Diaphyseal Femur Fractures in Osteogenesis Imperfecta: Characteristics and Relationship With Bisphosphonate Treatment

ORIGINAL ARTICLE JBMR Diaphyseal Femur Fractures in Osteogenesis Imperfecta: Characteristics and Relationship With Bisphosphonate Treatment Pamela Trejo, FranScois Fassier, Francis H Glorieux, and Frank Rauch Shriners Hospital for Children and McGill University, Montreal, Quebec, Canada ABSTRACT Several recent case reports have suggested that bisphosphonate treatment in individuals with osteogenesis imperfecta (OI) is causally related to atypical femur fractures. However, it is not known whether atypical femur fractures are actually more frequent in patients who have received bisphosphonates. In the present study, we retrospectively analyzed 166 femur fractures in 119 children with a diagnosis of OI that had not undergone intramedullary rodding procedures. A total of 130 fractures in 90 patients occurred in femurs with preexisting deformities (age at fracture between 1 month and 19.9 years; 43 girls). Because deformities are a typical cause of fracture in OI, deformed femurs were excluded from the analysis of atypical fractures. However, it was noted that in deformed femurs a transverse fracture pattern (one of the criteria of atypical fractures) was associated with a moderate to severe OI phenotype and not related to bisphosphonate treatment. Of the 36 fractures that occurred in nondeformed femurs (30 individuals; age at fracture between 1 month and 17.4 years; 13 girls), 11 (in nine children) occurred during bisphosphonate treatment. Three of these fractures (27%) resembled atypical femur fractures. Among the 25 femur fractures (23 patients) that occurred in the absence of prior bisphosphonate treatment, 8 (22%) resembled atypical femur fractures. Logistic regression analysis showed that bisphosphonate treatment history was not associated with the occurrence of atypical fractures. In contrast, the presence of moderate to severe OI (defined as any OI type other than OI type I) was strongly associated with atypical femur fractures. Thus, we observed an atypical appearance in about a quarter of nondeformed femur fractures that occurred in children with OI. Such atypical femur fractures seemed to be related to the severity of OI rather than to bisphosphonate treatment history. 2016 American Society for Bone and Mineral Research. KEY WORDS: BISPHOSPHONATES; CHILDREN; FRACTURES; OSTEOGENESIS IMPERFECTA Introduction Osteogenesis imperfecta (OI) is a hereditary disorder that is characterized by bone fragility and is often associated with extraskeletal findings such as blue/grey sclera, dental abnormalities, and joint hyperlaxity. (1) In the large majority of individuals with an OI phenotype the disease-causing mutation resides in COL1A1 or COL1A2, the genes that code for collagen type I. (2) On the clinical level, the most common OI phenotypes are OI types I, III, and IV. OI type I is the mildest form of OI, does not lead to bone deformities, and usually is associated with normal mobility. OI type III is the most severe form of OI compatible with survival, and is associated with skeletal deformities and limited ambulation. OI type IV is of moderate severity, intermediate between OI types I and III. (1) Three additional rare OI types (OI types V, VI, and VII) often resemble OI type IV regarding the severity of bone fragility, but have specific clinical characteristics and are caused by mutations in IFITM5, SERPINF1, and CRTAP, respectively. (3) Children with OI can benefit from multidisciplinary treatment approaches that depend on the severity of the phenotype. (1) Intramedullary rodding surgery is frequently necessary to straighten deformed femurs and tibias of children with more severe forms of OI. (4) For the past two decades, intravenous bisphosphonate infusions have been widely used to increase bone mineral density (BMD) and to decrease fracture rates. (5) When given long-term during the growing years, this approach improves spine development and to some extent decreases the rate of long-bone fractures. (6) A number of recent case reports, first on adults and later also on children, have described atypical femur fractures (AFFs) in the subtrochanteric and diaphyseal area as a potential complication of bisphosphonate treatment in OI and other rare heritable bone fragility disorders (7 14) (Table 1). However, the diagnostic category of AFF has been defined for postmenopausal women with osteoporosis, (15) where the typical fracture pattern is well characterized. In OI and similar disorders, there is much less information on the natural history of fracture epidemiology and Received in original form October 30, 2016; revised form December 19, 2016; accepted December 21, 2016. Accepted manuscript online December 26, 2016. Address correspondence to: Frank Rauch, MD, Shriners Hospital for Children, 1003 Boulevard Decarie, Montreal, Quebec, Canada H4A 0A9. E-mail: frauch@shriners.mcgill.ca Additional Supporting Information may be found in the online version of this article. Journal of Bone and Mineral Research, Vol. 32, No. 5, May 2017, pp 1034 1039 DOI: 10.1002/jbmr.3071 2016 American Society for Bone and Mineral Research 1034

Table 1. Literature Reports on Atypical Fractures in OI and Other Forms of Heritable Bone Fragility Sex OI type a Age (years) Bisphosphonate exposure Reported fracture Meier and colleagues (7) (2012) F I 75 Alendronate po for 3 years Short oblique fracture of femoral shaft Manolopoulos and colleagues (8) (2013) F IV 69 Pamidronate i.v. for 3 years; Incomplete subtrochanteric fracture zoledronic acid i.v. for 2 years Holm and colleagues (9) (2014) F I 64 Alendronate po for 3 years; Transverse fracture of femoral shaft pamidronate i.v. for 1 year; zoledronic acid for 1 year Etxebarria-Foronda and Carpintero (10) (2015) M IV 21 Pamidronate i.v. for 3 years; Short oblique fracture of femoral shaft alendronate orally for 5 years Carpintero and colleagues (11) (2015) M IV 13 Pamidronate i.v. for 2.5 years Transverse and oblique (of tibia and fibula; no femur fracture) Vasanwala and colleagues (12) (2016) F IV 11 Pamidronate i.v. for 5 years Transverse femoral shaft fractures van de Laarschot and Zillikens (13) (2016) M IJO 18 Pamidronate i.v. for 7 years; Subtrochanteric transverse fractures risedronate po for 2 years Boyce and colleagues (14) (2016) F IJO 16 Pamidronate i.v. for 2 years Subtrochanteric stress fracture po ¼ per os; i.v. ¼ intravenous; IJO ¼ idiopathic juvenile osteoporosis. a Also includes other rare heritable bone fragility disorders such as IJO. fracture patterns. It is therefore not clear that fractures resembling the AFF of postmenopausal osteoporosis are actually atypical in these rare disorders. This was highlighted in a recent case report on a man with osteoporosis pseudoglioma syndrome due to a LRP5 mutation who had an atypical subtrochanteric femur fracture even though he had not been exposed to bisphosphonates. (16) Similarly, transverse diaphyseal femur fractures have been among the most common fractures in OI, even before the bisphosphonate era. (17) Therefore, the relationship between bisphosphonate treatment and AFF in OI is not clear. In the present study we assessed the characteristics of subtrochanteric and diaphyseal femur fractures in children with OI and evaluated the potential effect of intravenous bisphosphonate treatment. We aimed at examining clinical situations that have some resemblance to AFF in postmenopausal osteoporosis. Consequently, our analysis focused on fractures that had occurred in femurs which had not previously undergone intramedullary rodding procedures and that had not been deformed prior to fracture. Fractures in deformed femurs were evaluated separately. Patients and Methods Study population This is a retrospective chart review of patients who were followed at the Shriners Hospital for Children Canada in Montreal. We reviewed the charts of all patients with a diagnosis of OI who had been evaluated before they reached 20 years of age and who had a radiographic record of at least one femur fracture. We included femur fractures in the present analysis when charts indicated that (1) radiographic documentation was available and (2) no surgical intervention had previously been performed on the femur (eg, no intramedullary rodding procedure). The retrospective observation period ranged from August 1986, which corresponds to the date of the first femur fracture listed in our database, until March 2016, when the data was locked for the purpose of the present analysis. To assess for atypical characteristics we used the criteria suggested by an American Society for Bone and Mineral Research Task force. (15) According to these criteria, atypical femur fractures are said to be present when at least four of five major features are present: (1) caused by minimal or no trauma; (2) located at the lateral cortex or complete; (3) transverse fracture pattern; (4) minimally comminuted or noncomminuted; and (5) the presence of a localized periosteal or endosteal thickening of the lateral cortex. Because fractures after minimal trauma is the hallmark of OI, trauma severity was assumed to have been minimal unless the documentation in the chart provided evidence to the contrary. All radiographs were digitized and identifying information was removed prior to evaluation. The radiographs of all femur fractures were evaluated by two pediatric OI specialists (PT and FR). In cases of interobserver disagreement, a third observer (FF) evaluated the radiographs for adjudication. Blinding of evaluators with regard to bisphosphonate treatment status could not be performed, because bisphosphonate infusions in growing children lead to sclerotic metaphyseal lines that are readily discernible on the radiographs. The study was approved by the Institutional Review Board of McGill University. Because this was a retrospective chart review, individual consent was not required. Measures Dual-energy X-ray absorptiometry was performed in the anteroposterior direction at the lumbar spine (L 1 L 4 ) using a Hologic QDR Discovery device (Hologic Inc., Waltham, MA, USA). Lumbar spine areal BMD results were transformed to age- and gender-specific Z-scores using published reference data. (18,19) Sequence analysis of OI-associated genes was performed as recently described. (2) Statistical analysis The chi square test was used to analyze group differences in the distribution of dichotomous variables. Group differences in continuous variables were compared by t test or Wilcoxon test, as appropriate. Logistic regression was used to evaluate the relationship between atypical fractures and clinical characteristics. Results were expressed as odds ratios (ORs) with 95% confidence intervals (95% CIs). The effect of potential predictor Journal of Bone and Mineral Research FEMUR FRACTURES IN OI 1035

Fig. 1. Number of evaluated femur fractures and their classification. variables was initially assessed in univariate models and then in multivariate models (all variables entered). In order to assess the effect of OI severity, OI type I (mild OI) was selected as reference category, and all other OI types were grouped together as moderate to severe OI, because the number of subjects of individual OI types was insufficient for separate analysis. All tests were two-sided using a 5% significance level. Calculations were performed using SPSS version 22 for Windows (IBM Corp., Armonk, NY, USA). Results During the observation period, 691 patients under 20 years of age and diagnosed with OI were evaluated at our institution. Of these, 132 individuals sustained a total of 195 fractures in femurs that had not undergone surgical interventions and for which radiographic documentation at the time of the fracture was available (Fig. 1). We excluded 29 fractures from further analysis because the radiographic image of the fracture was unclear, the fracture location was not subtrochanteric or diaphyseal, or the X-rays had been obtained when the healing process was too far advanced to characterize the fracture pattern. Of the remaining fractures, 130 (in 90 patients; median age at fracture 2.3 years; range, 0.1 to 19.9 years; 43 girls) had occurred in femurs with major anatomical abnormalities (developmental bone deformities such as bowing, Fig. 2A D). Femur deformities dramatically increase stress in the bone tissue (20) and consequently are a frequent and typical cause of fracture in children with OI. We therefore excluded fractures that had occurred in the context of femur deformities from the analysis of atypical fractures, but evaluated the fracture pattern in these deformed femurs separately. All of these fractures were complete and none was comminuted. The fracture pattern was transverse in 111 (85%) fractures, oblique in 17 (13%) fractures, and spiral in two (2%) fractures. Transverse fractures in deformed femurs occurred more often in the presence of a moderate to severe phenotype (defined as any OI type other than OI type I) than in mild OI (ie, OI type I). The presence of transverse fractures was not associated with bisphosphonate treatment, sex, age, or lumbar spine areal BMD Z-score (Table 2). Thirty-six fractures in nondeformed femurs were eventually available for the analysis of atypical fractures (Fig. 3A H). Individual data are shown in Supporting Table 1. These 36 fractures had occurred in 30 patients (13 girls, 17 boys) who had been diagnosed with OI type I (n ¼ 14), OI type IV (n ¼ 9), OI type V (n ¼ 1), OI type VI (n ¼ 5), or OI type VII (n ¼ 1). Sequencing of OI-associated genes had revealed disease-causing heterozygous mutations in COL1A1/COL1A2 (n ¼ 21 patients) or IFITM5 (n ¼ 1), and biallelic mutations in SERPINF1 (n ¼ 5) or CRTAP (n ¼ 1); in two patients no mutation was found. The median age at fracture was 2.7 years (range, 1 month to 17.4 years). All 36 fractures were complete, 31 (86%) were Fig. 2. Anteroposterior (A, C) and lateral (B, D) view of diaphyseal femur fractures in patients with femur deformities. (A, B) spiral fracture in a 1.5-year-old boy with OI type I due to a stop mutation in COL1A1, after 8 months of zoledronic acid treatment. (C, D) 1.6-year-old boy with OI type IV due to a c-propeptide mutation in COL1A1 after 2 months of pamidronate treatment. 1036 TREJO ET AL. Journal of Bone and Mineral Research

Table 2. Clinical Characteristics Associated With Fractures in Deformed Femurs Characteristic Transverse fracture (n ¼ 111) Non-transverse fracture (n ¼ 19) univariate multivariate Age (years), median (range) 2.1 (0.1 19.9) 3.7 (1.5 16.5) 1.0 (0.9 1.1) 1.0 (0.8 1.1) Sex (male/female), n 57/54 12/7 0.6 (0.2 1.7) 0.6 (0.2 1.9) OI severity (mild/moderate to severe), n 4/107 5/14 10 (2.3 40) 11 (2.5 49) Lumbar spine areal BMD (Z-score), mean SD 3.6 1.8 3.5 1.8 1.0 (0.7 1.3) 0.9 (0.6 1.3) Use of bisphosphonates prior the fracture (yes/no), n 76/35 11/8 1.6 (0.6 4.3) 2.3 (0.7 7.7) Comparison between fractures with a transverse pattern and fractures with other patterns (oblique, spiral). OR ¼ odds ratio calculated by logistic regression. displaced, 30 (83%) occurred in the mid third of the diaphysis, 12 (33%) were transverse, 14 (39%) were oblique/minimallyoblique, 10 (28%) were spiral, three (8%) were associated with a localized periosteal or endosteal thickening of the lateral cortex, and none were comminuted. Twenty-five fractures (in 23 patients) were observed in the absence of prior bisphosphonate exposure, and eight of these fractures (22%) resembled AFF (Fig. 1). Eleven fractures occurred during bisphosphonate treatment, and three of these fractures (27%) resembled AFF. The 11 fractures resembling AFF during bisphosphonate treatment had occurred in nine patients (five boys, four girls) who were diagnosed with OI type I (n ¼ 1), OI type IV (n ¼ 4), OI type V (n ¼ 1), or OI type VI (n ¼ 3). The fractures without atypical characteristics occurred in 21 patients (12 boys, nine girls), with the following diagnostic distribution: OI type I (n ¼ 13), OI type IV (n ¼ 5), OI type VI (n ¼ 2), or OI type VII (n ¼ 1). Univariate logistic regression analysis showed that the occurrence of a fracture with atypical characteristics was significantly associated with a moderate to severe phenotype. This relationship was still significant after multivariate analysis (included predictors: age, sex, lumbar spine areal BMD Z-score, use of bisphosphonate). However, neither univariate nor Fig. 3. Anteroposterior and lateral view of diaphyseal femur fractures without previous deformities of the femur, in the absence of prior bisphosphonate treatment (A D) and after bisphosphonate treatment (E H). (A, B) Five-year-old girl with OI type IV due to a glycine substitution in COL1A2. (C, D) Fiveyear-old boy with OI type V due to a mutation in IFITM5.(E, F) Seven-year-old girl with OI type VII due to a homozygous hypomorphic mutation in CRTAP, after 2.3 years of pamidronate treatment. (G, H) Six-year-old boy with OI type VI due to a homozygous mutation in SERPINF1 after 3.2 years of pamidronate treatment. Journal of Bone and Mineral Research FEMUR FRACTURES IN OI 1037

multivariate analysis showed a significant association between femur fractures with atypical appearance and bisphosphonate treatment, lumbar spine areal BMD Z-score, age, or sex (Table 3). Discussion In this study we observed atypical characteristics in 22% of femur fractures that had occurred in the absence of bisphosphonate treatment and in 27% of femur fractures that had occurred after the first dose of bisphosphonate treatment. The occurrence of atypical femur fractures was associated with the severity of OI but not with bisphosphonate use, sex, or lumbar spine areal BMD Z-score. Similarly, the fracture pattern that we observed in deformed femurs depended on the severity of OI but not on bisphosphonate treatment history. The notion that bisphosphonate treatment in OI is a cause of atypical fractures stems from a number of case reports which described a variety of bone lesions in the subtrochanteric and diaphyseal femur (Table 1). None of these reports compared their findings with those in patients that were not treated with bisphosphonates. This makes it difficult to establish whether the reported observations are related to bisphosphonate treatment or the underlying genetic disorder. Our finding that femur fractures with atypical appearance in OI are quite frequent in the absence of bisphosphonate therapy may reflect the fact that OI bone is abnormal on several levels. Collagen type I, the most prevalent protein in bone, is directly or indirectly affected by the mutations causing OI, which is likely to change the mechanical properties of collagen type I. (21,22) On the bone material level, calcium content is increased, leading to increased hardness, independent of bisphosphonate treatment. (23) On the tissue level, cortical porosity is markedly increased in OI and contributes to low bone strength. (24) All of these factors are likely to alter the fracture pattern of OI bone. In children with moderate to severe OI, 10 of the 20 femur fractures (50%) that we evaluated had an atypical appearance on radiographs, compared to one of the 16 fractures (6%) in children with OI type I. The reasons for this marked difference are unclear at present, but it is intuitive to assume that bone material properties and bone microstructure are more disturbed in the more severe forms of OI than in OI type I, as this is the likely cause of the more severe phenotype. Such abnormalities on the material and tissue levels may also affect fracture patterns and thus the appearance of fractures on radiographs. In accordance with our findings, OI type IV was also the most frequent diagnosis in the literature reports about atypical femur fractures in OI (Table 1). Among the nine patients in the present study who had femur fractures with an atypical appearance, three patients (33%) had a diagnosis of OI type VI caused by biallelic SERPINF1 mutations, even though only about 2% of the OI population followed at our institution have biallelic SERPINF1 mutations. (2) This apparent overrepresentation of OI type VI in the group with atypical fractures may be in part explained by the fact that children with OI type VI tend to have a large number of femur fractures in the absence of deformities, (25) and thus met inclusion criteria for the present study. In comparison, all our patients with OI type III had femur deformities and therefore none of the patients with OI type III was included in the present analysis. OI type VI is also characterized by extremely high material bone density and abnormal mineral crystal arrangement. (26) These factors may contribute to the extreme brittleness of OI type VI bone and thus could affect the way in which fractures develop in individuals with OI type VI. It is important to take into account that the labeling of fractures as atypical can lead to an early discontinuation of bisphosphonate treatment in growing children with OI. Bisphosphonate treatment is beneficial, eg, by helping to reconstitute compressed vertebra, (6) and the interruption of this treatment before growth is completed can lead to an increased fracture risk. (27) Our study is limited by the fact that this is a retrospective chart and X-ray review and therefore not all the desirable information, such as pain prior to fracture or bone turnover activity and mechanism of the fracture, was available. The X-rays had been obtained for clinical decision making rather than for a standardized assessment of fracture pattern and therefore were not optimized for the purpose of the present study; in some instances only one radiographic view of the femur (anteroposterior or lateral) was available. Subjects with OI type III are not represented in the analysis of atypical fractures because OI type III is associated with bone deformities at birth. However, our assessment did not reveal an effect of bisphosphonate treatment history on the fracture pattern of deformed femurs, and this analysis included the OI type III cohort. Finally, this study focused on fracture pattern as it appeared on the available radiographs. Because it is unknown how many fractures were sustained but not documented in this X-ray series, the present study cannot provide information on the incidence of femur fractures. In conclusion, we observed an atypical appearance in about a quarter of nondeformed femur fractures that occurred in children with OI. Such atypical femur fractures seemed to be related to the severity of OI rather than to bisphosphonate treatment history. In fractures that occurred in patients with femur deformities a transverse fracture pattern was also Table 3. Clinical Characteristics Associated With Fractures in Nondeformed Femurs Characteristic Atypical fracture (n ¼ 11) Non-atypical fracture (n ¼ 25) univariate multivariate Age (years), median (range) 3.7 (0.2 7.1) 2.3 (0.1 17.4) 1.0 (0.8 1.2) 1.1 (0.8 1.4) Sex (male/female), n 6/5 13/12 1.1 (0.3 4.6) 0.3 (0.0 2.7) OI severity (mild/moderate to severe), n 1/10 15/10 15 (1.7 136) 40 (2.4 651) Lumbar spine areal BMD (Z-score), mean SD 2.6 1.7 2.7 1.8 1.0 (0.7 1.6) 0.8 (0.4 1.5) Use of bisphosphonates prior the fracture (yes/no), n 3/8 8/25 0.8 (0.2 3.8) 0.4 (0.0 4.9) Comparison between fractures that fulfill criteria for atypical femur fractures and other femur fractures. OR ¼ odds ratio calculated by logistic regression. 1038 TREJO ET AL. Journal of Bone and Mineral Research

associated with disease severity. For further clarification, larger and prospective studies on fracture pattern and fracture epidemiology in OI seem warranted. Disclosures FF: Pega Medicale: Royalties. FHG: Novartis: Consulting fees and research grants; Amgen and Mereo Biopharma: Consulting fees and research grants. FR: Genzyme Inc: Advisory Board member; Novartis Inc: Study grant to institution; Alexion Inc: Study grant to institution. Acknowledgments This study was supported by the Shriners of North America. We thank Mark Lepik for the preparation of the figures. Authors roles: PT: data collection and analysis, evaluated X-rays, wrote the first draft of the manuscript; FF: contributed patient information, evaluated X-rays and reviewed manuscript content; FHG: contributed patient information and reviewed manuscript content; FR: contributed patient information, evaluated X-rays, reviewed manuscript content and accepts responsibility for data analysis. All authors read and approved of the final version of the manuscript. References 1. Trejo P, Rauch F. Osteogenesis imperfecta in children and adolescents new developments in diagnosis and treatment. Osteoporos Int. 2016;27:3427 37. 2. Bardai G, Moffatt P, Glorieux FH, Rauch F. DNA sequence analysis in 598 individuals with a clinical diagnosis of osteogenesis imperfecta: diagnostic yield and mutation spectrum. Osteoporos Int. 2016;27:3607 13. 3. Forlino A, Marini JC. Osteogenesis imperfecta. Lancet. 2016;387:1657 71. 4. Azzam KA, Rush ET, Burke BR, Nabower AM, Esposito PW. Mid-term results of femoral and tibial osteotomies and Fassier-Duval nailing in children with osteogenesis imperfecta. J Pediatr Orthop. Forthcoming. Epub 2016 Jul 02. DOI:10.1097/BPO.0000000000000824. 5. Glorieux FH, Bishop NJ, Plotkin H, Chabot G, Lanoue G, Travers R. Cyclic administration of pamidronate in children with severe osteogenesis imperfecta. N Engl J Med. 1998;339:947 52. 6. Palomo T, Fassier F, Ouellet J, et al. Intravenous bisphosphonate therapy of young children with osteogenesis imperfecta: skeletal findings during follow up throughout the growing years. J Bone Miner Res. 2015;30:2150 7. 7. Meier RP, Ing Lorenzini K, Uebelhart B, Stern R, Peter RE, Rizzoli R. Atypical femoral fracture following bisphosphonate treatment in a woman with osteogenesis imperfecta a case report. Acta Orthop. 2012;83:548 50. 8. Manolopoulos KN, West A, Gittoes N. The paradox of prevention bilateral atypical subtrochanteric fractures due to bisphosphonates in osteogenesis imperfecta. J Clin Endocrinol Metab. 2013;98:871 2. 9. Holm J, Eiken P, Hyldstrup L, Jensen JE. Atypical femoral fracture in an osteogenesis imperfecta patient successfully treated with teriparatide. Endocr Pract. 2014;20:e187 90. 10. Etxebarria-Foronda I, Carpintero P. An atypical fracture in male patient with osteogenesis imperfecta. Clin Cases Miner Bone Metab. 2015;12:278 81. 11. Carpintero P, Del Fresno JA, Ruiz-Sanz J, Jaenal P. Atypical fracture in a child with osteogenesis imperfecta. Joint Bone Spine. 2015;82:287 8. 12. Vasanwala RF, Sanghrajka A, Bishop NJ, Hogler W. Recurrent proximal femur fractures in a teenager with osteogenesis imperfecta on continuous bisphosphonate therapy: are we overtreating? J Bone Miner Res. 2016;31:1449 54. 13. van de Laarschot DM, Zillikens MC. Atypical femur fracture in an adolescent boy treated with bisphosphonates for X-linked osteoporosis based on PLS3 mutation. Bone. 2016;91:148 51. 14. Boyce AM, Collins MT, Tosi LL, Gafni RI. A subtrochanteric femoral stress fracture following bisphosphonate treatment in an adolescent girl. Horm Res Paediatr. Forthcoming. Epub 2016 Jul 6 DOI:10.1159/ 000447425. 15. Shane E, Burr D, Abrahamsen B, et al. Atypical subtrochanteric and diaphyseal femoral fractures: second report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2014;29:1 23. 16. Alonso N, Soares DC, E VM, Summers GD, Ralston SH, Gregson CL. Atypical femoral fracture in osteoporosis pseudoglioma syndrome associated with two novel compound heterozygous mutations in LRP5. J Bone Miner Res. 2015;30:615 20. 17. Dent JA, Paterson CR. Fractures in early childhood: osteogenesis imperfecta or child abuse? J Pediatr Orthop. 1991;11:184 6. 18. Kalkwarf HJ, Zemel BS, Yolton K, Heubi JE. Bone mineral content and density of the lumbar spine of infants and toddlers: influence of age, sex, race, growth, and human milk feeding. J Bone Miner Res. 2013;28:206 12. 19. Zemel BS, Kalkwarf HJ, Gilsanz V, et al. Revised reference curves for bone mineral content and areal bone mineral density according to age and sex for black and non-black children: results of the bone mineral density in childhood study. J Clin Endocrinol Metab. 2011;96:3160 9. 20. Fritz JM, Guan Y, Wang M, Smith PA, Harris GF. A fracture risk assessment model of the femur in children with osteogenesis imperfecta (OI) during gait. Med Eng Phys. 2009;31:1043 8. 21. Misof K, Landis WJ, Klaushofer K, Fratzl P. Collagen from the osteogenesis imperfecta mouse model (oim) shows reduced resistance against tensile stress. J Clin Invest. 1997;100:40 5. 22. Chen F, Guo R, Itoh S, et al. First mouse model for combined osteogenesis imperfecta and Ehlers-Danlos syndrome. J Bone Miner Res. 2014;29:1412 23. 23. Weber M, Roschger P, Fratzl-Zelman N, et al. Pamidronate does not adversely affect bone intrinsic material properties in children with osteogenesis imperfecta. Bone. 2006;39:616 22. 24. Albert C, Jameson J, Smith P, Harris G. Reduced diaphyseal strength associated with high intracortical vascular porosity within long bones of children with osteogenesis imperfecta. Bone. 2014;66:121 30. 25. Homan EP, Rauch F, Grafe I, et al. Mutations in SERPINF1 cause osteogenesis imperfecta type VI. J Bone Miner Res. 2011;26:2798 803. 26. Fratzl-Zelman N, Schmidt I, Roschger P, et al. Unique micro- and nano-scale mineralization pattern of human osteogenesis imperfecta type VI bone. Bone. 2015;73:233 41. 27. Rauch F, Munns C, Land C, Glorieux FH. Pamidronate in children and adolescents with osteogenesis imperfecta: effect of treatment discontinuation. J Clin Endocrinol Metab. 2006;91:1268 74. Journal of Bone and Mineral Research FEMUR FRACTURES IN OI 1039