Acetetabular impaction grafting series

Acetabular component loosening and pelvic osteolysis continue to be a significant clinical challenge in revision hip arthroplasty. We present results of 339 cases of acetabular reconstruction with impacted allograft.

All patients who under went acetabular reconstruction with impaction allograft between July 1995 and July 1999 were included. Clinical and radiographic data was collected prospectively.

There were 339 patients and the average age was 71 years. Most were first time revisions (201) but the group includes 2^nd, 3^rd and 4^th revisions. There were also 34 two-stage revisions for infection and 44 primary arthroplasties.

There were multiple surgeons with 2/3 being consultants and 1/3 fellows. Pre and post-operative clinical assessment included Oxford and Harris hip scores, and a modified Charnley score for pain, function and range of movement.

Pre-operative radiographs were classified with the Paprosky classification. Follow up radiographs were assessed for graft thickness, component migration, graft reabsorption and lucent lines. Bone graft preparation was also recorded.

There were 10 grade I, 205 grade II, and 103 grade III defects with 3 pelvic discontinuities. Reconstruction methods included impaction only, rim and/or medial mesh, KP plate fixation and reinforcement cages.

Follow up average was 6.1 years (4.3 – 8.4) and no patient was lost. Infection was identified in 13 patients (5 recurrent 89% eradication and 8 new 2.6%). There were 6 nerve injuries, 2 remain unresolved and 13 patients dislocated (3.8%). There have been 46 deaths in the group with 3 being peri-operative.

There have been 17 re-operations, 11 for aseptic loosening. 2 KP plates fractured, 1 cage migrated and 7 large rim meshes and 1 medial mesh failed. Radiological failure was defined as 5mm migration, greater than 5 ° angulation,or progressive movement after 2 years. With these criteria there were a further 27 failures, although most of these are doing well clinically. Nine other KP plates fractured (total 20.7% fractured), ten other patients with large rim meshes migrated (total 28.3% migration), 2 medial, 2 rim and medial meshes and 4 impaction only patients migrated.

Survivorship with any re-operation as an end point is 94.9% and with acetabular a septic loosening as an end point the survivorship is 96.7%.

The oxford hip score fell from 42.1 pre-op to 24.69 post-op (p=0.001) and the Harris hip pain and function from 17.13 and 17.24 to 35.83 and 28.44 respectively (p=0.001).

Average graft thickness in zone 1 was 1.6cm (0.4-6) zone 2 was 1.2cm and zone 3 0.9cm.

Factors associated with aseptic loosening include use of a large rim mesh particularly with an allograft thickness of >2cm (p=0.001)

We conclude that impaction allografting is a reliable method for acetabular reconstruction giving excellent clinical results and survivorship figures.  Careful consideration should be given when allograft thickness will be >2cm and a large rim mesh is required as this has been found to increase the chances of aseptic loosening or radiological ‘failure’. Under these circumstances the authors recommend the use of a cage, with the possible addition of a block allograft.