Development of an in vivo bone fatigue damage model using axial compression of the rabbit forelimb

Abstract

Many nontraumatic fractures seen clinically in patients with metabolic bone disorders or on antiresorptive treatment show an increased incidence of microdamage accumulation and impaired intracortical remodeling. However, the lack of basal remodeling and Haversian bone in rodents limits their translatability in studying bone damage repair mechanisms. The work presented here demonstrates the development of the forelimb loading model in rabbits, the smallest mammal with intracortical Haversian remodeling. The forelimbs of post-mortem female New Zealand white rabbits were loaded in axial end compression to determine their basic monotonic and fatigue properties. Following time zero characterization, stress fractures were created in vivo and animals were allowed to recover for a period of two to five weeks. The rabbit forelimb when loaded in axial compression demonstrates a consistent mid-diaphyseal fracture location characterized by a local mixed compression-bending loading environment. Forelimb apparent stiffness, when fatigue loaded, demonstrates a progressive increase until macrocrack formation, at which time apparent stiffness rapidly declines until failure. Stress fractures in the rabbit ulna display robust periosteal expansion and woven bone formation two weeks following fracture. Subsequent healing at five weeks post-fracture is marked by woven bone densification, resorption and intracortical remodeling along the stress fracture line. The rabbit forelimb fatigue model is a promising new platform by which bone׳s response to damage may be studied.

Keywords: Bone damage; Bone fatigue; Bone mechanics; In vivo forelimb loading; Rabbit ulna.

Figure 1

A: MicroCT image of an isolated forelimb depicts our axial end compression loading setup and the structure of the forelimb bones. However, all loading was performed with the forelimb still attached to the body and with soft tissue structure intact. B: Representative microCT image of the forelimb cross section at the longitudinal site of peak axial strain demonstrating the location of the three gauges used for ulnar strain analysis. For strain gauge testing only, soft tissues, including the periosteum, was excised around the central third of the forelimb to allow for proper strain gauge adherence.

Figure 2

Strain data recorded from gauges 1, 2 and 3 were plotted as a function of loading force. Gauges 1 and 2 experienced compressive strains whereas gauge 3 experienced only tensile strains. The change in strain magnitude and sign across the ulna’s cross-section demonstrate the local bending environment produced by forelimb axial compression. The nonlinear force-strain behavior at higher forces is not attributed to bone plastic deformations or viscoelastic effects, but rather load dependent alterations in forelimb apparent stiffness.

Figure 3

A: Forelimb apparent stiffness (Fmax−Fmin/Dmax−Dmin) for each cycle was calculated following fatigue to failure tests and plotted versus residual cycle # (current cycle/total cycles). The forelimb apparent stiffness at cycle 10 (K_10) was used as a reference value. Forelimb apparent stiffness increased throughout cycling until macrocrack formation (K_Event1), at which time forelimb apparent stiffness rapidly declined until failure (K_Fail). K_Secondary and K_Tertiary were the recorded stiffness values halfway between K_10 and K_Event1 and K_Event1 and K_Fail respectively. B: On average, forelimb apparent stiffness from cycle 10 progressively increased by 30% until macrocrack formation, at which time specimen apparent stiffness rapidly declined until failure. Arrows A-D demonstrate the stage of fatigue life that specimens in Figure 4 were loaded to based on forelimb apparent stiffness changes (K/K_10).

Figure 4

Macrodamage was visualized by microCT in specimens loaded to sub-failure progressive changes in forelimb apparent stiffness (K/K10). K10 was the forelimb apparent stiffness recorded at the tenth cycle of loading and K was the stiffness recorded at the last cycle before loading ceased. Samples loaded to forelimb apparent stiffness values before Event1 (K/K10 ≥1.3) had no visible macrodamage (Panel A – 35 micron resolution). Samples loaded to progressive changes in forelimb apparent stiffness past Event1 (K/K10 < 1.3) demonstrated increasingly more visible macrodamage in the ulna (Panel B–D – 17.5 micron resolution).

Figure 5

A: MicroCT of the rabbit ulna two weeks following stress fracture formation demonstrates robust woven bone formation at locations of peak strain found during strain analysis. B: Dynamic histomorphometry of stress fracture shows that woven bone callus formation occurs predominantly between week one and two following stress fracture. C: MicroCT of the rabbit ulna five weeks following stress fracture shows prominent callus densification and remodeling. D: Dynamic histomorphometry demonstrates areas of bone resorption and formation along the cortical fracture site and within the woven bone callus.