- Email: [email protected]
Symptomatic atypical femoral fractures are related to underlying hip geometry
Bone 63 (2014) 1–6
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Original Full Length Article
Symptomatic atypical femoral fractures are related to underlying hip geometry David P. Taormina a, Alejandro I. Marcano a, Raj Karia a, Kenneth A. Egol a,b, Nirmal C. Tejwani a,⁎ a b
Hospital for Joint Diseases at NYU Langone Medical Center, NYU School of Medicine, 301 East 17th Street, Suite 1402, New York, NY 10003, USA Jamaica Hospital Medical Center, 8900 Van Wyck Expressway, Jamaica, NY 11418, USA
a r t i c l e
i n f o
Article history: Received 4 October 2013 Revised 13 February 2014 Accepted 14 February 2014 Available online 21 February 2014 Edited by: David Burr Keywords: Atypical femur fracture Bisphosphonate Hip morphology
a b s t r a c t The benefits of bisphosphonates are well documented, but prolonged use has been associated with atypical femur fractures. Radiographic markers for fracture predisposition could potentially aid in safer medication use. In this case–control designed study, we compared hip radiographic parameters and the demographic characteristics of chronic bisphosphonate users who sustained an atypical femoral fracture with a group of chronic bisphosphonate users who did not sustain an atypical femur fracture and also a group who sustained an intertrochanteric hip fracture. Radiographic parameters included were neck-shaft angle (NSA), hip-axis length (HAL) and center-edge angle (CE). Multivariate regression was used to evaluate the relationship between radiographic measures and femur fracture. Receiver-operating characteristic analysis determined cut-off points for neck-shaft angle and risk of atypical femur fracture. Ultimately, pre-fracture radiographs of 53 bisphosphonate users who developed atypical fracture were compared with 43 asymptomatic chronic bisphosphonate users and 64 intertrochanteric fracture patients. Duration of bisphosphonate use did not statistically differ between users sustaining atypical fracture and those without fracture (7.9 [±3.5] vs. 7.7 [±3.3] years, p = 0.7). Bisphosphonate users who fractured had acute/varus pre-fracture neck-shaft angles (p b 0.001), shorter hip-axis length (p b 0.01), and narrower center-edge angles (p b 0.01). Regression analysis revealed associations between neckshaft angle (OR = 0.89 [95% CI = 0.81–0.97; p = 0.01), center edge angle (OR = 0.89 [95% CI = 0.80–0.99]; p = 0.03), and BMI (OR = 1.15 [95% CI = 1.02–1.31; p = 0.03) with fracture development. ROC curve analysis (AUC = 0.67 [95% CI = 0.56–0.79]) determined that a cut-off point for neck-shaft angle b 128.3° yielded 69% sensitivity and 63% specificity for development of atypical femoral fracture. Ultimately, an acute/varus angle of the femoral neck, high BMI, and narrow center-edge angle were associated with development of atypical femur fracture in long-term bisphosphonate users. Patients on long-term bisphosphonates should be regularly radiographically evaluated in order to assess for potential risk of atypical fracture. © 2014 Elsevier Inc. All rigths reserved.
Introduction Nearly half of all women and up to a quarter of men aged 50 years or older go on to sustain osteoporosis-related fractures, as reported by studies performed within the United States, UK, and Scandinavia [1]. In 1995, the US Food and Drug Administration approved the use of alendronate therapy in patients at risk for osteoporosis-related fractures, a significant advancement in medical prophylaxis [2–4]. Alendronate is a secondgeneration bisphosphonate that induces osteoclast apoptosis by hampering protein prenylation and, ultimately, cholesterol synthesis via the mevalonate pathway [2,5]. Bisphosphonates have proven to decrease markers of bone resorption, increase bone mineralization, and provide
⁎ Corresponding author at: 301 East 17th Street, Suite 1402, New York, NY 10003, USA. E-mail addresses: [email protected] (D.P. Taormina), [email protected] (A.I. Marcano), [email protected] (R. Karia), [email protected] (K.A. Egol), [email protected] (N.C. Tejwani).
http://dx.doi.org/10.1016/j.bone.2014.02.006 8756-3282/© 2014 Elsevier Inc. All rigths reserved.
an increase in bone mineral density (BMD) in the lumbar spine and upper femur over the course of 3 years of treatment [2,5]. Owing to these advantageous outcomes, bisphosphonates have become a treatment of choice for decreasing fracture risk and preventing postmenopausal bone loss [5,6]. Recent publications have identified an association of long-term bisphosphonate therapy with the development of atypical subtrochanteric and diaphyseal femoral fractures [3,7,8]. One explanation of this adverse outcome has been that the suppression of chronic bone turnover may impede local remodeling capabilities in the setting of expanding insufficiency micro-fractures [2,9–11]. Bisphosphonate administration, nonetheless, continues as a mainstay in the management of osteoporosis [12–15]. However, the patients who sustain bisphosphonate-associated femur fractures face significant sequelae and a slow recovery [2,16,17]. Few advancements have been made in either the understanding of the pathomechanics of these injuries and the identification of the patients at risk for atypical femur fractures. There is a paucity of scientific studies focusing on lower limb geometry and those few available have
2
D.P. Taormina et al. / Bone 63 (2014) 1–6
been limited in their scope. The 2013 ASBMR Task Force on Atypical Femoral Fractures argued for an expansion of the knowledge base characterizing the geometric risk factors predisposing to atypical femoral fracture [11]. The purpose of this study is to describe the bony anatomy of a group of chronic bisphosphonate users who went on to develop incomplete or complete atypical femur fractures and to compare it with two groups: 1) a group of long-term bisphosphonate users who did not sustain atypical femoral fracture and 2) a general population intertrochanteric hip fracture group. Our null hypothesis was that the pre-injury hip and proximal femur anatomy of chronic bisphosphonate users does not predispose them to the development of subtrochanteric and diaphyseal atypical femur fractures. We theorize that hips are predisposed to atypical femur fracture by having more acute (varus) neck-shaft angles and greater hip-axis length. Materials and methods The research protocol and all procedures followed were in accordance with and approved by the Institutional Review Board. Between 2004 and 2013, 69 consecutive chronic bisphosphonate patients, each sustaining either unilateral or bilateral atypical femoral fractures, were treated at our institution after identification of radiographic findings consistent with complete or incomplete bisphosphonate-associated lesions (“Atypical Femoral Fracture” patients, AFF). Patient presentation or referral was based on previous identification of radiographic findings consistent with bisphosphonate associated changes or painful symptomatology which was followed-up with radiographic imaging to confirm the diagnosis. Registry data was collected by trained research personnel from the patient lists of two trauma fellowship-trained orthopedic surgeons [6,18,19]. At the time of presentation, bisphosphonates were discontinued in all patients. Subsequent surgical and medical management was performed independent of this study and in coordination with medical specialists who have substantial experience in managing these conditions. For inclusion within this study, the following criteria were applied: history of long-term bisphosphonate use (3 years or more), sustained a complete or an incomplete sub-trochanteric atypical femoral fracture, and had been seen within our multi-center academic medical University system for examination with radiographic evaluation of their hip prior to the development of atypical femur fracture and undergoing orthopedic surgery. Atypical femoral fractures were defined by the radiographic evidence of a transverse or short oblique fracture line, medial spike (in the case of most complete fractures), focal lateral cortical thickening, and a relative lack of comminution at the fracture site [11,20]. In the case of complete fracture, the fracture line extended through both cortices whereas in incomplete fractures, the fracture line originated at the lateral cortex. Although not required for inclusion, these findings were frequently associated with a generalized increase in diaphyseal cortical thickness, a bilateral manifestation of changes, prodromal pain, and delayed fracture healing. Study exclusionary criteria were a failure to obtain pre-lesion proximal femora radiographs, poor quality radiographs (including poorly exposed or poorly aligned limbs), prior hip trauma treated with implant, history of arthrodesis or arthroplasty, lesions associated with high velocity mechanism, or those with identifiable bone tumor or other remarkable pathology associated fracture. Among the 69 patients within the AFF registry group, all had involvement of (at least) one limb which was identified as having an incomplete or complete fracture, although 16 were ultimately excluded leaving a total of 53 satisfying all selection criteria. In this case–control designed study, the radiographic measurements of the AFF group were compared with two other patient groups 1) an independent asymptomatic long-term bisphosphonate user group who did not develop atypical femoral fractures (“No Femoral Fracture” patients, NFF) and 2) an independent general population intertrochanteric femur fracture group (ITF). The NFF registry was maintained by our medical endocrinology group and includes patients who were enrolled
between 2009 and 2012 and agreed to a regular follow-up regimen with radiographs obtained at each visit (findings published in 2012) [21]. The ITF registry group was prospectively enrolled between May, 2011 and February, 2012, at our institution and consists of patients with typical osteoporotic proximal femur fractures (intertrochanteric or subtrochanteric) who did not report bisphosphonate use. The same exclusionary criteria which were applied to the AFF group were also applied to the comparison groups [19]. During the course of our chart review, we attempted to procure radiographs taken from either or, when available, both femora of all patients. When satisfactory pre-lesion radiographs were available bilaterally, both sides were measured and the mean value was used for our analyses. Proper femoral rotation and radiographic alignment were evaluated based on the prominence of the lesser trochanter. All films used for measurements had been captured as part of a standard of care treatment within the image archiving computerized storage system (PACS, GE Waukeshau Wi, Seimens, Erhlanger, Germany) of our hospital system. The geometric hip measurements obtained included neck-shaft angle (NSA), hip-axis length (HAL), and the center-edge (CE) angle of Wiberg (Fig. 1) [22–24]. Digital measurements were made by three reviewers (two surgeons and a trained researcher) using standardized published techniques on 2-dimensional radiographs viewed through the PACS software suite. Reviewers were blinded at the time of measurements. NSA measurements were obtained by using the method described by Ogata et al. (1979) and expanded by Faulkner et al. in 1993 [22,24]. The HAL was measured as the distance from the greater trochanter to the inner pelvic brim along the neck-shaft axis line [22]. CE angles were measured by drawing a vertical line down to the center of the femoral head (determined by a super-imposed best-fit circle), with the angle completed by the extension of a line just tangential to the lateral edge of the acetabular roof [23]. Based on the inter-rater evaluation study published by Karanicolas et al. (2010), reliability of NSA and HAL measurements was compared by using triplicate measurements on 30 randomly selected radiographs [25]. Inter-rater analyses between the three reviewers revealed a high degree of measurement reliability (NSA ICC = 0.98, p b 0.001; HAL ICC = 0.99, p b 0.001).
Statistical analysis Basic patient demographics, duration of bisphosphonate use (for AFF and NFF groups), and the radiographic measures described were recorded. Missing data ranged from 1 to 9% for measures of height, BMI, bisphosphonate duration, and the radiographic measurements. Continuous variables (excluding duration of bisphosphonate use) were analyzed between the three-groups by using one-way ANOVA and categorical comparisons were made by using Chi-square analysis. Post-hoc analysis included Tukey pairwise comparisons to look at individual differences between the groups. Comparison of duration of bisphosphonate use was performed between the AFF and NFF groups by using a Student's t-test. Sample size calculation revealed that in order to detect a medium difference (effect size = 0.30) in NSA between the 3 groups, 37 femoral measurements per group would be needed to achieve 80% power with alpha of 0.05. Among the bisphosphonate users (AFF + NFF), the association of the three radiographic measurements with ultimate development of a complete or incomplete atypical femur fracture was analyzed by using a stepwise multivariate logistic regression model. The regression model included the covariates of NSA, HAL, CE angle, age, gender, BMI, and duration of bisphosphonate usage. ROC curve analysis was then performed by using the grouped AFF + NFF sample in order to explore possible cut-off points for NSA as it related to development of atypical femur fracture. All statistical tests were performed by using SPSS 20 (IMB, Chicago, IL) with a significance level of p = 0.05.
D.P. Taormina et al. / Bone 63 (2014) 1–6
3
The NSA was measured as the angle formed by ABC. AB = Radiographic center of the diaphysis. BC = Radiographic center of the femoral neck, drawn perpendicular to the narrowest portion, defined by the line DD’. The HAL (denoted EF) was measured as the distance from the greater trochanter to the inner pelvic brim along the neck-shaft axis line, perpendicularly bisecting segment DD’. The CE angle was measured by drawing a vertical line down to the center of the femoral head (G), with an angle completed by the extension of a line tangential to the lateral edge of the acetabular roof (H).
Fig. 1. Geometric hip measurements.
Results Fifty-three AFF patients with 63 femora radiographs were identified. Radiographic measurements of the AFF group were compared with 43 NFF patients with 80 femora radiographs and 64 ITF patients with 79 femora films. Patient demographic characteristics and results from statistically significant comparisons are displayed in Table 1. Mean duration of bisphosphonate use was nearly 8 years in both the AFF (7.9 [±3.5] years) and NFF (7.7 [±3.3] years) patient groups (p = 0.7). Mean NSA measurement for all three groups was 129.5 [±6.3] degrees and consistent with published population means [26]. The overall mean HAL was 125.6 [±13.0] mm and mean CE angle was 44.6[±5.5] degrees. In the three-way group comparison, the proximal femoral anatomy of the AFF group was significantly different from each of the NFF and ITF on nearly all geometric measures (Table 2). The AFF patients were found to be more varus (NSA = 126.4 [± 6.6] degrees), with a shorter hip axis (HAL = 120.3 [± 11.7] mm) and narrower CE angle (42.6 [± 6.2] degrees). A weak correlation between NSA and HAL existed within the bisphosphonate groups (AFF + NFF) (r = 0.34, p b 0.01). When comparing the relationship between NSA and HAL (Fig. 2), patients who developed fracture tended to cluster in the area
of varus NSA with shorter HAL. Sub-group analysis between the NFF and ITF groups on measures of NSA (p = 0.51), HAL (p = 0.73), and CE angle (p = 0.52) revealed no statistical differences. Table 1 Demographics and patient characteristics.
Patients #n Female (%)a
b
Age (years) Height (in.) BMI (kg/m2) Bisphosphonate duration (years)
AFF
NFF
ITF
53 96.2
43 90.7
64 78.1
Mean (μ) [sd]
Mean (μ) [sd]
Mean (μ) [sd]
69.5 [10.1] 62.6 [3.5] 25.8 [6.7] 7.9 [3.5]
68.8 [10.3] 63.1 [3.1] 23.8 [3.3] 7.7 [3.3]
79.9 [12.3]c,d 63.1 [4.0] 24.9 [5.1] n/a
AFF = Atypical Femoral Fracture patients; NFF = No Femoral Fracture patients; ITF = Intertrochanteric Fracture patients. Unless otherwise noted, there was no statistical difference in one-way ANOVA comparisons between the three groups at a p ≤ 0.05 significance level, nor in Tukey pairwise post-hoc comparisons. a Chi-square p = 0.01. b One-way ANOVA p b 0.001. c ITF vs. AFF pairwise comparison p b 0.001. d ITF vs. NFF pairwise comparison p b 0.001.
4
D.P. Taormina et al. / Bone 63 (2014) 1–6
Table 2 Radiographic measurement means [and SDs] as well as differences between group means [and 95% confidence intervals].
Neck-shaft angle (°) Hip-axis length (mm.) Center-edge angle (°)
AFF
NFF
ITF
Mean difference (AFF vs NFF)
Mean difference (AFF vs ITF)
p-value
126.4 [6.6]a,b 120.3 [11.7]c,d 42.6 [6.2]e
130.3 [5.8] 127.3 [12.4] 45.1 [4.8]
131.1 [5.6] 128.2 [13.4] 45.8 [5.0]
3.9 [0.9–7.0] 7.0 [0.6–13.5] 2.6 [0.1–5.3]
4.7 [1.9–7.4] 7.9 [2.1–13.8] 3.2 [−0.8–5.7]
b0.001 0.004 0.007
AFF = Atypical Femoral Fracture patients; NFF = No Femoral Fracture patients; ITF = Intertrochanteric Fracture patients unless otherwise noted, there was no statistical difference in Tukey pairwise post-hoc comparisons at a p ≤ 0.05 significance level. a AFF vs. NFF pairwise comparison p b 0.01. b AFF vs. ITF pairwise comparison p b 0.001. c AFF vs. NFF pairwise comparison p b 0.05. d AFF vs. ITF pairwise comparison p b 0.01. e AFF vs. ITF pairwise comparison p b 0.01.
Logistic regression model analysis using a grouped AFF + NFF sample revealed a significant association between NSA (OR = 0.89 [95% CI = 0.81–0.97; p = 0.01), CE angle (OR = 0.89 [95% CI = 0.80–0.99]; p = 0.03), and BMI (OR = 1.15 [95% CI = 1.02–1.31; p = 0.03) with development of fracture. Although hip-axis length was associated with the development of fracture when conducting univariate logistic regression, the significance was lost after adjusting for other covariates. ROC curve analysis (area under curve = 0.67 [95% CI = 0.56–0.79]) revealed that an NSA b 128.3° yielded a 69% sensitivity and 63% specificity to predict the development of a complete or incomplete atypical femur fracture. Discussion With this study we have demonstrated an association between acute (varus) neck-shaft angle proximal femur anatomy and the
development of atypical femoral fractures after long-term bisphosphonate use. While no definitive conclusions can be reached regarding the role that altered proximal femoral anatomy plays in the development of atypical fractures, the establishment of an association may help physicians safely manage patients who are prescribed bisphosphonates. There are a variety of treatment options available to the osteoporotic patient and these include bisphosphonates, estrogen replacement, selective estrogen receptor modulators, Denosumab and Teriparatide [27]. Bisphosphonates are unique in that they can induce prolonged suppression of bone turnover and sustained reduction of osteoclastic activity [27,28]. Chronic suppression of bone turnover, along with the addition of non-enzymatically linked collagen molecules is believed to limit the plasticity and heterogeneity of developing bone [29]. This chronic therapy, in turn, may become the façade for weakening bony
Among all bisphosphonate users, NSA (X-axis) versus HAL (Y-axis) was plotted independently for patients who developed either complete or incomplete atypical femur fracture (AFF group; diamond) and those who had not fractured (NFF group; circle). Patients with femora which fractured tended to cluster having both shorter HAL and narrower NSA. Femora which had obtuse NSA and with larger HAL tended not to fracture.
Fig. 2. Hip-axis length versus neck-shaft angle in chronic bisphosphonate users.
D.P. Taormina et al. / Bone 63 (2014) 1–6
architecture and increasing micro-lesion accumulation with the potential for ultimate progression to gross fracture [9,28,30]. Unfortunately, the ubiquity of bisphosphonate therapy for the treatment of osteoporotic patients estranges a cohort who may ultimately succumb to a dangerously silent milieu of micro-anatomic changes [28,31]. One large cohort study (Abrahamsen et al., 2010) estimated that the general population incidence of subtrochanteric and diaphyseal femoral shaft fractures, per 10,000 patient-years, increases from 13 for women and 6 for men to 31 for both genders after bisphosphonate use [32]. Patients with osteoporosis brave these odds hoping to stave off femoral neck, intertrochanteric, and vertebral fractures, although bisphosphonate-associated atypical femur fractures are far from inconsequential [15,33]. Several studies have reported the use of proximal femur geometrics as markers for risk of osteoporosis related fragility fracture. In 1999, Gnudi et al. published findings from 111 Caucasian post-menopausal women and 329 Caucasian controls. Their comparison found that valgus NSA (wide-angle) and longer HAL were associated with greater risk of femoral neck and trochanteric hip fracture [34]. In a longitudinal follow-up, Gnudi and colleagues (2012) re-affirmed these findings by using automated geometric measurements [35]. Our group of patients with bisphosphonate-associated atypical femur fracture seem distinctly different from those identified as at risk in studies of osteoporosis related fractures [34,35]. Our patient group had acute NSAs with short HALs. Given the pattern of the atraumatic atypical femur fracture, the increased lateral cortical loading of a varus hip seems consistent with the mechanism of injury development. Although there has been a moderate association established between the development of hip fractures (particularly femoral neck) and increasing HAL, there are no studies on associations of HAL with subtrochanteric fractures. While we found a weak correlation between varus NSA and shorter HAL, we are unable to conclude whether shorter HAL is a true risk for atypical fracture, a surrogate marker (related to NSA, femoral anteversion, or femoral bow), or the product of a genetic confounder. Consistent with the findings of Franceschetti et al., we also identified increasing BMI as a risk factor for the development of fracture in chronic bisphosphonate users [36]. There are several limitations to this study. While we took as many precautions as possible to de-identify patients during radiographic measurements, surgeons could have recalled the patients they treated, thus leading to bias. There remains the possibility of measurement error during radiographic evaluation. There is also potential for limb rotation to confound radiographic readings even with a standardized technique for positioning and radiographic assessment. Nonetheless, our studies mean measurements, particularly NSA (mean = 129.5), are nearly identical to other published findings (Elbuken reported a population mean amongst 18,943 individuals = 129.2°) [26]. While the full breath of research encompassed any patient presenting (or referred) with symptomatic bisphosphonate associated fractures, enrollment of AFFs was performed by using the case load of physicians at an urban academic medical center. It is difficult to discern the extent of demographic confounders and their effect on generalizing the study findings. While other studies have published differences between proximal and distal atypical femoral fractures [37], measurement of the distance between the lesser trochanter and fracture site was not included within our measurement protocol. Thus, future studies may seek to correlate proximal hip geometry with the site of atypical fracture. Conclusions The results of this study demonstrated that chronic bisphosphonate users who developed symptomatic atypical femoral fracture had more varus (acute) neck-shaft angles and shorter hip-axis length. Furthermore, a clear association between NSA and BMI with the development of atypical femur fracture was identified. Patients who are on bisphosphonates and have a more acute (varus) neck-shaft angle as
5
well as those who have higher BMI are more likely to develop symptomatic atypical femur fractures. This study, to the best of our knowledge, is the first to identify neck-shaft angle as a potential geometric risk factor associated with the development of symptomatic atypical femur fracture.
References [1] Johnell O, Kanis J. Epidemiology of osteoporotic fractures. Osteoporos Int 2005;16:S3–7. [2] Capeci CM, Tejwani NC. Bilateral low-energy simultaneous or sequential femoral fractures in patients on long-term alendronate therapy. J Bone Joint Surg 2009;91-A: 2556–61. [3] Lenart BA, Lorich DG, Lane JM. Atypical fractures of the femoral diaphysis in postmenopausal women taking alendronate. N Engl J Med 2008;358:1304–6. [4] Park-Wyllie LY, Mamdani MM, Juurlink DN, Hawker GA, Gunraj N, Austin PC, et al. Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. J Am Med Assoc 2011;305:783–9. [5] Koval KJ, Zuckerman JD. Fractures in the elderly. Philadelphia: Lippincott-Raven Publishers; 1998 15. [6] Egol KA, Park JH, Prensky C, Rosenberg ZS, Peck V, Tejwani NC. Surgical treatment improves clinical and functional outcomes for patients who sustain incomplete bisphosphonate-related femur fractures. J Orthop Trauma 2013;27:331–5. [7] Neviaser AS, Lane JM, Lenart BA, Edobor-Osula F, Lorich DG. Low-energy femoral shaft fractures associated with alendronate use. J Orthop Trauma 2008;22:346–50. [8] Lenart BA, Neviaser AS, Lymann S, Chang CC, Edobor-Osula F, Steele B, et al. Association of low-energy femoral fractures with prolonged bisphosphonate use: a case control study. Osteoporos Int 2009;20:1353–62. [9] Odvina CV, Zerwekh JE, Rao DS, Maalouf N, Gottschalk FA, Pak CYC. Severely suppressed bone turnover: a potential complication of alendronate therapy. J Clin Endocrinol Metab 2005;90:1294–301. [10] Armamento-Villareal R, Napoli N, Panwar V, Novack D. Suppressed bone turnover during alendronate therapy for high-turnover osteoporosis. N Engl J Med 2006;355:2048–50. [11] Shane E, Burr D, Abrahamsen B, Adler RA, Brown TD, Cheung AM, 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. [12] Hollick R, Reid D. Role of bisphosphonates in the management of postmenopausal osteoporosis: an update on recent safety anxieties. Menopause Int 2011;17:66–72. [13] Shane E, Burr D, Ebeling PR, Abrahamsen B, Adler RA, Brown TD, et al. American Society for B, Mineral R. Atypical Subtrochanteric and Diaphyseal Femoral Fractures: Report of a Task Force of the American Society for Bone and Mineral Research. J Bone Miner Res 2010;25:2267–94. [14] Black DM, Kelly MP, Genant HK, Palermo L, Eastell R, Bucci-Rechtweg C, et al. Fracture Intervention Trial Steering C, Committee HPFTS. Bisphosphonates and fractures of the subtrochanteric or diaphyseal femur. N Engl J Med 2010;362:1761–71. [15] Napoli N, Schwartz AV, Palermo L, Jin JJ, Wustrack R, Cauley JA, et al. Risk factors for subtrochanteric and diaphyseal fractures: the study of osteoporotic fractures. J Clin Endocrinol Metab 2013;98:659–67. [16] Egol KA, Park JH, Prensky C, Rosenberg ZS, Peck V, Tejwani NC. Surgical treatment improves clinical and functional outcomes for patients who sustain incomplete bisphosphonate-related femur fractures. J Orthop Trauma 2013;27:331–5. [17] Allison MB, Markman L, Rosenberg Z, Vieira RL, Babb J, Tejwani N, et al. Atypical incomplete femoral fractures in asymptomatic patients on long term bisphosphonate therapy. Bone 2013;55:113–8. [18] Egol KA, Park JH, Rosenberg ZS, Peck V, Tejwani NC. Healing delayed but generally reliable after bisphosphonate-associated complete femur fractures treated with IM nails. Clin Orthop Relat Res 2013 [Epub (ahead of print)]. [19] Marcano A, Taormina D, Egol KA, Peck V, Tejwani NC. Are race and sex associated with the occurrence of atypical femoral fractures? Clin Orthop Relat Res 2014;472:1020–7. [20] Rosenberg ZS, Vieira RLR, Chan SS, Babb J, Akyol Y, Rybak LD, et al. Bisphosphonaterelated complete atypical subtrochanteric femoral fractures: diagnostic utility of radiography. Am J Roentgenol 2011;197:954–60. [21] Vieira RLR, Rosenberg ZS, Allison MB, Im SA, Babb J, Peck V. Frequency of incomplete atypical femoral fractures in asymptomatic patients on long-term bisphosphonate therapy. Am J Roentgenol 2012;198:1144–51. [22] Faulkner KG, Cummings SR, Black D, Palermo L, Glüer C-C, Genant HK. Simple measurement of femoral geometry predicts hip fracture: the study of osteoporotic fractures. J Bone Miner Res 1993;8:1211–7. [23] Fredensborg N, The CE. Angle of normal hips. Acta Orthop 1976;47:403–5. [24] Ogata K, Goldsand EM. A simple biplanar method of measuring femoral anteversion and neck-shaft angle. J Bone Joint Surg 1979;61:846–51. [25] Karanicolas PJ, Bhandari M, Kreder H, Moroni A, Richardson M, Walter SD, et al. Collaboration for outcome assessment in surgical trials musculoskeletal G. Evaluating agreement: conducting a reliability study. J Bone Joint Surg 2009;91:99–106. [26] Elbuken F, Baykara M, Ozturk C. Standardisation of the neck-shaft angle and measurement of age-, gender-and BMI-related changes in the femoral neck using DXA. Singapore Med J 2012;53:587–90. [27] Cermak K, Shumelinsky F, Alexiou J, Gebhart MJ. Case reports: subtrochanteric femoral stress fractures after prolonged alendronate therapy. Clin Orthop Relat Res 2010;468:1991–6. [28] Giusti A, Hamdy NA, Papapoulos SE. Atypical fractures of the femur and bisphosphonate therapy. A systematic review of case/case series studies. Bone 2010;47:169–80.
6
D.P. Taormina et al. / Bone 63 (2014) 1–6
[29] Ettinger B, Burr DB, Ritchie RO. Proposed pathogenesis for atypical femoral fractures: lessons from materials research. Bone 2013;55:495–500. [30] Mashiba T, Hirano T, Turner CH, Forwood MR, Johnston CC, Burr DB. Suppressed bone turnover by bisphosphonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J Bone Miner Res 2000;15:613–20. [31] Iwamoto J, Sato Y, Takeda T, Matsumoto H. Hip fracture protection by alendronate treatment in postmenopausal women with osteoporosis: a review of the literature. Clin Interv Aging 2008;3:483–9. [32] Abrahamsen B, Eiken P, Eastell R. Cumulative alendronate dose and the long-term absolute risk of subtrochanteric and diaphyseal femur fractures: a register-based national cohort analysis. J Clin Endocrinol Metab 2010;95:5258–65.
[33] Schilcher J, Michaëlsson K, Aspenberg P. Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med 2011;364:1728–37. [34] Gnudi S, Ripamonti C, Gualtieri G, Malavolta N. Geometry of proximal femur in the prediction of hip fracture in osteoporotic women. Br J Radiol 1999;72:729–33. [35] Gnudi S, Sitta E, Pignotti E. Prediction of incident hip fracture by femoral neck bone mineral density and neck-shaft angle: a 5-year longitudinal study in postmenopausal females. Br J Radiol 2012;85:e467–73. [36] Franceschetti P, Bondanelli M, Caruso G, Ambrosio MR, Lorusso V, Zatelli MC, et al. Risk factors for development of atypical femoral fractures in patients on long-term oral bisphosphonate therapy. Bone 2013;56:426–31. [37] Koeppen VA, Schilcher J, Aspenberg P. Dichotomous location of 160 atypical femoral fractures. Acta Orthop 2013;84:561–4.
Contents lists available at ScienceDirect
Bone journal homepage: www.elsevier.com/locate/bone
Original Full Length Article
Symptomatic atypical femoral fractures are related to underlying hip geometry David P. Taormina a, Alejandro I. Marcano a, Raj Karia a, Kenneth A. Egol a,b, Nirmal C. Tejwani a,⁎ a b
Hospital for Joint Diseases at NYU Langone Medical Center, NYU School of Medicine, 301 East 17th Street, Suite 1402, New York, NY 10003, USA Jamaica Hospital Medical Center, 8900 Van Wyck Expressway, Jamaica, NY 11418, USA
a r t i c l e
i n f o
Article history: Received 4 October 2013 Revised 13 February 2014 Accepted 14 February 2014 Available online 21 February 2014 Edited by: David Burr Keywords: Atypical femur fracture Bisphosphonate Hip morphology
a b s t r a c t The benefits of bisphosphonates are well documented, but prolonged use has been associated with atypical femur fractures. Radiographic markers for fracture predisposition could potentially aid in safer medication use. In this case–control designed study, we compared hip radiographic parameters and the demographic characteristics of chronic bisphosphonate users who sustained an atypical femoral fracture with a group of chronic bisphosphonate users who did not sustain an atypical femur fracture and also a group who sustained an intertrochanteric hip fracture. Radiographic parameters included were neck-shaft angle (NSA), hip-axis length (HAL) and center-edge angle (CE). Multivariate regression was used to evaluate the relationship between radiographic measures and femur fracture. Receiver-operating characteristic analysis determined cut-off points for neck-shaft angle and risk of atypical femur fracture. Ultimately, pre-fracture radiographs of 53 bisphosphonate users who developed atypical fracture were compared with 43 asymptomatic chronic bisphosphonate users and 64 intertrochanteric fracture patients. Duration of bisphosphonate use did not statistically differ between users sustaining atypical fracture and those without fracture (7.9 [±3.5] vs. 7.7 [±3.3] years, p = 0.7). Bisphosphonate users who fractured had acute/varus pre-fracture neck-shaft angles (p b 0.001), shorter hip-axis length (p b 0.01), and narrower center-edge angles (p b 0.01). Regression analysis revealed associations between neckshaft angle (OR = 0.89 [95% CI = 0.81–0.97; p = 0.01), center edge angle (OR = 0.89 [95% CI = 0.80–0.99]; p = 0.03), and BMI (OR = 1.15 [95% CI = 1.02–1.31; p = 0.03) with fracture development. ROC curve analysis (AUC = 0.67 [95% CI = 0.56–0.79]) determined that a cut-off point for neck-shaft angle b 128.3° yielded 69% sensitivity and 63% specificity for development of atypical femoral fracture. Ultimately, an acute/varus angle of the femoral neck, high BMI, and narrow center-edge angle were associated with development of atypical femur fracture in long-term bisphosphonate users. Patients on long-term bisphosphonates should be regularly radiographically evaluated in order to assess for potential risk of atypical fracture. © 2014 Elsevier Inc. All rigths reserved.
Introduction Nearly half of all women and up to a quarter of men aged 50 years or older go on to sustain osteoporosis-related fractures, as reported by studies performed within the United States, UK, and Scandinavia [1]. In 1995, the US Food and Drug Administration approved the use of alendronate therapy in patients at risk for osteoporosis-related fractures, a significant advancement in medical prophylaxis [2–4]. Alendronate is a secondgeneration bisphosphonate that induces osteoclast apoptosis by hampering protein prenylation and, ultimately, cholesterol synthesis via the mevalonate pathway [2,5]. Bisphosphonates have proven to decrease markers of bone resorption, increase bone mineralization, and provide
⁎ Corresponding author at: 301 East 17th Street, Suite 1402, New York, NY 10003, USA. E-mail addresses: [email protected] (D.P. Taormina), [email protected] (A.I. Marcano), [email protected] (R. Karia), [email protected] (K.A. Egol), [email protected] (N.C. Tejwani).
http://dx.doi.org/10.1016/j.bone.2014.02.006 8756-3282/© 2014 Elsevier Inc. All rigths reserved.
an increase in bone mineral density (BMD) in the lumbar spine and upper femur over the course of 3 years of treatment [2,5]. Owing to these advantageous outcomes, bisphosphonates have become a treatment of choice for decreasing fracture risk and preventing postmenopausal bone loss [5,6]. Recent publications have identified an association of long-term bisphosphonate therapy with the development of atypical subtrochanteric and diaphyseal femoral fractures [3,7,8]. One explanation of this adverse outcome has been that the suppression of chronic bone turnover may impede local remodeling capabilities in the setting of expanding insufficiency micro-fractures [2,9–11]. Bisphosphonate administration, nonetheless, continues as a mainstay in the management of osteoporosis [12–15]. However, the patients who sustain bisphosphonate-associated femur fractures face significant sequelae and a slow recovery [2,16,17]. Few advancements have been made in either the understanding of the pathomechanics of these injuries and the identification of the patients at risk for atypical femur fractures. There is a paucity of scientific studies focusing on lower limb geometry and those few available have
2
D.P. Taormina et al. / Bone 63 (2014) 1–6
been limited in their scope. The 2013 ASBMR Task Force on Atypical Femoral Fractures argued for an expansion of the knowledge base characterizing the geometric risk factors predisposing to atypical femoral fracture [11]. The purpose of this study is to describe the bony anatomy of a group of chronic bisphosphonate users who went on to develop incomplete or complete atypical femur fractures and to compare it with two groups: 1) a group of long-term bisphosphonate users who did not sustain atypical femoral fracture and 2) a general population intertrochanteric hip fracture group. Our null hypothesis was that the pre-injury hip and proximal femur anatomy of chronic bisphosphonate users does not predispose them to the development of subtrochanteric and diaphyseal atypical femur fractures. We theorize that hips are predisposed to atypical femur fracture by having more acute (varus) neck-shaft angles and greater hip-axis length. Materials and methods The research protocol and all procedures followed were in accordance with and approved by the Institutional Review Board. Between 2004 and 2013, 69 consecutive chronic bisphosphonate patients, each sustaining either unilateral or bilateral atypical femoral fractures, were treated at our institution after identification of radiographic findings consistent with complete or incomplete bisphosphonate-associated lesions (“Atypical Femoral Fracture” patients, AFF). Patient presentation or referral was based on previous identification of radiographic findings consistent with bisphosphonate associated changes or painful symptomatology which was followed-up with radiographic imaging to confirm the diagnosis. Registry data was collected by trained research personnel from the patient lists of two trauma fellowship-trained orthopedic surgeons [6,18,19]. At the time of presentation, bisphosphonates were discontinued in all patients. Subsequent surgical and medical management was performed independent of this study and in coordination with medical specialists who have substantial experience in managing these conditions. For inclusion within this study, the following criteria were applied: history of long-term bisphosphonate use (3 years or more), sustained a complete or an incomplete sub-trochanteric atypical femoral fracture, and had been seen within our multi-center academic medical University system for examination with radiographic evaluation of their hip prior to the development of atypical femur fracture and undergoing orthopedic surgery. Atypical femoral fractures were defined by the radiographic evidence of a transverse or short oblique fracture line, medial spike (in the case of most complete fractures), focal lateral cortical thickening, and a relative lack of comminution at the fracture site [11,20]. In the case of complete fracture, the fracture line extended through both cortices whereas in incomplete fractures, the fracture line originated at the lateral cortex. Although not required for inclusion, these findings were frequently associated with a generalized increase in diaphyseal cortical thickness, a bilateral manifestation of changes, prodromal pain, and delayed fracture healing. Study exclusionary criteria were a failure to obtain pre-lesion proximal femora radiographs, poor quality radiographs (including poorly exposed or poorly aligned limbs), prior hip trauma treated with implant, history of arthrodesis or arthroplasty, lesions associated with high velocity mechanism, or those with identifiable bone tumor or other remarkable pathology associated fracture. Among the 69 patients within the AFF registry group, all had involvement of (at least) one limb which was identified as having an incomplete or complete fracture, although 16 were ultimately excluded leaving a total of 53 satisfying all selection criteria. In this case–control designed study, the radiographic measurements of the AFF group were compared with two other patient groups 1) an independent asymptomatic long-term bisphosphonate user group who did not develop atypical femoral fractures (“No Femoral Fracture” patients, NFF) and 2) an independent general population intertrochanteric femur fracture group (ITF). The NFF registry was maintained by our medical endocrinology group and includes patients who were enrolled
between 2009 and 2012 and agreed to a regular follow-up regimen with radiographs obtained at each visit (findings published in 2012) [21]. The ITF registry group was prospectively enrolled between May, 2011 and February, 2012, at our institution and consists of patients with typical osteoporotic proximal femur fractures (intertrochanteric or subtrochanteric) who did not report bisphosphonate use. The same exclusionary criteria which were applied to the AFF group were also applied to the comparison groups [19]. During the course of our chart review, we attempted to procure radiographs taken from either or, when available, both femora of all patients. When satisfactory pre-lesion radiographs were available bilaterally, both sides were measured and the mean value was used for our analyses. Proper femoral rotation and radiographic alignment were evaluated based on the prominence of the lesser trochanter. All films used for measurements had been captured as part of a standard of care treatment within the image archiving computerized storage system (PACS, GE Waukeshau Wi, Seimens, Erhlanger, Germany) of our hospital system. The geometric hip measurements obtained included neck-shaft angle (NSA), hip-axis length (HAL), and the center-edge (CE) angle of Wiberg (Fig. 1) [22–24]. Digital measurements were made by three reviewers (two surgeons and a trained researcher) using standardized published techniques on 2-dimensional radiographs viewed through the PACS software suite. Reviewers were blinded at the time of measurements. NSA measurements were obtained by using the method described by Ogata et al. (1979) and expanded by Faulkner et al. in 1993 [22,24]. The HAL was measured as the distance from the greater trochanter to the inner pelvic brim along the neck-shaft axis line [22]. CE angles were measured by drawing a vertical line down to the center of the femoral head (determined by a super-imposed best-fit circle), with the angle completed by the extension of a line just tangential to the lateral edge of the acetabular roof [23]. Based on the inter-rater evaluation study published by Karanicolas et al. (2010), reliability of NSA and HAL measurements was compared by using triplicate measurements on 30 randomly selected radiographs [25]. Inter-rater analyses between the three reviewers revealed a high degree of measurement reliability (NSA ICC = 0.98, p b 0.001; HAL ICC = 0.99, p b 0.001).
Statistical analysis Basic patient demographics, duration of bisphosphonate use (for AFF and NFF groups), and the radiographic measures described were recorded. Missing data ranged from 1 to 9% for measures of height, BMI, bisphosphonate duration, and the radiographic measurements. Continuous variables (excluding duration of bisphosphonate use) were analyzed between the three-groups by using one-way ANOVA and categorical comparisons were made by using Chi-square analysis. Post-hoc analysis included Tukey pairwise comparisons to look at individual differences between the groups. Comparison of duration of bisphosphonate use was performed between the AFF and NFF groups by using a Student's t-test. Sample size calculation revealed that in order to detect a medium difference (effect size = 0.30) in NSA between the 3 groups, 37 femoral measurements per group would be needed to achieve 80% power with alpha of 0.05. Among the bisphosphonate users (AFF + NFF), the association of the three radiographic measurements with ultimate development of a complete or incomplete atypical femur fracture was analyzed by using a stepwise multivariate logistic regression model. The regression model included the covariates of NSA, HAL, CE angle, age, gender, BMI, and duration of bisphosphonate usage. ROC curve analysis was then performed by using the grouped AFF + NFF sample in order to explore possible cut-off points for NSA as it related to development of atypical femur fracture. All statistical tests were performed by using SPSS 20 (IMB, Chicago, IL) with a significance level of p = 0.05.
D.P. Taormina et al. / Bone 63 (2014) 1–6
3
The NSA was measured as the angle formed by ABC. AB = Radiographic center of the diaphysis. BC = Radiographic center of the femoral neck, drawn perpendicular to the narrowest portion, defined by the line DD’. The HAL (denoted EF) was measured as the distance from the greater trochanter to the inner pelvic brim along the neck-shaft axis line, perpendicularly bisecting segment DD’. The CE angle was measured by drawing a vertical line down to the center of the femoral head (G), with an angle completed by the extension of a line tangential to the lateral edge of the acetabular roof (H).
Fig. 1. Geometric hip measurements.
Results Fifty-three AFF patients with 63 femora radiographs were identified. Radiographic measurements of the AFF group were compared with 43 NFF patients with 80 femora radiographs and 64 ITF patients with 79 femora films. Patient demographic characteristics and results from statistically significant comparisons are displayed in Table 1. Mean duration of bisphosphonate use was nearly 8 years in both the AFF (7.9 [±3.5] years) and NFF (7.7 [±3.3] years) patient groups (p = 0.7). Mean NSA measurement for all three groups was 129.5 [±6.3] degrees and consistent with published population means [26]. The overall mean HAL was 125.6 [±13.0] mm and mean CE angle was 44.6[±5.5] degrees. In the three-way group comparison, the proximal femoral anatomy of the AFF group was significantly different from each of the NFF and ITF on nearly all geometric measures (Table 2). The AFF patients were found to be more varus (NSA = 126.4 [± 6.6] degrees), with a shorter hip axis (HAL = 120.3 [± 11.7] mm) and narrower CE angle (42.6 [± 6.2] degrees). A weak correlation between NSA and HAL existed within the bisphosphonate groups (AFF + NFF) (r = 0.34, p b 0.01). When comparing the relationship between NSA and HAL (Fig. 2), patients who developed fracture tended to cluster in the area
of varus NSA with shorter HAL. Sub-group analysis between the NFF and ITF groups on measures of NSA (p = 0.51), HAL (p = 0.73), and CE angle (p = 0.52) revealed no statistical differences. Table 1 Demographics and patient characteristics.
Patients #n Female (%)a
b
Age (years) Height (in.) BMI (kg/m2) Bisphosphonate duration (years)
AFF
NFF
ITF
53 96.2
43 90.7
64 78.1
Mean (μ) [sd]
Mean (μ) [sd]
Mean (μ) [sd]
69.5 [10.1] 62.6 [3.5] 25.8 [6.7] 7.9 [3.5]
68.8 [10.3] 63.1 [3.1] 23.8 [3.3] 7.7 [3.3]
79.9 [12.3]c,d 63.1 [4.0] 24.9 [5.1] n/a
AFF = Atypical Femoral Fracture patients; NFF = No Femoral Fracture patients; ITF = Intertrochanteric Fracture patients. Unless otherwise noted, there was no statistical difference in one-way ANOVA comparisons between the three groups at a p ≤ 0.05 significance level, nor in Tukey pairwise post-hoc comparisons. a Chi-square p = 0.01. b One-way ANOVA p b 0.001. c ITF vs. AFF pairwise comparison p b 0.001. d ITF vs. NFF pairwise comparison p b 0.001.
4
D.P. Taormina et al. / Bone 63 (2014) 1–6
Table 2 Radiographic measurement means [and SDs] as well as differences between group means [and 95% confidence intervals].
Neck-shaft angle (°) Hip-axis length (mm.) Center-edge angle (°)
AFF
NFF
ITF
Mean difference (AFF vs NFF)
Mean difference (AFF vs ITF)
p-value
126.4 [6.6]a,b 120.3 [11.7]c,d 42.6 [6.2]e
130.3 [5.8] 127.3 [12.4] 45.1 [4.8]
131.1 [5.6] 128.2 [13.4] 45.8 [5.0]
3.9 [0.9–7.0] 7.0 [0.6–13.5] 2.6 [0.1–5.3]
4.7 [1.9–7.4] 7.9 [2.1–13.8] 3.2 [−0.8–5.7]
b0.001 0.004 0.007
AFF = Atypical Femoral Fracture patients; NFF = No Femoral Fracture patients; ITF = Intertrochanteric Fracture patients unless otherwise noted, there was no statistical difference in Tukey pairwise post-hoc comparisons at a p ≤ 0.05 significance level. a AFF vs. NFF pairwise comparison p b 0.01. b AFF vs. ITF pairwise comparison p b 0.001. c AFF vs. NFF pairwise comparison p b 0.05. d AFF vs. ITF pairwise comparison p b 0.01. e AFF vs. ITF pairwise comparison p b 0.01.
Logistic regression model analysis using a grouped AFF + NFF sample revealed a significant association between NSA (OR = 0.89 [95% CI = 0.81–0.97; p = 0.01), CE angle (OR = 0.89 [95% CI = 0.80–0.99]; p = 0.03), and BMI (OR = 1.15 [95% CI = 1.02–1.31; p = 0.03) with development of fracture. Although hip-axis length was associated with the development of fracture when conducting univariate logistic regression, the significance was lost after adjusting for other covariates. ROC curve analysis (area under curve = 0.67 [95% CI = 0.56–0.79]) revealed that an NSA b 128.3° yielded a 69% sensitivity and 63% specificity to predict the development of a complete or incomplete atypical femur fracture. Discussion With this study we have demonstrated an association between acute (varus) neck-shaft angle proximal femur anatomy and the
development of atypical femoral fractures after long-term bisphosphonate use. While no definitive conclusions can be reached regarding the role that altered proximal femoral anatomy plays in the development of atypical fractures, the establishment of an association may help physicians safely manage patients who are prescribed bisphosphonates. There are a variety of treatment options available to the osteoporotic patient and these include bisphosphonates, estrogen replacement, selective estrogen receptor modulators, Denosumab and Teriparatide [27]. Bisphosphonates are unique in that they can induce prolonged suppression of bone turnover and sustained reduction of osteoclastic activity [27,28]. Chronic suppression of bone turnover, along with the addition of non-enzymatically linked collagen molecules is believed to limit the plasticity and heterogeneity of developing bone [29]. This chronic therapy, in turn, may become the façade for weakening bony
Among all bisphosphonate users, NSA (X-axis) versus HAL (Y-axis) was plotted independently for patients who developed either complete or incomplete atypical femur fracture (AFF group; diamond) and those who had not fractured (NFF group; circle). Patients with femora which fractured tended to cluster having both shorter HAL and narrower NSA. Femora which had obtuse NSA and with larger HAL tended not to fracture.
Fig. 2. Hip-axis length versus neck-shaft angle in chronic bisphosphonate users.
D.P. Taormina et al. / Bone 63 (2014) 1–6
architecture and increasing micro-lesion accumulation with the potential for ultimate progression to gross fracture [9,28,30]. Unfortunately, the ubiquity of bisphosphonate therapy for the treatment of osteoporotic patients estranges a cohort who may ultimately succumb to a dangerously silent milieu of micro-anatomic changes [28,31]. One large cohort study (Abrahamsen et al., 2010) estimated that the general population incidence of subtrochanteric and diaphyseal femoral shaft fractures, per 10,000 patient-years, increases from 13 for women and 6 for men to 31 for both genders after bisphosphonate use [32]. Patients with osteoporosis brave these odds hoping to stave off femoral neck, intertrochanteric, and vertebral fractures, although bisphosphonate-associated atypical femur fractures are far from inconsequential [15,33]. Several studies have reported the use of proximal femur geometrics as markers for risk of osteoporosis related fragility fracture. In 1999, Gnudi et al. published findings from 111 Caucasian post-menopausal women and 329 Caucasian controls. Their comparison found that valgus NSA (wide-angle) and longer HAL were associated with greater risk of femoral neck and trochanteric hip fracture [34]. In a longitudinal follow-up, Gnudi and colleagues (2012) re-affirmed these findings by using automated geometric measurements [35]. Our group of patients with bisphosphonate-associated atypical femur fracture seem distinctly different from those identified as at risk in studies of osteoporosis related fractures [34,35]. Our patient group had acute NSAs with short HALs. Given the pattern of the atraumatic atypical femur fracture, the increased lateral cortical loading of a varus hip seems consistent with the mechanism of injury development. Although there has been a moderate association established between the development of hip fractures (particularly femoral neck) and increasing HAL, there are no studies on associations of HAL with subtrochanteric fractures. While we found a weak correlation between varus NSA and shorter HAL, we are unable to conclude whether shorter HAL is a true risk for atypical fracture, a surrogate marker (related to NSA, femoral anteversion, or femoral bow), or the product of a genetic confounder. Consistent with the findings of Franceschetti et al., we also identified increasing BMI as a risk factor for the development of fracture in chronic bisphosphonate users [36]. There are several limitations to this study. While we took as many precautions as possible to de-identify patients during radiographic measurements, surgeons could have recalled the patients they treated, thus leading to bias. There remains the possibility of measurement error during radiographic evaluation. There is also potential for limb rotation to confound radiographic readings even with a standardized technique for positioning and radiographic assessment. Nonetheless, our studies mean measurements, particularly NSA (mean = 129.5), are nearly identical to other published findings (Elbuken reported a population mean amongst 18,943 individuals = 129.2°) [26]. While the full breath of research encompassed any patient presenting (or referred) with symptomatic bisphosphonate associated fractures, enrollment of AFFs was performed by using the case load of physicians at an urban academic medical center. It is difficult to discern the extent of demographic confounders and their effect on generalizing the study findings. While other studies have published differences between proximal and distal atypical femoral fractures [37], measurement of the distance between the lesser trochanter and fracture site was not included within our measurement protocol. Thus, future studies may seek to correlate proximal hip geometry with the site of atypical fracture. Conclusions The results of this study demonstrated that chronic bisphosphonate users who developed symptomatic atypical femoral fracture had more varus (acute) neck-shaft angles and shorter hip-axis length. Furthermore, a clear association between NSA and BMI with the development of atypical femur fracture was identified. Patients who are on bisphosphonates and have a more acute (varus) neck-shaft angle as
5
well as those who have higher BMI are more likely to develop symptomatic atypical femur fractures. This study, to the best of our knowledge, is the first to identify neck-shaft angle as a potential geometric risk factor associated with the development of symptomatic atypical femur fracture.
References [1] Johnell O, Kanis J. Epidemiology of osteoporotic fractures. Osteoporos Int 2005;16:S3–7. [2] Capeci CM, Tejwani NC. Bilateral low-energy simultaneous or sequential femoral fractures in patients on long-term alendronate therapy. J Bone Joint Surg 2009;91-A: 2556–61. [3] Lenart BA, Lorich DG, Lane JM. Atypical fractures of the femoral diaphysis in postmenopausal women taking alendronate. N Engl J Med 2008;358:1304–6. [4] Park-Wyllie LY, Mamdani MM, Juurlink DN, Hawker GA, Gunraj N, Austin PC, et al. Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. J Am Med Assoc 2011;305:783–9. [5] Koval KJ, Zuckerman JD. Fractures in the elderly. Philadelphia: Lippincott-Raven Publishers; 1998 15. [6] Egol KA, Park JH, Prensky C, Rosenberg ZS, Peck V, Tejwani NC. Surgical treatment improves clinical and functional outcomes for patients who sustain incomplete bisphosphonate-related femur fractures. J Orthop Trauma 2013;27:331–5. [7] Neviaser AS, Lane JM, Lenart BA, Edobor-Osula F, Lorich DG. Low-energy femoral shaft fractures associated with alendronate use. J Orthop Trauma 2008;22:346–50. [8] Lenart BA, Neviaser AS, Lymann S, Chang CC, Edobor-Osula F, Steele B, et al. Association of low-energy femoral fractures with prolonged bisphosphonate use: a case control study. Osteoporos Int 2009;20:1353–62. [9] Odvina CV, Zerwekh JE, Rao DS, Maalouf N, Gottschalk FA, Pak CYC. Severely suppressed bone turnover: a potential complication of alendronate therapy. J Clin Endocrinol Metab 2005;90:1294–301. [10] Armamento-Villareal R, Napoli N, Panwar V, Novack D. Suppressed bone turnover during alendronate therapy for high-turnover osteoporosis. N Engl J Med 2006;355:2048–50. [11] Shane E, Burr D, Abrahamsen B, Adler RA, Brown TD, Cheung AM, 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. [12] Hollick R, Reid D. Role of bisphosphonates in the management of postmenopausal osteoporosis: an update on recent safety anxieties. Menopause Int 2011;17:66–72. [13] Shane E, Burr D, Ebeling PR, Abrahamsen B, Adler RA, Brown TD, et al. American Society for B, Mineral R. Atypical Subtrochanteric and Diaphyseal Femoral Fractures: Report of a Task Force of the American Society for Bone and Mineral Research. J Bone Miner Res 2010;25:2267–94. [14] Black DM, Kelly MP, Genant HK, Palermo L, Eastell R, Bucci-Rechtweg C, et al. Fracture Intervention Trial Steering C, Committee HPFTS. Bisphosphonates and fractures of the subtrochanteric or diaphyseal femur. N Engl J Med 2010;362:1761–71. [15] Napoli N, Schwartz AV, Palermo L, Jin JJ, Wustrack R, Cauley JA, et al. Risk factors for subtrochanteric and diaphyseal fractures: the study of osteoporotic fractures. J Clin Endocrinol Metab 2013;98:659–67. [16] Egol KA, Park JH, Prensky C, Rosenberg ZS, Peck V, Tejwani NC. Surgical treatment improves clinical and functional outcomes for patients who sustain incomplete bisphosphonate-related femur fractures. J Orthop Trauma 2013;27:331–5. [17] Allison MB, Markman L, Rosenberg Z, Vieira RL, Babb J, Tejwani N, et al. Atypical incomplete femoral fractures in asymptomatic patients on long term bisphosphonate therapy. Bone 2013;55:113–8. [18] Egol KA, Park JH, Rosenberg ZS, Peck V, Tejwani NC. Healing delayed but generally reliable after bisphosphonate-associated complete femur fractures treated with IM nails. Clin Orthop Relat Res 2013 [Epub (ahead of print)]. [19] Marcano A, Taormina D, Egol KA, Peck V, Tejwani NC. Are race and sex associated with the occurrence of atypical femoral fractures? Clin Orthop Relat Res 2014;472:1020–7. [20] Rosenberg ZS, Vieira RLR, Chan SS, Babb J, Akyol Y, Rybak LD, et al. Bisphosphonaterelated complete atypical subtrochanteric femoral fractures: diagnostic utility of radiography. Am J Roentgenol 2011;197:954–60. [21] Vieira RLR, Rosenberg ZS, Allison MB, Im SA, Babb J, Peck V. Frequency of incomplete atypical femoral fractures in asymptomatic patients on long-term bisphosphonate therapy. Am J Roentgenol 2012;198:1144–51. [22] Faulkner KG, Cummings SR, Black D, Palermo L, Glüer C-C, Genant HK. Simple measurement of femoral geometry predicts hip fracture: the study of osteoporotic fractures. J Bone Miner Res 1993;8:1211–7. [23] Fredensborg N, The CE. Angle of normal hips. Acta Orthop 1976;47:403–5. [24] Ogata K, Goldsand EM. A simple biplanar method of measuring femoral anteversion and neck-shaft angle. J Bone Joint Surg 1979;61:846–51. [25] Karanicolas PJ, Bhandari M, Kreder H, Moroni A, Richardson M, Walter SD, et al. Collaboration for outcome assessment in surgical trials musculoskeletal G. Evaluating agreement: conducting a reliability study. J Bone Joint Surg 2009;91:99–106. [26] Elbuken F, Baykara M, Ozturk C. Standardisation of the neck-shaft angle and measurement of age-, gender-and BMI-related changes in the femoral neck using DXA. Singapore Med J 2012;53:587–90. [27] Cermak K, Shumelinsky F, Alexiou J, Gebhart MJ. Case reports: subtrochanteric femoral stress fractures after prolonged alendronate therapy. Clin Orthop Relat Res 2010;468:1991–6. [28] Giusti A, Hamdy NA, Papapoulos SE. Atypical fractures of the femur and bisphosphonate therapy. A systematic review of case/case series studies. Bone 2010;47:169–80.
6
D.P. Taormina et al. / Bone 63 (2014) 1–6
[29] Ettinger B, Burr DB, Ritchie RO. Proposed pathogenesis for atypical femoral fractures: lessons from materials research. Bone 2013;55:495–500. [30] Mashiba T, Hirano T, Turner CH, Forwood MR, Johnston CC, Burr DB. Suppressed bone turnover by bisphosphonates increases microdamage accumulation and reduces some biomechanical properties in dog rib. J Bone Miner Res 2000;15:613–20. [31] Iwamoto J, Sato Y, Takeda T, Matsumoto H. Hip fracture protection by alendronate treatment in postmenopausal women with osteoporosis: a review of the literature. Clin Interv Aging 2008;3:483–9. [32] Abrahamsen B, Eiken P, Eastell R. Cumulative alendronate dose and the long-term absolute risk of subtrochanteric and diaphyseal femur fractures: a register-based national cohort analysis. J Clin Endocrinol Metab 2010;95:5258–65.
[33] Schilcher J, Michaëlsson K, Aspenberg P. Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med 2011;364:1728–37. [34] Gnudi S, Ripamonti C, Gualtieri G, Malavolta N. Geometry of proximal femur in the prediction of hip fracture in osteoporotic women. Br J Radiol 1999;72:729–33. [35] Gnudi S, Sitta E, Pignotti E. Prediction of incident hip fracture by femoral neck bone mineral density and neck-shaft angle: a 5-year longitudinal study in postmenopausal females. Br J Radiol 2012;85:e467–73. [36] Franceschetti P, Bondanelli M, Caruso G, Ambrosio MR, Lorusso V, Zatelli MC, et al. Risk factors for development of atypical femoral fractures in patients on long-term oral bisphosphonate therapy. Bone 2013;56:426–31. [37] Koeppen VA, Schilcher J, Aspenberg P. Dichotomous location of 160 atypical femoral fractures. Acta Orthop 2013;84:561–4.