovenia a 40-70% higher incidence of falls and fractures compared to patients from the Czech Republic and Latvia (p=0.04 res. p<.0001). In this study we found that a CrCl of <65ml/min is a significant risk factor for falls and fractures and are thereby inline with the results of other studies. We also found that the country of origin is also a significant risk factor for falls and fractures in elderly people. This result might be explained by the well-known national differences in the health and nutritional status of elderly people.

Disclosures: E. Schacht, Teva Pharmaceutical Industries 3.

This study received funding from: TEVA Pharmaceutical Industries.

W285

Detection of Prevalent Vertebral Fractures Using a Combination of Physical Examination Maneuvers. K. Siminoski¹, K. Lee*², H. Jen*², R. Warshawski*². ¹Radiology and Internal Medicine, University of Alberta, Edmonton, AB, Canada. ²Radiology, University of Alberta, Edmonton, AB, Canada.

Studies of several physical examination maneuvers have shown that they can be used to detect prevalent vertebral fractures when used individually. In this study, we have assessed the accuracy of vertebral fracture detection when the results of three such maneuvers are combined into a single spinal evaluation. Subjects were women referred for specialist assessment of osteoporosis (n=24l; average age 53.5 yrs; range: 18-92 years). Vertebral morphometry was performed on all subjects from T4 to L4. Prevalent vertebral fracture was defined as a vertebral height ratio < 0.80. One or more fractures were present in 24.7%; the average number of fractures among those with fractures was 2.2. Three physical parameters were determined: HHL (historical height loss, tallest recalled height minus current measured height using a wall-mounted stadiometer), WOD (wall-occiput distance, the distance between the back of the head and the wall, with the patient standing erect and with feet and buttocks touching the wall, facing forward in the Frankfort plane, quantified using a tape measure); and RPD (rib-pelvis distance, the distance between the top of the pelvis and the lowest rib in the mid-axillary line, in fingerbreadths). Points were assigned to each of the physical findings as follows: HHL: ≤ 2.0 cm = 0 points, 2.1 to 4 cm = 1 point, 4.1 to 6.0 cm = 2 points, 6.1 to 8.0 cm = 3 points, > 8 cm = 4 points; WOD: ≤ 3 cm = 1 point, 3.1 to 6.0 cm = 2 points, ≤ 6 cm = 3 points; RPD: 0 fb = 4 points, 1 fb = 3 points, 2 fb = 2 points, 3 fb = 1 point, 4 fb = 0 points. The values for the three maneuvers were added to produce a score for each patient (referred to as the Combined Score) and the accuracy of the Combined Score for vertebral fracture detection was assessed. The area under the receiver operating characteristics curve was 0.78 (95% confidence interval, 0.72-0.85; p<0.001). Accuracy results are shown in the table.

A Combined Score < 3 rules out the presence of vertebral fracture with moderate accuracy (negative predictive value of 84% in this population) while a Combined Score ≥ 6 rules in the presence of vertebral fracture with moderate accuracy (positive predictive value of 82%). Combining physical exam maneuvers that assess spinal deformity may lead to enhanced detection of prevalent vertebral fractures. 

Table  .  

LR= likelihood ratio; OR = odds ratio; Sens = sensitivity; Spec = specificity

Disclosures: K. Siminoski, None.

W286

Prognostic Factor on MRI Findings Inducing Prolonged Intractable Pain due to Pseudoarthrosis Following Osteoporotic Vertebral Fracture. T. Tsujio*¹, H. Nakamura*², H. Terai*¹, A. Matsumura*¹, M. Hoshino*¹, A. Suzuki*¹, K. Takayama*¹, K. Takaoka¹. ¹Orthopaedic Surgery, Osaka City University Medical School, Osaka, Japan. ²Orthopaedic Surgery, Osaka City General Hospital, Osaka, Japan.

Severe pain following osteoporotic vertebral fracture usually subsides in accordance with bony union of the fracture. In some cases, however, severe pain continues due to delayed union of the fracture site in the vertebral body. The purpose of this study was to evaluate the prognostic factor of osteoporotic vertebral fracture on MRI in the early stage after the injury.

Fifty-eight patients, 62 vertebrae, with fresh osteoporotic vertebral fracture were enrolled in the study. Seven were men and 51 were women. The age at the time of injury was 73.7 years in average. Those cases were followed for six months prospectively and were judged whether bone union was obtained based on vertebral cleft within the vertebral body. In those cases, MRI findings at the time of injury both on T1 and T2 weighted image were evaluated.

Bony union was obtained in 52 cases and pseudarthrosis was observed in 10 cases. On plain X-ray, the collapsing ratio of vertebral body at the time of injury was not different each other, but the posterior wall of vertebral body was frequency involved in the pseudarthosis group. Cases showing wide low intensity change in the vertebral body both on T1 and T2 weighted MRI tended to become pseudoarthrosis. Also cases having local high intensity change on T2-weighted MRI had a tendency to become pseudoarthrosis. Recently vertebroplasty has been a treatment option for pseudoarthrosis following osteoporotic vertebral fracture. If we can estimate the prognosis of the fracture at the time of injury, we can utilize this procedure in the early stage following the fracture and prevent the prolonged pain and disability of the patients. Based on the present study, involvement of posterior wall of vertebral body, large low intensity change both on T1- and T2-weighted MRI and partial high intensity change on T2-weighted MRI seems to be prognostic factors for pseudoarthrosis.

Disclosures: T. Tsujio, None.

W287

Osteoporotic Fragility Fractures in Older Adults Admitted in a French Primary Care Department: A Prospective Population-based Study. M. Rancier*¹, S. Petit*¹, A. Bellou*², V. Pascal-Vigneron*¹, N. de Talancé*¹, I. Chary-Valckenaire*³, H. Coudane*⁴, D. Mainard*⁵, G. Weryha¹. ¹Department of Endocrinology, CHU de Nancy, Vandoeuvre, France. ²Department of Endocrinology, Service d'Accueil des Urgences, Nancy, France. ³Department of Rheumatology, CHU de Nancy, Vandoeuvre, France. ⁴Department of Endocrinology, ATOL-Orthopedy, Nancy, France. ⁵Department of Endocrinology, COT- Orthopedy, Nancy, France

Primary care units diagnose most osteoporotic fractures. During one year we have recorded all low energy fractures in patients attending the primary care department of the University Hospital of Nancy (France). Patients are 45 years old and over.

Population: 929 women(73.8 years+/-13.5), 365 men (65.4 years +/-14.3). 28 died in the first month; 100 patieents were already followed by a rheumatologist. Complete exploration of osteoporosis was performed in 170 women and 72 men.

Fractures: Women: 150 wrist; 283 femoral neck; 52 vertebra; 120 humerus; 19 pelvis; 302 other. Men: 32 wrist; 99 femoral neck; 20 vertebra; 46 humerus; 14 pelvis; 133 other. Fourteen double fractures; 2 triple fractures. 72% of patients experienced their first osteoporotic fracture.

Age distribution of fractures: Data are summarized in table 2. In men, fracture distribution has a bimodal shape with higher incidence in younger ans older patients. In all age groups, only femoral neck fractures show the expected sharp increase in elderly.

DEXA (Z-score+/-SD): Performed in 170 women and 72 men.

Conclusion: Less than 8% of our patients are usually followed by a medical specialist of osteoporosis. Primary care units are the best place to record and patients with a recent osteoporotic fracture. The population profile is different from patient detected by DEXA in primary prevention. This assessment strongly supports the necessity of a multidisciplinary cooperation 

Table  .  

Disclosures: G. Weryha, None.

W288

High Leisure Time Physical Activity and High BMI Protect Against Fractures, but Heavy Work and Smoking Increase Risk — A 30-Year Follow-up of 7202 Men. L. W. Wilhelmsen*¹, K. L. L. Landin-Wilhelmsen², P. Trimpou*¹, A. Odén*¹. ¹Department of Medicine, Sahlgrenska University Hospital, Institution of Medicine, Göteborg, Sweden. ²Department of Medicine, Sahlgrenska University Hospital, Section for Endocrinology, Göteborg, Sweden.

The purpose was to study life style factors for fracture risk in a random population of 7202 men aged 46-56 years at start and followed for > 30 years (203 051 person-years). Physical activity graded from low (1) to heavy work or regular training (4). Stress was coded from never experienced stress (1) to continuous stress during last year (5) or last 5 years (6). Never smoked was coded (1), stopped smoking (2), smoking 1-14 cig:s/day (3), 15-24 cig:s/day (4) and 25 or more (5). Coffee in cups/day.

Osteoporotic fractures, n=1031, were retrieved from the hospital registers.

All variables were of significant importance for fractures except stress and coffee consumption. Fracture risk decreased with higher level of physical activity during leisure time, but increased with higher level at work (unexpected) and with increasing age. By use of Poisson regression, which included interaction between the predictor and follow up time, we could estimate how the importance of the predictor changed by time (age not included) 

Table  .  

The effect of changes for the entire population was calculated. At the beginning of follow-up the hazard rate (HR) for previous smokers versus never smoked was 1.05, and after 15 years 1.0. Thus the 15-year horizon was selected. 49.7% were smokers, and the calculated reduction of HR after 15 years if all smokers stopped smoking was 6.9%. Low physical activity was reported by 84.2%, and if they were assumed to train regularly the estimated reduction of HR after 15 years was 6.9%. Men with BMI lower than the three cut offs 24, 25 and 26 kg/m², were 43.0%, 57.3% and 70.0%. The calculated reductions after 15 years if they instead had BMI = 24, 25 and 26 kg/m² would be 3.1%, 4.9% and 7.3%, respectively.

In summary, leisure time physical activity and high BMI protect against fractures, but heavy work activity and smoking increase fracture risk among men between age 50 and 80 years.

Disclosures: L.W. Wilhelmsen, None.

W289

Physical Impact of Morphometric Vertebral Fractures in Men: Geelong Osteoporosis Study. J. A. Pasco, M. J. Henry*, S. Korn*, G. C. Nicholson, M. A. Kotowicz. Clinical & Biomedical Sciences, Barwon Health, The University of Melbourne, Geelong, Australia.

This cross-sectional study compared physical functioning in elderly men with and without morphometric vertebral fractures (MVFs). Using vertebral morphometry from lateral scans of the spine (Lunar Prodigy Pro, T10-12 and L1-4), moderate and severe wedge, biconcave or compression deformities (>25% reduction in any height and accompanying area) were identified in a population-based, age-stratified, random sample of 556 men (median age 75.0yr, range 60-93; 99% white) enrolled in the Geelong Osteoporosis Study.

The timed ‘Up-&-Go’¹ and functional reach² tests were performed as clinical measures of functional mobility and balance, respectively. Physical activity scores, incorporating household, sport and leisure activities, were determined by questionnaire³ and categorised into tertiles representing high, medium and low levels. All statistical analyses were age-adjusted.

Fifty-five men had MVFs, 26 lumbar and 35 thoracic. Age-stratified prevalence was 4.7% for 60-69yr, 10.0% for 70-79yr, 14.6% for 80+yr and prevalence for the entire group was 8.1% (age-standardised, Australia 2001). Ninety-three percent were unaware of their vertebral deformity. MVFs were associated with longer ‘Up-&-Go’ times (10.0 (95%CI 9.3, 10.7) vs 9.2 (9.0, 9.5) secs, P=0.03) and lower height-adjusted functional reach on the non-dominant side (31.0 (28.8, 33.2) vs 34.1 (33.4, 34.7) cm, P=0.008). No difference was detected on the dominant side. Using high physical activity scores as the referent group, there were increased odds for MVF in the medium [OR=2.45 (95%CI 1.04, 5.77)] and low [OR=2.69 (1.16, 6.27)] groups.

Despite most men being unaware of their condition, MVFs were associated with compromised functional mobility, balance and reduced levels of physical activity.

¹Podsiadlo D et al. JAGS 1991;39:142. ²Duncan P, et al. J Gerontology 1990;45:192³Voorrips L, et al. Med Sci Sports Exerc 1991;23:974

Disclosures: J.A. Pasco, None.

W290

Fracture and Sway Analyzed by Computed Posturography. T. Fujita¹, M. Ohue*², Y. Fujii³, A. Miyauchi³, Y. Takagi⁴, T. Sugishita*⁵. ¹Medicine, Calcium Research Institute, Katsturagi Hospital, Kishiwada, Japan. ²Orthppedic Surgery, Katsturagi Hospital, Osaka, Japan. ³Medicine, Calcium Research Institute, Kishiwada, Japan. ⁴Medicine, National Hospital System Hyogo Chuo Hospital, Hyogo, Japan. ⁵First Division, The Tazuke Kofukai Medical Research Institute, Osaka, Japan.

In a preliminary study comparing spinal and non-spinal fracture in 723 patients consulting Osteoporosis and Osteoarthritis Clinic of Katsuragi Hospital, augmented sway was suggested in spinal but not in non-spinal fracture patients. In order to define the role of sway in each sway factor more precisely, fracture state in 420 patients with complete BMD and sway analysis were classified into 4 groups; 271 (64%) had no fracture (O), 63 (15%) only non-spinal fracture, 58 (14%) only spinal fracture, and 288 (7%) both spinal and non-spinal fracture. Lumbar bone mineral density (LBMD) was measured by DXA (DPX) and distal radial trabecular bone density (t), midradial cortical bone density (c) and midradial cortical bone volume relative to the whole bone (cv) by pQCT (Norland-Stratec). Sway parameters; total length of gravity center track (LNG), track density per unit area indicating postural controlling efficiency (LNGA) and rectangular area covered by the track (REC) were measured by computed posturography (Anima) with eyes open and then closed. The difference among each group was analyzed by analysis of variance followed by multiple comparison using PLSD, Scheffe and Bonferroni-Dunn method. LMD was found to be significantly lower in B 0.810 ±0.03 gm/cm² (p=0.0075 and 0.0005) and S 0.851±0.02 gm/cm² (p=0.0075 and 0.0007) than in O 0.953 ±0.01 gm/cm² (Scheffe and Bonferroni-Dunn). LNG expressing the degree of sway was also significantly larger in B 105 ±11(p=0.0046) than O, but not significantly different from O in S, 88 ±5 (p=0.1973), suggesting a close association of the increase of sway with spinal fracture. On eye closure, insufficient postural control effciency in the group with spinal fracture alone became clearer, compared with the group with both spinal and non-spinal fracture. Persistently significant negative regression was noted on pQCT -measured bone factors LNG, REC and LNG on eye closure suggesting sway aggravation on decrease of these pQCT-measured factors, whereas no significant regression of sway factors except for REC on eye closure suggesting a decreases sway on decreasing LBMD was noted, indicatings different influence on sway. Computed posturography appears to be useful in evaluating sway as a risk factor for fall and fracture.

Disclosures: T. Fujita, None.

W291

Does Low Subjective Well-Being Predict Falls and Fractures in Postmenopausal Women?, R. J. Honkanen¹, H. T. Koivumaa-Honkanen*², M. T. Tuppurainen¹, H. P. Kröger³. ¹BCRU, Clinical Research Center, University of Kuopio, Kuopio, Finland. ²Psychiatry, Kuopio University Hospital, Kuopio, Finland. ³Surgery, Kuopio University Hospital, Kuopio, Finland.

Some studies suggest that depression might be associated with osteoporosis. A 4-item life satisfaction (LS) scale is strongly correlated with depressive symptoms (r=0.7). It is associated with and predicts morbidity, mortality and health behaviour. The purpose was to examine if subjective well-being as measured with LS scale predicts falls and fractures in postmenopausal women.

The study population was formed of the 9403 OSTPRE cohort women (born in 1932-41) who responded to all the four OSTPRE enquiries in 1989, 1994, 1999 and 2004. The follow-up time was 5 years: 1999-2004. Self-reported follow-up fractures were validated by perusal of patient records: the number of women with fractures was 835, including 322 women with distal forearm fracture (DFF) and 28 women with hip fracture. A total of 3312 women reported a fall during the past 12 months in 2004. LS scale (range 4-20, mean 7.95 (SD 2.7) categorized as satisfied 4-6, intermediate 7-11, dissatisfied 12-20) was formed with four questions about happiness, interest in life, loneliness and ease of living. Logistic regression was used as the statistical method.

At baseline, the strongest correlate of LS was self-rated health (r=0.336)(p<0.001). Also work disability and number of health disorders were associated with LS (p<0.001). Old fractures before year 1989 were a correlate of dissatisfaction (r=0.035, p=0.001), whereas fractures in 1989-99 were not and former DFF correlated with satisfaction r=-0.024, p=0.022).

Life dissatisfaction predicted falls: falling risk was 19% higher in the intermediate group and 56% higher in the dissatisfied than in the satisfied (p<0.001). After adjusting for weight, height, calcium intake, fracture history, time of menopause, HRT and number of health disorder (or self-rated health) these risk estimates were only slightly reduced. Life dissatisfaction did not predict fractures in general or any specific fracture type except hip fracture. As a continuous (but not categorized) variable it predicted hip fracture (p=0.005) with a 16.5% increase in risk for a unit increment of LS scale. This prediction weakened but did not vanish after adjusting for weight, height, fracture and falling history, HRT, time of menopause, calcium intake and number of health disorders. Adding self-rated health to the model changed the association nonsignificant (p=0.119).

The results suggest that low subjective well-being (as an indicator of depression) predicts falls and hip fracture but not fractures in general in late postmenopausal.women before old age.

Disclosures: R.J. Honkanen, None.

W292

Low-Dose Estrogen Oral Contraceptive Use and Bone Mineral Density in Adolescents and Young Adult Women. L. Ichikawa*¹, D. Scholes¹, A. Z. LaCroix², S. M. Ott³. ¹Group Health Cooperative, Seattle, WA, USA. ²Fred Hutchinson Cancer Research Ctr, Seattle, WA, USA. ³Univ of WA, Seattle, WA, USA.

Oral contraceptive (OC) use may adversely impact bone mineral density (BMD) in young women. Use of low-estrogen dose OC has recently been increasing, but few studies have examined the association of these OC formulations and BMD in young women. We report results from a cross-sectional study of OC use and BMD, comparing adolescents 14-18 years old and young adult women 19-30 years old. Study invitation letters were sent to potential study participants who were selected from the computerized databases of Group Health, a Washington state HMO, based on age and OC prescriptions. We enrolled 606 women who were 20 mcg ethinyl estradiol (EE) OC users (n=148), 30-35 mcg EE OC users (n=241), or OC non-users (n=217) at the time of their visit. OC non-users may have had some past OC use but it was limited to no use in the last 2 years and less than 8 years lifetime use. BMD (DEXA) was measured at the hip, spine, and whole body. Women were asked about their contraceptive history and OC brand and start and stop dates were collected. Data on factors related to bone health were collected using health and food frequency questionnaires. OC users were more likely to be White and smokers. Young adult women had less calcium intake and physical activity and more smokers compared to adolescent women. Average current, continuous OC use was 6 months (range 0.1-37) among adolescents and 13 months (range 0.1-135) among young adults. OC use is often intermittent and life-time exposure was 8 months (range 0.3-53) in adolescents and 42 months (range 0.3-172) in young adults. We found no association between OC group and BMD at any anatomic site among adolescents after adjusting for age, race, BMI, calcium intake, physical activity, period regularity, and current smoking status. For young adults, the adjusted mean spine BMD by OC group was significantly different with the 20 mcg EE OC users 3.5% lower compared to 30-35 mcg EE OC users (p=0.03) and 3.3% lower compared to OC non-users (p=0.04) (Table). Patterns were similar but non-significant for the hip and whole body. Results did not differ when OC non-users were restricted to those who never used in the past and the amount of past OC use of a different formulation was restricted to less than one year. Our data suggest the use of 20 mcg EE OC may suppress bone mass accrual in young adult women. 

Table  .  

Disclosures: L. Ichikawa, None.

This study received funding from: US National Institutes of Health (NICHD).

W293

The Association of Pulse Wave Velocity and Bone Mineral Density. Y. Kim, Y. Kim*, S. Lee*. Family Medicine, Pusan National University Hospital, Busan, Republic of Korea.

Background: Increased arterial stiffness and decreased bone mass is age-related disease. So we investigated association of arterial pulse wave velocity(PWV) and bone mineral density(BMD).

Subjects and Metonds: In this study, 304 postmenopausal women who underwent annual physical health check-ups recruited. Brachial-ankle PWV were measured using automated device. The BMD was measured using dual-energy X-ray absorptiometry.

Results: The mean age of the subjects was 57.3 ± 6.0 years.202 subjects who have normal BMD and 102 subjects who have osteopenic and osteoporotic BMD. There were no significant difference of PWV between subjects with normal BMD and with osteopenic and osteoporotic BMD(P>0.05, 1452.6 ± 228.0 cm/sec, 1495.3 ± 247.1 cm/sec). But after adjusting for age, subjects with osteopenic and osteoprotic BMD had a significantly higher PWV(R = 0.270, P<0.01).

Conclusions: Postmenopausal women with normal BMD had a lower arterial stiffness than with osteopenic and osteoporotic BMD. This result suggest normal BMD postmenopausal women may have a lower cardiovascular disease with osteopenic and osteoporotic BMD. Key words: postmenopausal women, BMD, PWV

Disclosures: Y. Kim, None.

W294

Clinical Risk Factors for Fracture Are Additive in Postmenopausal Women Who Are at Risk for Fracture in a Primary Care Setting. J. LaFleur*¹, C. McAdam-Marx*¹, C. V. Asche*¹, S. Alder*², X. Sheng*², D. I. Brixner*¹, S. Silverman*³. ¹Department of Pharmacotherapy, University of Utah, Salt Lake City, UT, USA. ²Department of Family and Preventive Medicine, University of Utah, Salt Lake City, UT, USA. ³Cedars-Sinai/University of California, Los Angeles, CA, USA.

Background: Increasing number of clinical risk factors (CRFs) has been linked to higher fracture risk in postmenopausal (PM) patients. We characterized hip fracture (HF) risk associated with combining multiple CRFs among at-risk patients in a primary care setting.

Methods: A historical cohort study of primary care health records from 1995-2005 was conducted to identify PM patients at risk of fracture. Patients had a 395-day run-in period and an average follow-up of 29 months. CRFs of interest during run-in were age, body mass index (BMI), bone mineral density (BMD), fracture since age 50, and maternal osteoporosis history. Other CRFs including race, drug exposures, and comorbid diseases were also evaluated. Cox models were used to estimate the risks of HF over time for each CRF in univariate and multivariable models. Risks for patients with combined CRFs were calculated.

Results: We identified 50,783 patients. Mean age was 66.8 years (SD 8.6). BMD T-scores were present in 6.8% of which 52.4% and 25.9% had T-scores in the osteopenia and osteoporosis ranges, respectively. Most patients were overweight or obese (55.8% with BMI ≥25); 2.8% of the population was underweight (BMI<18.5). A prior fracture since age 50 was identified in 3.5% of which 41.9% and 21.1% had a prior vertebral fracture (VF) and HF, respectively. Maternal history was not available. Race was identified for 30.9% of which 82.8% were Caucasian.

Incident HFs occurred in 0.76% with a mean time-to-event of 23 months. In univariate analyses, increasing 5-year age increments, lower BMI, prior fracture, and race were statistically significant predictors of HF risk. BMD was not significant, but the number of patients with BMD data was small. Of the primary CRFs, only age, BMI, and prior fracture were used in the multivariable models due to a lack of univariate significance or missing data. Risks for HF were markedly higher for older patients with both low BMI and prior fractures versus younger patients with higher BMI and no fracture. Underweight patients age 70-74 with no prior fracture, a prior VF, or a prior HF had 42, 143, and 155 times the risk, respectively, compared to a younger obese woman with no fracture. These risks increased to 86, 290, and 315 for patients age 80 and above.

Conclusion: We studied the utility of CRFs for predicting HF risk in a primary care setting using historical data. CRFs for HF included age, BMI, and prior fragility fracture. The CRFs were additive, thus combinations of CRFs could be useful for identifying PM women at high risk for HF.

Disclosures: J. LaFleur, NPS Pharmaceuticals, Inc. 2.

This study received funding from: NPS Pharmaceuticals, Inc.

W295

Factors Associated with Lack of Osteoporosis Care at the Time of Fragility Fracture. J. J. Laughren*, B. G. Escott, S. E. Ward*, V. Elliot-Gibson*, D. E. Beaton*, E. R. Bogoch. Mobility Research, St. Michael's Hospital, Toronto, ON, Canada.

Purpose: Previous work has identified low rates of preventive OP care when people arrive for treatment of an OP related fracture. Identification of those least likely to be on care could help streamline or prioritize care in a busy fracture clinic. The purpose of this study is to identify the factors associated with not being on OP care at the time of presentation with a fragility fracture.

Methods: Cross sectional survey of women > 40 years and men > 50 years presenting to fracture clinic with a fragility fracture. Data collected at baseline via a self-reported survey included socio-demographics, previous osteoporosis diagnosis, treatment and risk factors as well as self-reported osteoporosis awareness and health beliefs. A patient was identified as receiving treatment for osteoporosis if they were taking any anti-resorptive or anabolic bone agent at the time of presentation. Potential predictors were evaluated first at a univariate, unadjusted odd-ratios of being on care and then in a multivariable logistic regression. The data was were randomly split in two prior to analysis, and with the modeling rerun in a the validation set of data. Results were compared for stability between multivariable models. Adjusted odds ratios are reported from the validation model.

Results: 508 people participated in this study. The average age was 68.5 years and there were 111 (21.9%) males. 165 patients (32.5%) were on care for OP at the time of presentation. Multivariable analysis revealed that being male (adjusted OR: 3.39; 95% CI: 1.22-9.43), not perceiving that OP could have caused the fracture (adjusted OR: 6.39; 95% CI: 2.96-13.81) and not perceiving that one has low bone mass (adjusted OR: 10.78; 95% CI: 4.10-28.36) were predictive of a lack of osteoporosis care in both design and validation multivariable models. While the effect of age<60 and previous history of a fracture were not stable in both design and validation models, their effect was in the same direction. After multivariable modeling, fracture site was not predictive of being on osteoporosis care.

Conclusions: Male patients and patients with poor awareness of osteoporosis and its risks are less likely to be receiving treatment for osteoporosis at the time of fracture. While men are at lower risk of osteoporosis than women, our findings may suggest that care should be exercised not to overlook these patients should be screened at time of presentation. Patients' poor awareness of OP and its risks may be both a cause and effect of their lower rates of treatment. While patients already on treatment may be more aware of OP and it's risks as a result of exposure and education, patients with poor beliefs may render thembe less willing to accept treatment if offered by GP or professional.

Disclosures: J.J. Laughren, None.

W296

Age Increases the Risk of Subsequent Fractures Following an Initial Fracture among Women in Managed Care Population. R. Lindsay¹, N. N. Borisov*², M. Steinbuch*². ¹Helen Hayes Hospital, West Haverstraw, NY, USA. ²Procter & Gamble Pharmaceuticals, Mason, OH, USA.

The study objective was to assess the risk of subsequent fracture with age among women 65 years and older utilizing integrated administrative medical claims databases (Ingenix Lab/Rx^TM and Medstat MarketScan®).

A retrospective cohort study was conducted among women who presented with a new fracture between July 1, 2000 and December 31, 2003. The study included non-traumatic closed fractures at nine sites: hip, femur, tibia/fibula, humerus, clavicle, pelvis, forearm, wrist, and spine. The cohort was followed for at least 12 months after the initial fracture to identify subsequent fractures and to evaluate the proportion of women who received treatment for osteoporosis (nasal calcitonin, raloxifene, alendronate, or risedronate).

From a total of 448,769 eligible women, 19,554 women were identified with a new fragility fracture during the study period. Nonvertebral fractures represented 77% of the all index fragility fractures in this population. The average follow-up time after the index fracture was 24 months (range 12 months to 54 months). During this period, 72% did not receive any treatment for osteoporosis. The occurrence of subsequent fragility fractures increased from the youngest to oldest age cohorts, peaking at 4% vertebral and 10% nonvertebral fractures.

Subsequent fractures and treatment pattern after a fragility fracture by age groups 

Table  .  

*significantly different from age group 65-74 at p < 0.05 **significantly different from age groups 65-74 and 75-84 at p < 0.05

In this study, age was observed to have a significant effect on risk of a subsequent fracture after the initial fracture. The risk of subsequent fractures (vertebral and nonvertebral) increased with age. The majority of women remained untreated after their initial fracture.

Disclosures: R. Lindsay, Procter & Gamble Pharmaceuticals 5.

W297

Association of Self-Reported Frailty with Bone Mineral Density (BMD) and Incident Fractures in an Elderly Population. S. Ma, J. Oyler*, T. J. Yokes. Medicine, University of Chicago, Chicago, IL, USA.

Although BMD declines with chronological age, little is known about the association between BMD and “biological age” or frailty. The Vulnerable Elders Survey (VES-13) measures self-reported functional impairment and is widely used to assess frailty. We examined whether the VES-13 can predict low calcaneal BMD and incident fractures in the elderly.

207 elderly community subjects (age 65-95 yrs, mean 77±7 yrs, 170 women, 37 men, 127 African-Americans[AA], 80 Caucasians[CA]) had calcaneal BMD measured by the PIXI densitometer, and frailty assessed by the VES-13. Besides the VES-13 score (range 0-10), a modified score was developed that excluded age, a known strong predictor of BMD. The modified total score (range 12-60) was the sum of the points for general health, 6 functional activities and 5 activities of daily living (ADL), with higher scores reflecting greater impairment. Functional and ADL subscores were created by adding the points for the respective functional and ADL items. The two subscores were added to create a Physical subscore reflecting physical impairment without age or general health. Six years after the initial assessment, each subject or relative was contacted by phone to ascertain incident fractures and/or death. 114 of the 207 subjects completed the follow-up.

At the initial evaluation, BMD significantly correlated with age (r=-0.32), weight (r=0.56) and height (r=0.42), was higher in men and in AA. In multivariate linear regression with BMD as the outcome, only the modified total score (p=0.013) and subscores (p=0.007 for Physical) remained significant after controlling for age, weight, sex and race. Based on the model, the difference in BMD between subjects with the lowest and highest Physical subscores was 0.102 g/cm², corresponding to 1.275 T-score units and a doubling of the fracture risk.

At the follow-up evaluation, 13 of 114 subjects were deceased and 17 had incident fractures. Death was associated with higher baseline frailty using the VES-13 score (p=0.01) and even more significantly with the Physical subscore (p=0.001), providing validation for both scoring methods. Fractures were associated with lower baseline heel BMD (OR=1.65 for fracture per 1 unit T-score decrease, p=0.02). The positive association between fractures and Physical subscore was not statistically significant (p=0.06), likely due to small numbers.

We conclude that frail elderly subjects have lower calcaneal BMD than expected for their age, sex, race and weight. Assessing frailty with an easily-obtained, self-reported measure like the VES-13 can identify elderly community subjects with lower BMD who would benefit most from an aggressive approach to the diagnosis and treatment of osteoporosis.

Disclosures: S. Ma, None.

W298

Epidemiology of Chronic Obstructive Pulmonary Disease and Osteoporosis: The E.O.L.O. Study. S. Maggi*¹, P. Siviero*¹, S. Gonnelli², S. Battucci*³, D. De Feo*³, G. Guglielmi⁴, L. Sartori⁵, R. Nuti², G. Crepaldi⁵. ¹Aging, National Research Council, Padova, Italy. ²Internal Medicine, University of Siena, Siena, Italy. ³Procter & Gamble, Rome, Italy. ⁴Department of Radiology and Division of Endocrinology, Scientific Institute Hospital “Casa Sollievo della Sofferenza, San Giovanni Rotondo, Italy. ⁵Internal Medicine, University of Padua, Padova, Italy.

Chronic obstructive pulmonary disease (COPD) is a complex disease and a major cause of morbidity and mortality worldwide. In the last few years a link has been identified between COPD and other systemic diseases, such as cardiovascular disease, diabetes and osteoporosis. Several factors contribute to the development of osteoporosis in COPD patients, such as smoking, Vit. D deficiency, low BMI, decreased free fatty mass, hypogonadism, immobility, use of glucocorticoids. (Iqbal et al., Chest 1999)

EOLO ((Evaluation of Obstructive Lung disease and Osteoporosis) is a study aimed at evaluating the prevalence of osteopenia, osteoporosis and vertebral fractures in a sample of patients affected by COPD treated or not with GCs. Fifty-nine Teaching and General Hospitals in Italy participated in the study. A total of 3030 outpatients with COPD, aged 50+ yrs, were enrolled. All subjects had undergone a BMD evaluation (QUS) and a latero-lateral chest X-ray for thoracic spine evaluation.

The results of this study showed:

• The prevalence rates of osteopenia and osteoporosis in COPD patients are higher than in the general population of same age and sex (Figure)

• The frequency of osteoporosis is associated with COPD severity, even after adjusting for age, sex, BMI, smoking and GCs use (Table)

Sex-specific multivariate analyses did not show differences between men and women in the association of predictors and outcomes.

This study confirms that COPD patients are at higher risk of osteoporosis compared to the general population. 

Table  .  

* Variables in the model: gender, age, BMI, smoking, GCs use, COPD

Disclosures: S. Maggi, None.

W299

Effects of Doxorubicin on Bone Mineral Content and Density in Rats. M. A. McNulty*¹, L. C. Sharkey*¹, C. C. Curran*², C. S. Carlson*¹. ¹Veterinary Population Medicine, University of Minnesota, Saint Paul, MN, USA. ²Biostatistics Design and Analysis Center, University of Minnesota, Minneapolis, MN, USA.

Doxorubicin (DOX) is a chemotherapeutic agent commonly used to treat a variety of cancers. This agent has known cardiotoxic effects and DOX treatment has been reported to be toxic to osteoblasts in vitro; however, little work has examined its effects on bone in vivo. The purposes of this study were 1) to evaluate the effects of 6 or 8 weeks of treatment with DOX on bone mineral content (BMC) and density (BMD) in rats; and 2) to determine if these effects were influenced by strain or gender. Six- to eight-week old male and female rats of 3 different strains (SHHF, SHR, and WKY) were treated with saline (Controls) or were injected subcutaneously with 2mg/kg of DOX once per week for 6 (DOX6) or 8 (DOX8) weeks. Each group contained 6 or 7 rats for a total of 120 animals. The rats were euthanized 12 weeks after the end of the treatment period and post-mortem dorsoventral whole body bone scans were taken using a GE Lunar Prodigy Densitometer. The spine of each animal (T2-L6) was analyzed using the GE Small Animal software to determine BMC & BMD values for this site. BMC and BMD data for treatment and strain were evaluated by ANOVA, and for gender, strain, and treatment after adjusting for pretreatment body weight using ANCOVA. BMC and BMD were significantly decreased by treatment in a dose dependent manner (P<0.001 for both BMC & BMD) and also were significantly affected by strain (P<0.001 for both BMC & BMD). The SHHF strain had significantly higher BMC & BMD values than both the SHR & WKY strains, and the SHR strain had slightly higher BMC values and significantly higher BMD values than the WKY strain. After adjusting for all other factors, the following were observed: 1) There was a significant effect of DOX treatment on BMC & BMD (p<0.001). The control group had a significantly higher mean BMC & BMD than the DOX6 group (P< 0.05), but not the DOX8 group; 2) There was a significant association between strain and mean BMC & BMD (P<0.001), with SHHF & SHR strains having higher mean values for both than the WKY strain (P<0.001); and 3) There was a significant gender by strain interaction (P<0.001). Specifically, males had on average significantly higher mean BMC & BMD values than females within strains SHHF (P<0.001 for both BMC & BMD) and SHR (P = 0.001 for BMC, P<0.001 for BMD). Within strain WKY, males had slightly lower mean BMC & BMD values, but these differences were not significant (P = 0.316 for BMC, P = 0.05 for BMD). These results demonstrate that DOX treatment in young rats results in significant reductions in BMC and BMD compared with saline-treated control rats and that the choice of rat strain and gender are important considerations.

Disclosures: M.A. McNulty, None.

This study received funding from: University of Minnesota Grant-in-Aid.

W300

Long-term Use of Oral Anticoagulants and Risk for Osteoporosis. E. Stenova*¹, J. Payer², Z. Killinger*², L. Baqi*². ¹1st Department of Internal Medicine, University Hospital In Bratislava, Bratislava, Slovakia. ²5th Department of Internal Medicine, University Hospital In Bratislava, Bratislava, Slovakia.

Introduction: Vitamin K allows for gamma-carboxylation of glutamyl residues, a conversion that activates clotting factors and bone proteins, including osteocalcin, which is a marker of osteoblastic activity. Vitamin K antagonists such as warfarin inhibit this process. Studies of oral anticoagulant use and BMD have produced conflicting results. There were usually retrospective, limited by small sample sizes with very short follow-up period and no control group.

Aim: The presented prospective study was conducted to demonstrate the effect of long-term oral anticoagulation therapy on bone mass.

Material and Methods: 54 patients / postmenopausal female 23, male 31/of age 50 years or older treated with warfarin of mean dosage 4,5mg per day were enrolled in this study. The mean INR was 2-2, 5. 62 age- and sex-matched volunteers were also randomized according to same exclusion and inclusion criteria as healthy controls, without warfarin treatment. Hip and spine bone density was measured by dual X-ray absorptiometry with Hologic Delphi scanner. Sera were analyzed for osteocalcin /OSC-S/, C-terminal cross-linking telopeptide of type I collagen /CTx-S/, alkaline phosphatase /ALP-S/ and calcium/Ca-S/. 24-hour urine samples were collected for detection of urinary calcium loss /Ca-dU/. The BMD and the concentrations of laboratory markers were measured at baseline and after 12 months.

Results: No differences between warfarin-treated and control groups in markers of calcium metabolism / Ca-S, Ca-dU/, ALP-S and BMD at baseline were observed. After 12 month there was also no significant difference of these markers between treated and untreated groups. The concentrations of CTx-S /p< 0,05/ and OSC-S /p<0,001/ were significantly lower in warfarin-treated group. We determined a significant elevation of Ca-S /p<0, 05/ and Ca-dU / p<0, 05/ after 12 month in warfarin-group. These results were not recognized in controls. There were no significant changes of BMD after 12 month in either group.

Conclusion: There were no changes of bone mineral density in both warfarin-treated and control group after 12 month. The significantly lower OSC and CTx serum levels suggest that vitamin K has an influence on bone turnover, but this effect is probably modest. The long-term anticoagulant treatment is associated with higher urinary calcium loss and elevation of calcium serum level. The mechanism of these laboratory findings is unclear and should be confirmed by other investigations including determination of vitamin D level.

Disclosures: J. Payer, None.

W301

The Osteoporosis Self-Assessment Tool vs. Alternative Triage Tests: a Comparative Systematic Review of Accuracy. B. Rud¹, J. Hilden*², L. Hyldstrup¹, A. Hróbjartsson*³. ¹Osteoporosis Unit 545, Hvidovre University Hospital, Copenhagen, Denmark. ²Department of Biostatistics, Faculty of Health Sciences, University of Copenhagen, Copenhagen, Denmark. ³The Nordic Cochrane Centre, Rigshospitalet, Copenhagen, Denmark.

Previous studies have suggested that the Osteoporosis Self-Assessment Tool (OST), which is an index based on age and weight, may be as accurate as alternative triage tests in selecting postmenopausal women with low BMD for assessment by DXA. Our aim was to systematically compare the accuracy of OST and alternative triage tests in postmenopausal women.

We searched PubMed, Embase, Web of Science, citation lists and conference proceedings for studies that compare the accuracy of OST and alternative triage tests using a paired design. Our main measure of accuracy was the diagnostic odds ratio (DOR). By using the Moses-Littenberg method we summarised DOR in pair wise meta-analyses for OST and alternative triage tests when three or more studies were available, otherwise we performed qualitative summaries. We evaluated methodological quality by using the QUADAS checklist.

We identified 21 relevant studies of which 16 provided data for the pair wise metaanalyses. In whites, regardless of measurement regions and threshold for low BMD, summary estimates of DOR (sDOR) for OST were consistently at least as high as sDOR for the clinical decision rules SCORE, ORAI and SOFSURF (p-values > 0.02). By contrast, sDOR was almost twice as high for Stiffness Index assessed by quantitative ultrasonography using Achilles devices (GE-lunar) than for OST with respect to lumbar spine T-score ≤ −2.5 (p = 0.005). Similar results were found across measurement regions and thresholds for low BMD, although meta-analyses were unfeasible because studies were few. A few studies in whites suggested that weight is only slightly less accurate than OST regardless of how low BMD is defined. Between-study heterogeneity in estimates of DOR varied considerably between triage tests, measurement regions and thresholds for low BMD. Studies were few in Asian and black women and meta-analyses were unfeasible, but OST and alternative triage test were in general equally accurate. Methodological quality was generally low; no studies compared the accuracy of the triage tests prospectively as used by clinicians in settings where decisions about DXA referral are common.

Only the QUS variable Stiffness Index was consistently more accurate than OST in white women; however the transferability of our findings to clinical settings is uncertain due to low methodological study quality.

Disclosures: B. Rud, None.

W302

Dental Status, Low Bone Mass, and other Osteoporosis-related Conditions as Predictors of Adverse Outcomes in an Elderly Population. E. Musacchio*¹, E. Perissinotto*², P. Binotto*², F. Silva-Netto*¹, M. C. Corti*³, G. Baggio*⁴, S. Zambon*¹, E. Manzato*¹, G. Crepaldi⁵, L. Sartori¹. ¹Clinica Medica I, Dpt of Medical and Surgical Sciences, University of Padova, Padova, Italy. ²Dpt of Environmental Medicine and Public Health, University of Padova, Padova, Italy. ³Direzione Sanitaria, ULSS 16, Padova, Italy. ⁴Azienda Ospedaliera, Padova, Italy. ⁵Institute of Neurosciences, CNR, Padova, Italy.

We have previously reported that, in an aging (65+ years old) Italian population of both sexes cross-sectionally evaluated (ProVA Study cohort), the condition of low remaining number of teeth alone identified a larger proportion of impaired subjects as compared to the presence of osteoporosis alone. We now further investigated this issue through the data resulting from the longitudinal phase of ProVA consisting of two follow up at 5 (F1) and 7 (F2) years. The prevalence of edentulism in our population was 43.7% at baseline (n=3058), 46.2% at F1 (n=2176), and 44.9% at F2 (n=160), while the prevalence of subjects with 20+ teeth was 15.8%, 13.0% and 13.1% respectively. Logistic regression analyses with stepwise forward selection were performed in order to estimate the independent contribution of number of teeth, low bone mass and other osteoporosis-related conditions — after adjustment for confounders and including in the model variables hypothesized to be in the causal pathway — at baseline, in predicting physical impairment (PI), mortality, femoral fracture and institutionalization at follow up. With respect to PI, a low basal T-score was not critical, while edentulism appeared to be an independent risk factor with an OR=1.65 (CI 95%: 1.19-2.27). A stronger significant association (OR=3.38) was found considering separately edentulous subjects that did not wear prosthesis. Interestingly, analysis at F2, which was in general consistent with that F1, revealed an effect of low vitamin D levels (OR=3.69) not present at F1. With respect to mortality, low T-score and vitamin D levels had OR=1.38 and 1.99 respectively, while for low functional remaining teeth, OR was 1.45. Incident femoral fractures were associated with low T-score (OR=2.35) and previous femoral fracture (OR=3.39), but not with the number of teeth. Institutionalization was associated with previous falls (OR=2.85). Our data support the hypothesis that, in the elderly, oral status is a more reliable tool than osteoporosis in predicting impaired physical activity and mortality. Oral condition assessment could then have a role in the daily clinical practice for the identification of subjects at risk.

Disclosures: L. Sartori, None.

W303

The Direct Assessment of Non-vertebral Fracture in Community Experience (DANCE) Study: Baseline Demographics and Reasons for Initiating Teriparatide Therapy. A. Sebba*¹, R. Sierra-Zorita*², P. Miller*³, P. Chen*⁴, K. Taylor*⁴, M. Wong*⁴, K. Krohn*⁴. ¹Arthritis Associates, Palm Harbor, FL, USA. ²Univ of Puerto Rico School of Medicine, Hato Rey, PR, USA. ³Colorado Center for Bone Research, Lakewood, CO, USA. ⁴Eli Lilly and Company, Indianapolis, IN, USA.

Randomized, placebo-controlled clinical trials have shown the efficacy of teriparatide (TPTD) in treating osteoporosis (OP) and preventing fractures, but therapeutic decisions may differ in community practice. The ongoing, prospective, observational DANCE study examines the long-term effectiveness, safety, and tolerability of TPTD in a heterogeneous “real world” population of patients with co-morbidities, severe OP, and/or prior OP therapy, in a community setting. Study investigators prescribe TPTD 20 μg/d for up to 24 months to patients, and follow them for another 24 months. For the 4057 patients enrolled in DANCE, baseline demographics (Table) and reasons for initiating TPTD are described here.

Physicians cited reasons for initiating TPTD therapy in 4010 patients, of which 3335 (83.2%) had ≥1 fracture risk factors. The most frequent reasons included very low BMD (61.2%), previous self-reported osteoporotic fracture (33.8%), general frailty (20.7%), advanced age (18.8%), propensity to fall (17.7%), and family history of OP fractures (17.3%). In 2938 patients who previously used OP therapies, 596 patients (20.3%) had intolerance, and 2041 (69.5%) had an inadequate response, which included a decline (44.5%) or no change (10.9%) in BMD, or occurrence of new fractures (20.8%). It appears that in addition to BMD, several other clinical risk factors are important in the physicians' decisions to initiate TPTD therapy. Data from DANCE may clarify the rationale used by physicians in the community for assessing antiresorptive therapy and initiating TPTD therapy, to facilitate an evidence-based method of patient management.

Table Table. Baseline Demographics of DANCE Study Cohort (Mean ± SD)

Disclosures: A. Sebba, Eli Lilly and Company 2.

This study received funding from: Eli Lilly and Company.

W304

Peripheral Artery Disease and Osteoprosis in Older Adults. D. von Mühlen*, M. A. Allison*, S. K. Jassal*, E. Barrett-Connor. Family and Preventive Medicine, UCSD, La Jolla, CA, USA.

Increased rates of bone loss, osteoporosis and osteoporotic fractures have been reported in patients with cardiovascular disease, suggesting a relationship between osteoporosis and atherosclerosis. Limited information is available on whether low bone mass is associated with vascular disease. We examined the cross-sectional and longitudinal association between peripheral artery disease (PAD) and bone health in a sample of community-dwelling older adults. We studied 525 men and 806 women, aged 30 to 97 (mean age=73.8, SD=9.2) who attended a baseline clinic visit in 1992-1996, when ankle-brachial index (ABI) and BMD were measured, and spine x-rays were taken. In 1999-2002, 321 men and 517 women attended a follow-up visit, when BMD was re-assessed and incident OP fractures were queried. PAD defined by an ABI ≤ 0.90 was present in 15.4% of the women and 13.3% of the men. Compared to those without PAD, participants with PAD were older (77 vs. 73 years old, p<0.001), leaner (BMI = 24.5 vs. 25.3, p=0.02), and more likely to have hypertension (77% vs. 70% p=0.05), diabetes (12% vs. 7%, p=0.02) and to be current smokers (12% vs. 6%, p=0.002). They were also more sedentary (38% vs.25%, p<0.001) and reported less frequent alcohol intake (38% vs. 47%, p=0.01) than participants without PAD. Among those with PAD, the prevalence of osteoporosis by T score at the femoral neck and hip was significantly higher in women (59% vs. 49% and 25% vs. 17% respectively, p<0.05), but not men. Women with PAD also had a significantly higher rate of bone loss at the hip (-0.85%/year vs. −0.52%/year, p=0.05). However, the associations between PAD and bone mass did not persist after adjusting analyses for age and/or body mass index. At baseline more women than men had a vertebral and/or non-vertebral osteoporotic fractures (13% vs.8% and 12% vs. 7% respectively, all p's<0.01). After a mean follow up of 3.7 (SD=0.9) years there were no sex differences in the incidence of non-vertebral fractures (8.3% in women and 8.9% in men). PAD was not associated with prevalent or incident osteoporotic fractures in either men or women. In conclusion, we found an association between PAD with osteoporosis and bone loss in women, but not men. However, the associations were not independent of age, body weight or other confounders.

Disclosures: D. von Mühlen, None.

W305

High Prevalence of Abnormal Lumbar Spinal DXA Scans in Men Receiving Methadone Maintenance Therapy for Opiate Dependence. K. M. Wesa*¹, L. S. Hafner*², D. P. McGuire*³, C. L. Smith¹, J. J. Monk*⁴, G. A. Carlson*¹, K. E. Ensrud*⁵, R. H. Grimm*¹. ¹Department of Medicine, Hennepin County Medical Center, Minneapolis, MN, USA. ²Department of Experimental and Clinical Pharmacology, University of Minnesota, Minneapolis, MN, USA. ³Thomson West, Eagan, MN, USA. ⁴Bone & Mineral Metabolism, Davita, Minneapolis, MN, USA. ⁵VA Medical Center & University of Minnesota, Minneapolis, MN, USA.

One area of chronic opiate use that has not been extensively examined is the correlation between narcotic-induced hypogonadism and associated side effects such as osteoporosis in men receiving methadone maintenance therapy (MMT). This is a case-control study comparing lumbar spine densitometry by DXA scan and sex-hormone levels in 29 men receiving MMT for opiate dependence and 28 age-matched controls not on MMT.

Serum testosterone (T), estradiol (E), lutenizing hormone (LH), sex-hormone binding globulin, albumin, 25(OH)vitamin D, 1,25(OH)2vitamin D and lumbar spinal bone mineral density (BMD) were measured in all participants. Serum samples were obtained 24 hours post methadone dosing for those on MMT.

Both groups were similar regarding age, physical activity, calcium intake and BMI. Mean length of time on MMT was 5.4 years (1–25, ± 5.6) with a mean daily methadone dose of 92 mg (13–150, ± 29). 

Table  .  

For the cases length of time receiving MMT was negatively correlated with serum T (p=0.035). Across all subjects there was a significant negative correlation between T level and both length of time receiving MMT (p=0.001) and daily methadone dose (p=0.001). Serum T was positively correlated with E (p=0.003) across all subjects as was E level and L2-4 BMD (p=0.014). The cases had 9% lower BMD compared with the controls (p=0.029). Vitamin D deficiency was present in 25 (86%) of the cases and 25 (89%) of the controls. Controlling for tobacco exposure, daily methadone dose strongly predicts E (p=0.001) but not T (p=0.092) or BMD (p=0.068); and years of MMT strongly predicts T (p=0.000) but not E (0.261) or BMD (0.446). Controlling for alcohol use, daily methadone dose strongly predicts E (p=0.001) but not T (p=0.065) or BMD (p=0.198); and years of MMT strongly predicts T (p=0.000) but not E (p=0.339) or BMD (p=0.668).

There are many osteoporosis risk factors which are amenable to modification in the MMT population. Routine screening of sex-hormone levels, Vitamin D status and bone densitometry should be strongly considered and smoking cessation encouraged.

Disclosures: K.M. Wesa, None.

This study received funding from: Minneapolis Medical Research Foundation.

W306

Hip Geometry and Bone Fragility Among Postmenopausal Women with Rheumatoid Arthritis. N. C. Wright¹, J. Lisse*¹, T. J. Beck², T. Bassford*¹, A. Z. LaCroix*³, J. A. Cauley⁴, C. E. Lewis⁵, S. B. Going*¹, Z. Chen¹. ¹Univ. of Arizona, Tucson, AZ, USA. ²Johns Hopkins University, Baltimore, MD, USA. ³Fred Hutchinson Cancer Research Center, Seattle, WA, USA. ⁴Univ. of Pittsburgh, Pittsburgh, PA, USA. ⁵Univ. of Alabama-Birmingham, Birmingham, AL, USA.

Bone strength, comprised of bone mineral properties and bone geometry, is one of the primary determinants of osteoporotic fractures. Rheumatoid Arthritis (RA) has also been shown to be an independent risk factor for osteoporotic fractures. The overall goal of this analysis is to investigate the impact of hip geometry on bone fragility of postmenopausal women with RA. Participants of the Women's Health Initiative Observational Study (WHI-OS) from the BMD clinical centers were included in the analysis. Arthritis status was self-reported at baseline, identifying 335 (5.8%) women with RA and 2,701 (47.1%) women without any form of arthritis. BMD and hip geometry parameters (cross sectional area (CSA), outer diameter, section modulus (SM), cortical thickness, and buckling ratio (BR)) at three hip regions (narrow neck, intertrochanter, and shaft) were derived from baseline and follow-up dual energy x-ray absorptiometry scans using hip structural analysis programs developed by Beck and colleagues. After adjusting for age, height, and weight, white women with RA (n=220) had significantly higher mean CSA, outer diameter, and SM at the narrow neck, and higher mean CSA and SM at the shaft compared to white women without arthritis (n=2,105) at baseline. No statistically significant differences by RA status were seen in African American women at baseline (RA n=69, non n=310), although having RA was associated with lower mean BMD, CSA, and SM in the narrow neck and intertrochanter, and higher means at the shaft. After adjusting for age, height, weight, and the baseline parameter, RA was associated with decreases in hip strength from baseline to year 3 in African American but not white women (Table 1: coefficients and 95% confidence intervals (CI)). Though decreases were seen in both groups, the findings suggest that RA may play a larger role in hip strength in African American women. Further investigation is needed to examine how variables such as hormone use, physical activity, and medication usage affect the relationships between hip geometry and RA.

Table Table 1. Associations between RA and changes in narrow neck hip strength from baseline to year 3

(year 3 — baseline hip geometry) = β1,RA+ β2baseline hip geometry+ β3age+ β4height+ β5weight+ ϵ *p<0.10; **p<0.05

Disclosures: N.C. Wright, None.

This study received funding from: NIAMS R01-AR049411.

W307

Is Osteoporosis Related to Future Incidence of Osteoarthritis Over 10 Years, or Vice-Versa?, N. Yoshimura, S. Muraki, H. Oka*, A. Mabuchi*, H. Kawaguchi, K. Nakamura. 22nd Century Medical Center, Univ. of Tokyo, Tokyo, Japan.

To learn the contribution of the presence of osteoporosis (OP) or osteoarthritis (OA), two major disorders causing disability of the elderly, to future incidence of the other disorder, a 10-year population-based epidemiological study was conducted in a cohort of a mountain area. Among the entire 1,543 participants (716 men, 827 women), 400 subjects (200 men, 200 women) were selected randomly but evenly from each generation, and evaluations of OP and OA were performed by BMD and X-ray examinations, respectively, in 1990 as the baseline study. BMD measurement was performed by DXA (Lunar DPX) at the antero-posterior projection of lumbar vertebrae (L2-4) and the proximal femur, and was repeated after 3, 7, and 10 years. The presence of OP was diagnosed by the WHO criteria. X-ray examination was performed on antero-posterior and lateral images of thoracolumbar vertebrae (Th5-L5), and was repeated after 10 years. The presence of OA was determined according to the Kellgren/Lawrence (KL) grade (0-4) at intervertebral spaces from Th5/6 to L5/S1, and those with at least one space of KL>=3 were diagnosed as OA. The cumulative incidences of OA at intervertebral spaces during 10 years for the subjects without OA at the baseline in their 40s, 50s, 60s and 70s were 15.6%, 24.1%, 24.2% and 34.4% for men, and 27.1%, 32.6%, 38.5% and 58.3% for women, respectively. A logistic regression analysis was carried out utilizing the incidence of OA (1: Yes, 0: No) as an objective factor and the presence of OP at the lumbar spine or the femoral neck (1: OP, 0: osteopenia or normal range) as an explanatory factor after adjustment for age. There was no significant relationship between the presence of OP at the baseline and the incidence of OA in men or women (OP at the lumbar spine; men: odds ratio [OR]=2.52, 95% confidential interval [CI]=0.57-11.3, P=0.22; women: OR=0.56, 95% CI=0.23-1.35, P=0.19), (OP at the femoral neck; men: OR=0.72, 95% CI=0.29-1.79, P=0.48; women: OR=0.95, 95% CI=0.37-2.41, P=0.91). Similarly, the presence of OA at the baseline was not significantly related to the incidence of OP during a 10 year period. In conclusion, based on the observation over 10 years of a population-based cohort, the presence of OP did not predict the future incidence of OA, or vice-versa, suggesting that these disorders have independent backgrounds.

Disclosures: N. Yoshimura, None.

W308

Effects of Treatment of Ovariectomized Adult Rhesus Monkeys with Parathyroid Hormone 1-84 on Trabecular and Cortical Bone Structure and Biomechanical Properties of the Proximal Femur. J. Fox¹, M. A. Miller¹, R. R. Recker², C. H. Turner³, S. Y. Smith⁴. ¹NPS Pharmaceuticals, Salt Lake City, UT, USA. ²Creighton University, Omaha, NE, USA. ³Indiana-Purdue University, Indianapolis, IN, USA. ⁴Charles River Laboratories, Montreal, PQ, Canada.

Treatment of monkeys and humans with parathyroid hormone 1-84 (PTH) stimulates skeletal remodeling which increases trabecular (Tb) BMD and decreases cortical (Ct) BMD at locations where these bone types predominate. We have also previously reported that daily PTH treatment (5, 10 or 25 μg/kg) of ovariectomized (OVX) rhesus monkeys for 16 months increased bone stiffness at a lumbar vertebra, and decreased stiffness at the femoral diaphysis which led to increased work-to-failure (the energy required to fracture). We now report the effects of PTH treatment on bone structure and biomechanical properties at the proximal femur, a mixed trabecular and cortical bone site, in these monkeys. PTH reversed the OVX-induced decrease in DXA BMD within 3 to 7 months at the proximal femur, femoral neck, and distal femur. pQCT confirmed a significant dose-related decrease in Ct.BMD and increase in Tb.BMD at the total proximal femur, and proximal and distal femoral metaphyses of PTH-treated monkeys. The decrease in Ct.BMD resulted primarily from increased cortical area, because cortical bone mineral content was unaffected by PTH. Histomorphometry revealed significant dose-related increases in trabecular bone formation rate, volume, and number at the proximal femur. Trabecular thickness was unaffected by PTH. Osteoblast and osteoid surface were increased by PTH, but osteoid thickness was unaffected. Activation frequency and eroded surface were increased significantly by PTH, but osteoclast surface was unchanged. PTH did not affect periosteal or Haversian BFR at the femoral neck; however, cortical porosity was slightly, but statistically significantly higher in the 25 μg/kg group (2.7% vs 1.5% in OVX controls). PTH treatment had no significant effects on stiffness or peak load measured using a shear test, but work-to-failure was significantly higher. Thus, PTH 1-84 treatment induced changes in trabecular and cortical bone at the proximal femur that were similar to those that occur at skeletal sites where each bone type predominates and ultimately increased the total energy required to break the proximal femur.

Disclosures: J. Fox, NPS Pharmaceuticals 1, 3.

This study received funding from: NPS Pharmaceuticals.

W309

Pharmacokinetics and Pharmacodynamics of Human Parathyroid Hormone 1-84 in Adult Ovariectomized Rhesus Monkeys. J. Fox¹, S. Y. Smith², D. S. Wells*¹. ¹NPS Pharmaceuticals, Salt Lake City, UT, USA. ²Charles River Laboratories, Montreal, PQ, Canada.

When administered by daily injection, human parathyroid hormone 1-84 (PTH) acts as a potent anabolic agent in bone of animals and humans, and also markedly decreases the incidence of vertebral fractures in postmenopausal women with osteoporosis. We have previously reported that daily PTH administration (5, 10 or 25 μg/kg) for 16 months increases bone formation and BMD and improves the biomechanical properties of vertebral bodies of ovariectomized adult rhesus monkeys. We now report the pharmacokinetics of PTH and the pharmacodynamic responses that result from PTH exposure in these animals. Blood samples were collected before and for 24 hours after PTH injection on day 1, and again during months 3, 6, 12, and 16 of the study, and analyzed for PTH and Ca levels. Predose serum 25(OH)D and 1,25(OH)₂D levels and urine Ca (normalized to creatinine levels) were also measured. A dose-related linear increase in PTH C_max and exposure (area under the concentration-time curve) occurred. PTH doses of 5, 10 or 25 μg/kg produced plasma PTH exposures that were 2.1-, 4.1- and 13.2-fold greater than occurred in women given a 100 μg dose. T_max occurred later with the high dose (1.1 vs. 0.6 hr). Predose PTH levels were restored by ≤6 hr with the 5 and 10 μg/kg doses but remained elevated at 6 hr with the 25 μg/kg dose. Serum Ca levels increased acutely in a dose-related and consistent manner following PTH administration at each time point throughout the study. When averaged across all time points, the serum Ca T_max occurred at 3.7, 4.8, and 7.3 hr after dosing, with a C_max increase of 0.6, 1.0 and 1.8 mg/dL with the 5, 10, and 25 μg/kg doses, respectively. No sustained increases in predose Ca occurred with the 5 and 10 μg/kg doses whereas an increase of ∼0.6 mg/dL occurred with the 25 μg/kg dose. Serum 25(OH)D levels were variable but tended to be lower in PTH-treated animals; the decrease was significant with the 25 μg/kg dose at months 6 and 12. Serum 1,25(OH)₂D levels were also variable but tended to be higher in PTH-treated animals, particularly with the 25 μg/kg dose at month 6. There were no increases in fasting predose urine Ca levels with the 5 and 10 μg/kg PTH doses, but sustained increases occurred with the 25 μg/kg dose. In summary, PTH exposures 4-fold those observed in women did not result in sustained increases in serum or urinary Ca in adult rhesus monkeys.

Disclosures: J. Fox, NPS Pharmaceuticals 1, 3.

This study received funding from: NPS Pharmaceuticals.

W310

Intratrabecular Tunneling Increases Trabecular Number at Multiple Skeletal Locations in Ovariectomized Rhesus Monkeys Treated with Parathyroid Hormone 1-84. M. A. Miller¹, S. P. Bare*², R. R. Recker², S. Y. Smith³, J. Fox¹. ¹NPS Pharmaceuticals, Salt Lake City, UT, USA. ²Creighton University, Omaha, NE, USA. ³Charles River Laboratories, Montreal, PQ, Canada.

We have reported that daily treatment of ovariectomized (OVX) adult rhesus monkeys with human parathyroid hormone 1-84 (PTH) (5, 10, or 25 μg/kg) for 16 months increased trabecular bone volume (BV/TV), number (Tb.N) and connectivity at lumbar vertebra-3 (L3) and thoracic vertebra-10. We proposed that the increased Tb.N and connectivity was achieved by stimulation of intratrabecular tunneling, a remodeling event orientated parallel to the long axis of a trabecula at a non-nodal location. Bone formation followed resorption thus maintaining normal trabecular thickness (Tb.Th). Collectively, these features are important determinants of bone strength. Using histomorphometry to determine frequency of events, we have now quantified intratrabecular tunneling at L3 and extended it to investigate the effects of PTH treatment on trabecular bone at the proximal femur, distal radius and iliac crest of these animals. At L3, tunneling frequency was low in control sham and OVX animals (∼0.05/mm²) but increased significantly in PTH-treated animals (0.27, 0.49 and 0.95/mm² with 5, 10 and 25 μg/kg doses, respectively). Very similar tunneling frequencies were observed at the other 3 trabecular bone sites in all treatment groups. Iliac crest biopsies were also collected at baseline and after 6 months of treatment and showed significant time- and dose-related increases in tunnels. For example, in the 10 μg/kg PTH dose group the tunneling frequency was 0.03, 0.36 and 0.64/mm² at baseline, month 6 and month 16, respectively. Although the pattern and magnitude of response varied slightly from site to site, the PTH-induced intratrabecular tunneling significantly increased Tb.N, as well as BV/TV and bone formation rate at all sites. A modest but statistically significant increase in Tb.Th occurred only at the iliac crest. In summary, intratrabecular tunneling is rare in control monkeys, but increased substantially with PTH 1-84 treatment. This phenomenon provides a plausible explanation for the PTH-induced increase in Tb.N observed at all trabecular bone locations in OVX monkeys.

Disclosures: J. Fox, NPS Pharmaceuticals 1, 3.

This study received funding from: NPS Pharmaceuticals.

W311

The Skeletal Effects of Teriparatide in Glucocorticoid-Treated Mice. K. S. Howe¹, J. H. Long*¹, T. J. Wronski², U. T. Iwaniec³, R. T. Turner³, R. W. Braith*¹. ¹Applied Physiology and Kinesiology, University of Florida, Gainesville, FL, USA. ²Department of Physiological Sciences, University of Florida, Gainesville, FL, USA. ³Nutrition and Exercise Sciences, Oregon State University, Corvallis, OR, USA.

Synthetic analogs of glucocorticoids (GC) are widely used in treating many inflammatory diseases. However, they result in serious side effects, including osteoporosis. The purpose of this study was to determine whether treatment with teriparatide (PTH 1-34) can prevent or reverse the detrimental effects of GC on bone. Seven month old male Swiss-Webster mice were randomized by weight into 4 groups: (1) vehicle only (8 w CNTL); (2) GC for 8 w (GC8); (3) simultaneous GC and teriparatide administration for 8 w (GC-PTH8); or (4) GC for 4 w followed by GC+ teriparatide for 4 w (GC4/GC-PTH4). Either vehicle, prednisolone (2.1 mg/kg/d), or prednisolone and teriparatide (40ug/kg/d) were administered subcutaneously 6 d/w. Flurochrome markers, demeclocycline (15 mg/kg) and calcein (15 mg/kg), were injected prior to sacrifice to label mineralizing bone. Femurs and lumbar vertebrae were harvested at the end of the 8-week study and processed for cancellous bone histomorphometry. Data were analyzed using the Kruskal-Wallis test followed by a non-parametric posthoc test.

GC treatment did not induce cancellous bone loss in adult male mice. However, GC inhibited and teriparatide increased osteoblast surface (Ob. S)and mineralizing surface (MS). Teriparatide prevented bone changes associated with GC exposure (Table). Although 8 weeks of teriparatide treatment increased BV/TV in the distal femur and lumbar vertebrae (data not shown), the relative effects were greater in the distal femur (+100%, p = 0.02) than in the lumbar spine (+42%, p = 0.01). Four and 8 w of teriparatide significantly increased Ob. S, MS, mineral apposition rate (MAR), and bone formation rate/bone surface (BFR/BS). In conclusion, teriparatide effectively prevented the inhibitory effects of GC on cancellous bone formation in the distal femur. 

Table  .  

Data are reported as mean ± standard deviation. a = p ≤ 0.05 versus 8 w CNTL; b = p ≤ 0.05 versus GC8; c = p ≤ 0.05 versus GC4/GC-PTH4.

Disclosures: K.S. Howe, None.

W312

RAP-011 (a Soluble Activin Receptor Type IIA) Increases Bone Mineral Density in Combination with Prior Antiresorptive Therapy. M. Mangini*, M. Cornwall-Brady*, A. Pullen*, K. Halley*, T. Monnell*, J. Milling*, B. Haigis*, R. Kumar*, K. Underwood*, R. S. Pearsall. Acceleron Pharma, Cambridge, MA, USA.

RAP-011 is a soluble activin receptor type IIA (ActRIIA) fused to a murine IgGl-Fc region. Previously, we demonstrated that treatment with RAP-011, an activin antagonist, reversed bone loss in ovariectomized mice (Pearsall et al, J Bone Min. Res. 21(S1) 2006). We have further investigated the use of RAP-011 in combination with bisphosphonate treatment in ovariectomized mice.

To investigate the bone effects of RAP-011, 8 week old female C57BL/6 mice (N=40) were ovariectomized and allowed to lose bone for 8 weeks prior to treatment. When the mice were 16 weeks old half the mice were given a single dose of zoledronic acid (ZOL, IP, 20 ug/kg). Three days later, mice began RAP-011 (IP, 1 or 10 mg/kg, biw) treatment for 8 weeks either individually or after the initial ZOL injection. Whole body DXA and pQCT scanning was performed at baseline, 4 weeks and at the conclusion of the study. In the ovariectomized mice, treatment with RAP-011 showed a significant (p≤0.01) increase in BMD (+11%) than either ZOL (+5%) or PBS treated mice (+1%). In addition, pre-treatment with ZOL followed by RAP-011