OBJECTIVETo discuss the characteristics and reconstruction strategies of acetabular bone deficiency in total hip arthroplasty in order to find the optimal resolution for this clinical puzzle.

METHODSClinical and radiological materials of 37 patients (37 hips) with acetabular deficiency who underwent primary or revision total hip arthroplasty were analyzed retrospectively from May 1998 to August 2008. According to the common classification system for acetabular bone deficiency-the AAOS classification system, the acetabular bone defect was reconstructed using morselized or structural bone grafting, accessory acetabular hardware (ring, cup or cage) together with cemented or cementless acetabular component respectively. All patients were followed up regularly to assess their hip function improvement by measuring the Harris hip score and to find out the grafting bone healing, absorption and position of the prosthesis by plain radiographic examination.

RESULTSThere were 11 primary and 26 revision total hip arthroplasty. There were 9 hips of AAOS-type I defect, 13 of type II and 15 of type III. Morselized bone grafting was used in 24 cases, bulk structural bone grafting in 6 cases and mixed bone grafting in 7 cases. Twenty-one patients used cementless prosthesis and 16 used cemented prosthesis. The average follow-up period was 53.7 months (6.5 to 130.5 months), the average preoperative Harris hip score was 42 +/- 8 with contrast to 87 +/- 5 of post-operation. Statistically significant difference was found according to t-test with P value less than 0.05. Major or complete healing of grafting bone was obtained at the latest follow-up.

CONCLUSIONSAcetabular bone defect during primary or revision total hip arthroplasty should be reconstructed according to its characteristics and severity. For relatively simple AAOS type I and type II cases, morselized or bulk structural bone grafting together with cementless prosthesis could achieve favorable initial stability, but for more severe AAOS type III cases, bulk structural bone grafting with accessory reinforcement hardware would be necessary. Though the early and mid-term results of the above strategies are satisfied, the long term outcome still require further study.