Measurement of blood flow to the skeleton is technically challenging. The specific problems of measuring blood flow that are particular to bone are:

i) there are 206 separate bones in the skeleton;

Because of this heterogeneity of the tissue, it is also important to specify precisely the region of bone that is being measured, and this problem accounts for some of the discrepancies in values of bone blood flow quoted in the literature. From a practical orthopaedic perspective, techniques to measure regional blood flow are normally more informative than measurements of total skeletal blood flow.

In experimental studies, the microsphere technique has been used most widely for the quantitative measurement of bone blood flow, and is regarded as the gold standard. Particles of the order of 15 microns in diameter are injected into the ventricle and trapped in the microcirculation during a single passage. The distribution of microspheres in the body is proportional to the distribution of cardiac output, and if a reference arterial blood sample is taken during injection of the microspheres, then blood flow may be calculated. Microspheres are normally labeled with a radioactive tracer or a colored dye, and microsphere number is estimated from assays of the attached label.

The microsphere technique is a specific example of indicator fractionation, and clinically indicator fractionation can be applied using imaging techniques such as magnetic resonance imaging (MRI) or positron emission tomography (PET). MRI-based techniques are based on gadolinium contrast agents, and PET uses positron-emitting isotopes such as oxygen-15 labelled water, fluorine-18 ion, or ¹⁸F-fluorodeoxyglucose. Positron-emitting isotopes are short-lived, and need to be produced daily by a cyclotron, limiting the general utility of the technique. However, dynamic PET measurements with fluorine-18 have been used to assess simultaneously both bone blood flow and bone formation rates.

Blood flow can also be estimated from velocity measurements, e.g. electromagnetic flowmetry, laser Doppler, and ultrasound Doppler. Laser Doppler measurements require contact between the probe and the tissue being measured, and have applications in experimental studies of vascular reactivity in bone. Although ultrasound is reflected very effectively from bone surfaces, ultrasound Doppler has been used to image the lumber arteries in patients with degenerative disc disease.

Bone, like other tissues in the body, is relatively transparent to light in the near-infra red, but there are specific absorption peaks for deoxy- and oxy-hemoglobin. This is the basis of near infra-red spectroscopy for perfusion measurements. However, because of the complexities of light scattering in tissue, spatial resolution is poor. Measurements in the proximal tibia are quite straightforward, and we are currently using this technique in studies of bone loss in spinal cord injury patients.

Introduction: Duplex ultrasound has recently been used to demonstrate inflammatory hyperaemia in arteries supplying inflamed joints in RA, bursitis, and tendonitis. The technique has yet to be applied to examine blood flow in lumbar arteries in LBP patients, though we have previously shown its feasibility in healthy subjects. Our aim was to determine if there are differences in the flow characteristics of lumbar arteries in patients with LBP that may be reflective of pathology.

Materials and Methods: Sixty four patients with LBP (21–82 years) and 30 volunteers with no history of LBP (19–82 years) were studied. Sacral and lumbar arteries at L5 to L1 were identified and hemodynamic data was obtained using Duplex ultrasound. Angle corrected measurements of blood flow peak systolic velocity (PSV) were obtained at all lumbar levels and the aorta.

Results: Mean lumbar artery PSV was normalized with mean aorta PSV for patient and control groups. Reference range (mean ± 1.96SD) for normal lumbar artery PSV was defined from the control data and the proportion of patients with abnormally high PSV determined.

Discussion: Blood flow velocity in lumbar arteries of LBP patients is significantly higher compared with asymptomatic controls (p< 0.01). Approximately 40% of the LBP group have abnormally high lumbar artery flow velocity and the proportion of abnormal values increases at lower levels. Abnormally high velocity flow suggests the presence of an inflammatory component in the lumbar spinal structures. This technique has important applications in improving diagnostic specificity and assessing outcome of treatment in patients with LBP.

Introduction and Aims: We have developed a novel murine open tibial fracture model to compare the vascularity of muscle and fasciocutaneous flaps during fracture healing and investigate their role in angiogenesis.

Method: Flaps were emulated by insertion of a piece of sterile, inert material (Polytetrafluoroethylene, PTFE), at the fracture site to exclude either muscle posteriorly (fasciocutaneous flap) or skin and fascia anteriorly (muscle flap). Animals were harvested at days three, five, seven, nine and 14 post-fracture. Immunohistochemistry was performed on specimens, to estimate vascularity using an antibody to factor VIII, which selectively demonstrates vascular endothelium. Vascular densities were determined within the muscle and fasciocutaneous tissues adjacent to the fracture sites. Vascular Endothelial Growth Factor (VEGF) was measured by ELISA in tissue specimens. Immunohistochemistry was performed to qualitatively assess distribution of VEGF.

Results: Significantly greater vascular densities per unit area were observed in fasciocutaneous flaps at all time points compared to muscle flaps (p< 0.0001). VEGF levels peaked at day seven post-fracture, fell at day nine, and increased again at day 14. This time-dependent variation was statistically significant (p< 0.02). However, there was no significant difference between muscle and fasciocutaneous flaps. Maximal staining for VEGF occurred on the deep surface of the flaps adjacent to the fracture site. We found that fasciocutaneous flaps have significantly higher vascular densities compared to muscle flaps during early fracture healing.

Conclusion: Our results contradict the widely held view that muscle flaps are superior. However, there was no significant difference between levels of the pro-angiogenic factor VEGF within the flaps. This would suggest that both flaps are equally effective in supplying the factors necessary for new vessel formation. Our data supports the continuing use of muscle and fasciocutaneous flaps in the clinical setting.

We have studied the ability of a range of antibiotics to penetrate intervertebral disc tissue in vitro, using a mouse disc model. Equilibrium concentrations of antibiotics incorporated into the entire disc were determined by bioassay using a microbial growth-inhibition method. Uptake was significantly higher with positively-charged aminoglycosides compared with negatively-charged penicillins and cephalosporins. Uncharged ciprofloxacin showed an intermediate degree of uptake. Our results support the hypothesis that electrostatic interaction between charged antibiotics and negatively-charged glycosaminoglycans in the disc is an important factor in antibiotic penetration, and may explain their differential uptake.

The tibial nutrient artery supplies 62% of cortical blood flow in the diaphysis and normal blood flow is centrifugal (Willans 1987). Intramedullary reaming destroys the nutrient artery and injures the endosteal surface of the cortex. Trueta (1974) suggested that the direction of blood flow can reverse from centrifugal to centripetal after loss of the endosteal supply. We examined this hypothesis by measuring cortical and periosteal blood flow after intramedullary reaming of the tibia in eight sheep, using 57Co radiolabelled microspheres. The unreamed contralateral tibiae served as a control group. Thirty minutes after reaming there was no significant change in cortical blood flow, but a sixfold increase in the periosteal flow. Our study confirms Trueta's hypothesis; after trauma or in other pathological states, flow can become centripetal.

We examined the effect of periosteal devascularisation upon the early healing of osteotomies of sheep tibiae held in an instrumented external fixation system with an axial stiffness of 240 N/mm. At 14 days, cortical blood flow measured by the microsphere technique was 19.3 ml/min/100g in the well-vascularised osteotomies, but only 1.7 ml/min/100g in the devascularised osteotomies, despite an increase in medullary flow (p less than 0.0005). Delay in healing of the devascularised osteotomies was suggested by an in vivo monitoring system and confirmed by post-mortem mechanical testing. We suggest that the osteogenic stimulus of dynamic external fixation is dependent on the early restoration of cortical blood flow in devascularised fractures.

There has been a long-standing debate as to whether medullary or periosteal flow is the dominant vascular supply during the healing of diaphyseal fractures. We used radioactive microspheres to quantify blood flow to the canine tibia two weeks after an osteotomy. There was a significant contribution from the periosteum to the blood supply of healing cortical bone after nutrient artery ligation, with a reversal of flow from a centrifugal to a centripetal direction. Our study has confirmed the qualitative observations of Trueta (1974) regarding the significant recruitment of vessels from surrounding soft tissue during fracture healing. We have not studied the later stages of healing.

We have investigated the effect of currents induced by electromagnetic fields on the healing of the tibia of sheep after osteotomy, using objective and quantifiable criteria wherever possible. A battery-powered, induction apparatus was developed and was enclosed within the cast applied to the limb, so that the treated fractures received pulsed magnetic fields for 24 hours a day while the animals were freely mobile. In all, 13 sheep were treated and 13 were used as controls. The response was assessed by radiography of the limb and of the excised bone, by histology, including measurement of the areas of callus, fibrocallus and cortical bone, and by measurement of the uptake and extraction of bone-seeking mineral. All the bones healed and no statistically significant differences between the treated animals and the controls were discovered except (at only P less than 0.05) in the uptake of bone-seeking mineral; this increased more rapidly in treated animals over the two to three weeks after osteotomy, although at six weeks the uptake in both groups was the same.