Vertebral morphometry in children and adolescents with osteogenesis imperfecta: Effect of intravenous pamidronate treatment

Bone 39 (2006) 901 906 www.elsevier.com/locate/bone Vertebral morphometry in children and adolescents with osteogenesis imperfecta: Effect of intravenous pamidronate treatment Christof Land, Frank Rauch, Craig F. Munns, Solmaz Sahebjam, Francis H. Glorieux Genetics Unit, Shriners Hospital for Children, 1529 Cedar Avenue, Montréal, Québec, Canada H3G 1A6 McGill University, Montreal, Quebec, Canada Received 1 February 2006; revised 31 March 2006; accepted 5 April 2006 Available online 26 May 2006 Abstract Results in small patient series suggest that cyclical intravenous treatment with pamidronate can lead to reshaping of compressed vertebral bodies in children and adolescents with osteogenesis imperfecta (OI), but more detailed analyses are lacking. In this study of patients with moderate to severe OI (age range 0.1 to 16.7 years), we used vertebral morphometry to longitudinally assess changes in lumbar vertebral shape before (n = 17 patients) and during 2 to 4 years of pamidronate treatment (n = 72 patients). Anterior, posterior and midpoint vertebral heights of lumbar vertebrae L1 to L4 were determined on lateral lumbar spine X-rays and were related to vertebral body length in the antero-posterior direction. Before pamidronate treatment, vertebral body height ratios did not change significantly, but the mean concavity index (defined as the ratio between midpoint and posterior vertebral body heights) decreased by 22% (P = 0.002). Pamidronate treatment was associated with an increase in vertebral height ratio at each of the 12 sites that were analyzed. Consequently, patients who had received pamidronate for an average of 3 years had less compressed vertebrae than a historical control group of patients who had the same OI type, age and sex but who had not received pamidronate. Multiple regression analysis revealed that age was negatively and lumbar spine areal bone mineral density z score was positively associated with vertebral shape at baseline. The main determinant of treatment response was the severity of vertebral deformities at baseline. These results suggest that vertebral deformations worsen in patients with moderate to severe OI who do not receive medical treatment and that pamidronate helps to reverse this trend. In moderate to severe forms of OI, pamidronate should be started as early as possible to treat or to prevent vertebral deformations. 2006 Elsevier Inc. All rights reserved. Keywords: Bisphosphonates; Children; Metabolic bone disease; Osteogenesis imperfecta; Osteoporosis Introduction Osteogenesis imperfecta (OI) is a heritable disorder characterized by increased bone fragility and low bone mass. The most commonly used classification distinguishes four clinical types [1]. OI type I comprises patients with absence of bone deformities. Type II is lethal in the perinatal period. OI type III is the most severe form in children surviving the neonatal time. Patients with moderate bone deformities and variable short stature are classified as OI type IV. In the majority of patients with OI, the disease can be linked to mutations in one of the two genes coding for collagen type I alpha chains [1]. Corresponding author. Fax: +1 514 842 5581. E-mail address: frauch@shriners.mcgill.ca (F. Rauch). Apart from long-bone fractures and deformities, vertebral compression fractures are a common finding in moderate to severe OI. A recent study on pediatric OI patients not receiving medical treatment reported that 30% of OI type I patients and 79% of patients with OI types III or IV had either biconcave or wedged vertebra [2]. Collapsed vertebrae predispose to the development of scoliosis, which affects the majority of patients with OI types III and IV [3,4]. Kyphoscoliosis, in turn, can lead to secondary impairment of cardiorespiratory function, which is a major cause of premature death in OI [5]. Cyclical intravenous treatment with the bisphosphonate pamidronate has a beneficial effect in children and adolescents with severe OI [1]. It has been reported that this treatment increases cortical thickness of long bones, decreases fracture rates and improves mobility. Areal bone mineral density (BMD) at the lumbar spine usually increases markedly after pamidronate 8756-3282/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2006.04.004

902 C. Land et al. / Bone 39 (2006) 901 906 treatment is started. In addition, a number of investigators have reported the impression that compressed vertebral bodies regain a more normal shape during pamidronate treatment [6 9]. Indeed, lumbar vertebral size, as evaluated on antero-posterior dual-energy X-ray absorptiometry (DXA) scans, was larger in OI patients who had received pamidronate than in untreated patients [10,11]. Nevertheless, estimating vertebral body size from anteroposterior DXA scans is a rather crude method. More detailed information on vertebral size and shape can be gained with vertebral morphometry [12]. This method has been used in a large number of studies on adult osteoporosis but has rarely been applied in pediatric research. Sumnik et al. observed reshaping of lumbar vertebrae in five children with primary osteoporosis who were treated with pamidronate [13]. A recent small placebo-controlled study of pediatric OI patients found that 1 year of pamidronate treatment was associated with faster vertebral size increases [14]. In a cross-sectional study, we observed that young OI patients who had received pamidronate treatment in the first 2 years of life had better vertebral shape than untreated patients of the same age [11]. More detailed analyses in larger patient groups are lacking. In the present study, we used vertebral morphometry to investigate changes in lumbar vertebral shape in untreated children and adolescents with OI as well as in patients who received pamidronate. In addition, the influence of OI type, age, sex, weight, lumbar spine areal BMD, bone metabolism and height velocity on vertebral shape was assessed. Subjects and methods Study population The study cohort consisted of children and adolescents with OI types I, III and IV who received pamidronate therapy at the Shriners Hospital for Children in Montreal, Canada. Patients who were diagnosed with OI types V, VI and VII according to our expanded classification were not included in this evaluation [1]. Patients were eligible for pamidronate treatment if they had long-bone deformities or had suffered three or more fractures per year (including vertebrae) during the previous 2 years. Patients were included in the main study group if they had received pamidronate for at least 2 years and if lateral lumbar spine radiographs were available at the start of pamidronate treatment and after 2.0 to 4.0 years of therapy. Of the 115 patients with OI types I, III and IV who had received pamidronate for at least 2 years, adequate radiographic documentation was available in 72 cases (37 girls, 35 boys; OI type I, n = 14; OI type III, n = 21; OI type IV, n = 37). These patients constituted the treatment group. Collagen type I mutations were found in 61 of these 72 patients. In seven patients, no collagen type I mutation was detectable by sequence analysis, and four patients had not been tested. For 17 patients of the treatment group (5 girls, 12 boys; OI type I, n =5;OI type III, n = 5; OI type IV, n = 7), the radiological documentation also included a pretreatment period of at least 1 year. In these patients, the first lateral lumbar spine radiograph had been obtained an average of 3.0 years (range 1.2 8.9 years) before the commencement of pamidronate treatment, when the patients had a mean age of 4.6 years (range 0.1 12.9 years). Radiographs of these patients were used to evaluate vertebral shape changes in the absence of medical therapy. Pamidronate-associated changes in the entire treatment group were assessed in two different ways. First, results after two to four treatment years were compared with baseline results. Second, a cross-sectional analysis was performed, in which results after 2 to 4 years of pamidronate treatment were compared to those of a Fig. 1. Method of quantitative vertebral morphometry on lateral spine radiographs. Six locations were marked per vertebra, corresponding to the four corners of the vertebral body and the midpoints of the end plates. Anterior height (AH), posterior height (PH) and middle height (MH) of each lumbar vertebra were measured and expressed relative to lower vertebral length (LL). The concavity index was calculated as MH/PH. historical control group. These patients were matched for OI type and age but had not received medical treatment at the time when the radiograph was obtained. For six patients in the treatment group, no untreated matching partner was available, so that 66 matched pairs could be assessed. Pamidronate treatment was later started in all but six patients of the historical control group. Data of 42 patients from that group were also used as baseline data for the longitudinal analysis. A collagen type I mutation was found in 52 of the 66 patients in the control group. Treatment Pamidronate was administered intravenously on three consecutive days in all patients. As described in detail elsewhere, the timing and dosage of these 3-day cycles varied with age, but the yearly dose of pamidronate remained at 9 mg per kg throughout the treatment period [1]. Calcium and vitamin D intake was maintained adequate according to the recommended daily allowance. Anthropometric and biochemical measurements Height was measured using a Harpenden stadiometer (Holtain Limited, Crymych, UK). Weight was determined using digital electronic scales for infants and mechanical scales for children (Healthometer, Bridgeview, IL, USA). Height and weight measurements were converted to age- and sex-specific z scores on the basis of reference data published by the Centers for Disease Control and Prevention [15]. Mean height velocity during pamidronate treatment (given in cm/year) was calculated, dividing the height change from baseline by the length of the follow up period. Urinary cross-linked N- telopeptides of type I collagen (NTX) were quantified by enzyme-linked immunosorbent assay (Osteomark, Ostex International Inc., Seattle, Washington, USA) using the second void sample of the morning. Results for urinary NTX/creatinine ratios in OI patients were expressed as a percentage of age- and sex-specific mean values using published reference data [16]. Radiological analyses All radiographs were taken using standardized techniques with a Siemens Multix H radiological unit (Siemens AG, Erlangen, Germany) onto Kodak medical film (Eastman Kodak Co., Rochester, NY, USA). Vertebral morphometry was performed on lateral spine radiographs for the region from lumbar vertebra L1 to L4 as described by Smith-Bindman et al. [17]. Six landmarks were identified at the silhouette of each vertebra, corresponding to the four corners of the vertebral body and the midpoints of the endplates (Fig. 1). The point-to-point distances were measured using a dial reading caliper (General tools, New York, USA) with a precision of 0.1 mm. From these points, anterior, posterior and midheight as well as lower length were measured. To account for magnification, all

C. Land et al. / Bone 39 (2006) 901 906 903 interobserver comparisons for vertebral heights ranged from 3.3% (anterior height) to 5.3% (mid-height). The intraobserver and interobserver comparisons for the concavity index were 5.3 and 6.9%, respectively. Antero-posterior areal BMD and projection area of L1 to L4 were determined by dual-energy X-ray absorptiometry (DXA), using a Hologic QDR 2000W or 4500A device (Hologic Inc., Waltham, MA, USA). Areal BMD results were transformed to age-specific z scores combining reference data from Salle et al. and data provided by the densitometer manufacturer [19]. Statistics Differences in cross-sectional comparisons between groups were tested for significance using unpaired t tests or Mann Whitney's U tests, as appropriate. Paired t tests were used to analyze longitudinal changes during pamidronate treatment. All tests were two-tailed, and throughout the study, P b 0.05 was considered significant. Associations are given as Pearson correlation or Spearman rank correlation, as appropriate. Differences between different types of OI at baseline or during pamidronate treatment were tested for significance using analysis of variance with Bonferroni's adjustment for multiple comparisons. Determinants of vertebral morphometry data were assessed using stepwise multiple regression analysis. For this analysis, OI types were recorded according to disease severity (type I = 1, type IV = 2, type III = 3). All calculations were performed using SPSS software version 11.5 for Windows (SPSS, Inc., Chicago, IL, USA). Fig. 2. Schematic representation of changes of the average vertebral body shape in 17 OI patients prior to pamidronate and after 2 to 4 years of pamidronate treatment. The heights of each vertebra are given as the vertebral height ratios. Significant changes between corresponding anatomical sites when compared with the previous time point are indicated by asterisks. *P b 0.05; **P b 0.005; ***P b 0.001. measurements were expressed relative to lower vertebral length (as this is the measure that is probably least influenced by compression fractures) and are given as vertebral height ratios. The concavity index was calculated as the ratio between mid-height and posterior height, averaged for L1 to L4 [18]. The radiographic lateral projection area of vertebral bodies was calculated as the surface area of the polygons that were defined by the six measurement points. The reproducibility of the measurements was evaluated in 12 patients (48 vertebral bodies). The coefficient of variation for intraobserver comparisons of vertebral heights ranged from 1.7% (anterior height) to 2.6% (mid-height). The Results Vertebral shape changes in the absence of medical treatment were evaluated in 17 patients. During a mean follow up period of 3.0 years, no significant changes occurred in vertebral body height ratios, but the concavity index decreased from 0.95 ± 0.19 (mean ± SD) to 0.78 ± 0.15 (P = 0.002). When pamidronate treatment was instituted, almost all vertebral height ratios and the concavity index increased (Figs. 2 and 3). A more detailed analysis of treatment-associated changes in the entire treatment group (n = 72) showed the expected decrease in bone metabolism and an increase in areal BMD and anteroposterior projection area of L1 to L4 (Table 1). At the same time, vertebral body height ratios and the concavity index increased, Fig. 3. Lateral lumbar spine radiographs of a boy with OI type III at age 1.0 year (A), 9.8 years (B) and 13.0 years (C). Pamidronate treatment was started at the age of 10.1 years. Vertebral bodies flattened before treatment were started (A and B). This was partially corrected after 3 years of pamidronate treatment (C).

904 C. Land et al. / Bone 39 (2006) 901 906 Table 1 Clinical characteristics at the start of pamidronate treatment and after 2 to 4 years of pamidronate therapy in 72 OI patients n Start of PAM 2 4 years PAM Age (years) 72 5.7 (0.1 16.7) 8.8 (2.1 19.3) Height 72 5.2 ± 2.9 4.1 ± 2.2 b0.001 Weight 72 2.7 ± 1.6 1.9 ± 1.5 b0.001 untx/creatinine (nmol/mmol) 69 894 ± 448 312 ± 136 b0.001 untx/creatinine (in % of age and sex-specific reference mean) 67 114 ± 73 61 ± 43 b0.001 P Lumbar spine densitometry Bone mineral content (g) 72 5.0 ± 4.6 15.0 ± 9.7 b0.001 Areal bone mineral density 72 0.22 ± 0.10 0.50 ± 0.14 b0.001 (g/cm 2 ) Areal bone mineral density 72 5.6 ± 1.3 2.2 ± 1.2 b0.001 Projection area (cm 2 ) 72 19.2 ± 8.7 28.8 ± 10.8 b0.001 Values are given as mean ± SD or as median (range). PAM, pamidronate treatment. leading to an 87% increase in the projection area of vertebral bodies on lateral spine radiographs (Table 2). A close relationship was found between the lateral and antero-posterior projection areas, as assessed by vertebral morphometry and DXA, respectively (Fig. 4). Results after 2 to 4 years of pamidronate treatment were compared to those of historical control patients who were matched for OI type, sex and age, but who had not received medical treatment at the time when the radiograph was obtained (Table 3). For six of the treated patients, no matching partner was available, so that 66 matched pairs were assessed. Vertebral height ratios and concavity index were higher in the treated group, with the exception of posterior height ratios at L3 and L4 (Fig. 5). Determinants of baseline vertebral morphometry results were assessed in the treatment group (n = 72) using multiple regression analyses. Morphometry results for L1 to L4 were Table 2 Longitudinal study. Vertebral morphometry at the start of pamidronate treatment and after 2 to 4 years of pamidronate therapy in 72 OI patients n Start of PAM 2 4 years PAM Anterior height ratio L1 70 0.54 ± 0.24 0.69 ± 0.17 b0.001 L2 71 0.66 ± 0.17 0.74 ± 0.16 b0.001 L3 71 0.71 ± 0.13 0.78 ± 0.14 b0.001 L4 72 0.76 ± 0.14 0.81 ± 0.13 0.03 Mid-height ratio L1 70 0.53 ± 0.31 0.71 ± 0.18 b0.001 L2 71 0.57 ± 0.27 0.74 ± 0.17 b0.001 L3 71 0.60 ± 0.25 0.75 ± 0.15 b0.001 L4 72 0.60 ± 0.25 0.75 ± 0.14 b0.001 Posterior height ratio L1 70 0.67 ± 0.18 0.79 ± 0.15 b0.001 L2 71 0.66 ± 0.18 0.79 ± 0.14 b0.001 L3 71 0.64 ± 0.20 0.78 ± 0.13 b0.001 L4 72 0.64 ± 0.21 0.76 ± 0.12 b0.001 Concavity index L1 L4 72 0.83 ± 0.22 0.94 ± 0.10 b0.001 Projection area (cm 2 ) 71 8.8 ± 4.7 16.5 ± 8.7 b0.001 Values are given as mean ± SD. PAM, pamidronate treatment. Fig. 4. Relationship between antero-posterior projection area of L1 to L4 as determined by DXA and projection area of L1 to L4 in a lateral direction as measured on radiographs. The regression equation is as follows: DXA (cm 2 )= 9.35 + 1.20 vertebral morphometry (cm 2 ). averaged for this analysis and used as dependent variables. OI type, sex, age, weight and areal BMD z scores as well as urinary NTX at the start of pamidronate treatment were entered into the model as independent determinants. These analyses revealed that areal BMD z score was positively associated with all of the examined vertebral shape parameters (Table 4). A negative correlation with age was found for all shape measures with the exception of anterior height. OI type provided additional independent information only for mid-height and concavity index. To elucidate independent determinants of the changes in vertebral height ratios from baseline (Δ vertebral height ratio) during pamidronate treatment, OI type, sex, age, weight, areal Table 3 Cross-sectional study: characteristics of patients who had received 2 4 years of pamidronate therapy and of untreated OI control patients who were matched for age, OI type and sex Pamidronate group OI control group P OI type (I/III/IV) 11/21/34 11/21/34 Sex (female/male) 35/31 35/31 Age (years) 8.0 ± 4.1 8.0 ± 4.1 0.96 Anthropometry Height 4.1 ± 2.1 5.1 ± 3.0 0.04 Weight 1.9 ± 1.6 2.6 ± 1.5 0.01 Lumbar spine densitometry Bone mineral content (g) 13.8 ± 7.9 6.2 ± 4.7 b0.001 Projection area (cm 2 ) 27.4 ± 8.9 21.9 ± 7.9 b0.001 Areal bone mineral density 0.48 ± 0.12 0.25 ± 0.10 b0.001 (g/cm 2 ) Areal bone mineral density 2.2 ± 1.2 5.4 ± 1.3 b0.001 Vertebral morphometry Concavity index L1 L4 0.95 ± 0.10 0.74 ± 0.24 b0.001 Projection area (cm 2 ) 14.8 ± 6.5 10.8 ± 6.9 0.001 Values are given as mean ± SD.

C. Land et al. / Bone 39 (2006) 901 906 905 Fig. 5. Schematic representation of average vertebral body shape after 2 to 4 years of pamidronate treatment in 66 children with OI ( treated ) compared to 66 untreated patients who were matched for age, OI types and sex ( controls ). The height of each vertebra is given as the vertebral height ratio. Significant differences between results at the same anatomical sites are indicated by asterisks. *P b 0.05; **P b 0.005; ***P b 0.001. Table 4 Multiple regression models to assess determinants of vertebral morphometry results at the start of pamidronate treatment Dependent variable Parameter Regression coefficient P R 2 Anterior height Intercept 1.009 ratio Areal BMD 0.063 b0.001 Corrected model b0.001 0.29 Mid-height ratio Intercept 1.49 Age (years) 0.027 b0.001 Areal BMD 0.094 b0.001 OI type 0.120 0.008 Corrected model b0.001 0.53 Posterior height Intercept 1.27 ratio Age (years) 0.013 0.001 Areal BMD 0.10 b0.001 Corrected model b0.001 0.44 Concavity Index Intercept 1.47 Age (years) 0.025 b0.001 Areal BMD 0.051 0.02 OI type 0.104 0.01 Corrected model b0.001 0.43 The mean results of L1 to L4 were used for each dependent variable. R 2 is the coefficient of determination of the whole model. Table 5 Multiple regression models to assess determinants of changes in vertebral morphometry results from baseline during pamidronate treatment (Δ vertebral height ratio) Dependent variable Δ Anterior height ratio Δ Mid-height ratio Δ Posterior height ratio Δ Concavity index BMD z scores, baseline urinary NTX and height velocity during pamidronate treatment as well as the baseline results for each dependent variable were entered into the model (Table 5). Baseline results were negatively associated with treatmentassociated changes in all vertebral shape measures. In addition, age at treatment start predicted changes in anterior and posterior height ratios, while height velocity and sex correlated with changes in vertebral concavity. Urinary NTX and areal BMD z scores were not associated with changes in vertebral shape. Discussion Parameter Regression coefficient P R 2 Intercept 0.49 Baseline anterior 0.68 b0.001 height Age (years) 0.008 0.004 Corrected model b0.001 0.52 Intercept 0.55 Baseline mid-height 0.67 b0.001 Corrected model b0.001 0.75 Intercept 0.53 Baseline posterior 0.68 b0.001 height Age (years) 0.007 0.01 Corrected model b0.001 0.69 Intercept 0.70 Baseline concavity 0.90 b0.001 index Height velocity 0.015 0.002 Sex 0.05 0.04 Corrected model 0.83 The mean results of L1 to L4 were used for each dependent variable. R 2 is the coefficient of determination of the whole model. In this study, we found that lumbar spine vertebral bodies of children and adolescents with moderate to severe OI had increasingly abnormal shape when no medical treatment was given. After pamidronate therapy was instituted, vertebral shape improved. The extent of vertebral deformities at baseline was the main determinant of the changes occurring during treatment. The worsening of vertebral shape in the absence of medical treatment was documented in two different ways. First, vertebral bodies became more concave in the subgroup of patients in whom sufficient pretreatment follow up was available for direct observation. Second, multiple regression analysis of baseline results in the entire treatment group revealed a significant negative association between age and most vertebral shape measures. This shows that with increasing age vertebral bodies flatten and become more concave, when no medical treatment is given. Apart from age, OI type and areal BMD were independently associated with vertebral shape at baseline. The relationship between OI type and vertebral compressions is expected as patients with the more severe OI types III and IV have more vertebral deformities than patients who suffer from OI type I [2,3]. The fact that areal BMD was positively correlated with vertebral height even when OI type was accounted for suggests that areal BMD is lower in more severely compressed vertebrae, regardless of OI type. Identical observations were made in a

906 C. Land et al. / Bone 39 (2006) 901 906 recent study [20]. However, these data do not prove that areal BMD can be used to predict vertebral deformities as the DXA scans and lateral spine radiographs were obtained at the same time. Pamidronate treatment resulted in a significant reshaping of vertebral bodies. Vertebral height relative to vertebral length increased at each of the 12 sites that were analyzed. Consequently, patients who had received pamidronate for an average of 3 years had less compressed vertebrae than patients who had the same OI type, age and sex but who had not received pamidronate. These observations are in accordance with earlier reports [10,14]. Interestingly, the main determinant of treatment response was the severity of vertebral deformities at baseline, whereas areal BMD did not contribute to explain the treatment effect. The association between vertebral heights and changes during treatment was negative, showing that more compressed vertebrae gain more than vertebrae that have a more normal shape. It is noteworthy that height velocity was positively associated with improvements in vertebral concavity, suggesting that the reshaping of the vertebral bodies during pamidronate treatment depends on growth. As most of our patients were still growing, this study does not shed light on the question whether vertebral reshaping is possible after final height has been achieved. As a side topic of the present investigation, we examined the relationship between two methods to estimate bone size, DXA and vertebral morphometry. DXA is often used to document improvements in vertebral size during bisphosphonate treatment [8 10]. Our results validate the use of DXA for this purpose as a close relationship between the two methods was found. However, it should be noted that the two methods measure different things as the DXA method used in this study includes the intervertebral disk space into bone projection area. In addition, DXA scans were performed in the antero-posterior direction, whereas vertebral morphometry was done on lateral views of the spine. This and earlier studies on vertebral shape in young OI patients have important clinical implications [11,14]. First, patients with moderate to severe OI not receiving medical treatment experience worsening of vertebral deformations over time. Second, pamidronate treatment helps to rebuild deformed vertebrae in children and adolescents with OI. Third, vertebral shape is better preserved in children who start pamidronate early in life [11].It must be stressed that the present study did not include patients with milder forms of OI type I. Many of these mildly affected patients do not have vertebral compression fractures, and therefore vertebral shape improvement is not a treatment goal in such patients. However, with regard to moderately to severely affected OI patients, the present observations suggest that intravenous pamidronate treatment should be started as early as possible to treat or prevent vertebral deformations. Acknowledgments We thank Mark Lepik for preparation of the figures. This study was supported by the Shriners of North America. F.R. is a Chercheur-Boursier Clinicien of the Fonds de la Recherche en Santé du Québec. This study was supported by the Shriners of North America. References [1] Rauch F, Glorieux FH. Osteogenesis imperfecta. Lancet 2004;363:1377 85. [2] Kok DJ, Uiterwaal CS, Van Dongen AJ, Kramer PP, Pruijs HE, Engelbert RH, et al. The interaction between Sillence type and BMD in osteogenesis imperfecta. Calcif Tissue Int 2003;73:441 5. [3] Ishikawa S, Kumar SJ, Takahashi HE, Homma M. Vertebral body shape as a predictor of spinal deformity in osteogenesis imperfecta. J Bone Jt Surg Am 1996;78:212 9. [4] Engelbert RH, Gerver WJ, Breslau-Siderius LJ, van der Graaf Y, Pruijs HE, van Doorne JM, et al. Spinal complications in osteogenesis imperfecta: 47 patients 1 16 years of age. Acta Orthop Scand 1998;69:283 6. [5] McAllion SJ, Paterson CR. Causes of death in osteogenesis imperfecta. J Clin Pathol 1996;49:627 30. [6] 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. [7] Astrom E, Soderhall S. Beneficial effect of long term intravenous bisphosphonate treatment of osteogenesis imperfecta. Arch Dis Child 2002;86:356 64. [8] Zacharin M, Bateman J. Pamidronate treatment of osteogenesis imperfecta Lack of correlation between clinical severity, age at onset of treatment, predicted collagen mutation and treatment response. J Pediatr Endocrinol Metab 2002;15:163 74. [9] Arikoski P, Silverwood B, Tillmann V, Bishop NJ. Intravenous pamidronate treatment in children with moderate to severe osteogenesis imperfecta: assessment of indices of dual-energy X-ray absorptiometry and bone metabolic markers during the first year of therapy. Bone 2004;34:539 46. [10] Rauch F, Plotkin H, Zeitlin L, Glorieux FH. Bone mass, size, and density in children and adolescents with osteogenesis imperfecta: effect of intravenous pamidronate therapy. J Bone Miner Res 2003;18:610 4. [11] Munns CF, Rauch F, Travers R, Glorieux FH. Effects of intravenous pamidronate treatment in infants with osteogenesis imperfecta: clinical and histomorphometric outcome. J Bone Miner Res 2005;20:1235 43. [12] Ferrar L, Jiang G, Adams J, Eastell R. Identification of vertebral fractures: an update. Osteoporos Int 2005;16:717 28. [13] Sumnik Z, Land C, Rieger-Wettengl G, Korber F, Stabrey A, Schoenau E. Effect of pamidronate treatment on vertebral deformity in children with primary osteoporosis. A pilot study using radiographic morphometry. Horm Res 2004;61:137 42. [14] Letocha AD, Cintas HL, Troendle JF, Reynolds JC, Cann CE, Chernoff EJ, et al. Controlled trial of pamidronate in children with types III and IV osteogenesis imperfecta confirms vertebral gains but not short-term functional improvement. J Bone Miner Res 2005;20:977 86. [15] Ogden CL, Kuczmarski RJ, Flegal KM, Mei Z, Guo S, Wei R, et al. Centers for disease control and prevention 2000 growth charts for the United States: improvements to the 1977 National Center for Health Statistics version. Pediatrics 2002;109:45 60. [16] Bollen AM, Eyre DR. Bone resorption rates in children monitored by the urinary assay of collagen type I cross-linked peptides. Bone 1994;15:31 4. [17] Smith-Bindman R, Cummings SR, Steiger P, Genant HK. A comparison of morphometric definitions of vertebral fracture. J Bone Miner Res 1991;6:25 34. [18] Eastell R, Cedel SL, Wahner HW, Riggs BL, Melton III LJ. Classification of vertebral fractures. J Bone Miner Res 1991;6:207 15. [19] Salle BL, Braillon P, Glorieux FH, Brunet J, Cavero E, Meunier PJ. Lumbar bone mineral content measured by dual energy X-ray absorptiometry in newborns and infants. Acta Paediatr 1992;81:953 8. [20] Kok DJ, Uiterwaal CS, Van Dongen AJ, Kramer PP, Pruijs HE, Engelbert RH, et al. The interaction between Sillence type and BMD in osteogenesis imperfecta. Calcif Tissue Int 2003;73:441 5.