Our aim was to examine the potential of autologous perichondral tissue to form a meniscal replacement. In 18 mature sheep we performed a complete medial meniscectomy. The animals were then divided into two groups: 12 had a meniscal replacement using strips of autologous perichondral tissue explanted from the lower rib (group G) and six (group C) served as a control group without a meniscal replacement. In all animals restriction from weight-bearing was achieved by means of transection and partial resection of tendo Achillis. Six animals (four from group G and two from group C) were each killed at 3, 6 and 12 months. The grafts and the underlying articular cartilage were removed and studied by gross macroscopic examination, light microscopy, SEM, polarised light examination, and by biomechanical tests.

In all the transplanted animals a new perichondral meniscus developed. After three months the transplants resembled normal menisci in size and thickness, while in the control animals only small rims of spontaneously grown tissue were seen. Microscopically, the perichondral menisci showed a normal orientation of collagen fibres and normal cellular characteristics, but in the central region, areas of calcification disturbed the regular tissue differentiation. Healing tissue in control animals lacked the normal fibre orientation and cellularity. SEM of perichondral menisci showed surface characteristics similar to those of normal sheep menisci without fissures and lacerations; the control specimens had these defects. The femoral and tibial cartilage in contact with the new menisci had normal surface characteristics apart from one animal with slight surface irregularities. Control animals showed superficial lesions after three months which increased at six to 12 months postoperatively. Microangiography of the newly grown tissue demonstrated a less intense vascularisation after three months when compared with normal menisci.

The failure stress and tensile modulus of perichondral menisci were significantly lower than those of normal contralateral menisci, and spontaneously regenerated tissue in meniscectomised animals had even lower values. There were no significant differences in values between newly grown perichondral menisci and spontaneously grown tissue.

Using a new, non-invasive method, we measured the patellofemoral force (PFF) in cadaver knees mounted in a rig to simulate weight-bearing. The PFF was measured from 20° to 120° of flexion before and after implanting three designs of knee prosthesis.

Medial unicompartmental arthroplasty with a meniscal-bearing prosthesis and with retention of both cruciate ligaments caused no significant change in the PFF. After arthroplasty with a posterior-cruciate-retaining prosthesis and division of the anterior cruciate ligament, the PFF decreased in extension and increased by 20% in flexion. Implantation of a posterior stabilised prosthesis and division of both cruciate ligaments produced a decrease in the PFF in extension but maintained normal load in flexion.

There was a direct relationship between the PFF and the angle made with the patellar tendon and the long axis of the tibia. The abnormalities of the patellar tendon angle which resulted from implantation of the two total prostheses explain the observed changes in the PFF and show how the mechanics of the patellofemoral joint depend upon the kinematics of the tibiofemoral articulation.

Chondrocytes of the growth plate are generally assumed to undergo apoptosis, but the mechanisms which induce this cell death are not known. The Fas receptor is a mediator of the apoptotic signal in some systems. We studied its expression in situ in growth plates of rabbits aged from five to 20 weeks. In addition, we investigated the immunolocalisation in the growth plates of the bone proteins, osteonectin and osteocalcin, and the changes in their expression with age.

The Fas-positive chondrocytes were found mostly in the hypertrophic zone, as were the osteonectin-positive and osteocalcin-positive cells. The percentage of Fas-positive cells increased with age whereas little change was found in the number of osteonectin-positive and osteocalcin-positive chondrocytes.

Many of the Fas-positive chondrocytes were also TUNEL-positive. This strongly suggests that apoptosis in the growth plate is mediated through the Fas system.

Double immunostaining for osteocalcin and Fas showed that not all hypertrophic chondrocytes were of the same cell type. Some chondrocytes stained for osteocalcin only, others for Fas only, while some were positive for both.

In a series of 12 patients with inoperable gastric carcinoma who had treatment with a synthetic matrix metalloproteinase inhibitor (Marimastat) for more than one month, six developed a frozen shoulder or a condition resembling Dupuytren’s disease.

This suggests that the matrix metalloproteinases, a family of naturally occurring proteinases, may be involved in the pathogenesis of these two conditions. Our observation opens avenues for further research which could lead to local or systemic therapeutic interventions for frozen shoulder and Dupuytren’s disease.

Our aim was to determine if the serum levels of bone-resorbing cytokines (IL-1β, TNF-α, IL-6, GM-CSF) are altered in patients with aseptic loosening of a total hip prosthesis, and if such levels are influenced by the type of implant. We determined cytokine levels in sera from 35 patients before revision for failed total hip arthroplasty and compared them with those in 25 healthy donors. We also assessed the soluble receptor of interleukin-2 (sIL-2r) in serum as an indication of a specific immune reaction against the implant.

Our findings showed that the sIL-2r and TNF-α serum level did not change. The IL-6 level was not significantly altered, but was higher in patients with TiAlV prostheses than in those with a CrCoMo implant and in patients with cemented prostheses. The IL-1β level was found to be higher in those with a TiAlV cemented prosthesis than in the control group (p = 0.0001) and other groups of patients (p = 0.003 v uncemented TiAlV, p = 0.01 v cemented CrCoMo, p = 0.001 v uncemented CrCoMo). The GM-CSF level significantly increased in patients compared with healthy subjects (p = 0.008), and it was higher in those with cemented than with uncemented implants (p = 0.01). Only patients with cementless CrCoMo prostheses had levels of GM-CSF similar to those of the control group. The highest GM-CSF concentrations were observed in patients treated with non-steroidal anti-inflammatory drugs (NSAIDs) in the last months before revision (p = 0.04). In addition, when massive osteolysis was observed, the level of GM-CSF tended to decrease to that of the control group.

We examined macroscopically and microscopically 55 cadaver rotator-cuff tendons attached to their humeral heads to determine the distance between the edge of the articular cartilage and the tendon insertion of the supraspinatus (the width of the sulcus) and the score of regressive changes at the sulcus. In 33 specimens we measured the tensile strength. The width of the sulcus was correlated with the score of regressive changes and with the ultimate tensile strength of the supraspinatus tendon.

The width of the sulcus correlated positively with the score of regressive changes (r = 0.66, p < 0.0001), but there was a negative correlation between the latter and the ultimate tensile strength (r = −0.81, p = 0.001) and between the width of the sulcus and the ultimate tensile strength (r = −0.74, p = 0.004).

We believe that the width of the sulcus is a simple and useful clinical indicator of the integrity and the tensile strength of the supraspinatus tendon.

We examined solvent-dried, gamma-irradiated (SD-R) allografts and fresh-frozen (FF) allografts mechanically and morphologically. Before transplantation, FF grafts were more than six times stronger than SD-R grafts. After four weeks, the tensile strength was about the same in both groups. At 24 weeks only collagen fibrils of small diameter were observed in the SD-R grafts while in FF grafts fibrils of small and intermediate diameter were seen. Clinically, we suggest that SD-R grafts could be used as a favourable alternative to FF grafts if care was taken regarding their initial mechanical weakness.

A foreign-body-type host response can contribute to the induction and release of collagenolytic tissue-destructive enzymes of pathogenetic significance. Our aim was to analyse collagenase-3 in two conditions with putative involvement of foreign-body reactions. Synovial membrane-like tissue samples were obtained from cases of aseptic loosening of a total hip replacement (THR) and osteoarthritis (OA).

The reverse transcription polymerase chain reaction (RT-PCR) disclosed that all the samples from patients contained collagenase-3 mRNA compared with only three out of ten control samples. The identity of the RT-PCR amplification product was confirmed by nucleotide sequencing. Immunohistochemical staining showed that collagenase-3 was present in endothelial cells, macrophages and fibroblasts, including those found in the synovial lining. This finding was confirmed by avidin-biotin-peroxidase complex-alkaline phosphatase-anti-alkaline phosphatase double staining and the specificity of the staining by antigen preabsorption using recombinant human collagenase-3.

Collagenase-3 was released into the extracellular space and thus found in the synovial fluid in all patient samples as shown by Western blotting. The similar extent of collagenase-3 expression in aseptic loosening and OA compared with the low expression in control synovial membrane suggests involvement of a similar, foreign-body-based pathogenetic component in both. Comparative analysis of collagenase-3 and of foreign particles indicates that paracrine factors rather than phagocytosis per se are responsible for the induction of collagenase-3.

We suggest that due to its localisation and substrate specificity, collagenase-3 may play a significant pathogenetic role in accelerating tissue destruction in OA and in aseptic loosening of a THR.

We observed the healing process under rigid external fixation after Salter-Harris type-1 or type-2 physeal separation at the proximal tibia in immature rabbits.

Metaphyseal vessels grew across the gap with little delay; the site of separation then came to lie in the metaphysis and was bridged by endochondral ossification. Union was achieved within two days in all rabbits.

Progression of endochondral ossification repaired the separated physis, thus showing ‘primary healing of physeal separation’. This depends on accurate reduction and stable fixation to allow the survival of vessels across the gap.

It has been thought that lumbar intervertebral discs were innervated segmentally. We have previously shown that the L5-L6 intervertebral disc in the rat is innervated bilaterally from the L1 and L2 dorsal root ganglia through the paravertebral sympathetic trunks, but the pathways between the disc and the paravertebral sympathetic trunks were unknown.

We have now studied the spines of 17 rats to elucidate the exact pathways. We examined serial sections of the lumbar spine using immunohistochemistry for calcitonin gene-related peptide, a sensory nerve marker. We showed that these nerve fibres from the intervertebral disc ran through the sinuvertebral nerve into the rami communicantes, not into the corresponding segmental spinal nerve.

In the rat, sensory information from the lumbar intervertebral discs is conducted through rami communicantes. If this innervation pattern applies to man, simple decompression of the corresponding nerve root will not relieve discogenic pain. Anterior interbody fusion, with the denervation of rami communicantes, may be effective for such low back pain.

We exposed human macrophages isolated from the peripheral blood of healthy donors to metal and bone-cement particles from 0.2 to 10 μm in size.

Zymography showed that macrophages exposed to titanium alloy and polymethylmethacrylate (PMMA) particles released a 92- and 72-kDa gelatinase in a dose- and time-dependent manner. Western immunoblotting confirmed that the 92- and 72-kDa gelatinolytic activities corresponded to matrix metalloproteinase-9 and matrix metalloproteinase-2 (MMP-9, MMP-2), respectively. Western immunoblotting also indicated that titanium alloy and PMMA particles increased the release of MMP-1. Northern blotting showed elevated mRNA signal levels for MMP-1, MMP-2, and MMP-9 after exposure to both types of particle. Collagenolytic activity also increased in the macrophage culture medium in response to both types of particle.

Our findings support the hypothesis that macrophages release MMPs in proportion to the amount of particulate debris within periprosthetic tissues.

We studied the morphometry of 35 femora from 31 female patients with developmental dysplasia of the hip (DDH) and another 15 from 15 age- and sex-matched control patients using CT and three-dimensional computer reconstruction models. According to the classification of Crowe et al 15 of the dysplastic hips were graded as class I (less than 50% subluxation), ten as class II/III (50% to 100% subluxation) and ten as class IV (more than 100% subluxation).

The femora with DDH had 10 to 14° more anteversion than the control group independent of the degree of subluxation of the hip. In even the most mildly dysplastic joints, the femur had a smaller and more anteverted canal than the normal control. With increased subluxation, additional abnormalities were observed in the size and position of the femoral head. Femora from dislocated joints had a short, anteverted neck associated with a smaller, narrower, and straighter canal than femora of classes I and II/III or the normal control group.

We suggest that when total hip replacement is performed in the patient with DDH, the femoral prosthesis should be chosen on the basis of the severity of the subluxation and the degree of anteversion of each individual femur.

The tissues surrounding 65 cemented and 36 cementless total joint replacements undergoing revision were characterised for cell types by immunohistochemistry and for cytokine expression by in situ hybridisation.

We identified three distinct groups of revised implants: loose implants with ballooning radiological osteolysis, loose implants without osteolysis, and well-fixed implants. In the cemented series, osteolysis was associated with increased numbers of macrophages (p = 0.0006), T-lymphocyte subgroups (p = 0.03) and IL-1 (p = 0.02) and IL-6 (p = 0.0001) expression, and in the cementless series with increased numbers of T-lymphocyte subgroups (p = 0.005) and increased TNFα expression (p = 0.04). For cemented implants, the histological, histochemical and cytokine profiles of the interface correlated with the clinical and radiological grade of loosening and osteolysis.

Our findings suggest that there are different biological mechanisms of loosening and osteolysis for cemented and cementless implants. T-lymphocyte modulation of macrophage function may be an important interaction at prosthetic interfaces.

Haemophilic arthropathy is characterised by iron deposits in synovial tissues. We investigated the suggestion that iron plays an important role in synovial changes.

We obtained synovial tissue from six patients with haemophilia during arthroplasty, finding that brown haemosideritic tissue was often adjacent to tissue with a macroscopically normal appearance in the same joint. Samples from both types of synovial tissue were analysed histologically and biochemically to determine catabolic activity. Macroscopically haemosideritic synovium showed a significantly higher inflammatory activity than that with a normal appearance. Cultures of abnormal synovial tissue gave a significantly enhanced production of IL-1, IL-6 and TNFα compared with cultures of synovial tissue with a normal appearance. In addition, the supernatant fluids from the cultures showed greater catabolic activity from haemosideritic tissue, as determined by the inhibition of the synthesis of articular cartilage matrix.

We conclude that in patients with haemophilic arthropathy, local synovial iron deposits are associated with increased catabolic activity.

We have developed a new drug delivery system using porous apatite-wollastonite glass ceramic (A-W GC) to treat osteomyelitis. A-W GC (porosity, 70% and 20% to 30%), or porous hydroxyapatite (HA) blocks (porosity 35% to 48%) used as controls, were soaked in mixtures of two antibiotics, isepamicin sulphate (ISP) and cefmetazole (CMZ) under high vacuum.

We evaluated the release concentrations of the antibiotics from the blocks. The bactericidal concentration of ISP from A-W GC was maintained for more than 42 days, but that from HA decreased to below the detection limit after 28 days. The concentrations of CMZ from both materials were lower than those of ISP. An in vivo study using rabbit femora showed that an osseous concentration of ISP was maintained at eight weeks after implantation. Osteoconduction of the A-W GC block was good.

Four patients with infected hip arthroplasties and one with osteomyelitis of the tibia have been treated with the new delivery system with excellent results.

We aimed to determine whether extracorporeal shock waves of varying intensity would damage the intact tendo Achillis and paratenon in a rabbit model. We used 42 female New Zealand white rabbits randomly divided into four groups as follows: group a received 1000 shock-wave impulses of an energy flux density of 0.08 mJ/mm², group b 1000 impulses of 0.28 mJ/mm², group c 1000 impulses of 0.60 mJ/mm², and group d was a control group.

Sonographic and histological evaluation showed no changes in group a, and transient swelling of the tendon with a minor inflammatory reaction in group b. Group c had formation of paratendinous fluid with a significant increase in the anteroposterior diameter of the tendon. In this group there were marked histological changes with increased eosin staining, fibrinoid necrosis, fibrosis in the paratenon and infiltration of inflammatory cells.

We conclude that there are dose-dependent changes in the tendon and paratenon after extracorporeal shock-wave therapy and that energy flux densities of over 0.28 mJ/mm² should not be used clinically in the treatment of tendon disorders.

We studied peridural fibrosis in 16 dogs after laminectomies at the L2, L4 and L6 levels. They received either a free fat graft, a biodegradable mechanical barrier (polyethylene oxide (PEO)/polybutylene terephthalate (PBT) copolymer), or no treatment. The animals were killed after 4, 12, 26 and 52 weeks.

Histomorphometry showed extensive and consistent peridural fibrosis in control and PEO/PBT groups. Fat grafts produced significantly less fibrous tissue, but the presence of the fat graft in the bony defect prevented closure. Degradation of the PEO/PBT barrier resulted in the formation of more fibrous tissue.

We conclude that up to one year a free fat graft is effective in reducing the amount of peridural scarring.

Ultra-high-molecular-weight polyethylene (UHMWPE) components for total joint replacement generate wear particles which cause adverse biological tissue reactions leading to osteolysis and loosening. Sterilisation of UHMWPE components by gamma irradiation in air causes chain scissions which initiate a long-term oxidative process that degrades the chemical and mechanical properties of the polyethylene. Using a tri-pin-on-disc tribometer we studied the effect of ageing for ten years after gamma irradiation in air on the volumetric wear, particle size distribution and the number of particles produced by UHMWPE when sliding against a stainless-steel counterface.

The aged and irradiated material produced six times more volumetric wear and 34 times more wear particles per unit load per unit sliding distance than non-sterilised UHMWPE. Our findings indicate that oxidative degradation of polyethylene after gamma irradiation in air with ageing produces more wear.

We performed a biomechanical study on human cadaver spines to determine the effect of three different interbody cage designs, with and without posterior instrumentation, on the three-dimensional flexibility of the spine. Six lumbar functional spinal units for each cage type were subjected to multidirectional flexibility testing in four different configurations: intact, with interbody cages from a posterior approach, with additional posterior instrumentation, and with cross-bracing. The tests involved the application of flexion and extension, bilateral axial rotation and bilateral lateral bending pure moments. The relative movements between the vertebrae were recorded by an optoelectronic camera system.

We found no significant difference in the stabilising potential of the three cage designs. The cages used alone significantly decreased the intervertebral movement in flexion and lateral bending, but no stabilisation was achieved in either extension or axial rotation. For all types of cage, the greatest stabilisation in flexion and extension and lateral bending was achieved by the addition of posterior transpedicular instrumentation. The addition of cross-bracing to the posterior instrumentation had a stabilising effect on axial rotation. The bone density of the adjacent vertebral bodies was a significant factor for stabilisation in flexion and extension and in lateral bending.

Previous studies of the Ilizarov procedure have concentrated on musculoskeletal assessments rather than the opinions of patients. In a prospective trial of 25 consecutive patients, we evaluated the effect of Ilizarov reconstruction of post-traumatic deformity on general health status using the SF36 and Nottingham Health Profile (NHP).

The patients had very low preoperative scores, which remained low during treatment and correction, but increased postoperatively. The mean overall SF36 score improved from 36 ± 3 to 58 ± 7 (p = 0.031) and the NHP score from 39 ± 11 to 67 ± 10 (p = 0.002). The improvements in scores were not limited to the physical components and were equal or better than the improvements reported for other orthopaedic procedures, including total joint arthroplasty.

Ilizarov-type reconstruction of deformity of the lower limb not only restores bony configuration, but also produces a large improvement in the general health status of patients.

We used radiostereometric analysis to compare wear rates between uncemented porous sockets and cemented all-polyethylene sockets in a series of 102 hips randomised for either a Harris-Galante or a Charnley cup. Wear was evaluated in 95 hips at a mean of five years (2 to 7). All hips had a cemented, 22 mm head mono-bloc Charnley stem.

The mean annual wear rate was 0.09 mm in the Charnley sockets and 0.10 mm in the Harris-Galante sockets, with no statistically significant differences in wear, migration or rotation.

We conclude that, up to five years, the wear characteristics of the modular and porous Harris-Galante socket resemble that of the Charnley socket.

Chronic compartment syndrome (CCS) is usually considered to be due to ischaemia of muscle. We have attempted to use the direct measurement of muscle blood flow for diagnosis since the assessment of intracompartmental pressure does not provide accurate knowledge of the vascular state. We recorded simultaneously continuous measurements of the laser Doppler flow (LDF) in muscle and the intracompartment pressure (ICP) after exercise in seven patients with CCS, and in seven control subjects.

The mean ICP was 74.1 ± 4.4 mmHg in CCS patients and 24.2 ± 3.4 mmHg in control subjects one minute after exercise, decreasing to 34.6 ± 2.3 mmHg and 15.0 ± 1.6 mmHg at 20 min, respectively. The LDF was 0.80 ± 0.11 arbitrary units (AU) in control subjects and 1.09 ± 0.14 AU in CCS patients one minute after exercise, and 0.41 ± 0.11 AU and 0.27 ± 0.04 AU, respectively, at the end of the recovery period.

The ICP showed a progressive decrease over time in both groups. The LDF decreased sharply during the first minutes of recovery in control subjects, but in patients with CCS there was a delayed hyperaemic peak with blood flow reaching 0.84 ± 0.10 AU at nine minutes as against 0.33 ± 0 .06 AU for control subjects (p < 0.01). The ICP increased in both control subjects and CCS patients after exercise with no clear cut-off point between the groups. By contrast, changes in muscle blood flow over time were clearly different between control subjects and patients with CCS.

For this reason, LDF should be investigated further as a technique for the diagnosis of CCS.

We compared the ability of three different posterior cruciate ligament (PCL) reconstructions to restore normal anteroposterior laxity to the knee from 0 to 130° of knee flexion.

Cadaver knees were tested intact, after PCL rupture or after bone-patellar tendon-bone grafting. Grafts were performed isometrically or with a single bundle representing the anatomical anterior PCL fibre bulk (aPC) or with a double bundle that added the posterior PCL fibre bulk (pPC). The grafts were tensioned to restore normal knee laxity at 60° of flexion, except for the pPC which was tensioned at 130°.

The isometric graft led to overconstraint as the knee extended resulting in high graft tension in extension and excess laxity in flexion. The aPC graft matched normal laxity from 0 to 60° of flexion but was lax from 90 to 130° of flexion. Only the double-bundled graft could restore normal knee laxity across the full range of flexion.

Lesions within the articular cartilage layer of synovial joints do not heal spontaneously. Some repair cells may appear, but their failure to become established may be related to problems of adhesion to proteoglycan-rich surfaces. We therefore investigated whether controlled enzymatic degradation of surface proteoglycan molecules to a depth of about 1 μm, using chondroitinase ABC, would improve coverage by repair cells.

We created superficial lesions (1.0 × 0.2 × 5 mm) in the articular cartilage of mature rabbit knees and treated the surfaces with 1 U/ml of chondroitinase ABC for four minutes. The defects were studied by histomorphometry and electron microscopy at one, three and six months.

At one month, untreated lesions were covered to a mean extent of 28% by repair cells; this was enhanced to a mean of 53% after enzyme treatment. By three months, the mean coverage of both control and chondroitinase-ABC-treated defects had diminished dramatically to 0.2% and 13%, respectively, but at six months both untreated and treated lesions had a similar coverage of about 30%, not significantly different from that achieved in untreated knees at one month.

These findings suggest that, with time, chondrocytes near the surface of the defect may compensate for the loss of proteoglycans produced by enzyme treatment, thereby restoring the inhibitory properties of the matrix as regards cell adhesion. This supposition was confirmed by electron microscopy. Our results have an important bearing on attempts made to induce healing responses by transplanting chondrogenic cells or by applying growth factors.

We obtained simultaneous measurements of sagittal knee laxity in 12 consecutive patients after reconstruction of the anterior cruciate ligament (ACL), using the Stryker laxity tester and radiostereometric analysis (RSA).

The mean anteroposterior (AP) displacement when a 90 N load was applied in both directions was 5.3 ± 2.7 mm with RSA and 9.8 ± 1.6 mm with the external device (p < 0.001). The corresponding measurements at a load of 180 N were 5.7 ± 2.4 mm and 13.8 ± 3.7 mm, respectively (p < 0.001).

More than 50% of the sagittal knee movement, as measured by the external device at a load of 180 N, was not true femorotibial displacement of the joint but was due to soft-tissue deformation.

We studied the presence of sensory nerves by immunohistochemistry in the interface membranes of hip prostheses after aseptic loosening. Substance P (SP), neurokinin A (NKA) and calcitonin gene-related peptide (CGRP) were analysed as was protein gene product (PGP) 9.5, a general marker for nerve fibres.

We identified nerve fibres in all samples but differences in their density were found. SP- and NKA-positive fibres were predominantly non-vascular, forming varicose nerve terminals. CGRP-immunoreactive nerve fibres with varicose terminals were seen mostly close to blood vessels, but also as free nerve endings.

Sensory neuropeptides participate not only in nociception but also stimulate immune cells to release cytokines. The presence of sensory nerves in the interface membrane may reflect a pathophysiological response contributing to the aseptic loosening of hip prostheses.

To study the effect of ligament injuries and surgical repair we investigated the three-dimensional kinematics of the ankle joint complex and the talocrural and the subtalar joints in seven fresh-frozen lower legs before and after sectioning and reconstruction of the ligaments. A foot movement simulator produced controlled torque in one plane of movement while allowing unconstrained movement in the remainder.

After testing the intact joint the measurements were repeated after simulation of ligament injuries by cutting the anterior talofibular and calcaneofibular ligaments. The tests were repeated after the Evans, Watson-Jones and Chrisman-Snook tenodeses. The range of movement (ROM) was measured using two goniometer systems which determined the relative movement between the tibia and talus (talocrural ROM) and between the talus and calcaneus (subtalar ROM).

Ligament lesions led to increased inversion and internal rotation, predominantly in the talocrural joint. The reconstruction procedures reduced the movement in the ankle joint complex by reducing subtalar movement to a non-physiological level but did not correct the instability of the talocrural joint.

We examined the roles of methylmethacrylate (MMA) monomer and cementing technique in the formation, and haemodynamic outcome, of pulmonary fat emboli. The preparation of the femoral canal and the cementing technique were studied in four groups of adult dogs as follows: control (no preparation); lavage; cement pressurisation; and cement pressurisation after lavage. We measured the intramedullary pressure, pulmonary artery pressure (PAP), pulmonary capillary wedge pressure and bilateral femoral vein levels of triglyceride, cholesterol and MMA monomer at rest and after reaming, lavage, and cementing.

Femoral vein triglyceride and cholesterol levels did not vary significantly from resting levels despite significant elevations in intramedullary pressure with reaming, lavage and cementing (p = 0.001). PAP was seen to rise significantly with reaming (p = 0.0038), lavage (p = 0.0031), cementing (p = 0.0024) and cementing after lavage (p = 0.0028) while the pulmonary capillary wedge pressure remained unchanged.

MMA monomer was detected in femoral vein samples when cement pressurisation was used. Intramedullary lavage before cementing had no significant effect on the MMA level. Haemodynamic evidence of pulmonary embolism was noted with reaming and intramedullary canal preparation, irrespective of the presence of MMA monomer. We found no relationship between MMA monomer level and intramedullary pressure, PAP or pulmonary capillary wedge pressure.

Our findings suggest that the presence of MMA monomer in femoral venous blood has no effect on the formation of fat emboli or their pulmonary haemodynamic outcome during cemented hip arthroplasty.

Bone loss around replacement prostheses may be related to the activation of mononuclear phagocytes (MNP) by prosthetic wear particles. We investigated how osteoblast-like cells were regulated by human MNP stimulated by particles of prosthetic material.

Particles of titanium-6-aluminium-4-vanadium (TiAlV) stimulated MNP to release interleukin (IL)-1β, tumour necrosis factor (TNF)α, IL-6 and prostaglandin E₂ (PGE₂). All these mediators are implicated in regulating bone metabolism. Particle-activated MNP inhibited bone cell proliferation and stimulated release of IL-6 and PGE₂. The number of cells expressing alkaline phosphatase, a marker associated with mature osteo-blastic cells, was reduced. Experiments with blocking antibodies showed that TNFα was responsible for the reduction in proliferation and the numbers of cells expressing alkaline phosphatase. By contrast, IL-1β stimulated cell proliferation and differentiation. Both IL-1β and TNFα stimulated IL-6 and PGE₂release from the osteoblast-like cells.

Our results suggest that particle-activated mono-nuclear phagocytes can induce a change in the balance between bone formation and resorption by a number of mechanisms.

We used fresh small-fragment osteochondral allografts to reconstruct post-traumatic osteochondral defects in 126 knees of 123 patients with a mean age of 35 years. At a mean follow-up of 7.5 years (2 to 20), 108 knees were rated as successful (85%) and 18 had failed (15%).

The factors related to failure included age over 50 years (p = 0.008), bipolar defects (p < 0.05), malaligned knees with overstressing of the grafts, and workers’ compensation cases (p < 0.04). Collapse of the graft by more than 3 mm and of the joint space of more than 50% were seen more frequently in radiographs of failed grafts.

Our encouraging clinical results for fresh small-fragment osteochondral allografts show that they are indicated for unipolar post-traumatic osteochondral defects of the knee in young active patients.

We have studied damage to the tibial articular surface after replacement of the femoral surface in dogs. We inserted pairs of implants made of alumina, titanium and polyvinyl alcohol (PVA) hydrogel on titanium fibre mesh into the femoral condyles.

The two hard materials caused marked pathological changes in the articular cartilage and menisci, but the hydrogel composite replacement caused minimal damage. The composite osteochondral device became rapidly attached to host bone by ingrowth into the supporting mesh.

We discuss the clinical implications of the possible use of this material in articular resurfacing and joint replacement.

We assessed peripheral nerve function during and after lower-limb lengthening by callotasis in 14 patients with short stature, using motor conduction studies.

Four patients with short stature of varying aetiology showed unilateral and one showed bilateral weakness of foot dorsiflexion. Both clinical and electrophysiological abnormalities consistent with involvement of the peroneal nerve were observed early after starting tibial callotasis. There was some progressive electro-physiological improvement despite continued bone distraction, but two patients with Turner’s syndrome had incomplete recovery. A greater percentage increase in tibial length did not correspond to a higher rate of peroneal nerve palsy. The function of the posterior leg muscles and the conduction velocity of the posterior tibial nerve were normal throughout the monitoring period.

The F-wave response showed a longer latency at the end of the bone distraction than in basal conditions; this is probably related to the slowing of conduction throughout the entire length of the nerve.

We have used an experimental model employing the bent tail of rats to investigate the effects of mechanical forces on bones and joints. Mechanical strain could be applied to the bones and joints of the tail without direct surgical exposure or the application of pins and wires.

The intervertebral disc showed stretched annular lamellae on the convex side, while the annulus fibrosus on the concave side was pinched between the inner corners of the vertebral epiphysis. In young rats with an active growth plate, a transverse fissure appeared at the level of the hypertrophic cell layer or the primary metaphyseal trabecular zone. Metaphyseal and epiphyseal trabeculae on the compressed side were thicker and more dense than those of the distracted part of the vertebra.

In growing animals, morphometric analysis of hemiepiphyseal and hemimetaphyseal areas, and the corresponding trabecular bone density, showed significant differences between the compressed and distracted sides. No differences were observed in adult rats. We found no significant differences in osteoclast number between compressed and distracted sides in either age group. Our results provide quantitative evidence of the working of ‘Wolff’s law’.

The differences in trabecular density are examples of remodelling by osteoclasts and osteoblasts; our finding of no significant difference in osteoclast numbers between the hemiepiphyses in the experimental and control groups suggests that the response of living bone to altered strain is mediated by osteoblasts.

Corrective osteotomies are often planned and performed on the basis of normal anatomical proportions. We have evaluated the length and torsion of the segments of the lower limb in normal individuals, to analyse the differences between left and right sides, and to provide tolerance figures for both length and torsion.

We used CT on 355 adult patients and measured length and torsion by the Ulm method. We excluded all patients with evidence of trauma, infection, tumour or any congenital disorder.

The mean length of 511 femora was 46.3 ± 6.4 cm (±2sd) and of 513 tibiae 36.9 ± 5.6 cm; the mean total length of 378 lower limbs was 83.2 ± 11.4 cm with a tibiofemoral ratio of 1 to 1.26 ± 0.1. The 99th percentile level for length difference in 178 paired femora was 1.2 cm, in 171 paired tibiae 1.0 cm and in 60 paired lower limbs 1.4 cm.

In 505 femora the mean internal torsion was 24.1 ± 17.4°, and in 504 tibiae the mean external torsion was 34.9 ± 15.9°. For 352 lower limbs the mean external torsion was 9.8 ± 11.4°. The mean torsion angle of right and left femora in individuals did not differ significantly, but mean tibial torsion showed a significant difference between right (36.46° of external torsion) and left sides (33.07° of external torsion). For the whole legs torsion on the left was 7.5 ± 18.2° and 11.8 ± 18.8°, respectively (p < 0.001). There was a trend to greater internal torsion in femora in association with an increased external torsion in tibiae, but we found no correlation. The 99th percentile value for the difference in 172 paired femora was 13°; in 176 pairs of tibiae it was 14.3° and for 60 paired lower limbs 15.6°.

These results will help to plan corrective osteotomies in the lower limbs, and we have re-evaluated the mathematical limits of differences in length and torsion.

We tested in compression specimens of human proximal tibial trabecular bone from 31 normal donors aged from 16 to 83 years and determined the mechanical properties, density and mineral and collagen content.

Young’s modulus and ultimate stress were highest between 40 and 50 years, whereas ultimate strain and failure energy showed maxima at younger ages. These age-related variations (except for failure energy) were non-linear.

Tissue density and mineral concentration were constant throughout life, whereas apparent density (the amount of bone) varied with ultimate stress. Collagen density (the amount of collagen) varied with failure energy. Collagen concentration was maximal at younger ages but varied little with age.

Our results suggest that the decrease in mechanical properties of trabecular bone such as Young’s modulus and ultimate stress is mainly a consequence of the loss of trabecular bone substance, rather than a decrease in the quality of the substance itself. Linear regression analysis showed that collagen density was consistently the single best predictor of failure energy, and collagen concentration was the only predictor of ultimate strain.

We examined the cellular responses to various particles injected into the knees and the intramedullary femoral cavities of rats in the presence of polymethyl-methacrylate (PMMA) plugs.

The intra-articular particles were mainly ingested by synovial fibroblasts. Increased numbers of macrophages were not detected and there was only a slight increase in synovial thickness.

Cellular responses in the intramedullary space were similarly mild and bone resorption around the PMMA plug did not occur. Bone formation was inhibited only by polyethylene particles.

In contrast to current views, our study shows that wear particles per se do not initiate bone resorption.

We developed a rat model of limb lengthening to study the basic mechanism of distraction osteogenesis, using a small monolateral external fixator. In 11-week-old male rats we performed a subperiosteal osteotomy in the midshaft of the femur with distraction at 0.25 mm every 12 hours from seven days after operation.

Radiological and histological examinations showed a growth zone of constant thickness in the middle of the lengthened segment, with formation of new bone at its proximal and distal ends.

Osteogenic cells were arranged longitudinally along the tension vector showing the origin and the fate of individual cells in a single section. Typical endochondral bone formation was prominent in the early stage of distraction, but intramembraneous bone formation became the predominant mechanism of ossification at later stages. We also showed a third mechanism of ossification, ‘transchondroid bone formation’. Chondroid bone, a tissue intermediate between bone and cartilage, was formed directly by chondrocyte-like cells, with transition from fibrous tissue to bone occurring gradually and consecutively without capillary invasion.

In situ hybridisation using digoxigenin-11-UTP-labelled complementary RNAs showed that the chondroid bone cells temporarily expressed type-II collagen mRNA. They did not show the classical morphological characteristics of chondrocytes, but were assumed to be young chondrocytes undergoing further differentiation into bone-forming cells.

We found at least three different modes of ossification during bone lengthening by distraction osteogenesis. We believe that this is the first report of such a rat model, and have shown the validity of in situ hybridisation techniques for the study of the cellular and molecular mechanisms involved in distraction osteogenesis.

We have studied the three-dimensional geometry of the proximal humerus on human cadaver specimens using a digitised measuring device linked to a computer. Our findings demonstrated the variable shape of the proximal humerus as well as its variable dimensions. The articular surface, which is part of a sphere varies individually in its orientation as regards inclination and retroversion, and it has variable medial and posterior offsets.

These variations cannot be accommodated by the designs of most contemporary humeral components. Although good clinical results can be achieved with current modular and non-modular components their relatively fixed geometry prevents truly anatomical restoration in many cases.

To try to restore the original three-dimensional geometry of the proximal humerus, we have developed a new type of humeral component which is modular and adaptable to the individual anatomy. Such adaptability allows correct positioning of the prosthetic head in relation to an individual anatomical neck, after removal of the marginal osteophytes. The design of this third-generation prosthesis respects the four geometrical variations which have been demonstrated in the present study. These are inclination, retroversion, medial offset and posterior offset.

Transarticular screws at the C1 to C2 level of the cervical spine provide rigid fixation, but there is a danger of injury to a vertebral artery. The risk is related to the technical skill of the surgeon and to variations in local anatomy.

We studied the grooves for the vertebral artery in 50 dry specimens of the second cervical vertebra (C2). They were often asymmetrical, and in 11 specimens one of the grooves was deep enough to reduce the internal height of the lateral mass at the point of fixation to ≤2.1 mm, and the width of the pedicle on the inferior surface of C2 to ≤2 mm. In such specimens, the placement of a transarticular screw would put the vertebral artery at extreme risk, and there is not enough bone to allow adequate fixation.

Before any decision is made concerning the type of fixation to be used at C2 we recommend that a thin CT section be made at the appropriate angle to show both the depth and any asymmetry of the grooves for the vertebral artery.

We have developed a novel, two-layered, collagen matrix seeded with chondrocytes for repair of articular cartilage. It consists of a dense collagen layer which is in contact with bone and a porous matrix to support the seeded chondrocytes. The matrices were implanted in rabbit femoral trochleas for up to 24 weeks. The control groups received either a matrix without cells or no implant.

The best histological repair was seen with cell-seeded implants. The permeability and glycosaminoglycan content of both implant groups were nearly normal, but were significantly less in tissue from empty defects. The type-II collagen content of the seeded implants was normal. For unseeded implants it was 74.3% of the normal and for empty defects only 20%. The current treatments for articular injury often result in a fibrous repair which deteriorates with time. This bilayer implant allowed sustained hyaline-like repair of articular defects during the entire six-month period of observation.

We evaluated the effects of a serum protein coating on prosthetic infection in 29 adult male rabbits divided into three groups: control, albumin-coated and uncoated. We used 34 grit-blasted, commercially pure titanium implants. Eleven were coated with cross-linked albumin. All the implants were exposed to a suspension of Staphylococcus epidermidis before implantation.

Our findings showed that albumin-coated implants had a much lower infection rate (27%) than the uncoated implants (62%). This may be a useful method of reducing the infection of prostheses.

Abundant implant-derived biomaterial wear particles are generated in aseptic loosening and are deposited in periprosthetic tissues in which they are phagocytosed by mononuclear and multinucleated macrophage-like cells. It has been stated that the multinucleated cells which contain wear particles are not bone-resorbing osteoclasts. To investigate the validity of this claim we isolated human osteoclasts from giant-cell tumours of bone and rat osteoclasts from long bones. These were cultured on glass coverslips and on cortical bone slices in the presence of particles of latex, PMMA and titanium. Osteoclast phagocytosis of these particle types was shown by light microscopy, energy-dispersive X-ray analysis and SEM. Giant cells containing phagocytosed particles were seen to be associated with the formation of resorption lacunae. Osteoclasts containing particles were also calcitonin-receptor-positive and showed an inhibitory response to calcitonin.

Our findings demonstrate that osteoclasts are capable of phagocytosing particles of a wide range of size, including particles of polymeric and metallic bio-materials found in periprosthetic tissues, and that after particle phagocytosis, they remain fully functional, hormone-responsive, bone-resorbing cells.

Our aim was to analyse the influence of the size, shape and number of particles on the pathogenesis of osteolysis. We obtained peri-implant tissues from 18 patients having revision surgery for aseptically loosened Freeman total knee replacements (10), Charnley total hip replacements (3) and Imperial College/London Hospital double-cup surface hip replacements (5). The size and shape of the polyethylene particles were characterised using SEM and their concentration was calculated. The results were analysed with reference to the presence of radiological osteolysis.

The concentration of polyethylene particles in 6 areas with osteolysis was significantly higher than that in 12 areas without osteolysis. There were no significant differences between the size and shape of the particles in these two groups.

We conclude that the most critical factor in the pathogenesis of osteolysis is the concentration of polyethylene particles accumulated in the tissue.

We studied the origin of the anterior deltoid from the lateral third of the clavicle and the leading anterior edge of the acromion in 18 cadaver shoulders by anatomical and histological methods.

The main origin of the deltoid was from the superior surface of the anterior acromion, but muscle and tendinous attachments were also seen on the entire anterior surface of the acromion, its anteroinferior surface and on the whole width of the anterior surface of the clavicle.

Mock arthroscopic acromioplasty was shown to detach deltoid fibres from the anterior surfaces, leaving the superior attachment in continuity. Potentially, arthroscopic subacromial and clavicular resection can detach deltoid fibres originating from the anterior and anteroinferior surfaces of the acromion and clavicle and thus weaken the anterior deltoid.

From a prospective, cross-sectional survey of 402 patients who had a total hip (THR) or a total knee (TKR) replacement for idiopathic osteoarthritis (OA) at a major centre, we determined the prevalence of these replacements for idiopathic OA in their 1171 siblings and 376 spouses. Using spouses as controls, the relative risk of THR in siblings was 1.86 (95% CI 0.93 to 3.69). The relative risk for TKR in siblings v spouses was 4.8 (95% CI 0.64 to 36.4) whereas the risk for the combined outcome measure of THR or TKR was 2.32 (95% CI 1.22 to 4.43) when siblings and spouses over 64 years of age were compared. Using a threshold liability model (Falconer), the heritability of end-stage OA of the hip was estimated at 27%.

The increased risks of joint replacement for severe, idiopathic OA which we found in siblings suggest that genetic influences are important in end-stage OA of the hip and knee.

In a prospective study of 14 patients undergoing total hip replacement we have used dual-energy X-ray absorptiometry (DEXA) to investigate remodelling of the bone around two different designs of cementless femoral prosthesis. The bone mineral density (BMD) was measured at 12-weekly intervals for a year. Eight patients (group A) had a stiff, collarless implant and six (group B) a flexible isoelastic implant.

Patients in group A showed a decrease in BMD from 14 weeks after operation. By 12 months, the mean loss in BMD was 27%, both medially and laterally to the proximal part of the implant. Those in group B showed an overall increase in BMD which reached a mean of 12.6% on the lateral side of the distal portion of the implant.

Our results support the current concepts of the effects of stem stiffness and flexibility on periprosthetic remodelling.

Spinal nerve roots often sustain compression injuries. We used a Wistar rat model of the cauda equina syndrome to investigate such injuries. Rapid transient compression of the cauda equina was produced using a balloon catheter. The results were assessed by daily neurological examination and somatosensory evoked potential (SEP) recording before surgery and ten weeks after decompression.

Compression of the spinal nerves induced changes in the SEP which persisted for up to ten weeks after decompression, but it had no effect on the final neurological outcome. Our study shows the importance of early surgical decompression for cauda equina syndrome.

The clinical use of hydroxyapatite (HA) coating is controversial especially in regard to the long-term performance of the coating and the effects of resorption. In each of 15 consenting patients we inserted two implants, coated with either HA or fluorapatite (FA) into the iliac crest. They were harvested at a mean of 13.6 ± 0.6 months after surgery.

Histological examination showed that bone ongrowth on the HA-coated implants was significantly greater (29%) than that on the FA-coated implants. When bone was present on the coating surface the HA coating was significantly thicker than the FA coating. When bone marrow was present, the HA coating was significantly thinner than the FA coating. The reduction in coating thickness when covered by bone or bone marrow was 23.1 ± 9.7 μm for HA and 5.1 ± 1.7 μm for FA (p < 0.01) suggesting that FA is more stable than HA against resorption by bone marrow.

The findings suggest that in man the osteoconductive properties of HA coating are superior to those of FA. Resorption rates for both coatings were approximately 20% of the coating thickness per year. Bone ongrowth appears to protect against resorption whereas bone marrow seems to accelerate resorption. No adverse reaction was seen in the surrounding bone.

It has been suggested that the endurance of cemented femoral reconstructions in total hip arthroplasty is affected by the creep of acrylic cement, but it is not known to what extent cement creeps under loading conditions in vivo, or how this affects load transfer. We have simulated the long-term creep properties of acrylic cement in finite-element models of femoral stem constructs and analysed their effects. We investigated whether subsidence rates measured in vivo could be explained by creep of acrylic cement, and if polished, unbonded, stems accommodated creep better than bonded stems.

Our findings showed that polished prostheses subsided only about 50 μm as a result of cement creep. The long-term prosthetic subsidence rates caused by creep of acrylic cement are therefore very small and do not explain the excessive migration rates which have sometimes been reported. Cement creep did, however, relax cement stresses and create a more favourable stress distribution at the interfaces. These trends were found around both the bonded and unbonded stems. Our results did not confirm that polished, unbonded, stems accommodated creep better than bonded stems in terms of cement and interface stress patterns.