We have reviewed 29 patients (30 hips) who had undergone revision total hip arthroplasty using a Freeman metal-backed acetabular component and acetabular impaction allografting. The mean follow-up was for 15.3 years (12 to 17).

Five patients (5 hips) died with the prosthesis in situ and four (4 hips) were lost to follow-up. Twelve hips had failed and in the remaining nine there were minor symptoms. The mean time to failure requiring further surgery was nine years. Excluding patients who were lost to follow-up or had died, 72% of the hips were radiologically loose at the last review. The commonest pattern in those requiring revision was failure of the reinforcement ring in adduction with remodelling of the medial wall.

Of the nine patients who had not undergone revision, one with bilateral replacements had no current radiographs and only three of the remaining seven replacements had no radiological signs of loosening.

The short-term results for this technique have been reported to be satisfactory, but in the long term they are not. The factors associated with failure include the design of the prosthesis, which has been implicated in disappointing long-term results when used in primary arthroplasty, but not with the frequency of failure found in this series. It seems that the reliance on peripheral screw fixation over a bed of allograft without bridging the graft does not provide sufficient stability to allow incorporation of the graft.

We measured the proximal migration of 265 acetabular cups over seven years and correlated the findings with clinical outcome and acetabular revision for aseptic loosening. Cups which eventually became aseptically loose were shown to migrate more rapidly than successful cups. The average proximal migration at two years postoperatively for four groups of cups showed a monotonic relationship to the acetabular revision rate for aseptic loosening at 6.5 years. We conclude that acetabular cups which develop aseptic loosening as evidenced by pain, revision or screw fracture show increased proximal migration by one year, and that the 'migration rate' at two years can be used to predict the acetabular revision rate from aseptic loosening at 6.5 years.

We studied the wear generated by motion between polished and shot-blasted titanium-alloy (Ti-6Al-4V) or cobalt-chrome alloy (Co-Cr) surfaces and cortical bone in vitro. Semicircular sections of human proximal femoral cortex were reamed to fit metal cylinders of each alloy. The cylinders were then fitted in the bone, loaded and rotated in physiological saline. Ti-alloy resulted in more wear both of the bone and of the metal than did Co-Cr alloy. Metal wear was reduced and bone wear was increased by shot-blasting, a procedure which introduces surface residual stresses and roughens the metal surface. We conclude that when there is gross motion between a metal implant and bone, Ti-alloy is likely to generate more wear debris than Co-Cr alloy. The least wear both of bone and of metal was produced by polished Co-Cr.

The vertical migration of four configurations of a proximal femoral prosthesis, followed for up to nine years, was measured on standard radiographs. The same implant was used without cement (group 1) and with cement (group 2). The migration of both groups was linear from six months onwards. The mean migration rate and the incidence of late aseptic loosening were both greater in group 1. Survival analysis of the two groups, however, showed no statistically significant difference. In both groups, hips later destined for revision migrated more rapidly from the initial postoperative period onwards, than did the remainder. A threshold migration of 1.2 mm/year during the first two years after implantation detected hips likely to fail with a specificity of 86% and a sensitivity of 78%. This 'migration test' was applied to the results in two further groups of patients in which a modified femoral prosthesis had been implanted without hydroxyapatite coating (group 3) and with hydroxyapatite coating (group 4). The test distinguished between the four groups and suggested that at least two fixation procedures should be abandoned. We conclude that vertical migration measured on standard radiographs in the first two years after implantation can be used to predict late aseptic loosening. New prosthetic configurations should be evaluated by migration measurements before their general release. Our observations support the view that one cause of late aseptic loosening is imperfect initial fixation.

We have studied 27 tibial prostheses retrieved from knee replacements after 1 to 9 years. In 22 the femoral components were of cobalt-chrome, in five polyacetal. The design of the components gave a nominal contact area of 320 mm2 on each condyle. The tibial component was of high-density polyethylene (HDP) at least 6 mm thick, and not heat-treated. In the metal/HDP prostheses the average wear rate was 0.025 mm/year. The relative wear on the medial and lateral sides was related to the leg axis. None of the retrieved prostheses showed any severe disruption of their surface. The polyacetal/HDP prostheses showed similar wear with a statistically insignificant trend towards slower penetration. We conclude that the rate of wear of HDP in a conforming tibiofemoral bearing with a fixed tibial component at least 6 mm thick and not heat-treated is slow enough to be safe in clinical practice.

We studied two groups of femoral hip prostheses: 43 TiAlV ridged press-fit stems, and 26 with similar stems coated with hydroxyapatite on the proximal half. At one year, radiological measurement showed a mean downward migration of 0.99 mm for the TialV prostheses and 0.12 mm for the HA-coated prostheses (p = 0.0002). Hydroxyapatite coating appeared to provide effective bio-active supplementary fixation.

We studied the effect of a layer of cement placed under the tibial component of Freeman-Samuelson total knee prostheses with a metal back and an 80 mm intramedullary stem, using roentgen stereophotogrammetry to measure the migration of the tibial component during one year in 13 uncemented and 16 cemented knees. The addition of cement produced a significant reduction in migration at one year, from a mean of 1.5 mm to one of 0.5 mm (p less than 0.01), including a significant reduction in pure subsidence. One year postoperatively the clinical results were similar between the groups, but, at three years, one uncemented knee had required revision.

We determined the cumulative survival rates, in arthroplasty of the knee, of three designs of tibial component, using a change of position on standard radiographs or revision for aseptic loosening as criteria of failure. The average migration of each of the three designs in the first postoperative year is known from roentgen stereophotogrammetric analysis reported by other authors. The ranking order of the components as judged by cumulative survival is the same as that determined by early migration. This finding supports the view that the measurement of early migration can predict late aseptic loosening and therefore that such measurements are clinically of value.

One-stage reimplantation for the salvage of infected total knee arthroplasty in 18 patients was reviewed at an average follow-up of five years. There had been one recurrence and one new infection, both in rheumatoid patients with another focus of infection. In four other patients the clinical result was impaired by pain after walking (2) and limited flexion (2). Our results suggest that one-stage reimplantation is a reasonably reliable procedure for the management of a loose infected prosthesis.

Tissues from five patients who underwent revision operations for failed total hip replacements were found to contain large quantities of particulate titanium. In four cases this metal must have come from titanium alloy screws used to fix the acetabular component; in the fifth case it may also have originated from a titanium alloy femoral head. Monoclonal antibody labelling showed abundant macrophages and T-lymphocytes, in the absence of B-lymphocytes, suggesting sensitisation to titanium. Skin patch testing with dilute solutions of titanium salts gave negative results in all five patients. However, two of them had a positive skin test to a titanium-containing ointment.

We studied the effect of a metal tray with an intramedullary stem on the micromotion of the tibial component in total knee arthroplasty. Of 32 uncemented Freeman-Samuelson knee arthroplasties performed in London and Gothenburg, nine had a metal backing and stem added to the tibial component. Micromotion of the tibial components, expressed as migration and inducible displacement, was analysed using roentgen stereophotogrammetric analysis up to two years follow-up. The addition of a metal back and a 110 mm stem to the standard polyethylene component significantly reduced both migration over two years and inducible displacement.

Individual components of a total hip replacement are difficult to evaluate and quantify. We have studied the assessment of the acetabular component, and conclude that the measurement of migration allows the comparison of implants, although there is no established link between migration and significant loosening. A method of measurement based on clinical radiographs has been developed, and its limitations estimated. The accuracy of the technique was calculated to be +/- 3 mm.

Torsional instability of femoral components has not received much attention, and is difficult to detect in conventional radiographs. To test this we designed a system to apply a load in an anteroposterior direction to the head of a femoral component, implanted into a cadaveric femur. Rotation within the bone was measured, using a purpose built transducer, with and without preservation of the neck, with and without cement, and with longitudinal ridges but no cement. The results show that torsional instability may be a problem in uncemented replacement. Preservation of the femoral neck and the use of a ridged prosthesis increases resistance to rotation. Rotational movements occurring in vivo during such activities as climbing stairs and rising from the seated position may contribute to mechanical loosening.

Thirty-seven patients with extensive acetabular defects due to loose implants had revisions with uncemented components, the acetabulum being augmented with homograft bone. In six of these, a histological study of graft incorporation was made. At a mean follow-up of 1.5 years 34 patients were free of pain and 35 could walk for 30 minutes or longer. No graft had obviously sequestrated. Two components had radiological evidence of migration but remain asymptomatic. We conclude that cementless revision surgery with homograft supplementation of the acetabulum is clinically successful in the short-term. The long-term outcome is unknown.

Mechanical and biomechanical testing of a new bone cement suggests that improved load transfer to the proximal femur could be achieved with the combination of a cement having a lower modulus, a greater ductility and a lower creep resistance than polymethylmethacrylate and a suitably shaped femoral component.

A review of patients with an infected resurfacing prosthesis is presented. Eight patients with a loose infected prosthesis were treated by a one-stage exchange arthroplasty; six others with a well-fixed infected prosthesis were treated by drainage and antibiotics. All eight treated by exchange arthroplasty remained free of infection as did five of those treated by drainage. In four of these last five patients, the prosthesis was inserted without cement; the possible role of polymethylmethacrylate in the persistence of infection is discussed.

We report a retrospective study of 62 total ankle arthroplasties performed between 1972 and 1981. Forty-one of these have been reviewed clinically after an average follow-up of five and a half years; only 13 can be described as satisfactory. The complications encountered in all 62 arthroplasties are detailed, the most significant being superficial wound healing problems, talar collapse, and loosening of the components; 13 prosthetic joints have already been removed and arthrodesis attempted. The management of the complications is discussed. In view of the high complication rate and the generally poor long-term clinical results, we recommend arthrodesis as the treatment of choice for the painful stiff arthritic ankle, regardless of the underlying pathological process.

Fifty-three failed knee replacements were revised using minimally constrained implants with smooth uncemented intramedullary stems and metal-backed tibial components. Polymethylmethacrylate was used only to replace lost bone near the surface of the implant. Excluding four knees which had serious postoperative complications, 91% had successful relief of pain, 84% had over 90 degrees of movement and 80% could walk for more than 30 minutes. Review of the radiographs showed that there were no progressive lucencies at the interface between bone and cement, and no subsidence of components or changes in alignment. At the uncemented stem-to-bone interface, thin white lines developed near the metal, and their significance is discussed. This revision technique is an effective treatment for aseptic failure of primary total knee arthroplasty.

The results of resurfacing or not resurfacing the articular cartilage of the patella were reviewed in 71 knees--47 with rheumatoid arthritis and 24 with osteoarthritis--which had been replaced with the ICLH prosthesis. Two rheumatoid and osteoporotic patellae developed transverse fractures with separation and were excluded from the review. The results showed that resurfacing the patella greatly reduced peripatellar pain with no counter-balancing disadvantage.

Clinical experience of the ICLH method of resurfacing the hip now spans 10 years. The first 36 arthroplasties, performed between 1972 and 1974, have been previously reported. This review covers in detail 204 such arthroplasties performed between 1975 and 1979 using a standard operative technique and prosthesis; these hips were consecutive and entirely unselected. Follow-up averaged 3.2 years (range two to six years). Successful hips are comparable to hips successfully replaced with stemmed prostheses. The rates of perioperative complications and failure due to infection, ectopic ossification, fracture of the femoral neck, and dislocation are favourably comparable to rates reported for stemmed total hip arthroplasties. The rate of aseptic loosening (17 per cent) greatly exceeded that found for stemmed total hip arthroplasties. Errors of patient selection, operative technique and prosthetic design have been identified. It is concluded that, although resurfacing arthroplasty of the hip may be an appealing procedure with theoretical advantages in certain instances, patient selection is essential, operative technique is demanding and changes in the design of the prosthesis are required. The procedure is still in evolution and therefore it should not yet be generally employed.

The occurrence of a radiolucent line at the interface of bone and cement in total joint prostheses is a frequently observed, although little understood, phenomenon. Because of an operative technique utilised in instances of bone loss, we have, within a single implant mass used in each of a series of 18 total knee replacements, been able to observe two separate interfaces, one between bone and cement and the other between bone and cobalt chrome. The average period of observation was 32 months. All of the knees except one demonstrated a lucency at the bone-cement interface; only one of the knees had a similar lucency at the bone-CoCr interface. One of the knees was studied histologically. In the light of the universal observation of macrophages at bone-cement interfaces and the recent finding that osteoclasts are derived from macrophages, these observations are significant in relation to the aetiology of bone-cement lucencies.

The essentially satisfactory results from the ICLH implant as used until 1975 were marred by examples of loosening and sinking of the tibial implant, by patellar pain of varying severity, by wear of the tibial implant caused by fragments of cement and by failure consistently to control the alignment of the leg. This report describes the methods now being used to overcome these complications and gives an account of the success so far achieved.

Cadaveric knees replaced with the Geomedic, ICLH, Marmor and Total Condylar prostheses were tested in axial compression, in rotation and in hyperextension in order to observe the strength of fixation of the tibial components. In axial compression the strengths at failure varied widely, both with any one prosthesis and between prostheses. This is attributed largely to the strength of the cancellous bone of the tibia, which was measured in each case and also varied widely. Three natural knees failed at loads of 7300, 7600 and 8300 newtons respectively, whereas the strengths of replaced knees ranged from 3000 to 15750 newtons. At least one example of each design failed at less than 7300 newtons, suggesting little or no reserve of strength. The strength of fixation was greater when the tibial prosthesis was large enough to rest on the whole cross-section of the tibia. In rotation the three prostheses embodying rollers in troughs were stiffer than the Marmor which had a nearly flat tibial-bearing surface. The presence or absence of the cruciate ligaments had a negligible effect on torsional stiffness. In hyperextension, knees replaced with the ICLH, Marmor and Total Condylar prostheses failed by rupture of the posterior capsule at moments of about 60 newton-metres, compared with about 100 for natural knees. With the Marmor prosthesis the anterior cruciate ligament was avulsed at about 20 newton-metres compared with about 75 in natural knees, suggesting that in this respect the retention of the cruciate ligaments contributes little. None of the four knees tested after inserting a Geomedic prosthesis showed strengths as high as those replaced with the other three designs.

A multi-centre clinical trial of ICLH (Freeman-Swanson) arthroplasty has been in progress since 1971. In this paper the results up to two years after operation are reported in seventy-one knees displaying at least 30 degrees of fixed flexion, 25 degrees of valgus or 20 degrees of varus, before operation. It has been found that knees displaying 70 degrees of fixed flexion, 70 degrees of valgus, 30 degrees of varus or 50 degrees of valgus/varus instability can be satisfactorily aligned and stabilised with acceptable function. Three knees required revision. The other complications are listed and were unremarkable in nature. These results depend upon the prosthesis and upon the operative technique. The latter avoids damage to healthy bone but does involve the replacement of the tissues in the midline of the knee.

It has been shown that in the puppy, two infarcts separated by an interval of four weeks produce a disorder of long duration which results in flattening and broadening of the femoral head and which reproduces the radiological changes seen in Perthes' disease in man. The histological appearances produced by two infarcts are characteristic. In this study the histological appearance of fifty-seven femoral head biopsy specimens in Perthes' disease in man have been studied. In 51 per cent of hips histopathological changes characteristic of double infarction were present, and there were grounds for postulating that double infarction might eventually occur in all cases. The findings support the concept that the deformation of the femoral head and the chronicity of Perthes' disease in man may be due at least as much or even more to repeated episodes of infarction and the ensuing abnormalities of growth as to mechanical factors related to weight-bearing.