Osteoporos Int (2013) 24:2671–2680 DOI 10.1007/s00198-013-2374-x
ORIGINAL ARTICLE
Micro-morphological properties of osteons reveal changes in cortical bone stability during aging, osteoporosis, and bisphosphonate treatment in women A. Bernhard & P. Milovanovic & E. A. Zimmermann & M. Hahn & D. Djonic & M. Krause & S. Breer & K. Püschel & M. Djuric & M. Amling & B. Busse
Received: 4 January 2013 / Accepted: 9 April 2013 / Published online: 30 April 2013 # International Osteoporosis Foundation and National Osteoporosis Foundation 2013
Abstract Summary We analyzed morphological characteristics of osteons along with the geometrical indices of individual osteonal mechanical stability in young, healthy aged, untreated osteoporotic, and bisphosphonate-treated osteoporotic women. Our study revealed significant intergroup differences in osteonal morphology and osteocyte lacunae indicating different remodeling patterns with implications for fracture susceptibility. Introduction Bone remodeling is the key process in bone structural reorganization, and its alterations lead to changes in bone mechanical strength. Since osteons reflect different bone remodeling patterns, we hypothesize that the femoral
A. Bernhard and P. Milovanovic contributed equally and therefore share first authorship. A. Bernhard : P. Milovanovic : E. A. Zimmermann : M. Hahn : M. Krause : S. Breer : M. Amling : B. Busse (*) Department of Osteology and Biomechanics, University Medical Center Hamburg-Eppendorf, Lottestr. 59, 22529 Hamburg, Germany e-mail:
[email protected] P. Milovanovic : D. Djonic : M. Djuric Laboratory for Anthropology, Institute of Anatomy, School of Medicine, University of Belgrade, Dr. Subotica 4/2, 11 000 Belgrade, Serbia E. A. Zimmermann Lawrence Berkeley National Laboratory, University of California, Berkeley, One Cyclotron Road, Berkeley 94720 CA, United States K. Püschel Department of Forensic Medicine, University Medical Center Hamburg-Eppendorf, Martinistr. 52, 20246 Hamburg, Germany
cortices of females under miscellaneous age, disease and treatment conditions will display distinct osteonal morphology and osteocyte lacunar numbers along with different mechanical properties. Methods The specimens used in this study were collected at autopsy from 35 female donors (young group, n=6, age 32±8 years; aged group, n=10, age 79±9 years; osteoporosis group, n = 10, age 81 ± 9 years; and bisphosphonate group, n=9, age 81±7 years). Von Kossa-modified stained femoral proximal diaphyseal sections were evaluated for osteonal morphometric parameters and osteocyte lacunar data. Geometrical indices of osteonal cross-sections were calculated to assess the mechanical stability of individual osteons, in terms of their resistance to compression, bending, and buckling. Results The morphological assessment of osteons and quantification of their osteocyte lacunae revealed significant differences between the young, aged, osteoporosis and bisphosphonate-treated groups. Calculated osteonal geometric indices provided estimates of the individual osteons’ resistance to compression, bending and buckling based on their size. In particular, the osteons in the bisphosphonate-treated group presented improved osteonal geometry along with increased numbers of osteocyte lacunae that had been formerly impaired due to aging and osteoporosis. Conclusions The data derived from osteons (as the basic structural units of the cortical bone) in different skeletal conditions can be employed to highlight structural factors contributing to the fracture susceptibility of various groups of individuals. Keywords Cortical bone . Femur . Individual osteon’s stability . Osteocyte lacunae . Osteonal morphometry
2672
Introduction Aging and osteoporosis are commonly associated with increased bone fragility [1, 2]. Due to demographic changes and an aging population, the bone fracture risk and incidence of fractures are even estimated to further increase in the next years [1, 3]. Whereas in the year 1990, 1.6 million hip fractures were reported worldwide, current reports predict 6.3 million hip fractures worldwide in the year 2050 [1, 2]. In 2005, nonvertebral fractures accounted for 73 % of total fractures in the elderly population implying that skeletal regions containing significant contributions from cortical bone suffer from age- and osteoporosis-related fragility [2, 3]. Beyond the trabecular architecture and the cortical thickness that can be associated with the mechanical properties of the bone [4, 5], the patterns of the cortical microstructure in the femur and their alteration with age represent a significant factor in regard to fracture risk [6]. The femoral cortex is composed of numerous osteons, which are also known as the bone structural units [7]. The number of osteons and their morphological characteristics have been reported to represent an important determinant of cortical bone strength [8, 24]. However, although osteonal morphology has been investigated in various species, age groups, skeletal sites, as well as for forensic and archaeological purposes [8–13], the data on femoral osteons obtained from osteoporotic cases [6, 14] as well as subsequent pharmacologic osteoporosis treatment are rare and unestablished. Basically, both the whole femur and individual osteons can be regarded as hollow tubes, which allows a first-order estimation of the individual osteon’s mechanical resistance based on the geometrical cross-sectional measurements revealing their resistance to compression (cross-sectional area), bending (section modulus), and buckling (buckling ratio (BR)). Whereas such mechanical properties have already been determined at the whole bone level [15] using hip structure analysis software (HSA) developed by Beck and colleagues [16], analogous mechanical properties have not yet been determined for the individual osteons. As bone remodeling is the key process in bone structural reorganization and its alterations in various conditions may lead to changes in bone mechanical strength [17], we specifically aim to assess changes in the individual osteons morphological parameters to deduce in turn the effect on the mechanical properties in individual osteons. Because the osteonal structure reflects the bone remodeling pattern [6, 8, 10, 18–21], we hypothesize that the femoral cortices of women with various bone conditions will display distinct osteonal morphology and osteocyte lacunar numbers along with different mechanical properties. Therefore, in this study, we report osteocyte lacunar indices and osteonal structural indices in relation to calculated estimates of osteonal stability
Osteoporos Int (2013) 24:2671–2680
in young, aged, untreated osteoporotic, and bisphosphonatetreated osteoporotic women.
Materials and methods Characteristics of cases All bone samples were obtained from 35 female donors. The specimens were taken from the proximal femoral diaphyses during autopsy at the Department of Forensic Medicine, University Medical Center Hamburg-Eppendorf. The circumstances leading to death were traffic accidents, assaults, suicides, and other unnatural or unexpected causes. Based on patients’ records and autopsy data, individuals who suffered from cancer, renal diseases, primary hyperparathyroidism, or Paget’s disease or showed any other signs or symptoms of bone diseases apart from postmenopausal osteoporosis in the appropriate groups were excluded from the study. Based on autopsy reports, patients’ histories and osteoporosis diagnostic criteria provided by the World Health Organization (i.e., bone mineral density at least 2.5 standard deviations below the young mean as measured by dual-energy X-ray absorptiometry and/or the presence of a fragility fracture), the specimens were divided into four study groups: Group 1.
Young cases (young group, n=6) with a mean age of 32±8 years Group 2. Aged cases (aged group, n=10) with a mean age of 79±9 years Group 3. Cases with treatment-naïve osteoporosis (OP group, n=10) with a mean age of 81±9 years Group 4. Cases subject to osteoporosis treatment with bisphosphonates (BP group, n=9) with a mean age of 81±7 years The latter group received the third-generation bisphosphonate alendronate at either 10 mg/day or 70 mg/week. These cases were treated with bisphosphonates for 6±1.6 years. Specimen preparation Horizontal cross-sections with a thickness of 4 mm were cut from the proximal femoral diaphysis (Fig. 1) with a diamond belt saw (Exakt, Norderstedt, Germany) and fixed in formalin (3.5 %), as reported previously [22]. The undecalcified specimens were dehydrated by means of an ascending ethanol series (70, 80, 2×96, and 3×100 %) and further infiltrated with plastic embedding medium (Technovit 7200; Heraeus/Kulzer, Wehrheim/Ts., Germany). The infiltration was performed in steps with the following volumetric ethanol/Technovit ratios: 70:30, 50:50, 30:70, and 0:100. Finally, Technovit with
Osteoporos Int (2013) 24:2671–2680
2673
lacunar number per osteon (N.Ot.Lc, in No.). Using those measured values, the following parameters [6, 8, 20] were calculated: osteon area (On.Ar, in square micrometers), Haversian canal area (HC.Ar, in square micrometers), osteon bone area (On.B.Ar, in square micrometers), mean osteonal wall thickness (On.W.Th, in micrometers), osteocyte lacunar density (N.Ot.Lc/On.B.Ar, in No. per square micrometer), and percentage of osteon refilling (On.B.Ar/On.Ar, in percent). Assessment of the osteons’ biomechanical properties Fig. 1 Preparation of femoral specimen resected from the proximal femoral diaphysis and histomorphometric evaluation of osteonal characteristics
benzoyl-peroxide was used to continue the infiltration process for another 10 days. Ground tissue specimens were made based on Donath’s grinding technique [22, 23]. After the grinding and polishing process, the specimens were stained with von Kossa-modified stain in accordance with previous studies [6, 24]. Specimen analysis The von Kossa-modified stained histological specimens were analyzed to determine osteonal structure and cell indices in all study groups. To assess whether the osteons’ characteristics depend also on location, the femoral crosssections were subdivided in four regions: dorsal, lateral, medial, and ventral [6, 24]. A total of 60 osteons (fifteen per region) were selected randomly in the mid-cortex of each patient and analyzed under microscopic magnification (×100) (Fig. 1). Assuming simplified circular shapes for osteons and Haversian canals, corresponding areas were calculated from the diameters (Fig. 1). The diameters were marked manually and automatically measured using image analysis software (ImageJ, 1.45q, National Institutes of Health, USA). Only osteons were counted that were located to 100 % in the measuring squares and fulfilled the criteria proposed in previous studies [6, 14]. Namely, excluding criteria for the measurement of the osteons were: (a) unidentifiable osteon boundaries (cement lines), (b) incompletely visible osteons within the measuring field, (c) non-circular osteonal shape, and (d) Haversian canal merged with other canals (i.e., Volkmann’s canal and/or other Haversian canals), as well as merged resorbing osteons that formed the so-called superosteons [25, 26]. In accordance with the nomenclature committee of the American Society of Bone and Mineral Research [27], the following parameters were directly measured: osteon diameter (On.Dm, in micrometers), Haversian canal diameter (HC.Dm, in micrometers), and osteocyte
Previous studies named osteon morphological characteristics as important determinants of bone strength albeit without numerical values (e.g., 6, 20); in our study, to estimate the role of morphology of individual osteons on the bone’s mechanical stability, we have approximated osteons as hollow tubes to extract certain geometric mechanical properties from the osteonal cross-sections, such as: section modulus (Z), cross-sectional area (CSA) and buckling ratio (BR). In contrast to the macroscopic level where HSA was applied to estimate those parameters as indicators of bone resistance to bending, compression, and buckling [15], analogous mechanical properties have not yet been determined for individual osteons. Section modulus is an indicator of bending strength, and in this study the section modulus of an osteon was determined according to the following formula Z=π*(On.Dm4 −HC.Dm4)/(32*On.Dm). Cross-sectional bone area is considered as an indicator of compressive strength [15] and in the case of an osteon, this parameter would correspond to the osteon bone area. The BR of a tube cross-section is a rough measure of its buckling risk and can be determined as the ratio between the tube outer radius and wall thickness [15]. The BR of an osteon was calculated according to the equation: BR=On.Dm/(2*W.Th). Statistical analysis One sample Kolmogorov–Smirnov test was applied for assessing the normality of the data distribution. Analysis of variance (ANOVA) was applied to assess the influence of the study group (defined as between-subject factor), the region (within-subject factor), and the interaction between the group and region. After assessing overall effects of a factor by means of ANOVA, post hoc multiple comparison procedures with Bonferonni correction were performed to determine individual differences between the groups. Linear regression analysis was performed to evaluate the relationship between osteocyte lacunar density and osteon wall thickness. The Statistical Package for Social Sciences version 15 (SPSS Inc., Chicago, IL) was used for statistical analysis at the 0.05 level of significance.
89.09f (3.05) 74.94c, f (4.43)
No significant differences between the particular groups g
Significant differences between aged and BP group
Significant differences between aged and OP group
Significant differences between OP and BP group f
e
d
Significant differences between young and OP group
Significant differences between young and aged group
Significant differences between young and BP group c
b
Post hoc comparisons between each two groups:
a
Data are presented as mean (SD). OP untreated osteoporotic group, BP bisphosphonates-treated group
380.33e (54.38) 34,650.33f (5,364.77) 12.76e, f (2.63)
88.55a, b, c (11.09) 66.63d (7.71) 53.05d, f (6.28) 374.75 (68.05) 307.83e (43.27) 367.20 (45.95)
N.Ot.Lc/On.B.Ar On.W.Th (μm) (No./μm2) Ot.Lc.N (No.) On.B.Ar (μm2) HC.Dm (μm) HC.Ar (μm2)
39,086.22f (7,169.63) 65.89 (14.33) 4,435.67 (1,981.76) 215.78e, f (20.38) BP
ANOVA showed slightly significant intergroup differences in Haversian canal diameter (p=0.045) and Haversian canal area (p=0.048) in the mid-cortex. However, significant differences were not detectable in pairwise post hoc comparisons between the groups showing that all groups together contribute to the observed significance of the model, while they do not express statistically significant individual differences under strict Bonferroni correction (p>0.05) (Table 1).
Young 233.14a, b (23.25) 45,405.46a, b (9,143.73) 56.03 (10.17) 3,021.72 (1,109.51) 42,383.9a, b (8,796.44) 15.61a, b (4.55) Aged 193.35e (10.98) 31,699.65 (3,305.29) 60.08 (10.88) 3,613.61 (1,382.93) 28,085.94 (3,427.84) 8.36e (1.65) f f f OP 179.7 (15.35) 24,741.04 (4,591.09) 73.60 (14.18) 5,423.5 (2,112.58) 22,601.9 (3,795.79) 7.63f (1.44)
Group-specific changes in Haversian canal dimensions
On.Ar (μm2)
Our study revealed that osteon diameter, osteon area, and osteonal bone area significantly differed between the groups (p
