Introduction Impaction grafting in revision hip arthroplasty has now been used with a number of cemented stem designs. Follow-up has been short/intermediate with variable results due to pronounced stem subsidence and incomplete cement mantles. This study investigated the performance of a cementless, HA-coated stem used with the impaction grafting technique.

Methods We performed revision hip arthroplasty on a series of 16 stems for mechanical loosening (majority type II) in 14 patients (mean age 64 years). An uncemented Anatomic® stem was inserted into a neomedullary canal of impacted fresh frozen allograft bone. Stem migration was assessed with RSA. Graft and host cortical bone remodeling were evaluated radiographically as was the quality of the impaction grafting and amount of radiographic bone ingrowth. Bone mineral density (BMD) was assessed with DEXA at two and 10 years. Harris Hip Score was recorded for clinical assessment.

Results At six months the stems had subsided 0.78 mm (−7.78 – 0.22). Thereafter, the majority stabilized (6 months vs. 12 years: p=0.3). The graft had a homogenous distribution proximally, but to a lesser extent distally. Remodeling of graft was frequently seen at two years, especially proximally. There were no signs of graft resorption and no change in BMD in any Gruen zone between two and 10 years (P=0.15 – 0.54). About one third of patients had evidence of cortical restitution. In most cases there were no radiolucencies at all between the HA-fiber mesh and the bone. Slight cortical erosion at the stem tip was observed in nine cases but none progressed after two years. These parameters (remodeling, ingrowth and radiolucencies) changed minimally between the two and 12 years. At two years, the Harris Hip (50 to 85) and pain scores (20 to 40) had increased (p = 0.001). Thereafter the scores remained stationary. One hip dislocated during the first year. Heterotopic ossification occurred in three cases. One case underwent two stage revision at 30 months for late infection. One fracture was detected post-operatively. This stem subsided 16.2 mm at two years requiring revision. One patient died due to unrelated causes. Early migration was seen. Radiographs suggest much of the graft was substituted with living bone with no deterioration between two and 12 years.

Conclusions Our findings are sufficiently encouraging to initiate trials of uncemented stems with extended HA coatings.

In relation to the conduct of this study, one or more of the authors is in receipt of a research grant from a non-commercial source.