Atypical Femoral Fractures—Ongoing and History of Bone-Specific Therapy, Concomitant Diseases, Medications, and Survival

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health, vol. -, no. -, 1e9, 2015 Ó Copyright 2015 by The International Society for Clinical Densitometry 1094-6950/-:1e9/$36.00 http://dx.doi.org/10.1016/j.jocd.2015.05.070

Original Article

Atypical Femoral FracturesdOngoing and History of Bone-Specific Therapy, Concomitant Diseases, Medications, and Survival Christian Muschitz,*,1 Heinrich W. Thaler,2 Hans Peter Dimai,3 Heinrich Resch,1 Roland Kocijan,1 Martina Kostic,1 Corinna Geiger,1 Alexandra Kaider,4 Gabriela K. Muschitz,5 Michael Szivak,6 and Peter Pietschmann7 1

Medical Department II, St. Vincent Hospital, Academic Teaching Hospital of the Medical University of Vienna, Vienna, Austria; 2Trauma Center Meidling, Vienna, Austria; 3Division of Endocrinology and Metabolism, Department of Internal Medicine, Medical University of Graz, Graz, Austria; 4Center for Medical Statistics, Informatics, and Intelligent Systems, Medical University of Vienna, Vienna, Austria; 5Division of Plastic and Reconstructive Surgery, Department of Surgery, Medical University of Vienna, Vienna, Austria; 6Department of Medical Documentation and Statistics, Austrian Trauma Insurance Agency (AUVA), Vienna, Austria; and 7Department of Pathophysiology and Allergy Research, Center for Pathophysiology, Infectiology, and Immunology, Medical University of Vienna, Vienna, Austria

Abstract Although atypical femoral fractures (AFFs) are generally rare events; several studies have indicated a potential link between AFF and long-term bone-specific therapies (BSTs). The aim of this study was to analyze the frequency of AFF and potential associations with prior or ongoing BST. A total of 8851 Caucasian female and male patients with de novo hip fractures treated in the largest Austrian level 1 trauma center from 2000 to 2013 were selected. Of the total, 194 patients with a de novo low-traumatic subtrochanteric or shaft fractures were identified: 35 atypical and 159 typical fractures. Of these patients, concomitant diseases, medication, previous fractures, and survival data were retrieved and analyzed. Female patients in both groups were significantly older. The median survival was significantly shorter in patients with AFF (9 vs 18 months; p ! 0.0001). Cardiovascular disease, sarcopenia, chronic kidney disease, type 2 diabetes, smoking (past or current history), and prevalent fragility fractures were more frequent in AFF patients, as well as the concomitant use of phenprocoumon, furosemide, and sulfonylurea. Although the number of patients with current BST was less in (14.5%) both groups, more patients in the AFF group were previously treated with BST (71% vs 49%; p 5 0.016), and they received these therapies for a longer time period. A combination of severe comorbidities, long-term pharmaceutical therapies, and a history of previous or ongoing BST was associated with an increased individual risk for AFF. Key Words: Antiresorptives; atypical femoral fractures; concomitant medications; osteoporosis; survival.

Introduction

Received 05/01/15; Revised 05/22/15; Accepted 05/27/15. A plenary oral presentation of this study was delivered at the Austrian National Osteoporosis Congress in St. Wolfgang on 9th May, 2014. A plenary poster was also presented at the American Society for Bone and Mineral Research 2014 meeting in Houston, TX, USA, on September 12, 2014. *Address correspondence to: Christian Muschitz, MD, Medical Department II, St. Vincent Hospital, Academic Teaching Hospital of the Medical University of Vienna, Stumpergasse 13, Vienna A-1060, Austria. E-mail: [email protected]

Osteoporosis is strongly associated with low-trauma fractures, mortality, and economic consequences. Hip fractures typically occur in the femoral neck or intertrochanteric region but less frequently in the subtrochanteric or shaft region of the femur (1,2). Atypical femoral fractures (AFFs) generally emerge after a minimal trauma with a characteristic radiographic pattern. AFFs may be stress or insufficiency fractures, potentially arising from increased tensile forces (3e5). Case 1

2 reports and observational studies have raised concerns about a potential link between long-term bisphosphonate (BP) use, diverse medications, and AFF (6e9). A general warning for all BPs was pronounced by the Food and Drug Administration and the European Medicines Agency in 2010 and 2011 (10,11). Although randomized trials have shown that treatment with BPs reduces the risk of osteoporotic fractures, longterm use may suppress bone remodeling sufficiently to negatively interfere with the process of microdamage repair in a subset of patients (6,12,13). Studies examining the potential association between AFF risk and long-term BP use are conflicting owing to the differences in size, design, outcome, and definition of AFF (2,14e16). Concomitant diseases and long-term medications influence endocrinological pathways and bone metabolism and might increase the individual risk for an AFF. The main purpose of this study was to compare patients with AFF and typical fractures with regard to prior or ongoing bone-specific therapies (BSTs). Secondary objectives included the evaluation of potential associations with comorbidities, concomitant medications and laboratory values, and the survival in these 2 groups.

Methods Data Source and Study Patients This study was conducted at the Hospital of Meidling in Vienna, Austria. This hospital is the largest level 1 trauma center with O60,000 new ambulatory and inpatient patients per year (catchment area, 1.9 million people). Patients eligible for the present study were identified by International Classification of Disease 10 coding. All patients with de novo hip fractures treated between January 2000 and December 2013 with documented recent low-traumatic subtrochanteric or femoral shaft fractures and corresponding X-ray and complete medical data sets were selected. Patient-related data were gathered at the time of the fracture in the hospital as well as by contacting the treating general practitioners to retrieve the respective duration of any medication or BST and further outcome after discharge from the trauma center. Demographic data (age, gender), data related to the incident, serum values, current medication, comorbidities, and the survival after fracture or the respective cause of death were collected. Areal bone mineral density (aBMD) measurements taken less than 1 year before the fracture were also evaluated. Sarcopenia was defined as muscle atrophy combined with progressive loss of muscle function and frailty and/or unintentional nonedematous weight loss of O5% over 6 months. To select appropriate potential clinical risk factors for analysis, we screened the following different established research tools and risk calculators for osteoporosis and fracture risk before the initiation of the study: the FRAX calculator, the QFracture-2013 risk calculator, and the current osteoporosis guidelines of the Austrian-German-Swiss Osteology Association (DVO Osteologie).

Muschitz et al. Fracture Assessment All subtrochanteric femoral fractures were independently and consecutively reanalyzed by 2 experienced blinded radiologists from 2 different centers. X-ray results of all de novo low-traumatic femoral subtrochanteric or femoral shaft fractures were categorized based on the latest major criteria defined by the American Society for Bone and Mineral Research (3). The most frequent reason for classifying the fractures as ‘‘typical’’ was an extension of the fracture line into the region of the lesser trochanter. The second reason was the existence of multiple bone fragments or a more oblique configuration of the fracture line. Optional minor criteria were not available in all patients and are therefore not reported.

Ethical Aspects This study was conducted in line with the Declaration of Helsinki and gained approval from the local ethics committee (AUVA-EC-number: 21/2011).

Statistical Analysis Continuous variables are described by the median (interquartile range). Group comparisons of continuous variables were performed using the 2-sample t test. In case of nonnormally distributed data, the nonparametric Wilcoxon rank sum test was used. Differences with respect to binary variables were tested using the chi-square test or, as appropriate, the Fisher exact test. A stepwise logistic regression analysis was applied to compare patients with AFF to patients with typical fractures in a multivariate, adjusted sense. Thereby, to reduce a possible overestimation of any effect, all concomitant diseases and medications were considered as variables in a forward selection approach to discriminate between these 2 patients groups, adjusted for age and gender. The Firth correction was used in logistic regression models for bias reduction of the estimates. The Kaplan-Meier method was applied to estimate the survival probabilities and the log-rank test was calculated to compare the survival curves. Univariate and multivariate Cox regression models were performed to evaluate the unadjusted and age- and sex-adjusted difference in mortality in AFF patients compared with patients with typical femoral fractures. Two-sided p values !0.05 were considered as indicating statistical significance. All our statistical calculations were based on patients, not on fractures. If a patient had an atypical and a typical fracture simultaneously, this patient was considered as an AFF patient. The SAS software (version 9.3; SAS Institute Inc., 2002-2010; Cary, NC) was used for data analyses.

Results Between 2000 and 2013, a total of 8851 patients with de novo hip fractures were treated in the trauma center of Meidling, thereof 3337 (37.7%) with femoral neck fractures, 3233 (36.6%) with intertrochanteric fractures, and 1288 (14.5%) patients with subtrochanteric or femoral shaft fractures. The percentage of low-traumatic fractures was 73.2%

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

Atypical Femoral Fractures and Concomitant Diseases for the femoral neck, 76.4% for the intertrochanteric region, and 27.1% for the investigated subtrochanteric or shaft region. Figure 1 shows the patient flow generating the study population, with 35 AFFs (18%) and 159 typical fractures (82%) diagnosed. The gender distribution was similar in both groups (odds ratio [OR], 1.03), but female patients were significantly older than males regardless of the fracture type (median age: 83 vs 73 years; p ! 0.001; Fig. 2). The mean aBMD T-scores, available in approximately 75% of all patients, were in the osteopenic or close to the osteoporotic range with lower vertebral T-scores in AFF patients ( p 5 0.011; Table 1). Patients with AFF had significantly lower hemoglobin or thrombocyte levels but higher values of alkaline phosphatase, C-reactive protein (CRP), and creatinine. Serum calcium levels and 25-OH vitamin D levels were lower in AFF patients (2.24 vs 2.33 mmol/L; p 5 0.013), and intact parathyroid hormone (iPTH) was higher (67 vs 56 pg/ mL; p 5 0.006). Markers of bone resorption and bone formation were only available in a small number of patients (12.5% and 21.3%, respectively). Comorbidities such as chronic heart failure (OR, 3.53; 95% confidence interval [CI], 1.66e7.52; p ! 0.005), hyperlipidemia (OR, 2.23; 95% CI, 1.04e4.77; p 5 0.035), hypothyroidism (OR, 2.58; 95% CI, 1.12e5.94; p 5 0.023), type 2 diabetes (OR, 2.39; 95% CI, 1.11e5.12; p 5 0.002), chronic kidney disease (CKD) grade 3 or 4 (OR, 3.60; 95% CI, 1.64e7.87; p ! 0.005), and sarcopenia (OR, 2.68; 95% CI,

3 1.27e5.72; p 5 0.009) were more frequent in patients with AFF ( p ! 0.05 for all; Table 2). Hypothyroidism was more frequent in the female population (21.2% vs 2.9%; p 5 0.012), whereas epilepsy (1.9% vs 8.8%; p 5 0.068) and type 2 diabetes (41.2% vs 24.4%; p 5 0.046) were more frequent in the male study population. There were more patients with AFF who concomitantly took angiotensin-conerting-enzyme (ACE) inhibitors, furosemide, sulfonylurea, acetylsalicylic acid ( p ! 0.05 for all), or longacting oral anticoagulant phenprocoumon and former or current smokers ( p ! 0.0001for both; Table 3). AFF patients had more vertebral ( p 5 0.063), nonvertebral ( p 5 0.003), or hip fractures ( p ! 0.005) in their medical history. Furthermore, patients with AFF had significantly more bilateral hip fractures (22.9% vs 11.3%; p ! 0.005). The median survival in patients with a typical subtrochanteric or femoral shaft fracture was 18 months compared to 9 months in the AFF population ( p ! 0.0001; Fig. 3), with an age- and sex-adjusted hazard ratio of 3.74 (95% CI, 2.46e5.56) for AFF. The most frequent causes of death were terminal cardiac insufficiency, pneumonia, urosepsis, and renal failure without significant differences in both populations (Table 3). Seventy-four percent of patients with AFF and 84% with typical fractures did not receive any antiresorptive (oral or intravenous BPs, denosumab, raloxifene), dual-acting

Fig. 1. Study population flow. BST, bone-specific therapy.

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

4

Muschitz et al.

Fig. 2. Box plot of age distribution in female and male patients with subtrochanteric and femoral shaft fractures (n 5 194). (strontium ranelate), or anabolic therapy (teriparatide) at the time of fracture (Table 4). More than 70% of the AFF patients had a medical history of BST compared to 49% ( p 5 0.016) without any BST. The

most frequently used antiresorptive regimen in both groups was weekly alendronate (37.1% and 18.2%). Strontium ranelate was used previously in 17.1% of the AFF patients and 4.4% of patients with typical fractures; intravenous ibandronate was taken by 2.9% with AFF compared to 10.7% of patients with typical fractures. Raloxifene use was 8.6% and 3.8% in the AFF and non-AFF group, respectively. Two patients with typical fractures had a history of denosumab use. The duration of these therapies varied substantially in both groups, with a longer median application time in the AFF population ( p 5 0.012). The mean duration since discontinuation of these therapies to fracture was 19.5  18.8 months in the AFF group vs 11.6  17.4 months ( p 5 0.007; Table 4). After Firth’s bias correction procedure, hypothyroidism (OR, 2.82), advanced stages of CKD (OR, 4.22), furosemide (OR, 5.92), and phenprocoumon (OR, 5.59), but not age, gender, or diabetes mellitus, were significant contributing factors in AFF patients (Table 5). There was also a trend for sulfonylurea (OR, 3.19; p 5 0.066) to be associated with AFF. Both univariate and multivariate analyses did not reveal any statistically significant effect of the BST status in discriminating between AFF and typical femoral fracture patients.

Table 1 Patient Characteristics, Areal Bone Mineral Density (aBMD), and Laboratory Findings at the Time of Fracture Fracture type, median (interquartile range) Characteristic Age (yr) Study population Females Males BMI (kg/m2) aBMD L1eL4 (available: 72%; 81%) Hip (available: 75%; 82%) Laboratory Hemoglobin (g/dL) Thrombocytes (g/L) Alkaline phosphatase (IU/L) CRP (mg/L) Glucose (mg/dL) HbA1c (%) Creatinine (mg/dL) PTT (s) TSH (mU/mL) 25-OH vitamin D (ng/mL) Calcium (mmol/L) iPTH (pg/mL)

Atypical (n 5 35)

Typical (n 5 159)

82 83 73 22.3

82 83 71 22.8

(79 to 86) (81 to 87) (62 to 79) (19.6 to 23.9)

2.3 ( 2.6 to 2.3 ( 3.0 to 9.1 205 99 89 130 7.5 1.08 27.9 1.45 16.7 2.24 67

1.0) 1.8)

(8.8 to 9.6) (131 to 232) (69 to 122) (79 to 101) (113 to 175) (7.2 to 8.5) (0.93 to 1.99) (25.5 to 30.3) (0.64 to 2.19) (11.5 to 17.6) (2.20 to 2.33) (52.8 to 87)

(74 to 88) (77 to 90) (63 to 79) (19.6 to 24.6)

2.7 ( 3.3 to 2.6 ( 3.1 to 12.0 228 70 55 128 7.6 0.96 26.1 2.01 19.0 2.33 56

p Value 0.722 0.641 1.0 0.247

1.9) 1.9)

0.011 0.233 !0.001 0.009 0.005 !0.001 0.189 0.135 0.010 0.009 0.097 0.001 0.013 0.006

(11.3 to 13.4) (182 to 275) (59 to 99) (34 to 89) (113 to 145) (6.9 to 8.2) (0.81 to 1.32) (24.2 to 29.2) (1.29 to 2.45) (14.1 to 23.0) (2.21 to 2.42) (49.1 to 67.6)

Note: The aBMD was assessed at a median of 12 months before fracture and vitamin D, calcium, and iPTH at a median of 9 months before and after fracture. Abbr: BMI, body mass index; PTT, partial thromboplastin time; TSH, thyroid-stimulating hormone. Significant p-values are bold. Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

Atypical Femoral Fractures and Concomitant Diseases

5

Table 2 Diseases and Current Medications: Odds Ratios of AFF (Comparing ‘‘Yes’’ vs ‘‘No’’), Chi-Square or Fisher Exact Test Fracture type Diseases and medications Concomitant diseases, n (%) Arterial hypertension Arteriosclerosis COLD/asthma Dementia Epilepsy Chronic heart failure Coronary heart disease Cardiomyopathy Atrial fibrillation Hyperlipidemia Hypothyroidism Hypothyreosis Stroke Rheumatoid arthritis Sarcopenia Type 2 diabetes Chronic kidney disease III/IV Medication, n (%) Current BST ACE inhibitors Beta blockers Furosemide Hydrochlorothiazide Statins Phenprocoumon Proton pump inhibitors H2-receptor antagonists Insulin Metformin Sulfonylurea Acetylsalicylic acid Clopidogrel SSRI Tricyclic antidepressants Antiepileptic drugs Levothyroxine Thyreostatic drugs Methotrexate Glucocorticoids TNF-a inhibitors Calcium and/or vitamin D

Atypical, n 5 35 31 19 6 13

Typical, n 5 159

Odds ratio

p Value

(88.6) (54.3) (17.5) (37.1) 0 (54.3) (42.9) (5.7) (34.3) (42.9) (5.7) (31.4) (20.0) (5.7) (45.7) (42.9) (68.6)

118 58 20 44 6 40 58 10 37 40 11 24 19 10 38 38 60

(74.2) (36.6) (12.6) (27.7) (3.8) (25.2) (36.5) (6.3) (23.3) (25.2) (6.9) (15.1) (12.0) (6.3) (23.9) (23.9) (37.4)

2.69 2.07 1.44 1.54 n.e. 3.53 1.31 0.90 1.72 2.23 0.82 2.58 1.84 0.90 2.68 2.39 3.60

0.069 0.051 0.583 0.266 0.594 !0.005 0.481 1.000 0.175 0.035 1.000 0.023 0.264 1.000 0.009 0.002 !0.005

(25.7) (57.1) (34.3) (68.6) (34.3) (65.7) (28.6) (71.4) (8.6) (20.0) (31.4) (22.9) (17.1) (5.7) (48.6) (20.0) (5.7) (31.4) (14.3) (2.9) (2.9) 0 18 (51.4)

26 55 47 58 49 82 11 99 12 19 28 9 60 9 59 18 8 41 10 10 4 1 58

(16.4) (34.6) (29.6) (36.5) (30.9) (51.6) (6.9) (62.2) (7.6) (12.0) (17.6) (5.7) (30.7) (5.6) (37.1) (11.3) (5.0) (25.8) (6.3) (6.3) (2.5) (0.6) (36.5)

1.77 2.52 1.24 3.80 1.17 1.80 5.38 1.51 1.15 1.84 2.14 4.94 0.34 1.01 1.60 1.96 1.14 1.32 2.48 0.44 1.14 n.e. 1.84

0.192 0.013 0.582 !0.005 0.689 0.129 !0.001 0.307 0.737 0.270 0.065 0.001 0.020 1.000 0.208 0.170 1.000 0.495 0.153 0.692 1.000 1.000 0.101

19 15 2 12 15 2 11 7 2 16 15 24 9 20 12 24 12 23 10 25 3 7 11 8 6 2 17 7 2 11 5 1 1

Abbr: BST, bone-specific therapy; n.e., not evaluated. Significant p-values are bold.

Discussion In this retrospective observational study, our findings indicate that AFFs are rare events, accounting for 0.39% of all hip fractures in this population and for 2.7% of all

subtrochanteric or femoral shaft fractures, which is in line with previously published data (17). We made a distinction between ongoing and previous therapy with diverse BST in both genders and included several potentially associated

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

6

Muschitz et al. Table 3 Fracture Classification, Lifestyle Components, Cause of Fracture, and Death Divided by Fracture Type Fracture type, n (%)

Parameter

Atypical (n 5 35)

Typical (n 5 159)

29 (82.9) 6 (17.1)

131 (82.4) 28 (17.6)

35 (100) 0 8 (22.9)

150 (94.3) 9 (5.7) 18 (11.3)

27 (77.2) 20 (57.1)

49 (30.8) 92 (57.9)

20 (57.1) 11 (31.4) 4 (11.4)

113 (71.5) 35 (22.2) 10 (6.3)

32 (91.4) 1 (2.9) 2 (5.7)

156 (98.1) 0 3 (1.9)

Gender Population Female gender Male gender Fracture classification Complete Incomplete Bilateral fracture Lifestyle Smoking (former/current) Alcohol (former/current) Residential status At home Nursing home Hospital Cause of fracture Fall from standing height Atraumatic No recall, reported atraumatic Prevalent fragility fractures Vertebral Nonvertebral Hip Any fracture Cause of death Terminal cardiac insufficiency Pneumonia Urosepsis Pulmonary embolism Myocardial infarction Insult Renal failure

Odds ratio

p Value

1.03

0.948

n.e.

0.368

4.63

!0.005

7.58 0.97

!0.0001 0.096 0.232

0.073

18 15 12 32

(51.4) (42.9) (34.3) (91.4)

55 31 15 86

(34.6) (19.5) (9.4) (54.1)

(38.2) (26.5) (17.7) (8.8) 0 1 (2.9) 2 (5.9)

73 20 15 11 7 7 9

(50.7) (13.9) (10.4) (7.4) (4.9) (4.9) (6.3)

13 9 6 3

2.00 3.10 5.01 9.05

0.063 0.003 !0.005 !0.0001 n.d.

Abbr: n.d., not determined; n.e., not evaluated. Significant p-values are bold.

diseases and medications in our analysis including the survival. Recent literature describes increases in the incidence of subtrochanteric or femur shaft fractures based on the duration of BPs with a 3-fold increase in year 5 compared to the first year (18). Other findings state that patients with AFF tended to be younger and more likely to be Asian, have osteoporosis, and use BPs. (19). In our study population, the percentage of any current BST and the respective duration were comparable, suggesting no association with BPs in this context. In contrast to this, the proportion of patients with AFF and a history of BP use were significantly higher. In addition, AFF patients had a longer mean duration of previous therapy, but the mean duration since discontinuation of BP was higher in

the AFF group. Alendronate was one of the first approved antiresorptives and this is probably the reason why more patients received this medication compared to other BSTs. Suppressed bone formation together with reduced osteoblast activity and matrix synthesis is considered a possible pathomechanism for increased nonvertebral fracture risk (20). A study in postmenopausal women with a mean age of 71 years evaluated bone microarchitecture in glucocorticoid-na€ıve patients with rheumatoid arthritis or diabetes who had sustained an AFF during long-term BP treatment. No association with altered microarchitecture features in the peripheral skeleton was observed suggesting other individual predisposing or contributing factors (21). In our study, female patients were older. Age and gender were also observed in a study on subtrochanteric fragility

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

Atypical Femoral Fractures and Concomitant Diseases

Fig. 3. Kaplan-Meier plot of survival in patients with atypical and typical subtrochanteric and femoral shaft fractures ( p ! 0.0001). The age- and sex-adjusted hazard ratio for patients with atypical femoral fractures was 3.74 (95% confidence interval [CI], 2.46e5.56). fractures that included 20,000 patients, which suggested that male patients were more likely to be younger (22). The medical history of our patients revealed a higher proportion of AFF patients with chronic heart failure, hyperlipidemia, hypothyroidism, type 2 diabetes, and advanced stages of CKD. Inadequate aldosterone and parathyroid hormone secretion is strongly linked to the development and progression of cardiovascular disease, which was observed in 54% of our AFF patients compared to 25% in the other group (23). In treatment-na€ıve patients with stable cardiovascular disease, osteoprotegerin negatively correlates with aBMD, which may be an endogenous rescue mechanism for bone strength (24). A pathophysiological association between chronic heart failure and catabolic bone remodeling with severely altered osteoprotegerin or RANKL levels was recently published (25). In this study, significantly more patients with AFF used the long-acting oral anticoagulant phenprocoumon concomitantly. This medication is known to reduce bone mass and bone formation (26,27). By this, phenprocoumon might aggravate bone loss and increase the risk for AFF in patients with osteoporosis. The proportion of persons with advanced stages of CKD and BP pretreatment was significantly higher in patients with AFF, suggesting a strong effect of CKD (28,29). Almost twice as many patients with AFF had an ongoing therapy with furosemide, reduced serum calcium levels, and increased parathyroid hormone levels with known effects on calcium metabolism (30,31). Significantly more patients with AFF suffered from type 2 diabetes. Insufficient glycemic control is associated with

7 cortical porosity and impaired stem cell mobilization (32,33). More than 4 times as many AFF patients had an ongoing antidiabetic therapy with sulfonylurea. Studies evaluating a potential association between these antidiabetics and fracture risk are sparse and yield inconsistent results (34,35). The proportion of patients with diagnosed hypothyroidism was higher in the AFF group, suggesting a low bone turnover (36). The number of patients with a history of former or current smoking was more than double in the AFF group. BMD is a measure of bone strength and might also indicate lifestyles that preserve bone mass or quality in the elderly people (37). Sarcopenia was considerably more frequent in the AFF group. This also negatively influences bone metabolism and bone quality and might contribute to the development of an AFF (38). We noted a significant difference in age between male and female patients. The median age of our male population (regardless of the type of fracture) was about 1 decade lower than that of the female population and is partly in line with previously published data (39). Male patients in the AFF group were less likely to be currently taking BST, but their history of any previous BST was similar compared to the female population. A possible explanation could be the higher awareness of osteoporosis in the female population and hence a better monitoring and treatment compared to males. Gender distribution was similar in both groups, but the absolute number of males was small in the AFF group. Neither age nor gender was a significant contributing factor for an AFF. But in contrast, hypothyroidism, advanced stages of CKD, as well as furosemide, sulfonylurea, and phenprocoumon intake, but not any ongoing or history of BST, were associated with AFF. Patients with AFFs had a significantly shorter survival. Besides the severity of the hip fracture, a history of a previous fracture seems to influence the survival of elderly patients (40). Forty-six percent of patients with typical fractures had no previous fracture, whereas this was only observed in 9% of the AFF population. Other major factors for early death in both populations were terminal cardiac insufficiency, pneumonia, and urosepsis, presumably associated with the severe concomitant diseases or comedications of all patients, and to a greater extent in the AFF population. This study has limitations, which might potentially mask or conceal some observed effects. It was not possible to retrieve the duration of concomitant diseases and concomitant medications other than BST. Adherence of patients regarding their medication could not be ensured. Owing to the less number of AFF patients and the lack of a control group without fractures, we could not calculate a risk model with corrections for multiple comparisons to evaluate the effect of each possible risk factor. Altogether, our study confirmed that AFFs are rare events in Caucasian female and male patients with osteoporotic hip fractures.

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

8

Muschitz et al. Table 4 Current or Previous Bone-Specific Therapies in Patients With Atypical and Typical Subtrochanteric or Femoral Shaft Fractures Fracture type Atypical (n 5 35)

Therapy Current bone-specific therapy, n (%) No specific medication Alendronate Risedronate Ibandronate (intravenous) Zoledronic acid Denosumab Raloxifene Strontium ranelate Teriparatide Duration of ongoing therapy (mo), median (interquartile range) History of bone-specific therapy, n (%) No specific medication Alendronate Risedronate Ibandronate (intravenous) Zoledronic acid Denosumab Raloxifene Strontium ranelate Teriparatide Duration of previous therapy (mo), median (interquartile range) Months discontinued before fracture (mo), median (interquartile range)

Typical (n 5 159)

p Value 0.192a

26 (74.3) 3 (8.6) 0 3 (8.6) 2 (5.7) 0 0 1 (2.9) 0 0 (0e9) 10 13 2 1

3 6 36 15

(28.6) (37.1) (5.7) (2.9) 0 0 (8.6) (17.1) 0 (0e64) (0e32)

133 12 2 5 2 3 1 1

(83.7) (7.6) (1.3) (3.1) (1.3) (1.9) (0.6) (0.63) 0 0 (0e0)

81 29 10 17 6 2 6 7 1 0 0

0.210 0.016a

(50.9) (18.2) (6.3) (10.7) (3.8) (1.3) (3.8) (4.4) (0.63) (0e48) (0e16)

0.012 0.007

Significant p-values are bold. a Global p value comparing ‘‘no medication’’ vs ‘‘any medication.’’

We conclude that a combination of severe comorbidities, long-term pharmaceutical therapies, prior fragility fractures, and a history of previous or ongoing long-term BST could increase the individual risk for AFF. Table 5 Adjusted Odds Ratios (After Bias Correction) From the Multivariate Logistic Regression Model for Patients With Atypical Femoral Fractures Parameter

Odds ratio

95% CI

p Value

Age Gender Hypothyroidism Chronic kidney disease III/IV Furosemide Sulfonylurea Phenprocoumon Tricyclic antidepressants

0.99 1.86 2.82 4.22 5.92 3.19 5.59 4.96

0.94e1.04 0.54e6.36 1.02e7.81 1.67e10.67 2.29e15.28 0.93e11.00 1.86e16.85 1.42e17.37

0.613 0.324 0.046 0.002 0.001 0.066 0.002 0.012

Abbr: CI, confidence interval.

Acknowledgments All authors made substantial contributions to either the conception, design, acquisition of data, and analysis and interpretation of data. All authors approved the final version of the submitted article.

References 1. Johnell O, Kanis JA, Oden A, et al. 2004 Mortality after osteoporotic fractures. Osteoporos Int 15:38e42. 2. Michelson JD, Myers A, Jinnah R, et al. 1995 Epidemiology of hip fractures among the elderly. Risk factors for fracture type. Clin Orthop Relat Res (311):129e135. 3. Shane E, Burr D, Abrahamsen B, et al. 2014 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 29:1e23. 4. Morin S, Godbout B, Wall M, et al. 2013 Lower limb geometrical parameters in the pathogenesis of bisphosphonateassociated atypical femur fractures. J Bone Miner Res 28(1 Suppl). Abstract 1095. Available at: http://www.asbmr.org/ education/2013-abstracts. Accessed: July 15, 2014. 5. Martelli S, Pivonka P, Kersh M, et al. 2013 Atypical femoral fractures are associated with high cyclic tensile strain regions during walking. J Bone Miner Res 28(1 Suppl). Abstract MO

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015

Atypical Femoral Fractures and Concomitant Diseases

6. 7. 8. 9. 10.

11. 12. 13.

14.

15. 16.

17. 18.

19.

20. 21.

22.

0074. Available at: http://www.asbmr.org/education/2013-abstra cts. Accessed: July 15, 2014. Odvina CV, Levy S, Rao S, et al. 2010 Unusual mid-shaft fractures during long-term bisphosphonate therapy. Clin Endocrinol (Oxf) 72:161e168. Lenart BA, Neviaser AS, Lyman S, et al. 2009 Association of low-energy femoral fractures with prolonged bisphosphonate use: a case control study. Osteoporos Int 20:1353e1362. Meier RP, Perneger TV, Stern R, et al. 2012 Increasing occurrence of atypical femoral fractures associated with bisphosphonate use. Arch Intern Med 172:930e936. Park-Wyllie LY, Mamdani MM, Juurlink DN, et al. 2011 Bisphosphonate use and the risk of subtrochanteric or femoral shaft fractures in older women. JAMA 305:783e789. FDA - US Food and Drug Administration - posted: 10/13/2010. Available at: http://www.fda.gov/Safety/Medwatch/SafetyInfor mation/SafetyAlertsforHumanMedicalProducts/ucm229244.htm. Accessed June 15, 2015. European Medicines Agency (EMA) - posted: 04/14/2011. Available at: http://www.ema.europa.eu/docs/en_GB/document_library/Pre ss_release/2011/04/WC500105281.pdf. Accessed June 15, 2015. Siris ES, Pasquale MK, Wang Y, et al. 2011 Estimating bisphosphonate use and fracture reduction among US women aged 45 years and older, 2001-2008. J Bone Miner Res 26:3e11. Wells G, Cranney A, Peterson J, et al. 2008 Risedronate for the primary and secondary prevention of osteoporotic fractures in postmenopausal women. Cochrane Database Syst Rev (1): CD004523. Gedmintas L, Solomon DH, Kim SC. 2013 Bisphosphonates and risk of subtrochanteric, femoral shaft, and atypical femur fracture: a systematic review and meta-analysis. J Bone Miner Res 28:1729e1737. Girgis CM, Sher D, Seibel MJ. 2010 Atypical femoral fractures and bisphosphonate use. N Engl J Med 362:1848e1849. G€uerri-Fernandez RC, Nogues X, Quesada Gomez JM, et al. 2013 Microindentation for in vivo measurement of bone tissue material properties in atypical femoral fracture patients and controls. J Bone Miner Res 28:162e168. Schilcher J, Micha€elsson K, Aspenberg P. 2011 Bisphosphonate use and atypical fractures of the femoral shaft. N Engl J Med 364:1728e1737. Wang J, Ward M, Bhattacharyya T, et al. 2013 Adherence to oral bisphosphonates and the risk of subtrochanteric or diaphyseal femur fractures among US Medicare beneficiaries with part D coverage. J Bone Miner Res 28(1 Suppl). Abstract LB-MO23. Available at: http://www.asbmr.org/education/2013-abstracts. Accessed: July 15, 2014. Adams A, Xue F, Wang J, et al. 2013 Atypical femoral fracture risk factors: a population-based case-control study. J Bone Miner Res 28(1 Suppl). Abstract FR 0342. Available at: http://www. asbmr.org/education/2013-abstracts. Accessed: July 15, 2014. Odvina CV, Zerwekh JE, Rao DS, et al. 2005 Severely suppressed bone turnover: a potential complication of alendronate therapy. J Clin Endocrinol Metab 90:1294e1301. Zanchetta MB, Diehl M, Buttazzoni M, et al. 2014 Assessment of bone microarchitecture in postmenopausal women on longterm bisphosphonate therapy with atypical fractures of the femur. J Bone Miner Res 29:999e1004. Wang Z, Bhattacharyya T. 2011 Trends in incidence of subtrochanteric fragility fractures and bisphosphonate use among the US elderly, 1996-2007. J Bone Miner Res 26:553e560.

9 23. Tomaschitz A, Ritz E, Pieske B, et al. 2014 Aldosterone and parathyroid hormone interactions as mediators of metabolic and cardiovascular disease. Metabolism 63:20e31. 24. Chen HY, Lin CC, Chiu YL, et al. 2012 Serum fetuin A and chemerin levels correlate with hepatic steatosis and regional adiposity in maintenance hemodialysis patients. PLoS One 7: e38415. 25. Leistner DM, Seeger FH, Fischer A, et al. 2012 Elevated levels of the mediator of catabolic bone remodeling RANKL in the bone marrow environment link chronic heart failure with osteoporosis. Circ Heart Fail 5:769e777. 26. Pietschmann P, Woloszczuk W, Panzer S, et al. 1988 Decreased serum osteocalcin levels in phenprocoumon-treated patients. J Clin Endocrinol Metab 66:1071e1074. 27. Van Haarlem LJ, Knapen MH, Hamulyak K, et al. 1988 Circulating osteocalcin during oral anticoagulant therapy. Thromb Haemost 60:79e82. 28. Salam SN, Eastell R, Khwaja A. 2014 Fragility fractures and osteoporosis in CKD: pathophysiology and diagnostic methods. Am J Kidney Dis 63:1049e1059. 29. Liu WC, Yen JF, Lang CL, et al. 2013 Bisphophonates in CKD patients with low bone mineral density. ScientificWorldJournal 2013:837573. 30. Law PH, Sun Y, Bhattacharya SK, et al. 2005 Diuretics and bone loss in rats with aldosteronism. J Am Coll Cardiol 46:142e146. 31. Drinka PJ, Krause PF, Nest LJ, et al. 2007 Determinants of parathyroid hormone levels in nursing home residents. J Am Med Dir Assoc 8:328e331. 32. Fadini GP, Ferraro F, Quaini F, et al. 2014 Concise review: diabetes, the bone marrow niche, and impaired vascular regeneration. Stem Cells Transl Med 3:949e957. 33. Puar TH, Khoo JJ, Cho LW, et al. 2012 Association between glycemic control and hip fracture. J Am Geriatr Soc 60: 1493e1497. 34. Lapane KL, Yang S, Brown MJ, et al. 2013 Sulfonylureas and risk of falls and fractures: a systematic review. Drugs Aging 30:527e547. 35. Fronczek-Sok o1 J, Pytlik M. 2014 Effect of glimepiride on the skeletal system of ovariectomized and non-ovariectomized rats. Pharmacol Rep 66:412e417. 36. Turner MR, Camacho X, Fischer HD, et al. 2011 Levothyroxine dose and risk of fractures in older adults: nested case-control study. BMJ 342:d2238. 37. Emaus N, Wilsgaard T, Ahmed LA. 2014 Impacts of body mass index, physical activity, and smoking on femoral bone loss: the Tromsø study. J Bone Miner Res 29:2080e2089. 38. Rizzoli R, Stevenson JC, Bauer JM, et al. 2014 The role of dietary protein and vitamin D in maintaining musculoskeletal health in postmenopausal women: a consensus statement from the European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis (ESCEO). Maturitas 79: 122e132. 39. Markman LH, Allison MB, Rosenberg ZS, et al. 2013 A retrospective review of patients with atypical femoral fractures while on long-term bisphosphonates: including pertinent biochemical and imaging studies. Endocr Pract 19: 456e461. 40. Ha YC, Baek JH, Ko YB, et al. 2014 High mortality and poor morbidity after hip fracture in patients with previous vertebral fractures. J Bone Miner Metab. [Epub ahead of print]. PMID: 25227286.

Journal of Clinical Densitometry: Assessment & Management of Musculoskeletal Health

Volume

-,

2015