Theo H. Smit

The use of a quadruped as an in vivo model for the study of the spine – biomechanical considerations

Abstract. Animal models in spine research are often criticized for being irrelevant to the human situation due to the horizontal position of the spine. Whether this is justified from a biomechanical point of view can be questioned. The purpose of the study reported here was to provide arguments that a quadruped can be a valuable in vivo model for the study of the spine in spite of its horizontal position. Relevant literature is reviewed, and biomechanical analyses were made of the standing and walking quadruped. Further, the vertebral trabecular bone architecture was quantitatively analysed by computer and interpreted in the light of Wolffs law. Due to the fact that spinal segments cannot withstand substantial bending moments, additional tensile forces from muscles and ligaments are necessary to control the posture of a quadruped spine. As a consequence, the spine is mainly loaded by axial compression. The trabeculae in a goats vertebral body were found to course horizontally between its anterior and posterior endplates, implying that the main load within the vertebral body was indeed an axial compression force. The density of the vertebrae of quadrupeds is higher than that of human vertebrae, suggesting that the quadruped has to sustain higher axial compression stresses. The quadruped spine is mainly loaded along its long axis, just like the human spine. The quadruped can thus be a valuable animal model for spine research. An important point of difference is the higher axial compression stress in quadrupeds, which leads to higher bone densities in the vertebrae. This puts some limitations on the transferability of the results of animal experiments to the human situation.

Strain-derived canalicular fluid flow regulates osteoclast activity in a remodelling osteon—a proposal

The concept of bone remodelling by basic multicellular units is well established, but how the resorbing osteoclasts find their way through the pre-existing bone matrix remains unexplained. The alignment of secondary osteons along the dominant loading direction suggests that remodelling is guided by mechanical strain. This means that adaptation (Wolffs Law) takes place throughout life at each remodelling cycle. We propose that alignment during remodelling occurs as a result of different canalicular flow patterns around cutting cone and reversal zone during loading. Low canalicular flow around the tip of the cutting cone is proposed to reduce NO production by local osteocytes thereby causing their apoptosis. In turn, osteocyte apoptosis could be the mechanism that attracts osteoclasts, leading to further excavation of bone in the direction of loading. At the transition between cutting cone and reversal zone, however, enhanced canalicular flow will stimulate osteocytes to increase NO production, which induces osteoclast retraction and detachment from the bone surface. Together, this leads to a treadmill of attaching and detaching osteoclasts in the tip and the periphery of the cutting cone, respectively, and the digging of a tunnel in the direction of loading.

Estimation of the poroelastic parameters of cortical bone.

Cortical bone has two systems of interconnected channels. The largest of these is the vascular porosity consisting of Haversian and Volkmanns canals, with a diameter of about 50 microm, which contains a.o. blood vessels and nerves. The smaller is the system consisting of the canaliculi and lacunae: the canaliculi are at the submicron level and house the protrusions of the osteocytes. When bone is differentially loaded, fluids within the solid matrix sustain a pressure gradient that drives a flow. It is generally assumed that the flow of extracellular fluid around osteocytes plays an important role not only in the nutrition of these cells, but also in the bones mechanosensory system. The interaction between the deformation of the bone matrix and the flow of fluid can be modelled using Biots theory of poroelasticity. However, due to the inhomogeneity of the bone matrix and the scale of the porosities, it is not possible to experimentally determine all the parameters that are needed for numerical implementation. The purpose of this paper is to derive these parameters using composite modelling and experimental data from literature. A full set of constants is estimated for a linear isotropic description of cortical bone as a two-level porous medium. Bone, however, has a wide variety of mechanical and structural properties; with the theoretical relationships described in this note, poroelastic parameters can be derived for other bone types using their specific experimental data sets.

Osteocyte morphology in fibula and calvaria --- is there a role for mechanosensing?

INTRODUCTION External mechanical forces on cells are known to influence cytoskeletal structure and thus cell shape. Mechanical loading in long bones is unidirectional along their long axes, whereas the calvariae are loaded at much lower amplitudes in different directions. We hypothesised that if osteocytes, the putative bone mechanosensors, can indeed sense matrix strains directly via their cytoskeleton, the 3D shape and the long axes of osteocytes in fibulae and calvariae will bear alignment to the different mechanical loading patterns in the two types of bone. MATERIALS AND METHODS We used confocal laser scanning microscopy and nano-computed tomography to quantitatively determine the 3D morphology and alignment of long axes of osteocytes and osteocyte lacunae in situ. RESULTS Fibular osteocytes showed a relatively elongated morphology (ratio lengths 5.9:1.5:1), whereas calvarial osteocytes were relatively spherical (ratio lengths 2.1:1.3:1). Osteocyte lacunae in fibulae had higher unidirectional alignment than the osteocyte lacunae in calvariae as demonstrated by their degree of anisotropy (3.33 and 2.10, respectively). The long axes of osteocyte lacunae in fibulae were aligned parallel to the principle mechanical loading direction, whereas those of calvarial osteocyte lacunae were not aligned in any particular direction. CONCLUSIONS The anisotropy of osteocytes and their alignment to the local mechanical loading condition suggest that these cells are able to directly sense matrix strains due to external loading of bone. This reinforces the widely accepted role of osteocytes as mechanosensors, and suggests an additional mode of mechanosensing besides interstitial fluid flow. The relatively spherical morphology of calvarial osteocytes suggests that these cells are more mechanosensitive than fibular osteocytes, which provides a possible explanation of efficient physiological load bearing for the maintenance of calvarial bone despite its condition of relative mechanical disuse.

Is BMU‐Coupling a Strain‐Regulated Phenomenon? A Finite Element Analysis

Histologically, two types of bone reconstruction are distinguished: modeling and remodeling. Modeling changes the amount of bone and determines its geometrical form in relation to the prevailing mechanical loads and their resulting deformation (strain). Remodeling renews existing bone in a sequence of resorption and formation. However, in both processes the cells responsible for resorption and formation are the same: osteoclasts and osteoblasts. We studied if there is a relation between the activity of these cells and the deformation of the local bone tissue during remodeling. Two finite element models were built on a microscopic, supracellular level: (1) a secondary osteon in cortical bone and (2) a Howships lacuna in a trabecula. Both models were loaded in the “natural,” that is, longitudinal direction. Equivalent strains were determined as a measure for the deformation of the bone tissue. In the first model, the strain field around the osteon showed a region of decreased deformation in front of the tunnel, just where osteoclasts excavate cortical bone tissue. Behind the cutting cone, elevated strain levels appear in the tunnel wall at locations where osteoblasts are active. The second model showed that a local excavation of a loaded trabecula leads to higher strains at the bottom of the lacuna, where resorption is stopped and osteoblasts are recruited to refill the gap. However, in the direction of loading reduced strain levels appear, just where resorption continues to proceed along the trabecular surface. We conclude that at the tissue level, strain distributions occur during the remodeling process that show a relationship to the activity of osteoblasts and osteoclasts. This suggests that BMU coupling, that is, the subsequent activation of osteoclasts and osteoblasts during remodeling, is a strain‐regulated phenomenon. (J Bone Miner Res 2000;15: 301–307)

Structure and function of vertebral trabecular bone.

Study Design. A combined morphologic and finite‐element study on vertebral trabecular bone. Objective. To relate the form and function of vertebral trabecular bone, in an attempt to better understand the mechanical function of a lumbar vertebra. Summary of Background Data. The architecture of bone is closely related to its mechanical function (Wolffs Law). In the human spine, vertebrae are subjected to a large variety of loads. Yet, these bones show a typical architecture, which means that they carry typical loads. Methods. Five trabecular bone cubes from specific sites of a lumbar vertebra were 3D‐reconstructed for computerized analysis. The architecture of the specimens was quantified by the bone volume fraction and a measure of anisotropy, the mean bone length. A finite element model was used to calculate internal stresses within a homogeneous vertebral body under basic loads. For each load case, bone volume fraction of the specimens was compared with the equivalent von Mises stress, and mean bone length was compared with the principal stress directions. Results. Bone volume fraction poorly related to the von Mises stress in the physiologic load case of axial compression. However, high bone volume fractions exist at locations where multiple load situations occur (e.g., near the pedicles and endplates). Remarkably, these sites also show finer architectures. Comparison of mean bone length with principal stresses revealed that the vertebral trabecular bone architecture particularly, but not entirely, corresponds to the stress field under axial compression. The horizontal struts near the endplates were found to be due to the function of the healthy intervertebral disc, and facetal joint loads introduce stress components that relate well with the bone structures near the pedicle bases. Conclusions. The trabecular bone architecture and the vertical orientation of the facet joints suggest that walking may be the main activity that determines the lumbar vertebral bone architecture.

The effect of cage stiffness on the rate of lumbar interbody fusion: an in vivo model using poly(l-lactic Acid) and titanium cages.

Study Design. A goat interbody fusion model using poly-(L-lactic acid) and titanium cages was designed to evaluate the effect of cage stiffness on lumbar interbody fusion. Objective. To investigate the effect of cage stiffness on the rate of interbody fusion. Summary of Background Data. Various types of cages considerably exceed the stiffness of vertebral bone, which ultimately may lead to postoperative complications. To avoid these complications, poly-(L-lactic acid) cages with limited stiffness have been designed. The mechanical integrity of the cages remains intact for at least 6 months. Methods. Interbody fusions were performed at L3–L4 of 15 Dutch milk goats, and one of three cages was randomly implanted: 1) a titanium cage (n = 3), 2) a stiff poly-(L-lactic acid) cage (n = 6), or 3) a flexible poly-(L-lactic acid) cage (n = 6). Interbody fusion was assessed radiographically by three independent observers 3 and 6 months after surgery. Results. At 3 months, all the poly-(L-lactic acid) specimens showed ingrowth of new bone, but with radiolucency in the fusion mass. At 6 months, solid arthrodesis was observed in four of six poly-(L-lactic acid) specimens, advanced ingrowth in one specimen, and infection in one specimen. Titanium cages showed ingrowth of bone, but with radiolucency in the fusion mass. Interbody fusion using poly-(L-lactic acid) cages showed a significantly higher rate statistically (P = 0.016) and more complete fusion than titanium cages of the same design. Conclusions. The reduced stiffness of poly-(L-lactic acid) cages showed enhanced interbody fusion, as compared with titanium cages after 6 months. Bioabsorbable poly-(L-lactic acid) cages thus may be a viable alternative to current interbody cage devices, thereby avoiding the concomitant problems related to their excessive stiffness. However, the bioabsorbability of the poly-(L-lactic acid) cages awaits investigation in a long-term study currently underway.

A case for strain-induced fluid flow as a regulator of BMU-coupling and osteonal alignment

Throughout life, human bone is renewed continuously in a tightly controlled sequence of resorption and formation. This process of bone remodeling is remarkable because it involves cells from different lineages, collaborating in so‐called basic multicellular units (BMUs) within small spatial and temporal boundaries. Moreover, the newly formed (secondary) osteons are aligned to the dominant load direction and have a density related to its magnitude, thus creating a globally optimized mechanical structure. Although the existence of BMUs is amply described, the cellular mechanisms driving bone remodeling—particularly the alignment process—are poorly understood. In this study we present a theory that explains bone remodelling as a self‐organizing process of mechanical adaptation. Osteocytes thereby act as sensors of strain‐induced fluid flow. Physiological loading produces stasis of extracellular fluid in front of the cutting cone of a tunneling osteon, which will lead to osteocytic disuse and (continued) attraction of osteoclasts. However, around the resting zone and the closing cone, enhanced extracellular fluid flow occurs, which will activate osteocytes to recruit osteoblasts. Thus, cellular activity at a bone remodeling site is well related to local fluid flow patterns, which may explain the coordinated progression of a BMU.

Bone cell responses to high-frequency vibration stress: does the nucleus oscillate within the cytoplasm?

Mechanosensing by cells directs changes in bone mass and structure in response to the challenges of mechanical loading. Low‐amplitude, high‐frequency loading stimulates bone growth by enhancing bone formation and inhibiting disuse osteoporosis. However, how bone cells sense vibration stress is unknown. Hence, we investigated bone cell responses to vibration stress at a wide frequency range (5–100 Hz). We used NO and prostaglandin E2 (PGE2) release, and COX‐2 mRNA expression as parameters for bone cell response since these molecules regulate bone adaptation to mechanical loading. NO release positively correlated whereas PGE2 release negatively correlated to the maximum acceleration rate of the vibration stress. COX‐2 mRNA expression increased in a frequency‐dependent manner, which relates to increased NO release at high frequencies, confirming our previous results. The negatively correlated release of NO and PGE2 suggests that these signaling molecules play different roles in bone adaptation to high‐frequency loading. The maximum acceleration rate is proportional to ω3 (frequency=ω/2π), which is commensurate with the Stokes‐Einstein relation for modeling cell nucleus motion within the cytoplasm due to vibration stress. Correlations of NO and PGE2 with the maximum acceleration rate then relate to nucleus oscillations, providing a physical basis for cellular mechanosensing of high‐frequency loading.—Bacabac, R. G., Smit, T. H., Van Loon, J. J. W. A., Doulabi, B. Z., Helder, M., Klein‐Nulend, J. Bone cell responses to high‐frequency vibration stress: does the nucleus oscillate within the cytoplasm? FASEB J. 20, 858–864 (2006)

Disorders in trunk rotation during walking in patients with low back pain: a dynamical systems approach

OBJECTIVE (1) To introduce an evaluation tool for the assessment of walking disorders in low back pain patients. (2) To investigate whether walking patterns in low back pain patients are different from those of control subjects. DESIGN Relative phase measures of movement coordination are applied in the assessment of trunk function in a small group of patients with non-specific low back pain and in control subjects. BACKGROUND Normal subjects change the coordination of pelvic and thoracic rotations from an in-phase to an out-of-phase pattern with increasing walking speed. Low back pain patients may have a reduced ability to counter rotate pelvis and thorax at higher walking speeds (from 1.0 m/s onwards) as a result of hyperstable coordination patterns. METHODS Six patients with non-specific low back pain and six healthy control subjects walked on a treadmill at comfortable walking speeds and during a systematic variation of the treadmill velocity. Coordination of arm and leg movements as well as of pelvic and thoracic rotations was analyzed using a relative phase algorithm. RESULTS AND CONCLUSIONS The comfortable walking speed was reduced in the patient group. In contrast to the control subjects, four of the six patients were not able to establish an out-of-phase coordination pattern between thorax and pelvis at higher walking speeds. This coincided with an increased stability of movement coordination, indicating guarded behavior. In addition, an increased asymmetry between the phase-relations of left and right side of the body was found in some of the patients.