Previous studies of failure mechanisms leading to revision total knee replacement (TKR) performed between 1986 and 2000 determined that many failed early, with a disproportionate amount accounted for by infection and implant-associated factors including wear, loosening and instability. Since then, efforts have been made to improve implant performance and instruct surgeons in best practice. Recently our centre participated in a multi-centre evaluation of 844 revision TKRs from 2010 to 2011. The purpose was to report a detailed analysis of failure mechanisms over time and to see if failure modes have changed over the past 10 to 15 years. Aseptic loosening was the predominant mechanism of failure (31.2%), followed by instability (18.7%), infection (16.2%), polyethylene wear (10.0%), arthrofibrosis (6.9%) and malalignment (6.6%). The mean time to failure was 5.9 years (ten days to 31 years), 35.3% of all revisions occurred at less than two years, and 60.2% in the first five years. With improvements in implant and polyethylene manufacture, polyethylene wear is no longer a leading cause of failure. Early mechanisms of failure are primarily technical errors. In addition to improving implant longevity, industry and surgeons must work together to decrease these technical errors. All reports on failure of TKR contain patients with unexplained pain who not infrequently have unmet expectations. Surgeons must work to achieve realistic patient expectations pre-operatively, and therefore, improve patient satisfaction post-operatively.

Cite this article: Bone Joint J 2014;96-B(11 Suppl A):101–4.

We investigated whether the indentation of bone cement spacers used in revision of infected joint arthroplasty with a MacDonald dissector increased the elution of antibiotic in vitro. A total of 24 cement discs containing either 0.17 g (0.88% w/w), 0.25 g (1.41% w/w), or 0.33 g (1.75% w/w) gentamicin of constant size were made. Of these, 12 were indented with the dissector. Each disc was immersed in ammonium acetate buffer in a sealed container, and fluid from each container was sampled at zero, one, three, six, 24, 48 and 72 hours and at one, and two weeks. The concentration of gentamicin in the fluid was analysed using high performance liquid chromatography mass spectrometry.

The fluid sampled at 72 hours from the indented discs containing 0.17 g gentamicin (0.88% w/w) contained a mean of 113 mcg/ml (90.12 to 143.5) compared with 44.5 mcg/ml (44.02 to 44.90) in the fluid sampled from the plain discs (p = 0.012). In discs containing 0.33 g gentamicin (1.75% w/w), the concentration eluted from the indented discs at 72 hours was a mean of 316 mcg/ml (223 to 421) compared with a mean of 118 mcg/ml (100 to 140) from the plain discs (p < 0.001).

At two weeks, these significant differences persisted. At nine weeks the indented discs eluted a greater concentration for all gentamicin doses, but the difference was only significant for the discs containing 0.17 g (0.88% w/w, p = 0.006). However if the area under the curve is taken as a measure of the total antibiotic eluted, the indented discs eluted more gentamicin than the plain discs for the 0.17 g (0.88% w/w, p = 0.031), the 0.25 g (1.41% w/w, p < 0.001) and the 0.33 g (1.75% w/w, p < 0.001) discs.

When preparing antibiotic spacers for use in staged revision arthroplasty surgery we recommend indenting the spacer with a MacDonald dissector to increase the elution of antibiotic.

Cite this article: Bone Joint J 2015;97-B:1519–24.

Previous classification systems of failure of limb salvage focused primarily on endoprosthetic failures and lacked sufficient depth for the effective study of the causes of failure. In order to address these inadequacies, the International Society of Limb Salvage (ISOLS) formed a committee to recommend revisions of the previous systems. The purpose of this study was to report on their recommendations. The modifications were prepared using an earlier, evidence-based model with subclassification based on the existing medical literature. Subclassification for all five primary types of failure of limb salvage following endoprosthetic reconstruction were formulated and a complementary system was derived for the failure of biological reconstruction. An additional classification of failure in paediatric patients was also described.

Limb salvage surgery presents a complex array of potential mechanisms of failure, and a complete and precise classification of types of failure is required. Earlier classification systems lacked specificity, and the evidence-based system outlined here is designed to correct these weaknesses and to provide a means of reporting failures of limb salvage in order to allow the interpretation of outcome following reconstructive surgery.

Cite this article: Bone Joint J 2014;96-B:1436–40.

Aseptic loosening of the acetabular component continues to be the most common indication for revision of total hip replacements in younger patients. Early in the evolution of the cemented hip, arthroplasty surgeons switched from removal to retention of the acetabular subchondral bone plate, theorising that unfavourable mechanical forces were the cause of loosening at the bone-cement interface.

It is now known that the cause of aseptic loosening is probably biological rather than mechanical and removing the subchondral bone plate may enhance biological fixation of cement to bone. With this in mind, perhaps it is time to revive removal of the subchondral bone as a standard part of acetabular preparation.

Revision total hip replacement (THR) for young patients is challenging because of technical complexity and the potential need for subsequent further revisions. We have assessed the survivorship, functional outcome and complications of this procedure in patients aged <  50 years through a large longitudinal series with consistent treatment algorithms. Of 132 consecutive patients (181 hips) who underwent revision THR, 102 patients (151 hips) with a mean age of 43 years (22 to 50) were reviewed at a mean follow-up of 11 years (2 to 26) post-operatively. We attempted to restore bone stock with allograft where indicated. Using further revision for any reason as an end point, the survival of the acetabular component was 71% (sd 4) and 54% (sd 7) at ten- and 20 years. The survival of the femoral component was 80% (sd 4) and 62% (sd 6) at ten- and 20 years. Complications included 11 dislocations (6.1%), ten periprosthetic fractures (5.5%), two deep infections (1.1%), four sciatic nerve palsies (2.2%; three resolved without intervention, one improved after exploration and freeing from adhesions) and one vascular injury (0.6%). The mean modified Harris Hip Score was 41 (10 to 82) pre-operatively, 77 (39 to 93) one year post-operatively and 77 (38 to 93) at the latest review.

This overall perspective on the mid- to long-term results is valuable when advising young patients on the prospects of revision surgery at the time of primary replacement.

Cite this article: Bone Joint J 2014;96-B:1047–51.

The original forged Müller straight stem (CoNiCr) has shown excellent ten- to 15-year results. We undertook a long-term survival analysis with special emphasis on radiological changes within a 20-year period of follow-up.

In all, 165 primary total hip replacements, undertaken between July 1984 and June 1987 were followed prospectively. Clinical follow-up included a standardised clinical examination, and radiological assessment was based on a standardised anteroposterior radiograph of the pelvis, which was studied for the presence of osteolysis, debonding and cortical atrophy.

Survival of the stem with revision for any reason was 81% (95% confidence interval (CI), 76 to 86) at 20 years and for aseptic loosening 87% (95% CI, 82 to 90). At the 20-year follow-up, 15 of the surviving 36 stems showed no radiological changes. Debonding (p = 0.005), osteolysis (p = 0.003) and linear polyethylene wear (p = 0.016) were associated with aseptic loosening, whereas cortical atrophy was not associated with failure (p = 0.008).

The 20-year results of the Müller straight stem are comparable to those of other successful cemented systems with similar follow-up. Radiological changes are frequently observed, but with a low incidence of progression, and rarely result in revision. Cortical atrophy appears to be an effect of ageing and not a sign of loosening of the femoral component.

Impaction bone grafting for the reconstitution of bone stock in revision hip surgery has been used for nearly 30 years. Between 1995 and 2001 we used this technique in acetabular reconstruction, in combination with a cemented component, in 304 hips in 292 patients revised for aseptic loosening. The only additional supports used were stainless steel meshes placed against the medial wall or laterally around the acetabular rim to contain the graft. All Paprosky grades of defect were included. Clinical and radiographic outcomes were collected in surviving patients at a minimum of ten years after the index operation. Mean follow-up was 12.4 years (sd 1.5) (10.0 to 16.0). Kaplan–Meier survival with revision for aseptic loosening as the endpoint was 85.9% (95% CI 81.0 to 90.8) at 13.5 years. Clinical scores for pain relief remained satisfactory, and there was no difference in clinical scores between cups that appeared stable and those that appeared radiologically loose.

Cite this article: Bone Joint J 2014;96-B:188–94.

This review examines the future of total hip arthroplasty, aiming to avoid past mistakes

The strain on clinic and surgeon resources resulting from a rise in demand for total knee replacement (TKR) requires reconsideration of when and how often patients need to be seen for follow-up. Surgeons will otherwise require increased paramedical staff or need to limit the number of TKRs they undertake. We reviewed the outcome data of 16 414 primary TKRs undertaken at our centre to determine the time to re-operation for any reason and for specific failure mechanisms. Peak risk years for failure were determined by comparing the conditional probability of failure, the number of failures divided by the total number of TKRs cases, for each year. The median times to failure for the most common failure mechanisms were 4.9 years (interquartile range (IQR) 1.7 to 10.7) for femoral and tibial loosening, 1.9 years (IQR 0.8 to 3.9) for infection, 3.1 years (IQR 1.6 to 5.5) for tibial collapse and 5.6 years (IQR 3.4 to 9.3) for instability. The median time to failure for all revisions was 3.3 years (IQR 1.2 to 8.5), with an overall revision rate of 1.7% (n = 282). Results from our patient population suggest that patients be seen for follow-up at six months, one year, three years, eight years, 12 years, and every five years thereafter. Patients with higher pain in the early post-operative period or high body mass index (≥ 41 kg/m²) should be monitored more closely.

Cite this article: Bone Joint J 2013;95-B:1484–9.

We have evaluated the in vivo migration patterns of 164 primary consecutive Charnley-Kerboull total hip replacements which were undertaken in 155 patients. The femoral preparation included removal of diaphyseal cancellous bone to obtain primary rotational stability of the stem before line-to-line cementing. We used the Ein Bild Roentgen Analyse femoral component method to assess the subsidence of the femoral component.

At a mean of 17.3 years (15.1 to 18.3) 73 patients were still alive and had not been revised, eight had been revised, 66 had died and eight had been lost to follow-up. The mean subsidence of the entire series was 0.63 mm (0.0 to 1.94). When using a 1.5 mm threshold, only four stems were considered to have subsided. Our study showed that, in most cases, a highly polished double-tapered stem cemented line-to-line does not subside at least up to 18 years after implantation.

We reviewed 142 consecutive primary total hip replacements implanted into 123 patients between 1988 and 1993 using the Exeter Universal femoral stem. A total of 74 patients (88 hips) had survived for ten years or more and were reviewed at a mean of 12.7 years (10 to 17). There was no loss to follow-up.

The rate of revision of the femoral component for aseptic loosening and osteolysis was 1.1% (1 stem), that for revision for any cause was 2.2% (2 stems), and for re-operation for any cause was 21.6% (19 hips). Re-operation was because of failure of the acetabular component in all but two hips.

All but one femoral component subsided within the cement mantle to a mean of 1.52 mm (0 to 8.3) at the final follow-up. One further stem had subsided excessively (8 mm) and had lucent lines at the cement-stem and cement-bone interfaces. This was classified as a radiological failure and is awaiting revision. One stem was revised for deep infection and one for excessive peri-articular osteolysis. Defects of the cement mantle (Barrack grade C and D) were found in 28% of stems (25 hips), associated with increased subsidence (p = 0.01), but were not associated with endosteal lysis or failure.

Peri-articular osteolysis was significantly related to the degree of polyethylene wear (p < 0.001), which was in turn associated with a younger age (p = 0.01) and male gender (p < 0.001).

The use of the Exeter metal-backed acetabular component was a notable failure with 12 of 32 hips (37.5%) revised for loosening. The Harris-Galante components failed with excessive wear, osteolysis and dislocation with 15% revised (5 of 33 hips). Only one of 23 hips with a cemented Elite component (4%) was revised for loosening and osteolysis.

Our findings show that the Exeter Universal stem implanted outside the originating centre has excellent medium-term results.

We report the effectiveness of revision of total elbow replacement by re-cementing. Between 1982 and 2004, 53 elbows in 52 patients were treated with re-cementing of a total elbow replacement into part or all of the existing cement mantle or into the debrided host-bone interface, without the use of structural bone augmentation or a custom prosthesis. The original implant revision was still in situ and functional in 42 of 53 elbows (79%) at a mean of 94.5 months (26 to 266) after surgery. In 31 of these 42 elbows (74%) the Mayo Elbow Performance Score was good or excellent. Overall, of the 53 elbows, 18 (34%) required re-operation, ten (19%) for loosening. A classification system was developed to identify those not suitable for revision by this technique, and using this we have showed that successful re-implantation is statistically correlated to properly addressing the bone deficiency for both the humeral (p = 0.005) and the ulnar (p = 0.039) components.

We evaluated the outcome of 41 consecutive Charnley low-friction arthroplasties (LFAs) performed by a single surgeon in 28 patients aged ≤ 35 years at operation between 23 and 36 years previously. There were 20 women and eight men with a mean age of 32 years (23 to 35) at surgery. Two patients (three hips) were lost to follow-up at 12 and 17 years post-operatively, respectively, and one patient (one hip) died at 13 years post-operatively. These patients were excluded from the final evaluation. The survival rate of the acetabular components was 92.7% (95% confidence interval (CI) 88.7 to 96.7) at ten years, 67.1% (95% CI 59.75 to 74.45) at 20 years and 53.2% (95% CI 45.3 to 61.1) at 25 years. For the femoral component the survival was 95.1% (95% CI 91.8 to 98.5) at ten years, 77.1% (95% CI 73.9 to 80.3) at 20 years and 68.2% (95% CI 60.7 to 75.8) at 25 years. The results indicate that the Charnley LFA remains a reasonable choice in the treatment of young patients and can serve for comparison with newer techniques and implants.

Cite this article: Bone Joint J 2013;95-B:1052–6.

We describe the survivorship of the Exeter femoral component in a District General Hospital. Between 1994 and 1996, 230 Exeter Universal cemented femoral components were implanted in 215 patients who were reviewed at a mean of 11.2 years (10 to 13). We used one acetabular implant, the Elite Ogee component, in 218 of the 230 hips.

During the period of this study 76 patients (79 hips) died. Of the remaining 139 patients (151 hips), 121 were able to attend for radiological analysis at a minimum of ten years. One patient was lost to follow-up.

No femoral component was revised for aseptic loosening. Three hips were revised for deep infection and six acetabular components required revision, four for loosening and two for recurrent dislocation. Taking the ‘worst-case scenario’ including the one patient lost to follow-up, the overall survival rate was 94.4% at 13 years.

Our results confirm excellent medium-term results for the Exeter Universal femoral component, implanted in a general setting. The excellent survival of this femoral component, when used in combination with the Ogee acetabular component, suggests that this is a successful pairing.

We evaluated the impact of pre-coating the tibial component with polymethylmethacrylate (PMMA) on implant survival in a cohort of 16 548 primary NexGen total knee replacements (TKRs) in 14 113 patients. In 13 835 TKRs a pre-coated tray was used while in 2713 TKRs the non-pre-coated version of the same tray was used. All the TKRs were performed between 2001 and 2009 and were cemented. TKRs implanted with a pre-coated tibial component had a lower cumulative survival than those with a non-pre-coated tibial component (p = 0.01). After adjusting for diagnosis, age, gender, body mass index, American Society of Anesthesiologists grade, femoral coupling design, surgeon volume and hospital volume, pre-coating was an independent risk factor for all-cause aseptic revision (hazard ratio 2.75, p = 0.006). Revision for aseptic loosening was uncommon for both pre-coated and non-pre-coated trays (rates of 0.12% and 0%, respectively). Pre-coating with PMMA does not appear to be protective of revision for this tibial tray design at short-term follow-up.

Cite this article: Bone Joint J 2013;95-B:367–70.

The ideal acetabular component is characterised by reliable, long-term fixation with physiological loading of bone and a low rate of wear. Trabecular metal is a porous construct of tantalum which promotes bony ingrowth, has a modulus of elasticity similar to that of cancellous bone, and should be an excellent material for fixation.

Between 2004 and 2006, 55 patients were randomised to receive either a cemented polyethylene or a monobloc trabecular metal acetabular component with a polyethylene articular surface. We measured the peri-prosthetic bone density around the acetabular components for up to two years using dual-energy x-ray absorptiometry.

We found evidence that the cemented acetabular component loaded the acetabular bone centromedially whereas the trabecular metal monobloc loaded the lateral rim and behaved like a hemispherical rigid metal component with regard to loading of the acetabular bone. We suspect that this was due to the peripheral titanium rim used for the mechanism of insertion.

We evaluated all cases involving the combined use of a subtrochanteric derotational femoral shortening osteotomy with a cemented Exeter stem performed at our institution. With severe developmental dysplasia of the hip an osteotomy is often necessary to achieve shortening and derotation of the proximal femur. Reduction can be maintained with a 3.5 mm compression plate while the implant is cemented into place. Such a plate was used to stabilise the osteotomy in all cases. Intramedullary autograft helps to prevent cement interposition at the osteotomy site and promotes healing. There were 15 female patients (18 hips) with a mean age of 51 years (33 to 75) who had a Crowe IV dysplasia of the hip and were followed up for a mean of 114 months (52 to 168). None was lost to follow-up. All clinical scores were collected prospectively. The Charnley modification of the Merle D’Aubigné-Postel scores for pain, function and range of movement showed a statistically significant improvement from a mean of 2.4 (1 to 4), 2.3 (1 to 4), 3.4 (1 to 6) to 5.2 (3 to 6), 4.4 (3 to 6), 5.2 (4 to 6), respectively. Three acetabular revisions were required for aseptic loosening; one required femoral revision for access. One osteotomy failed to unite at 14 months and was revised successfully. No other case required a femoral revision. No postoperative sciatic nerve palsy was observed.

Cemented Exeter femoral components perform well in the treatment of Crowe IV dysplasia with this procedure.

There are no long-term published results on the survival of a third-generation cemented total shoulder replacement. We describe a clinical and radiological study of the Aequalis total shoulder replacement for a minimum of ten years. Between September 1996 and May 1998, 39 consecutive patients underwent a primary cemented total shoulder replacement using this prosthesis. Data were collected prospectively on all patients each year, for a minimum of ten years, or until death or failure of the prosthesis. At a follow-up of at least ten years, 12 patients had died with the prosthesis intact and two had emigrated, leaving 25 available for clinical review. Of these, 13 had rheumatoid arthritis and 12 osteoarthritis. One refused radiological review leaving 24 with fresh radiographs.

Survivorship at ten years was 100% for the humeral component and 92% for the glenoid component. The incidence of lucent lines was low. No humeral component was thought to be at risk and only two glenoid components. The osteoarthritic group gained a mean 65° in forward flexion and their Constant score improved by a mean 41.4 points (13 to 55). The rheumatoid group gained a mean of 24° in flexion and their Constant score improved by 29.4 points. This difference may have been due to failure of the rotator cuff in 75% of the patients with rheumatoid arthritis.

Thus a third-generation total shoulder replacement gives an excellent result in patients with osteoarthritis and an intact rotator cuff. Patients with rheumatoid arthritis have a 75% risk of failure of the rotator cuff at ten years.

There have been comparatively few studies of the incidence of osteolysis and the survival of hybrid and cementless total hip replacements (THRs) in patients younger than 50 years of age. We prospectively reviewed 78 patients (109 hips) with a hybrid THR having a mean age of 43.4 years (21 to 50) and 79 patients (110 hips) with a cementless THR with a mean age of 46.8 years (21 to 49). The patients were evaluated clinically using the Harris hip score, the Western Ontario and McMaster Universities (WOMAC) osteoarthritis score and the University of California, Los Angeles (UCLA) activity score. Radiographs and CT scans were assessed for loosening and osteolysis. The mean follow-up was for 18.4 years (16 to 19) in both groups.

The mean post-operative Harris hip scores (91 points versus 90 points), the mean WOMAC scores (11 points versus 13 points) and UCLA activity scores (6.9 points versus 7.1 points) were similar in both groups. The revision rates of the acetabular component (13% versus 16%) and the femoral component (3% versus 4%), and the survival of the acetabular component (87% versus 84%) and the femoral component (97% versus 96%) were similar in both groups.

Although the long-term fixation of the acetabular metallic shell and the cemented and cementless femoral components was outstanding, wear and peri-acetabular osteolysis constitute the major challenges of THR in young patients.

The mechanical performance of the cement-in-cement interface in revision surgery has not been fully investigated. The quantitative effect posed by interstitial fluids and roughening of the primary mantle remains unclear. We have analysed the strength of the bilaminar cement-bone interface after exposure of the surface of the primary mantle to roughening and fluid interference. The end surfaces of cylindrical blocks of cement were machined smooth (Ra = 200 nm) or rough (Ra = 5 μm) and exposed to either different volumes of water and carboxymethylcellulose (a bone-marrow equivalent) or left dry. Secondary blocks were cast against the modelled surface. Monoblocks of cement were used as a control group. The porosity of the samples was investigated using micro-CT. Samples were exposed to a single shearing force to failure.

The mean failure load of the monoblock control was 5.63 kN (95% confidence interval (CI) 5.17 to 6.08) with an estimated shear strength of 36 MPa. When small volumes of any fluid or large volumes were used, the respective values fell between 4.66 kN and 4.84 kN with no significant difference irrespective of roughening (p > 0.05). Large volumes of carboxymethylcellulose significantly weakened the interface. Roughening in this group significantly increased the strength with failure loads of 2.80 kN (95% CI 2.37 to 3.21) compared with 0.86 kN (95% CI 0.43 to 1.27) in the smooth variant. Roughening of the primary mantle may not therefore be as crucial as has been previously thought in clinically relevant circumstances.