Abstract
Summary Statement
We analysed impaction bone grafting used together with cemented or uncemented fixation in acetabular revision surgery. The overall risk for re-revision did not differ between the cemented and uncemented group. However, aseptic loosening was more common in the cemented group.
Background
Several surgical techniques address bone defects in cup revision surgery. Bone impaction grafting, introduced more than thirty years ago, is a biologically and mechanically appealing method. The primary aim of this study was to evaluate the effect of bone impaction grafting when used with uncemented and cemented fixation in cup revision surgery. Uncemented cups resting on more than 50% host bone were used as controls.
Patient and Methods
Cup fixation was studied in ninety hips (eighty-two patients), revised due to loosening between 1993 and 1997. There were fifty-three isolated cup and thirty-seven total revisions. Patients were followed for thirteen years using conventional radiography, radiostereometry (RSA), Harris Hip score and a pain questionnaire. Peroperatively the surgeon assessed the acetabular bone bed vitality. In hips where the cup was judged to rest on > 50% vital bone (group I, n=43), an uncemented cup was used. If the cup was resting on ≤ 50% living bone, uncemented (group IIa, n=21,) or cemented (group IIb, n=26) technique was chosen, according to the surgeon's preference. The mean age of patients at index revision was 61±12 years, 56% were females. The most common index diagnosis was primary osteoarthritis (n=45) followed by rheumatoid arthritis (n=10).
Results
At thirteen years, acetabular component failure had necessitated a second revision in 6/7/8 hips in Groups I/IIa/IIb respectively. These re-revisions were performed 1–10 (mean 7.1) years after index revision. Moreover four cup / liner revisions were performed in hips with femoral loosening, not allowing further RSA measurements. These twenty-five hips were followed until re-revision. Deceased patients (n=21) and patients with deteriorating medical condition, not able to attend the follow-up (n=7), were censored in the survival statistics. Aseptic loosening was the most common reason of re-revision. However, in the uncemented groups (I/IIa), four cups were re-revised due to liner wear, osteolysis or instability. In the total study population, and up to two years, the median proximal migration was lowest in Group I followed by Group IIa and Group IIb (p≤0,006). At thirteen years the mean proximal migration was highest in Group IIb 1.29 mm (SD 1.23) followed by Group I 0.30 mm (SD 0.40) and Group IIa 0.22 mm (SD 0.22), p = 0.05. In cases subsequently re-revised because of loosening or with radiographically loose cups at the last follow-up, a higher proximal migration was observed compared to the non-revised and radiographically well-fixed group (up to seven years: p < 0.001; thirteen years: p=0.04).
Discussion/Conclusion
We found an increased risk for rerevision in cases with less than 50% host bone-implant contact. These cups showed high early proximal migration, measured by RSA, indicating poor initial fixation. Rate of re-revision due to any reason did not differ between cemented and uncemented cups. The cemented group (IIb) had a higher risk of being re-revised due to aseptic loosening. Poor bone stock, use of small bone chips, inferior impaction technique, and no or restricted contact with living bone are probable reasons for failures when extensive bone grafting is needed.
