It is well known that as the proportion of elderly individuals within the population has increased, so has the number of hip fractures, and that the incidence of hip fracture increases exponentially with age for both males and females (Hedlund et al., 1987; Holmberg and Thorngren, 1987; Kannus et al., 1996). Consequently, hip fractures can be a major cause of morbidity and mortality in elderly humans (Block and Stubbs, 1997) and are the focus of a great deal of research. Research on the etiology of hip fractures has mainly focused on the role of increased cancellous bone loss in the proximal femur with age. In addition to the effect of cancellous bone loss, the role of cortical bone in hip fractures has been recently examined. Both experimental evidence (Werner et al., 1988) and finite element analysis (Lotz et al., 1995) have suggested that the cortex of the femoral neck carries a substantial portion of the load, and thus age-related changes in the morphology and microstructure of the cortex may be a significant contributing factor in hip fractures.
In recent years, researchers have begun to examine age-related changes in the metaphyseal cortex of the proximal human femur. Investigations by Boyce and Bloebaum (1993) and Vajda and Bloebaum (1999) identified a hypermineralized tissue on periosteal surface of the neck and trochanteric region of the femur that increases in fractional area with advancing age in both males and females. This hypermineralized tissue has been identified as calcified fibrocartilage from capsular and tendon insertions on the femoral neck and trochanters by Shea et al. (2001). Calcified fibrocartilage appears to be the adhesive tissue in tendon, ligament, and capsular insertion sites on bone (Cooper and Misol, 1970; Benjamin et al., 1986; Benjamin and Ralphs, 1998). There are four distinct tissue zones at insertion sites: parallel-fibered collagen, fibrocartilage, calcified fibrocartilage, and bone (Fig. 1). Each zone has been shown histologically to be separated by a tidemark, with the calcified fibrocartilage/bone junction being interdigitated to secure their attachment (Benjamin and Ralphs, 1998).
Calcified fibrocartilage may potentially play a significant role in hip fracture due to both the biological and the mechanical properties of the tissue. An increase in the fractional area of calcified fibrocartilage has been shown to be associated with age-related thinning of cortical bone in the femoral neck. Ultimately, calcified fibrocartilage can comprise up to 60% of the fractional area of the femoral neck cortical shell with advanced age in both men and women (Vajda and Bloebaum, 1999). Additionally, an increase in the fractional area of calcified fibrocartilage has been associated with an increase in the brittleness and decrease in ultimate strength of the cortical bone in the femoral neck (Shea, 2002). Furthermore, calcified firbrocartilage has been shown to be an avascular tissue (Shea et al., 2002b); thus, as the fractional area of calcified fibrocartilage increases with age, larger proportions of the femoral neck in the elderly are composed of this brittle avascular tissue rather than vascularized cortical bone.
The importance of vascularity in the maintenance and repair of tissues is well known. In bone, vascular proliferation occurs at the site of remodeling osteons, indicating the importance of blood supply in the remodeling processes of cortical bone (Marotti, 1996). Additionally, several researchers have noted the increased presence of remodeling osteons after in vivo and in vitro fatigue loading, most likely to repair mechanically detrimental microcracks (Mori and Burr, 1993; Burr et al., 1998; Reilly and Currey, 2000). Since blood supply is important to the maintenance and remodeling capacity of cortical bone, the large areas of avascular calcified fibrocartilage present on the elderly femoral neck may predispose these regions to microcrack accumulation and ultimately fracture. This potential role in hip fracture could be minimized if the calcified fibrocartilage on the femoral neck has the capacity to remodel and heal similar to that of bone.
The mechanisms by which calcified fibrocartilage remodels or what capacity it has to remodel are currently unknown. Cooper and Misol (1970) noted that fibrocartilage cells surrounded by mineralized matrix survive and appear to control the matrix in a manner similar to osteocytes in bone matrix. They further noted that the structural differences in arrangement of the mineral in bone and fibrocartilage from tendon insertions might denote either “different mechanisms of mineralization or the same mechanism with subsequent mineral growth influenced by the amount and arrangement of collagen or mucopolysacharide or both” (Cooper and Misol, 1970: p. 18). Frost et al. (1960, 1961) indicated that both newly forming bone and calcifying cartilage are similarly labeled by tetracycline, while inactive surfaces are not. While the mechanism of calcified fibrocartilage maintenance and repair is undefined, the work of Cooper and Misol (1970) and Frost et al. (1960, 1961) suggests that the tissue may remodel and repair in a manner similar to bone.
The present investigation was designed to determine if calcified fibrocartilage remodels in a manner similar to that of cortical bone. It was hypothesized that the avascularity of calcified fibrocartilage would result in an absence of a remodeling capacity similar to cortical bone. The goal of this investigation was to quantify and compare the remodeling capacity of the underlying cortical bone compared to the calcified fibrocartilage of the proximal femur. Tetracycline labeling was used to test the hypothesis.
MATERIALS AND METHODS
An Institutional Animal Care and Use Committee-approved ovine model was chosen to assess the remodeling capacity of calcified fibrocartilage compared to underlying cortical bone because the model has been shown to be a suitable investigative correlate for human bone due to similar morphological features for both cortical bone and calcified fibrocartilage (Shea et al., 2002a).
Ten free-range adult female sheep, aged 3.5–4 years and with normal activity levels, were given two tetracycline injections on the 16th and 5th day prior to sacrifice, following an established protocol for the ovine model (Willie et al., 2004). After sacrifice, both the right and left femurs of each animal were dissected and excess soft tissue was removed. The right or left femur from each animal was randomly selected for dehydration and embedding. Each selected femur was then fixed in 70% ethyl alcohol for a minimum of 2 weeks, after which the remaining soft tissue was manually removed, taking care to avoid damaging the cortical surface. The whole femurs (n = 10) were then cut below the lesser trochanter with a band saw, and the proximal segment of each bone was dehydrated in ascending concentrations of ethyl alcohol, cleared in xylene, and embedded in methylmethacrylate following published protocols (Emmanual et al., 1987). Each embedded femur was sectioned in the greater trochanteric region perpendicular to the axis of the femoral neck at 5 mm increments to ensure the inclusion of regions with capsular, tendon, and/or ligament insertions, and thus underlying calcified fibrocartilage. The sections were cut with a high-speed water-cooled low-feed bone saw with a diamond-impregnated blade (Rockazona, Peoria, AZ). These sections were then manually ground to light on a rotary grinding wheel (Buehler, Lake Bluff, IL) and polished with 0.5 μm levigated alumina (LECO, ST. Joseph, MI) until scratch-free when viewed at 35× in a reflected light microscope. The final thickness of each thin section was approximately 50 μm. Each thin section (n = 30; 3 sections per femur) was then stained with a dilute (50%) basic fuchsin solution to differentiate the fibrous tissue, cortical bone, and calcified fibrocartilage. The following staining protocol was established and tested in pilot study to ensure proper staining of the tissues without obliteration of the tetracycline labels within the bone. The polished thin sections were soaked in a 5% HCL acid ethyl alcohol bath for not more than 5 min to allow for adequate staining of the tissues. The sections were then stained in the dilute basic fuchsin solution for 1–2 min until the tissues were appropriately stained.
Fluorescent light microscopy was used to quantify the number of double and single tetracycline labels in the cortical bone and calcified fibrocartilage for the total cross-sectional area of each thin section. Both double and single labels were quantified to assess remodeling activity of the tissues and to account for any possible differences in the mineral apposition rates between the tissues (Bloebaum et al., 1994).
The mean number and standard deviation of both double and single labels in the total cross-sectional area for each tissue were calculated and then statistically analyzed using a paired t-test to identify any significant differences between the number of labels in the calcified firbrocartilage and cortical bone.
A significantly greater number of both single and double labels, indicating active remodeling, were found in the cortical bone compared to the calcified fibrocartilage (P < 0.001; Fig. 2). The mean number of double labels in the cortical bone was 33 ± 21 per total cross-sectional area. The mean number of double labels in the calcified fibrocartilage was less than one per cross-sectional area, with a maximum of three labels in a single cross-section. Only 4 of the 30 sections displayed double labels in the calcified fibrocartilage region. The mean number of single labels in the cortical bone was 20 ± 15 per total cross-sectional area. The mean number of single labels in the calcified firbrocartilage was also less than one per cross-sectional area, with a maximum of two labels in a single cross-section. Only 6 of the 30 sections displayed single labels in the calcified fibrocartilage region. In the rare instance when a single or double label did appear within the region of calcified fibrocartilage, it was always associated with an isolated osteon interdigitated along the margin of the calcified fibrocartilage (Fig. 3). This indicated that the few tetracycline labels that were observed in the region of calcified fibrocartilage were regions where viable cortical bone was interdigitating with the calcified fibrocartilage, rather than active remodeling within calcified fibrocartilage. The statistically significant lack of tetracycline labeling in the calcified fibrocartilage in comparison to the cortical bone indicated that the tissue lacked the capacity to remodel through a mechanism similar to bone.
The results of this investigation demonstrated that while the cortical bone was actively remodeling, the calcified fibrocartilage appeared to have no remodeling capacity similar to that of cortical bone. This does not mean that the calcified fibrocartilage may not have some other currently unknown biological mechanism for remodeling and repairing microcracks that may accumulate. However, the data from this study, and the absence of blood supply (Shea et al., 2002b), suggest that the healing capacity of calcified fibrocartilage would be minimal. A limited remodeling capacity for calcified fibrocartilage has also been indicated in recent tendon healing studies (Leung et al., 2002; Weiler et al., 2002). In a rabbit model of bone to tendon healing, Leung et al. (2002: p. 600) noted that incorporation between the tendon and bone was poor with “no normal bone-tendon junctional zone with characteristic fibrocartilage zone” restored. Similarly, Weiler et al. (2002) found no mineralized fibrocartilage tissue adjacent to the fibrocartilge in the tendon-bone interface in a sheep model of tendon healing in a bone tunnel. The data from these studies and the data from this investigation suggest that calcified fibrocartilage has limited to no remodeling or healing capacity.
The potential lack of remodeling capacity within calcified fibrocartilage may have significant effects on the durability and structural integrity of the proximal femur with age. The cortical bone of the femoral neck and trochanteric regions has been shown to thin and become more porous with age (Boyce and Bloebaum, 1993; Vajda and Bloebaum, 1999), likely as a result of endosteal resorption and cortical bone remodeling. As the cortical bone thins, a larger proportion of the cortex shell along regions of the femoral neck and greater trochanter then consists of hypermineralized calcified fibrocartilage, which is known to be very friable (Shea et al., 2002b) and to have limited capacity to remodel and heal. As the volume fraction of this tissue increases with the decrease in viable cortical bone, the potential for calcified fibrocartilage to share more of the stresses and forces of the hip increases. Currey (1979, 1984) has indicated that highly minerialized bone tissue is extremely stiff, fractures in a brittle manner, and absorbs little energy before fracturing. The high mineral content of calcified fibrocartilage and the decrease in ultimate strength associated with the tissue, as shown by Shea et al. (2001) for human femoral neck samples, coupled with an increase in volume fraction with age (Vajda and Bloebaum, 1999), may make this region more susceptible to failure in the elderly.
Future studies should attempt to define what, if any, mechanism calcified fibrocartilage has to repair and remodel. Specifically, the capability of chondrocytes to remodel/repair the mineralized extracellular matrix needs to be examined. The regional distribution of calcified fibrocartilage and periosteum on the human femoral neck should also be examined to understand the healing capacity of the proximal femur. While there has been some controversy in the past as to the presence of a periosteum on the femoral neck (Sherman and Phemister, 1947; Frost, 1964, 1995; Bagi et al., 1997), Shea et al. (2002b) identified a well-vascularized periosteum on regions of the human femoral neck where there was not calcified fibrocartilage. The regional distribution and relative amounts of each tissue on the proximal femur are unknown, but almost certainly have an effect on the mechanics and healing capacity of the region and subsequently hip fracture. Additional future studies involving a fracture model to examine if and how calcified fibrocartilage will react and repair with fracture, and how it may affect the healing capacity of the underling bone, would also be beneficial to understanding the role of calcified fibrocartilage in the proximal femur hip fractures in the elderly population.
The authors thank Shannon Ricks and Rebekah Steren for laboratory assistance. The material in this study is based on work supported by the Office of Research and Development Medical Research Service, Department of Veterans Affairs Salt Lake City Health Care System, Department of Orthopedics, University of Utah School of Medicine, and the Albert and Margaert Hofmann Endowed Chair in Orthopedic Research.