Long-term reaction to bone cement in osteoporotic bone: new bone formation in vertebral bodies after vertebroplasty

Abstract

Elderly patients frequently suffer from osteoporotic vertebral fractures resulting in the need of vertebroplasty or kyphoplasty. Nevertheless, no data are available about the long-term consequences of cement injection into osteoporotic bone. Therefore, the aim of the present study was to evaluate the long-term tissue reaction on bone cement injected to osteoporotic bone during vertebroplasty. The thoracic spine of an 80-year-old female was explanted 3.5 years after vertebroplasty with polymethylmethacrylate. The treatment had been performed due to painful osteoporotic compression fractures. Individual vertebral bodies were cut in axial or sagittal sections after embedding. The sections were analysed using contact radiography and staining with toluidine blue. Furthermore, selected samples were evaluated with scanning electron microscopy and micro-compted tomography (in-plane resolution 6 µm). Large amounts of newly formed callus surrounding the injected polymethylmethacrylate were detected with all imaging techniques. The callus formation almost completely filled the spaces between the vertebral endplate, the cancellous bone, and the injected polymethylmethacrylate. In trabecular bone microfractures and osteoclast lacuna were bridged or filled with newly formed bone. Nevertheless, the majority of the callus formation was found in the immediate vicinity of the polymethylmethacrylate without any obvious relationship to trabecular fractures. The results indicate for the first time that, contrary to established knowledge, even in osteoporosis the formation of large amounts of new bone is possible.

Introduction

Osteoporosis is a highly focused issue in current scientific research and clinical treatment. The clinical and social impact of osteoporosis will increase due to the age distribution of the population in industrial countries, and along with the medical aspects its economic relevance is expected to become of major importance. In the United States, 2 million osteoporotic fractures at a cost of $17 billion were observed in 2005; the number of annual fractures and associated costs are predicted to rise by almost 50% by 2025 (Burge et al. 2007).

The manifestation of osteoporosis is frequently observed in the human spine (Goh et al. 2000) and almost all osteoporotic fractures involve the vertebral bodies. In patients with vertebral compression fractures, percutaneous vertebroplasty and kyphoplasty are frequently applied treatments, resulting in almost instant pain relief and functional improvement in the majority of cases (Alvarez et al. 2006). However, there is persisting controversy regarding the functional consequences of procedures that introduce polymethylmethacrylate (PMMA) to expand and stabilize fractured and compressed vertebral bodies (Rosen, 2005). Fractures in adjacent vertebrae due to increased mechanical stress are frequent complications following vertebroplasty or kyphoplasty, often occurring within the first 2 months after surgical treatment (Voormolen et al. 2006). It is thus surprising only a few studies have reported on the existence of refractures of the PMMA-treated vertebra itself following vertebroplasty or kyphoplasty (Gaughen et al. 2002). When partially filling a fractured vertebral body with PMMA in an osteoporotic patient, it seems reasonable to assume that the different stiffness of the cement and the surrounding osteoporotic bone will increase the risk of refracture of the remaining cancellous bone (Wagner & Baskurt, 2006).

The hallmark of osteoporosis is a progressive loss of bone mass, resulting in an increased susceptibility to fracture (Weycker et al. 2007). Significant new bone formation during the integration process of an exogenous material has not been reported thus far. Hence, such formation would help to explain the almost complete absence of refractures in the vertebral body after vertebroplasty or kyphoplasty.

Methods

Background

The thoracic spine of an 80-year-old female was explanted during a post-mortem investigation. The explantation of four thoracic vertebral bodies (Th7–Th10) was done 3.5 years after vertebroplasty with PMMA. The treatment had been performed due to painful osteoporotic compression fractures and had resulted in complete pain relief. After vertebroplasty the patient was treated with oral medication of bisphosphonates.

Methodology

After explantation, the vertebrae were separated and macerated. Three vertebrae were then embedded in London Resin (LR) White and cut in horizontal or sagittal directions. The sections were analysed using contact radiography and staining with toluidine blue. The forth vertebral body was cut into pieces with a maximum dimension of 60 × 7 × 7 mm to allow analysis of the interface between PMMA and the surrounding bone using scanning electron microscope (SEM) and micro-computed tomography (µCT; in-plane resolution 6 µm).

Results

Large quantities of diffuse bone formation were present in the immediate neighbourhood of the PMMA in horizontal sections (Fig. 1A). In sagittal sections contact radiography and histological observation revealed large amounts of bone within the spaces between the superior vertebral endplate and the PMMA (Figs 1B, 2A). The frequency and quantity of bone decreased in the dorsal and inferior parts of the vertebrae (Fig. 1B). Under contact radiography, the higher resolutions allowed us to identify the woven structure of the newly formed bone which filled the spaces between the trabeculae of the cancellous bone (Fig. 1C).

Fig. 1.

Contact radiography (A–C), µCT (D) and SEM (E) images. (A) Horizontal section through the 10th thoracic vertebra. PMMA cement surrounded by large amounts of woven bone (Green, PMMA cement; , woven bone). (B) Sagittal section through the 10th thoracic vertebra. Large amounts of woven bone fill the space between the vertebral endplate and the PMMA cement. The frequency and quantity of the woven bone formation decreases in the dorsal and peripheral regions. The black gap represents the part of the vertebra used for horizontal sections (Green, PMMA cement; , woven bone; →, cement injection canal). (C) Detailed view of woven bone, which fills the spaces between the trabeculae of the cancellous bone (, woven bone; *, trabecular bone). (D) Three-dimensional visualization of woven bone surrounding the PMMA cement (Green, PMMA cement; *, trabecular bone). (E) Dense network of woven bone in close vicinity to the PMMA cement, resulting in a covering of the cement and a thickening of the individual trabeculi (Green, PMMA cement; *, trabecular bone).

Fig. 2.

Toluidine blue-stained sections. (A) Woven bone fills almost all the spaces between the vertebral endplate, trabecular bone and PMMA cement (, vertebral endplate; , PMMA cement; , woven bone; *, trabecular bone). (B) Woven bone shows different staining behaviour in different regions (arrow). This indicates that it was formed over a longer period of time. No signs of bone resorption (i.e. osteoclast lacuna) can be detected on woven bone surfaces (, PMMA cement; , woven bone; *, trabecular bone). (C) Former resorption lacunae on trabecular bone, caused by osteoclasts, are now filled with woven bone (, woven bone; *, trabecular bone). (D) A former microfracture of trabecular bone is bridged with woven bone (, woven bone formation; *, trabecular bone).

Although the woven bone appeared morphologically to be deposited at different time points, no signs of ongoing bone resorption (i.e. osteoclast lacuna) could be detected throughout the specimen (Fig. 2B). In contrast, inside the central parts of trabecular bone, residues of lacunae which were already filled with woven bone could again be detected (Fig. 2C). Former microfractures were also filled and/or bridged with woven bone (Fig. 2D). Although observed in most parts of the vertebra, the majority of the woven bone was found in the immediate vicinity to the PMMA without any obvious relationship to trabecular bone fractures (Fig. 2A).

SEM and µCT enabled to image in three dimensions the woven bone structures at the interface with the PMMA. As observed with the other techniques, a dense network of woven bone was found in close proximity to the PMMA (Fig. 1D,E), with the cement being covered by a sheath of bone that bridged the gaps to the surrounding cancellous bone.

Discussion

Osteoporosis is a condition characterized by an imbalance between bone resorption and bone formation, leading to a reduction in skeletal mass (Burston et al. 1998) and to the accumulation of microdamage (Lee et al. 2003). Reduced bone strength combined with an increased rate of falls in the elderly are the main factors causing osteoporotic fractures in this age group. Patients with a history of vertebral fracture have a 2.3-fold increased risk of suffering future hip fractures (Sambrook & Cooper, 2006). Furthermore, all major osteoporotic fractures are associated with increased mortality (Center et al. 1999). Successful treatment is therefore a challenging task comprising many different factors.

In biomechanical terms, the huge difference in stiffness between the weak bone and the stiff osteosynthesis materials reflects the main challenge in the operative treatment of osteoporotic fractures, particularly in heavily loaded parts of the skeletal system. In the case of more or less permanently loaded vertebral fractures, decompression and stabilization with different kinds of metal implants are associated with high risks for the often elderly patients. Therefore, these treatments have generally been limited to cases with concurrent spinal instability or neurological deficit (Taylor et al. 2006). Vertebroplasty and kyphoplasty are minimally invasive procedures developed for the treatment of symptomatic (i.e. painful) vertebral compression fractures. Despite the undoubted clinical value of these procedures, controversy persists mainly because of limiting issues related to the different mechanical stiffness of PMMA and the adjoining bony structures (Wagner & Baskurt, 2006). Cancellous bone filled with PMMA is known to be 36 times stiffer than the surrounding vertebral cancellous bone (Baroud et al. 2003). Assuming that these biomechanical facts are associated with a high risk of refracture at the treated levels, the reported refracture rate is extremely low (Leslie-Mazwi & Deen, 2006).

Owinf to inhibition of osteoclast-mediated bone resorption, bisphosphonates are the most commonly used medication in osteoporosis (Reginster et al. 2006). In clinical studies the effects of bisphosphonates on the incidence of non-vertebral osteoporotic fractures vary from study to study, although bisphosphonates have demonstrated a substantial reduction in vertebral fracture risk in post-menopausal women suffering from osteoporosis (Adami, 2007).

The specific mechanisms by which bisphosphonates prevent fractures are not yet entirely understood (Solomon, 2002). Regarding their effect on osteoblasts, conclusions reported in the literature are inconsistent. Although bisphosphonates seem to stimulate osteoblast proliferation and differentiation in vitro (Im et al. 2004), the contrary effect was found in vivo in animal models (Iwata et al. 2006).

Osteoporotic woven bone formation in terms of single, nodular, rounded, fusiform, angulated or arched bridge lesions in cancellous bone have already been described (Hahn et al. 1995). These formations can measure about 500 µm in diameter and surround fractured trabeculae with microcallus (Fazzalari, 1993).

However, the massive callus formation found in the present case does not comply with the characterization of microcallus reported thus far. The callus observed here embeds the PMMA to a large extent, mainly without a direct affiliation to the trabecular bone and without any evidence of resorption. The patient's bisphosphonate medication could possibly explain the absence of recent signs of resorption. Besides the fact that a positive bisphophosphonate influence on bone proliferation is unlikely, the local interaction between callus formation and PMMA points towards local mechanical rather than systemic influences.

The most probable reason for the callus formation is therefore the differences in stiffness between PMMA and cancellous bone, leading to increased stress levels at the interface region between these two materials. The trabecular bone that is embedded within the cement can no longer fail. By contrast, the trabeculae outside the cement are without such support and have to withstand the full load without additional support. This may ulimately lead to trabecular microfractures or at least result in an increased mechanical stimulus, which leads to the observed woven bone formation. Therefore, this process can be regarded as an attempt to strengthen the interface between the injected PMMA cement and the surrounding cancellous bone. Such integration may lead to the formation of a functionally better adapted interface, resulting in a smoother transition between the material properties of osteoporotic bone and PMMA. The results from this study also suggest that, contrary to established knowledge, even in case of osteoporosis, the formation of large amounts of new bone is possible. Formation of callus, as observed in the present case, resulting from the integration of PMMA, could be an explanation for the positive outcome of vertebroplasty and kyphoplasty.