- Email: [email protected]
Association of the presence of bone bars on radiographs and hip fracture in postmenopausal Caucasian women
Clinical Radiology 67 (2012) 840e842
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Association of the presence of bone bars on radiographs and hip fracture in postmenopausal Caucasian women D.B. Sarver a, *, R. Lopez-Ben b, S.L. Morgan c, D. Rehder b, J.N. Duke d, N. Fineberg e, M.J. Pitt b a
University of Alabama School of Medicine, Birmingham, AL, USA Department of Radiology, University of Alabama School of Medicine, Birmingham, AL, USA c Division of Clinical Immunology and Rheumatology, Department of Medicine, University of Alabama School of Medicine, Birmingham, AL, USA d Department of Orthopaedics, University of Alabama School of Medicine, Birmingham, AL, USA e Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA b
article in formation Article history: Received 15 July 2011 Received in revised form 20 December 2011 Accepted 4 January 2012
AIM: To determine whether the presence of bone bars (BB) identified on anteroposterior hip radiographs are more prevalent in patients that have had a hip fracture as compared to patients without a fracture. MATERIALS AND METHODS: Ninety-two Caucasian women with a unilateral proximal femur fracture were retrospectively evaluated and randomly selected using radiology database records to comprise the investigational group. Ninety-eight age-matched Caucasian women without hip fracture were selected as a control group. Anteroposterior hip radiographs were evaluated for the presence of BBs by two musculoskeletal radiologists. Chi-square tests were used to assess whether fractures were more prevalent in patients with BB than those without BB. RESULTS: The patient population was comprised Caucasian women with a mean age of 79.8 6.4 years in the control group and 79.9 6.6 years in the investigational group. Regardless of the reader, BB were identified in a significantly higher percentage of women with a fracture (75 versus 39%, p < 0.001 or 53 versus 38%, p ¼ 0.041) as compared to those without a fracture. CONCLUSION: BB are associated with hip fracture. Their presence is a trigger for requesting a dual-energy x-ray absorptiometry (DXA) examination to confirm or refute a diagnosis of low bone mineral density (BMD) and a subsequent increased risk of fracture. Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction Linear bands of increased radiodensity, termed bone bars (BB), have previously been described in the osteopaenic skeleton of middle-aged and elderly patients.1 BB are struts
* Guarantor and correspondent: D.B. Sarver, University of Alabama School of Medicine, 1530 3rd Avenue South, FOT 1203 Birmingham, AL 35294e3412, USA. Tel.: þ1 205 617 2217. E-mail address: [email protected] (D.B. Sarver).
of normal trabecular bone that cross the medullary portions of the metaphysis and diaphysis at right angles to the long axis of the shaft.2 When the thickness of the bone bar exceeds approximately 0.3 mm, it becomes visible on a radiograph.1 BB are seen as punctate or short, linear, radiodense foci when their longitudinal axes are approximately parallel to the x-ray beam, as seen in a anteroposterior hip radiograph.1 Fig 1, an anteroposterior hip radiograph, shows the typical radiographic appearance of BB in the intertrochanteric region of the femur. BB are often seen on routine radiographs of older adult patients,
0009-9260/$ e see front matter Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2012.01.011
D.B. Sarver et al. / Clinical Radiology 67 (2012) 840e842
Figure 1 Bone Bars in intertrochanteric region of femur.
frequently in the intertrochanteric proximal femur and proximal tibial diametaphyses. A previous study has shown a positive association between BB and low bone mineral density (BMD; osteopaenia).1 BB present at the intertrochanteric region of the femur have a high predictive association for low BMD at the intertrochanteric region and also the femoral neck.3 Low BMD at a specific anatomical site is generally more predictive of fracture at that site than at other anatomic sites.4 The purpose of this investigation was to determine whether the presence of BB identified on anteroposterior hip radiographs are more prevalent in patients that have suffered a hip fracture (subtrochanteric, intertrochanteric, femoral neck).
Materials and methods This investigation was approved by the Institutional Review Board at the University of Alabama at Birmingham, USA. Patients were retrospectively evaluated and randomly selected using radiology database records and an electronic medical record information system. Inclusion criteria for the study included (1) postmenopausal Caucasian women aged 65 years or older; (2) patients with an anteroposterior radiograph of the hip taken from January 2004 to June of 2010 for reasons not associated with high-energy trauma (i.e., motor vehicle accident). Exclusion criteria included (1) previous medical conditions known to be associated with low BMD or fracture5; (2) patients who received medications known to affect BMD (osteoporosis
841
medications, glucocorticoids, hormone replacement therapies, chemotherapeutic agents, and certain anticonvulsant medications, except calcium and vitamin D); (3) patients with previous hip fracture besides the one noted for the study. After evaluation of the medical record, 92 patients with a unilateral proximal femur fracture (subtrochanteric, intertrochanteric, or femoral neck) fit the criteria and were used for the investigational group, while 98 age-matched patients without hip fracture were used for the control group. It was previously shown in a study of 134 postmenopausal women that BB present at the intertrochanteric region of the femur are bilateral 70e75% of the time.3 This allows for reasonable evaluation of BB in the contralateral hip if a hip fracture was present at the time of the radiograph. All radiographs (investigational and control) were manipulated so that only one randomly chosen side of the patient’s hip was visible to the evaluator. This allowed the evaluator to be blinded to the hip fracture present in the investigational group. The control and investigational radiographs were intermixed and randomized. The radiographs were independently evaluated by two experienced musculoskeletal radiologists [D.R. (10 years of experience) and R.L. (11 years of experience)]. The two observers met prior to the study and agreed upon uniform criteria [punctate clustered foci of increased radiodensities consistent with bone trabeculations greater than 0.3 mm in diameter within the intertrochanteric region of the proximal femur on the anteroposterior (AP) views] for the identification and evaluation of BB. The presence of BB at the intertrochanteric region of the proximal femur was scored as present or absent. Continuous data are presented as mean standard deviation. Discrete data are given as counts (%). The Kappa statistics were used to assess agreement on the presence of BB between readers. Chi-square tests were used to compare the incidence of hip fractures in patients with and without BB. p < 0.05 was considered significant.
Results The demographic characteristics of the cohort are presented in Table 1. The patient population was entirely postmenopausal Caucasian women with a mean age of 79.8 6.4 years (range 70e96 years) in the control group and 79.9 6.6 years (range 66e93 years) in the investigational group. Table 2 shows the presence of BB according to each evaluator. The kappa statistic for agreement on the presence of BB between readers was 0.41 (p < 0.001), which is considered moderate agreement. The percent agreement on BB between the two readers for all patients, patients Table 1 Demographic characteristics of the study population. Characteristic
Control group (no fracture)
Investigational group (hip fracture)
Caucasian female (n) Mean age (years) Age range (years)
98 79.8 6.4 (SD) 70e96
92 79.9 6.6 (SD) 66e93
842
D.B. Sarver et al. / Clinical Radiology 67 (2012) 840e842
Table 2 Bone bars prevalence. Musculoskeletal radiologist
Control group (no fracture)
Experimental group (fracture)
#1 #2
38 (39%) 37 (38%)
69 (75%) 49 (53%)
with fractures, and patients without fractures was 70, 67, and 74%, respectively. The prevalence of intertrochanteric BB in the control group ranged from 38e39%, depending on the reader. The prevalence of intertrochanteric BB in the investigational group ranged from 53e75%. Regardless of the reader, BB were identified in a significantly higher percentage of women with a hip fracture (75 versus 39%, p < 0.001 or 53 versus 38%, p ¼ 0.041) as compared to those without a hip fracture. The investigational group consisted of 35 femoral neck fractures, 51 intertrochanteric fractures, and six subtrochanteric fractures. Regardless of the evaluator, BB are seen in similar prevalence in patients with femoral neck and intertrochanteric fractures; no significant difference was found between locations (p ¼ 0.999 and p ¼ 0.664). There were too few patients with subtrochanteric fractures to evaluate any significance.
fracture. Thus the identification of BB in the non-fractured proximal femur of a patient is a trigger for requesting a dual-energy x-ray absorptiometry (DXA) examination with calculation of the FRAX 10 year absolute fracture risk to further assist in the diagnosis of osteoporosis and fracture risk. The present study is limited as the radiographs of the investigational group were not available prior to the incident of fracture. Thus the radiographs had to be evaluated on the contralateral side in order to blind the readers to the investigational group. This may introduce errors as BB in the intertrochanteric femur location are seen bilaterally 70e75% of the time according to a previous study.3 A more objective radiological definition of BB based on the presence or absence and size, extent, and location would enable more precise comparisons between studies. Future studies should evaluate BB in multiple locations and expand the population to include both genders and a larger range of ages and races. In addition the effect of medications, such as glucocorticoids, on BB should be evaluated.
Acknowledgement This study was supported by the NIH T35 research grant.
Discussion References In a previous study, evaluation of gross specimen sections of BB established that these radiodense foci are areas of thickened trabeculae extending from the cortex partially or completely across the medullary canal.1 It has been hypothesized that BB develop during insidiously progressing osteopaenia in mature skeletons as a response to mechanical stress.1 Alternatively, BB may have formed during skeletal growth but remain hidden and are unmasked by surrounding osseous resorption; a combination of the two hypotheses may be causative.1,2 The presence of BB is associated with lower BMD compared to patients without BB.3 The Singh Index established a visual method for evaluation of BMD based on the trabecular pattern of the proximal femur on AP pelvic radiographs.6 While some have confirmed its efficacy in the assessment of human femur bone strength, others have questioned its reliability in the age of computed radiography.7,8 In the present study the radiological presence of BB was significantly increased in post-menopausal Caucasian women who suffered a hip
1. Krusunoglu S, Pate D, Resnick D, et al. Bone reinforcement lines in chronic adult osteopenia: a hypothesis. Radiology 1986;158:409e15. 2. Pitt M. Osteopenic bone diseases. Orthop Clin N Am 1983;14:65e80. 3. Pitt MJ, Morgan SL, Lopez-Ben R, et al. Association of the presence of bone bars on radiographs and low bone mineral density. Skeletal Radiol 2011;40:905e11. 4. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996;312:1254e9. 5. Morgan SL, Saag KG, Julian BA, et al. Osteopenic bone disease. In: Koopman WJ, editor. Arthritis and allied conditions: a textbook of rheumatology. 14th ed. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 2449e513. 6. Singh M, Nagrath AR, Maini PS. Changes in trabecular pattern of the upper end of the femur as an index of osteoporosis. J Bone Joint Surg 1970;52 A:457e67. 7. D’Amelio P, Rossi P, Isaia G, et al. Bone mineral density and Singh index predict bone mechanical properties of human femur. Connect Tissue Res 2008;49:99e104. 8. Hauschild O, Ghanem N, Oberst M, et al. Evaluation of Singh index for assessment of osteoporosis using digital radiography. Eur J Radiol 2009;71:152e8.
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
Association of the presence of bone bars on radiographs and hip fracture in postmenopausal Caucasian women D.B. Sarver a, *, R. Lopez-Ben b, S.L. Morgan c, D. Rehder b, J.N. Duke d, N. Fineberg e, M.J. Pitt b a
University of Alabama School of Medicine, Birmingham, AL, USA Department of Radiology, University of Alabama School of Medicine, Birmingham, AL, USA c Division of Clinical Immunology and Rheumatology, Department of Medicine, University of Alabama School of Medicine, Birmingham, AL, USA d Department of Orthopaedics, University of Alabama School of Medicine, Birmingham, AL, USA e Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA b
article in formation Article history: Received 15 July 2011 Received in revised form 20 December 2011 Accepted 4 January 2012
AIM: To determine whether the presence of bone bars (BB) identified on anteroposterior hip radiographs are more prevalent in patients that have had a hip fracture as compared to patients without a fracture. MATERIALS AND METHODS: Ninety-two Caucasian women with a unilateral proximal femur fracture were retrospectively evaluated and randomly selected using radiology database records to comprise the investigational group. Ninety-eight age-matched Caucasian women without hip fracture were selected as a control group. Anteroposterior hip radiographs were evaluated for the presence of BBs by two musculoskeletal radiologists. Chi-square tests were used to assess whether fractures were more prevalent in patients with BB than those without BB. RESULTS: The patient population was comprised Caucasian women with a mean age of 79.8 6.4 years in the control group and 79.9 6.6 years in the investigational group. Regardless of the reader, BB were identified in a significantly higher percentage of women with a fracture (75 versus 39%, p < 0.001 or 53 versus 38%, p ¼ 0.041) as compared to those without a fracture. CONCLUSION: BB are associated with hip fracture. Their presence is a trigger for requesting a dual-energy x-ray absorptiometry (DXA) examination to confirm or refute a diagnosis of low bone mineral density (BMD) and a subsequent increased risk of fracture. Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction Linear bands of increased radiodensity, termed bone bars (BB), have previously been described in the osteopaenic skeleton of middle-aged and elderly patients.1 BB are struts
* Guarantor and correspondent: D.B. Sarver, University of Alabama School of Medicine, 1530 3rd Avenue South, FOT 1203 Birmingham, AL 35294e3412, USA. Tel.: þ1 205 617 2217. E-mail address: [email protected] (D.B. Sarver).
of normal trabecular bone that cross the medullary portions of the metaphysis and diaphysis at right angles to the long axis of the shaft.2 When the thickness of the bone bar exceeds approximately 0.3 mm, it becomes visible on a radiograph.1 BB are seen as punctate or short, linear, radiodense foci when their longitudinal axes are approximately parallel to the x-ray beam, as seen in a anteroposterior hip radiograph.1 Fig 1, an anteroposterior hip radiograph, shows the typical radiographic appearance of BB in the intertrochanteric region of the femur. BB are often seen on routine radiographs of older adult patients,
0009-9260/$ e see front matter Ó 2012 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.crad.2012.01.011
D.B. Sarver et al. / Clinical Radiology 67 (2012) 840e842
Figure 1 Bone Bars in intertrochanteric region of femur.
frequently in the intertrochanteric proximal femur and proximal tibial diametaphyses. A previous study has shown a positive association between BB and low bone mineral density (BMD; osteopaenia).1 BB present at the intertrochanteric region of the femur have a high predictive association for low BMD at the intertrochanteric region and also the femoral neck.3 Low BMD at a specific anatomical site is generally more predictive of fracture at that site than at other anatomic sites.4 The purpose of this investigation was to determine whether the presence of BB identified on anteroposterior hip radiographs are more prevalent in patients that have suffered a hip fracture (subtrochanteric, intertrochanteric, femoral neck).
Materials and methods This investigation was approved by the Institutional Review Board at the University of Alabama at Birmingham, USA. Patients were retrospectively evaluated and randomly selected using radiology database records and an electronic medical record information system. Inclusion criteria for the study included (1) postmenopausal Caucasian women aged 65 years or older; (2) patients with an anteroposterior radiograph of the hip taken from January 2004 to June of 2010 for reasons not associated with high-energy trauma (i.e., motor vehicle accident). Exclusion criteria included (1) previous medical conditions known to be associated with low BMD or fracture5; (2) patients who received medications known to affect BMD (osteoporosis
841
medications, glucocorticoids, hormone replacement therapies, chemotherapeutic agents, and certain anticonvulsant medications, except calcium and vitamin D); (3) patients with previous hip fracture besides the one noted for the study. After evaluation of the medical record, 92 patients with a unilateral proximal femur fracture (subtrochanteric, intertrochanteric, or femoral neck) fit the criteria and were used for the investigational group, while 98 age-matched patients without hip fracture were used for the control group. It was previously shown in a study of 134 postmenopausal women that BB present at the intertrochanteric region of the femur are bilateral 70e75% of the time.3 This allows for reasonable evaluation of BB in the contralateral hip if a hip fracture was present at the time of the radiograph. All radiographs (investigational and control) were manipulated so that only one randomly chosen side of the patient’s hip was visible to the evaluator. This allowed the evaluator to be blinded to the hip fracture present in the investigational group. The control and investigational radiographs were intermixed and randomized. The radiographs were independently evaluated by two experienced musculoskeletal radiologists [D.R. (10 years of experience) and R.L. (11 years of experience)]. The two observers met prior to the study and agreed upon uniform criteria [punctate clustered foci of increased radiodensities consistent with bone trabeculations greater than 0.3 mm in diameter within the intertrochanteric region of the proximal femur on the anteroposterior (AP) views] for the identification and evaluation of BB. The presence of BB at the intertrochanteric region of the proximal femur was scored as present or absent. Continuous data are presented as mean standard deviation. Discrete data are given as counts (%). The Kappa statistics were used to assess agreement on the presence of BB between readers. Chi-square tests were used to compare the incidence of hip fractures in patients with and without BB. p < 0.05 was considered significant.
Results The demographic characteristics of the cohort are presented in Table 1. The patient population was entirely postmenopausal Caucasian women with a mean age of 79.8 6.4 years (range 70e96 years) in the control group and 79.9 6.6 years (range 66e93 years) in the investigational group. Table 2 shows the presence of BB according to each evaluator. The kappa statistic for agreement on the presence of BB between readers was 0.41 (p < 0.001), which is considered moderate agreement. The percent agreement on BB between the two readers for all patients, patients Table 1 Demographic characteristics of the study population. Characteristic
Control group (no fracture)
Investigational group (hip fracture)
Caucasian female (n) Mean age (years) Age range (years)
98 79.8 6.4 (SD) 70e96
92 79.9 6.6 (SD) 66e93
842
D.B. Sarver et al. / Clinical Radiology 67 (2012) 840e842
Table 2 Bone bars prevalence. Musculoskeletal radiologist
Control group (no fracture)
Experimental group (fracture)
#1 #2
38 (39%) 37 (38%)
69 (75%) 49 (53%)
with fractures, and patients without fractures was 70, 67, and 74%, respectively. The prevalence of intertrochanteric BB in the control group ranged from 38e39%, depending on the reader. The prevalence of intertrochanteric BB in the investigational group ranged from 53e75%. Regardless of the reader, BB were identified in a significantly higher percentage of women with a hip fracture (75 versus 39%, p < 0.001 or 53 versus 38%, p ¼ 0.041) as compared to those without a hip fracture. The investigational group consisted of 35 femoral neck fractures, 51 intertrochanteric fractures, and six subtrochanteric fractures. Regardless of the evaluator, BB are seen in similar prevalence in patients with femoral neck and intertrochanteric fractures; no significant difference was found between locations (p ¼ 0.999 and p ¼ 0.664). There were too few patients with subtrochanteric fractures to evaluate any significance.
fracture. Thus the identification of BB in the non-fractured proximal femur of a patient is a trigger for requesting a dual-energy x-ray absorptiometry (DXA) examination with calculation of the FRAX 10 year absolute fracture risk to further assist in the diagnosis of osteoporosis and fracture risk. The present study is limited as the radiographs of the investigational group were not available prior to the incident of fracture. Thus the radiographs had to be evaluated on the contralateral side in order to blind the readers to the investigational group. This may introduce errors as BB in the intertrochanteric femur location are seen bilaterally 70e75% of the time according to a previous study.3 A more objective radiological definition of BB based on the presence or absence and size, extent, and location would enable more precise comparisons between studies. Future studies should evaluate BB in multiple locations and expand the population to include both genders and a larger range of ages and races. In addition the effect of medications, such as glucocorticoids, on BB should be evaluated.
Acknowledgement This study was supported by the NIH T35 research grant.
Discussion References In a previous study, evaluation of gross specimen sections of BB established that these radiodense foci are areas of thickened trabeculae extending from the cortex partially or completely across the medullary canal.1 It has been hypothesized that BB develop during insidiously progressing osteopaenia in mature skeletons as a response to mechanical stress.1 Alternatively, BB may have formed during skeletal growth but remain hidden and are unmasked by surrounding osseous resorption; a combination of the two hypotheses may be causative.1,2 The presence of BB is associated with lower BMD compared to patients without BB.3 The Singh Index established a visual method for evaluation of BMD based on the trabecular pattern of the proximal femur on AP pelvic radiographs.6 While some have confirmed its efficacy in the assessment of human femur bone strength, others have questioned its reliability in the age of computed radiography.7,8 In the present study the radiological presence of BB was significantly increased in post-menopausal Caucasian women who suffered a hip
1. Krusunoglu S, Pate D, Resnick D, et al. Bone reinforcement lines in chronic adult osteopenia: a hypothesis. Radiology 1986;158:409e15. 2. Pitt M. Osteopenic bone diseases. Orthop Clin N Am 1983;14:65e80. 3. Pitt MJ, Morgan SL, Lopez-Ben R, et al. Association of the presence of bone bars on radiographs and low bone mineral density. Skeletal Radiol 2011;40:905e11. 4. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone mineral density predict occurrence of osteoporotic fractures. BMJ 1996;312:1254e9. 5. Morgan SL, Saag KG, Julian BA, et al. Osteopenic bone disease. In: Koopman WJ, editor. Arthritis and allied conditions: a textbook of rheumatology. 14th ed. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 2449e513. 6. Singh M, Nagrath AR, Maini PS. Changes in trabecular pattern of the upper end of the femur as an index of osteoporosis. J Bone Joint Surg 1970;52 A:457e67. 7. D’Amelio P, Rossi P, Isaia G, et al. Bone mineral density and Singh index predict bone mechanical properties of human femur. Connect Tissue Res 2008;49:99e104. 8. Hauschild O, Ghanem N, Oberst M, et al. Evaluation of Singh index for assessment of osteoporosis using digital radiography. Eur J Radiol 2009;71:152e8.