Case Report 3:

A 46 year old man with confirmed bilateral osteoarthritis had a routine right hip resurfacing in March 2000 and this was followed up with a left hip resurfacing in November 2000. He had routine follow up but represented to the clinic with right hip pain and an audible clicking sound. Radiographs at the time of this episode were satisfactory and the patient was reassured. He returned in August 2006 with similar problems and a radiograph demonstrated loosening of the acetabular component. This was confirmed at revision of the implant in December 2006. There was no evidence of sepsis and the implants were exchanged for a hybrid hip arthoplasty. A year later at routine follow up the patient was comfortable with both hips.

Case Report 4:

In February 2002 a 46 year old man with left hip due to pain and osteoarthritis confirmed on radiograph, underwent routine left hip resurfacing. There were no immediate complications and he was discharged after a routine inpatient stay. He had normal follow up but he represented in March 2005 with increased left hip pain and a painful limp. A radiograph at this presentation demonstrated possible loosening of the femoral head component. Inflammatory levels were normal (CRP 2, ESR 5) and a Bone Scan showed increased uptake around the femoral component consistent with loosening. He had the implant revised in June 2006. During the loosening of the femoral component was confirmed.  The ace tabular component was sound and there was no evidence of sepsis. The decision was taken to exchange the femoral component for an uncemented stem. At follow up 1 year later the patient was pain free and walking without a limp.

Case Report 5:

A 60 year old woman with confirmed osteoarthritis on radiograph underwent left hip resurfacing in August 2001. At surgery a large split was noted on the posterolateral edge of the acetabulum, which on testing was non mobile. The acetabular component on fitting was stable. A large subarticular cyst was observed on the femoral head which was less than 25% of the articular surface with hard sclerotic bone lining the cyst. This was packed with acetabular reamings. There were no immediate post operative complications and the patient was discharged after routine inpatient stay.

The patient was readmitted after eleven months following a GP referral feeling unwell with pain radiating down the lateral aspect of the left thigh. Ultra sound demonstrated an ileo-psoas abscess. This responded to intra-venous antibiotics and E.Coli  was isolated on aspiration and sensitive to cefuroxime. She was discharged and reviewed in clinic several months later when her ESR was down to 50 and CRP was 6 and she had a satisfactory radiograph. In the following months her inflammatory markers fluctuated but she remained largely asymptomatic other than noting that her mobility was deteriorating. She underwent a left hip aspirate with marcain injection in September 2003. The aspirate was negative on culture but she started to complain of a persistent low grade pain and a Bone Scan was requested. Before it could be performed she was admitted with a suspected septic joint in October 2003. On admission her CRP was 86 and ESR113 and she was started on cefuroxime. An inpatient Bone Scan was performed and was unremarkable. She improved and was discharged and was listed to return for a 2 stage revision. The first stage was performed in December 2003. At operation pus was found on opening the capsule; the acetabulum was loose with global loss of bone. The femoral component was not loose, but there was evidence of neck erosion. Again E. coli was isolated on culture and sensitive to ciprofloxacin. Later it was surmised that a possible source of this infection was a diverticular abscess. At the second stage there was no sign of sepsis and a hybrid hip arthoplasty was used. At follow up 1 year later she was pain free and mobilising well.

Case Report 6:

65 year old man with long standing right hip pain with confirmed osteoarthritis on radiograph underwent a right hip resurfacing in June 2000. The procedure was unremarkable with no immediate complications and the patient was discharged following a routine inpatient stay. He was readmitted 8 months later with a history of right-sided sciatica, fever and increasing right hip pain.  On ultrasound an abscess was diagnosed around the right hip, which was washed out in an open procedure. The abscess extended around psoas and beneath the inguinal ligament. It was felt that it originated from a retrocaecal abscess secondary to appendicitis. On culture Staphylococcus aureus was grown which was sensitive to flucoxacillin and rifampicin. The infection settled and as the hip joint did not appear to be involved the patient was treated with expectant management and follow up in clinic regularly. 3 months later the patient was complaining of increased pain but inflammatory markers were only minimally raised. He was listed for a diagnostic aspiration. This was negative but his CRP was raised and a repeat radiograph showed that the acetabular component had migrated and he was listed for a 2 stage revision. This was carried out in June 2002. At surgery it was found that there was a collection of 100ml of pus posterior to the hip joint. No pus was found inside the hip joint itself. The acetabular component was loose and the femoral component was sound but a decision was taken to remove it due to the presence of pus. During his inpatient stay his ESR and CRP normalised and as a precaution he was put on oral co-amoxiclav. He had his second stage in August 2002 and had a cemented total hip arthoplasty. There were no complications and at 1 year follow-up he was mobilising well and pain free.

Case Report 7:

A 54 year old lady was referred with a 2 year history of right hip pain due to severe osteoarthritis, limiting her walking distance to 300 yards. Due to her age she underwent right hip resurfacing in January 1999 using a Cormet 2000 prosthesis. 2 years post surgery she represented with increasing right hip pain and reduced mobility. Inflammatory markers showed a CRP of 2 and ESR 25. A radiograph showed no signs of loosening. A Bone Scan showed increased activity around the acetabular component.  A hip arthogram showed no abnormal tracking of the contrast. Relief from the injection was short lived and as pain was not settling she underwent a single stage revision in September 2001. The findings at surgery were that the acetabular component was completely loose with some bone loss and the femoral component sound. These were both exchanged for a cemented total hip arthoplasty. All microbiology samples were culture negative. At I year review the patient was mobilising well and pain free.

Case Report 8:

A 57 year old lady with osteoarthritis in her right hip underwent a right hip resurfacing using Cormet 2000 prosthesis, in March 1999. At the time there were no complications and intra-operative findings were unexceptional. At her six month check she complained of pain mainly on her left hand side and she was referred for some physiotherapy. At 18 months she represented with impaired weight bearing on her right hip and pain on flexion of the hip. A radiograph showed the prosthesis in a satisfactory position with no evidence of loosening but with some bone resorption around the neck of femur. On review because the pain was increasing a Bone Scan was organised. The scan showed increased uptake around the acetabulum relating to a cystic region seen laterally behind the acetabular cup. Inflammatory markers taken at the time were normal. A repeat radiograph on follow-up showed the acetabular component loosened and rotated and she was listed for revision in December 2001. At surgery the aspirate from capsule was clear. Thickening of the capsule and a granulation membrane were noted. Culture however was negative.  The acetabular component was completely loose whilst the femur was solid but with areas of resorption in the medial cortex. There were two bony defects anteriorly and medially in the remnant of the acetabular component which were grafted with bone from the proximal femur cut. A plasma cup was used with 2 screws for stability in the acetabulum. The femur was cemented, with no immediate complications. At follow up 1 year later the patient was well pain free and mobilising unaided.

Case Report 9:

A 51 year old male underwent left hip resurfacing in January 2002, after presenting with left hip pain. Previously as a child he had DDH which was treated with an osteotomy. However this resulted in a permanent limp. The pain in his left hip developed in the preceding years prior to his presentation, and in view of his age resurfacing was preferred. The findings at the original procedure showed he had a shortened femoral neck and as expected the acetabulum was deficient elliptically in an anterior direction and anterolaterally after reaming. A 50mm acetabular cup was inserted with bone graft for stability. There were no immediate complications.

He presented 2 ½ years post operatively with increasing left hip pain. Radiographs at the time were unremarkable and inflammatory markers were normal. A Bone Scan demonstrated increased uptake around the femoral head consistent with loosening of the component. Revision surgery was performed in October 2004, which showed the femoral component was completely loose, with the cement mantle intact with no bone attached. There was excessive granulation tissue between the cement mantle and a large amount of bone loss consistent with avascular necrosis of the femoral head. The acetabular component was intact with no evidence of loosening, or sign of infection. As a result only the femoral component was exchanged, with a cemented Corin taper fit stem. At follow up 1 year later he was pain free and mobilising well.

Results :

OA= Osteoarthritis, DDH= developmental dysplasia of the hip, SUFE= Slipped Upper femoral eypiphesis, AL= Acetabular loosening, FL= Femoral Loosening, ALVAL= aseptic lymphocytic vasculitis associated lesion, THA= Total hip arthoplasty.

Discussion :

Femoral Loosening

Femoral loosening has been shown to be the next most common complication, in particular with metal on metal bearings. Amstutz et al⁷ looked at how differing techniques could prevent this.  They performed a prospective study comparing different techniques involved in preparing the femoral head. There study covered 600 consecutive cases. In the first   generation of cases they used no suction, few key holes in the femoral dome, the stem was not cemented, and they placed the femoral component in anatomical position. If they found cysts in the femoral head they only curetted them as per standard practise. In the second generation of cases dome suction was used when preparing the femoral component, the number of key holes was increased and the stem was cemented.  The femoral component was placed at angle of 140 degrees, and a high speed burr was used to remove all debris in any femoral cysts. Their results showed that with this difference in surgical technique there was a reduction in the amount of femoral loosening, this was with similar demographics in both patient groups. The limitations of this study were that it was based at one centre and with only one surgeon. It could be that the reduction in complications was related to improvement in surgical technique over time. The differences in preparation of the femoral head were introduced gradually rather than at one point in time. It does suggest however that careful preparation of the femoral head prior to cementation of the femoral component can reduce the amount of revisions secondary to femoral loosening. This is significant particularly since Morlock et al⁸ in found in 267 retrieved femoral heads only 31% were cemented according to recommendations.

In our 2 cases where there was femoral loosening there were no significant defects with the femoral head at the time of implantation but rather there had been loosening at the cement mantle leaving the component loose.  There are other possible reasons for this which include osteolysis which is the result of wear and tear at the articulating surfaces.  Ayers, Allen and Schoonmaker⁹ noted the effect of tumour necrosis factor (TNF) on bone resorption around cement in animal models. TNF lead to a decreased osteoid concentration and promoted bone resorption which could lead to failure of bone to bind with the cement mantle. This is very important to note particularly in hip resurfacing when one considers the large surface area involved. Some authors suggest that the orientation of the implants will have an effect on bone remodelling around the femoral component with Ong et al noting that femoral components put in valgus can increase the risk of loosening secondary to the loss of osseous support from stress shielding. However   Beaule et al¹⁰ noted that implants in varus have significantly higher chance of fracture.  Other papers also look at osteonecrosis as a cause mentioned previously in the paper of   C. P. Little et al³. Out of their 13 cases, 3 had femoral loosening. All of these showed evidence of osteonecrosis on histology. In our 2 cases there was no macroscopic evidence of osteo necrosis so consequently samples were not sent for histology.  It would be prudent in future cases to consider histology at the time of revision. Campbell, Mirra and Amstutz¹¹ hypothesised that the reason for osteonecrosis was thermal injury.  T.P Gross et al¹² have looked at this aspect as well with   uncemented femoral components in hip resurfacing with a patient group of 18 (20 hips) with one lost to follow up.  4 required revision due to reasons other than femoral loosening. With 15 hips surviving up to 7 years despite a small study it does suggest another option for hip resurfacing.

 Acetabular Loosening

Acetabular loosening is another reason for both revising hip resurfacing and hip arthoplasty.  The preferred option in resurfacing is the uncemented cup. Here, as in hybrid hips, one common cause of loosening is orientation of the acetabular component.  Morlork et al⁸ when looking at retrieved specimens in comparison with radiograph’s in situ noted that cup inclination could have a possible detrimental effect on metal wear and fatigue in the implant, particularly as resurfacing in metal on metal.   These cups are cementless. There is therefore a high reliance on biological fixation of the cup.  This in turn relies on initial stability and orientation of the component. This aspect has been looked at by others in particular they have studied different coatings on the cup.  Manley et al¹³ looked at 3 such different methods and concluded that hydroxyapatite coating alone on a smooth cup is not sufficient for fixation, but rather a porous coating or another adjunct to fixation is required such as the central peg in Cormet 2000.  This can lead to additional problems at revision such as reduced bone stock on removal of the component, as was the case with 2 out 3 of our cases with acetabular loosening. This has meant re-evaluating   the cup design in order to preserve acetabular bone stock.  Some thinner designs on the market to also allow for larger femoral heads, reducing the risk of dislocation as well. A complication of this noted by Ong et al¹⁴ was increased risk of cup deformation with diametrical pinching.  This lead to an adverse effect on the fluid- film lubrication of the metal on metal bearing and lead to increased wear. This has also been observed in hybrid hip arthoplasty.

 Infection

Infection remains a highly feared complication following all forms of arthroplasty.  However there are few papers looking at infection following hip resurfacing particularly considering the large metal surface area involved. With such a large metal and bone surface area there is the possibility for increased haematogenous spread of infection which can become   established in the hip. It is noteworthy that 2 out of the3 infections at Derby following hip resurfacing appeared to have originated from abdominal sources. These infections were all late presentations occurring 2 years after hip resurfacing.

 Pseudo-Tumours

Aseptic lymphocytic vasculitis (ALVAL) associated lesion as a cause for hip resurfacing failure is becoming more common as it also affects other metal on metal arthoplasty procedures¹⁵. It can be associated with inflammatory granulomatous masses termed pseudotumours, first described by Harris et al in the 1970s ¹⁶. Pseudotumours appear to follow a granulomatous reaction with the macrophage response excessive. Why there is a macrophage over stimulation is unclear.   Pundit et al ¹⁷ looked at 20 hips in 17 patients which were all similarly affected by phenomenon and found that metal wear present throughout each case.  However histologically as previously thought,  no gross metallosis was present clinically at revision. All 17 patients in this study were female, which raised the possibility that pre-operative sensitisation to metal could be a factor. ALVAL may be a normal body response to metal on metal prostheses with over stimulation of lymphocytes, leading to a hypersensitivity style response. Both these complications appear to be very rare with few cases reported around the world but certainly as the demand for metal on metal resurfacing increases, we may begin to see more of them. It is important to note that both appear to be immunologically mediated, which may require surgeons to be more selective in their choice of prosthesis particularly in people with known sensitivity since it is known that up to 15% of women are allergic to nickel.

 Fracture

One complication which we did not see in our series of cases is fracture of the neck of femur. Fracture is one of the most commonly quoted cause of failure in hip resurfacings with some studies quoting figures of up to 47% of revisions². Research into this suggests that there is a link to early fractures and osteonecrosis.  C. P. Little et al³ found that in 12 out of 13 femoral heads retrieved for revision, 8 had a sustained a fracture following resurfacing and there was evidence of osteonecrosis in 4 cases.  They suggested that this could be linked to the surgical approach which involves an extended posterior approach leading to the disruption of the extraosseous blood supply leading to ischaemia in the remaining bone. They found that this could occur post operatively since two of their samples came from fractures within a week of the implant being established.

The mainly limitations of this study is its small sample size, Others such as Shimmin and Back⁴ looked at the occurrence of neck fractures following resurfacing in Australia between 1999 and 2003 from around 3500 procedures performed by 89 different surgeons. They obtained results from 45 out of 50 fractures in their study. Their main conclusions were that notching of the neck, varus positioning of the implant and intraoperative technical problems lay behind femoral fractures. They found that there was little correlation between surgical experience and this complication. Another study by Marker et al⁵ looked at the incidence of femoral fractures in a number of procedures performed by the senior author between November 2000 and august 2006; out 550 procedures they had 14 femoral neck fractures. They queried if the experience of the surgeon was the reason for this complication. Their findings suggested that again the risk of fracture was multifactorial associated with high BMI, female, evidence of femoral head cysts at the original operation, intraoperative notching of the femoral head and there was some correlation with surgical experience.  Whilst this is good review   it is limited again by small sample size and only one surgical operator. Others suggest that femoral neck fracture does not necessarily warrant a revision procedure. Cossey et al⁶ looked at 407 patients who had consecutive primary hip resurfacing at 2 centres.  7 had periprosthetic fractures within 4 months following their operation and were successfully treated conservatively. They all presented with painful limp and had radiographs which demonstrated un-displaced fractures which were treated with non-weight bearing for 4-6 weeks followed by partial weight bearing for another 2-4 weeks.  They only found 1 patient with notching while the others had femoral components in valgus.

 Summary

There are a number of causes leading to failure of hip resurfacing procedures which require revision surgery. Many papers demonstrate that in particular patient group demographics and the type of original procedure have a significant bearing on whether or not the prosthesis will survive. As surgeons we need to be mindful of patient selection when considering hip resurfacing and be ready to consider other options if patient’s demographics predispose them to early failure. The importance of good surgical technique is crucial in determining the survival of the components. More research still needs to be carried out in particular focused at the causes of infection in resurfacing and the mechanisms and possible prevention of the immunologically mediated causes of failure.

Reference :

1. National Joint Registry. National Joint Registry for England and Wales 4^th annual report 2007.

2. Buergi M & Walter W. Hip Resurfacing Arthoplasty the Australian Experience. Journal of Arthoplasty (JA) 2007 Vol. 22: No. 7: Suppl 3: 61-65.

3. Little C, Ruiz A, Harding I, Mclardy-Smith P, Gundle R, Murray D, Athanasou N. Osteonecrosis In Retrieved Femoral Heads  After Failed Resurfacing Arthoplasty of the Hip. Journal of Bone and Joint Surgery (JBJS Br) 2005: 87-B: 320-3.

4. Shimmin A, Back D. Femoral Neck Fractures Following Birmingham Hip Resurfacing- A National Review of 50 Cases JBJS 2005: Vol. 87-B: 463.

5. Marker D, Seyler T, Jinnah R, Delanois R, Ulrich S. Mont M. Femoral Neck Fractures After  Metal-on-Metal Total Hip Resurfacing. JA 2007: Vol. 22: 66-71

6. Cossey A, Back D, Shimmin A, Et Al. The Non-operative Management of Periprosthetic Fractures Associated With The Birmingham Hip Resurfacing Prosthesis. JA 2005: Vol. 20: 358

7. Amstutz H, Le Duff M, Campbell P, Dorey F. The Effects of Technique Changes On Aspetic Loosening Of Femoral Component In Hip Resurfacing. The Results of 600 Conserve Plus With 3 To 9 Year Follow Up. JA 2007: Vol. 22: 481-489.

8. Morlock M, Bishop N, Zustin J, Hahn M, Ruther W, Amling M. Modes Of Failure After Hip Resurfacing: Morphological And Wear Analysis Of 267 Retrieval Specimens. JBJS 2005: Vol. 90-A: 89-95.

9. Allen M, Schoonmaker J, Ayer D. Tumour Necrosis Factor Alpha Induced Endosteal Bone Resorption In Rabbits. JBJS 2004: Vol. 86-B: 432.

10. Beaule P, Harvey N, Zaragoza E, Le Duff M, Dorey F. The Femoral Head/Neck Offset And Resurfacing. JBJS 2007: Vol. 89-B: 9-15.

11. Campbell P, Mira J, Amstutz H. Viability Of Femoral Heads Treated With Resurfacing Arthoplasty. JA 2000: Vol. 15: 120-122.

12. Gross T, Liv F. Metal-On-Metal Hip Resurfacing With An Uncemented Femoral Component A Seven Year Follow Up Study. JBJS 2008: Vol. 90-A: 32-37

13. Manley M, Capello W, D’ Antonio J, Eddin A, Geesink R. Fixation of Acetabular Cups Without Cement Total Hip Arthoplasty. A Comparison of 3 different Implant Surfaces at a Minimum Duration of Follow-up of Five Years. JBJS1998: Vol. 80-A: 1175-85.

14. Ong K, Manley M, Kurtz S. Have Contemporay Hip Resurfacing Designs Reached Maturity. A Review. JBJS 2008: Vol. 90-A: 81-88.

15. Minutes of the Committee on the safety of devices expert advisory group on metal wear and debris from Hip Implants. April 2007

16. Harris, W. H.; Schiller, A. L.; Scholler, J.-M.; Freiberg, R. A.; and and Scott, R.: Extensive localized bone resorption in the femur following total hip replacement. JBJS 1976: Vol. 58-A: 612-618.

17. Pandit H, Glyn-Jones S, Mclardy-Smith P, Gundle R, Whitwell D, Gibbons C, Ostlere S. Pseudotumours associated with metal-on-metal hip resurfacings. JBJS 2008: Vol.90-B: 847-51.