We assessed carpal collapse by measuring the capitate-radius (CR) distance on standard plain radiographs. This new method required validation of diagnostic accuracy, so we compared it with the method of Nattrass et al¹ known as revised carpal height (RCH).

We studied wrist radiographs from 16 normal subjects and 11 patients with unilateral Kienböck’s disease. We found that there was a significant difference in the left/right CR index between the normal wrists and those with Kienböck’s disease (p < 0.001). The use of left/right RCH index showed no significant difference (p = 0.30). Diagnostic accuracy was shown to be higher for the CR index using ROC curves.

We then assessed 40 normal wrists and found the mean CR index to be 0.999 ± 0.034, and suggest that values less than 0.92 are abnormal. The CR index can be used for diagnosis in unilateral carpal collapse, and for monitoring progress where the condition is bilateral.

Inadequate bone stock is often found in revision surgery of femoral components of total knee replacements. Our aim was to test the hypothesis that these remodelling patterns can be explained by stress shielding, and that prosthetic bonding characteristics affect maintenance of bone mass.

We made a three-dimensional finite-element model of an average male femur with a cemented femoral knee component. This model was integrated with iterative remodelling procedures. Two extreme prosthetic bonding conditions were analysed and gradual changes in bone density were calculated.

The long-term bone loss under the femoral knee component resembled clinical findings which confirms the hypothesis that stress shielding can cause distal femoral bone loss. Our study predicts, contrary to clinical findings, that an equilibrium situation is not reached after two years, but that bone resorption may continue. This hidden bone loss may be so drastic that large reconstructions are needed at the time of revision.

We obtained intervertebral discs with cartilage endplates and underlying cancellous bone at operation from patients with degenerative disc disease and then used immunohistochemical techniques to localise the nerves and nerve endings in the specimens. We used antibodies for the ubiquitous neuronal protein gene product 9.5 (PGP 9.5). Immunoreactivity to neuropeptide Y was used to identify autonomic nerves and calcitonin gene-related peptide (CGRP) and substance P to identify sensory nerves. Blood vessels were identified by immunoreactivity with platelet-endothelial cell-adhesion molecule (CD31; PECAM).

In a control group with no known history of chronic back pain, nerve fibres immunoreactive to PGP 9.5 and neuropeptide Y were most closely related to blood vessels, with occasional substance P and CGRP immunoreactivity. In patients with severe back pain and markedly reduced disc height, proliferation of blood vessels and accompanying nerve fibres was observed in the endplate region and underlying vertebral bodies. Many of these nerves were immunoreactive to substance P or CGRP, and in addition, substance P- and CGRP-immunoreactive nociceptors were seen unrelated to blood vessels. Quantification by image analysis showed a marked increase in CGRP-containing sensory nerve fibres compared with normal control subjects.

We speculate that a chemotactic response to products of disc breakdown is responsible for the proliferation of vascularity and CGRP-containing sensory nerves found in the endplate region and vertebral body adjacent to degenerate discs. The neuropeptides substance P and CGRP have potent vasodilatory as well as pain-transmitting effects. The increase in sensory nerve endings suggests increase in blood flow, perhaps as an attempt to augment the nutrition of the degenerate disc. The increase in the density of sensory nerves, and the presence of endplate cartilage defects, strongly suggest that the endplates and vertebral bodies are sources of pain; this may explain the severe pain on movement experienced by some patients with degenerative disc disease.

We analysed the bacterial contamination of 1999 bone allografts retrieved from 200 cadaver donors under sterile operating conditions. The effect of various factors on the relative risk of contamination was estimated using a multiple logistic regression model.

Organisms of low pathogenicity were cultured from 50% of the grafts and of high pathogenicity from 3%. The risk of contamination with low pathogenic organisms (mainly skin commensals) increased by a factor of 1.6 for each member added to the procurement team. The risk of contamination with high pathogenic organisms (mainly contaminants from the gastrointestinal tract) was 3.4 times higher in donors with a traumatic cause of death and 5.2 times higher in those with a positive blood culture. Preceding organ procurement did not significantly influence the risk of contamination. Rinsing the graft with an antibiotic solution was not an effective decontamination method.

The major source of contamination is exogenous and is strongly influenced by the procurement team. Contamination from endogenous sources can be controlled by donor selection. We discuss methods that can be used to decrease contamination and the rate of discarding of bone allografts.

We made a semiquantitative study of the comparative histology of pseudomembranes from 12 loose cemented ceramic-ceramic and 18 metal-polyethylene total hip replacements. We found no significant difference in cellular reaction between the two groups, but there was a major difference in the origin of the particulate debris. In the metal-polyethylene group, polyethylene of articular origin was predominant, while in the ceramic-ceramic group the cellular reaction appeared to be a response to zirconia ceramic particles used to opacify cement used for fixation.

Isolation and characterisation of the debris showed that the zirconia particles formed the greatest proportion (76%) in ceramic-ceramic hips, while alumina debris of articular origin formed only 12%.

Our study has indicated that aseptic loosening of ceramic cups is not due to a response to debris generated at the articular interface, but to mechanical factors which lead to fragmentation of the cement.

Estimates of knee joint loadings were calculated for 12 normal subjects from kinematic and kinetic measures obtained during both level and downhill walking. The maximum tibiofemoral compressive force reached an average load of 3.9 times body-weight (BW) for level walking and 8 times BW for downhill walking, in each instance during the early stance phase. Muscle forces contributed 80% of the maximum bone-on-bone force during downhill walking and 70% during level walking whereas the ground reaction forces contributed only 20% and 30% respectively.

Most total knee designs provide a tibiofemoral contact area of 100 to 300 mm². The yield point of these polyethylene inlays will therefore be exceeded with each step during downhill walking. Future evaluation of total knee designs should be based on a tibiofemoral joint load of 3.5 times BW at 20° knee flexion, 8 times BW at 40° and 6 times BW at 60°.

A three-dimensional computer model of a total hip replacement was used to examine the relationship between the position of the components, the range of motion and the prosthetic joint contact area. Horizontal acetabular positions with small amounts of acetabular and femoral anteversion provide the largest contact areas, but result in limited joint movement.

These data will allow surgeons to select implant positions that will provide the largest possible joint contact area for a given joint range of motion although these are conflicting goals. In some component positions a truncated spherical prosthetic head resulted in smaller contact areas than a completely spherical head.

We have tested the axial and torsional stability of femoral components after revision arthroplasty in a cadaver model, using impacted morsellised cancellous graft and cement. Each one of six matched pairs of fresh frozen human femora had either a primary or a revision prosthesis cemented in place. For the ‘revision’ experiments, all cancellous bone was removed from the proximal femur which was then over-reamed to create a smooth-walled cortical shell. An MTS servohydraulic test frame was used to apply axial and torsional loads to each specimen through the prosthetic femoral heads with the femur submerged in isotonic saline solution at 37°C.

The mean subsidence was 0.27 ± 0.17 mm for the primary and 0.52 ± 0.30 mm for the revision groups. The difference was statistically significant (p < 0.025), but the mean subsidence was < 1 mm in both groups. The mean maximum torque before failure was 42.9 ± 26.9 N-m for the primary and 34.8± 20.7 N-m for the revision groups. This difference was not statistically significant (p > 0.015).

Based on our results we suggest that revision of the femoral component using morsellised cancellous graft followed by cementing with a collarless prosthesis with a polished tapered stem restores the integrity of the proximal femur and provides immediate stability of the implant.

Local treatment with phenol is often used after intralesional excision of chondroblastomas and giant-cell tumours which involve bone near joints, and has been shown to reduce the rate of recurrence. The ideal concentration of phenol is uncertain, but may be important because of the high rate of absorption and toxicity. We have studied the effectiveness of different concentrations on standard sarcoma cell lines.

Our results suggest that a 5% solution of phenol is effective against dispersed single cells, and that higher concentrations give no significant advantage, but create problems due to lack of homogeneous mixing, temperature and safety.

We investigated the distribution of compressive ‘stress’ within cadaver intervertebral discs, using a pressure transducer mounted in a 1.3 mm diameter needle. The needle was pulled along the midsagittal diameter of a lumbar disc with the face of the transducer either vertical or horizontal while the disc was subjected to a constant compressive force. The resulting ‘stress profiles’ were analysed in order to characterise the distribution of vertical and horizontal compressive stress within each disc. A total of 87 discs from subjects aged between 16 and 87 years was examined.

Our results showed that age-related degenerative changes reduced the diameter of the central hydrostatic region of each disc (the ‘functional nucleus’) by approximately 50%, and the pressure within this region fell by 30%. The width of the functional annulus increased by 80% and the height of compressive ‘stress peaks’ within it by 160%. The effects of age and degeneration were greater at L4/L5 than at L2/L3, and the posterior annulus was affected more than the anterior. Age and degeneration were themselves closely related, but the stage of degeneration had the greater effect on stress distributions.

We suggest that structural changes within the annulus and endplate lead to a transfer of load from the nucleus to the posterior annulus. High ‘stress’ concentrations within the annulus may cause pain, and lead to further disruption.

Axial forces were measured during limb lengthening in a series of ten patients with varying pathologies in order to assess the mechanical characteristics of the distracted tissues and the levels of axial force to which soft tissues are subjected during leg lengthening.

The pattern of force was found to vary according to the underlying pathology. For post-traumatic shortening in adults both the peak and the resting forces rose steadily during lengthening reaching maximum forces of the order of 300 N. Patients with congenitally short limbs developed very high peak forces (in some cases over 1000 N) and also showed large amounts of force relaxation (typically 400 to 500 N).

When very high levels of force were recorded, there was a higher complication rate. In particular, there was a high instance of angular deformity. This occurred because the loads encountered resulted in failure of some of the external fixation frames.

The use of ultraclean air (UCA) in operating theatres reduces the infection rate after joint replacement but some cases of infection still occur. We investigated one possible source of contamination, namely the setting up of instruments in a conventional plenum-ventilated preparation room.

We measured bacterial fallout using agar settle plates and compared instruments set up in the preparation room with those set up in the UCA theatre, assessed the effect of covering instruments after preparation and compared fallout during their preparation with total fallout throughout the operation.

Our findings showed that covering the instruments reduced total bacterial fallout fourfold by reducing the exposure time, particularly during periods of increased activity and bacterial dispersal. Preparation in the UCA theatre and subsequent covering of the instruments reduced total fallout 28-fold. All measurable bacterial fallout occurred during the setting up and not during surgery.

We studied the reliability of the Singh classification of trabecular bone structure in the proximal femur as a measure of osteoporosis, using kappa statistics. Radiographs of fractures of the femoral neck or trochanteric region in 80 consecutive patients were assessed by six observers. The interobserver variation was large; only three of 72 radiographs were given the same classification by all six observers and the kappa values ranged from 0.15 to 0.54. The intraobserver variation showed substantial strength of agreement; kappa values ranged from 0.63 to 0.88.

In 77 patients dual-energy X-ray absorptiometry was used to measure bone mineral density. The results were compared with those of the Singh classification: we found no correlation.

We reviewed 194 revision arthroplasties of the hip and knee performed over a ten-year period. The results of intraoperative Gram staining were available in 169 (87%). Thirty-two were found to be infected (11 hips and 21 knees) and 137 had no evidence of infection.

Intraoperative Gram staining was negative in all 169 cases. The method therefore had a sensitivity of 0% for detecting infection. We conclude that the absence of organisms on intraoperative Gram staining during revision arthroplasty does not confirm the absence of infection.

Resistance to infection may be influenced by foreign bodies such as devices for fracture fixation. It is known that stainless steel and commercially-pure titanium have different biocompatibilities. We have investigated susceptibility to infection after a local bacterial challenge using standard 2.0 dynamic compression plates of either stainless steel or titanium in rabbit tibiae. After the wounds had been closed, various concentrations of a strain of Staphylococcus aureus were inoculated percutaneously.

Under otherwise identical experimental conditions the rate of infection for steel plates (75%) was significantly higher than that for titanium plates (35%) (p < 0.05).

Using a rat model, we created a bone-to-titanium interface and applied phagocytosable high-density polyethylene pArticles between the bone and implant, either initially or when the interface had matured. No fibrous membrane developed and no bone resorption was found.

If sliding movements were initiated at the interface after two weeks, there was formation of a fibrous membrane. The additional application of pArticles did not change the thickness of the membrane, and there were only minor qualitative changes. Creation of a membrane by movement followed by cessation of movement and the application of pArticles caused the membrane to persist, whereas in a pArticle-free control group bone-to-metal contact was re-established.

Our findings suggest that mechanical stimuli are of primary importance for prosthetic loosening, and that pArticles may modulate the later stages of the loosening process.

Bone growth into cementless prosthetic components is compromised by osteoporosis, by any gap between the implant and the bone, by micromotion, and after the revision of failed prostheses. Recombinant human transforming growth factor-β1 (rhTGF-β1) has recently been shown to be a potent stimulator of bone healing and bone formation in various models in vivo.

We have investigated the potential of rhTGF-β1, adsorbed on to weight-loaded tricalcium phosphate (TCP) coated implants, to enhance bone ongrowth and mechanical fixation. We inserted cylindrical grit-blasted titanium alloy implants bilaterally into the weight-bearing part of the medial femoral condyles of ten skeletally mature dogs. The implants were mounted on special devices which ensured stable weight-loading during each gait cycle. All implants were initially surrounded by a 0.75 mm gap and were coated with TCP ceramic.

Each animal received two implants, one with 0.3 μg rhTGF-β1 adsorbed on the ceramic surface and the other without growth factor. Histological analysis showed that bone ongrowth was significantly increased from 22 ± 5.6% bone-implant contact in the control group to 36 ± 2.9% in the rhTGF-β stimulated group, an increase of 59%. The volume of bone in the gap was increased by 16% in rhTGF-β1-stimulated TCP-coated implants, but this difference was not significant. Mechanical push-out tests showed no difference in fixation of the implant between the two groups. Our study suggests that rhTGF-β1 adsorbed on TCP-ceramic-coated implants can enhance bone ongrowth.

There have been conflicting reports on the effects of gamma irradiation on the material properties of cortical allograft bone. To investigate changes which result from the method of preparation, test samples must be produced with similar mechanical properties to minimise variations other than those resulting from treatment.

We describe a new method for the comparative measurement of bone strength using standard bone samples. We used 233 samples from six cadavers to study the effects of irradiation at a standard dose (28 kGy) alone and combined with deep freezing. We also investigated the effects of varying the dose from 6.8 to 60 kGy (n = 132).

None of the treatments had any effect on the elastic behaviour of the samples, but there was a reduction in strength to 64% of control values (p < 0.01) after irradiation with 28 kGy. There was also a dose-dependent reduction in strength and in the ability of the samples to absorb work before failure

We suggest that irradiation may cause an alteration in the bone matrix of allograft bone, but provided it is used in situations in which loading is within its elastic region, then failure should not occur.

The fatigue failure of bone cement, leading to loosening of the stem, is likely to be one mode of failure of cemented total hip replacements. There is strong evidence that cracks in the cement are initiated at voids which act as stress risers, particularly at the cement-stem interface. The preferential formation of voids at this site results from shrinkage during polymerisation and the initiation of this process at the warmer cement-bone interface, which causes bone cement to shrink away from the stem. A reversal of the direction of polymerisation would shrink the cement on to the stem and reduce or eliminate the formation of voids at this interface.

We have investigated this by implanting hip prostheses, at room temperature or preheated to 44°C, into human cadaver femora kept at 37°C. Two types of bone cement were either hand-mixed or vacuum-mixed before implantation. We found that the area of porosity at the cement-stem interface was dramatically reduced by preheating the stem and that the preheating temperature of 44°C determined by computer analysis of transient heat transfer was the minimum required to induce initial polymerisation at the cement-stem interface. Temperature measurements taken during these experiments in vitro showed that preheating of the stem caused a negligible increase in the temperature of the bone. Reduction of porosity at the cement-stem interface could significantly increase the life of hip arthroplasties.