Purpose of the study: Locked implants have yielded favourable results for femoral bone reconstruction for revision total hip arthroplasty, but defective integration has also been reported due to insufficient bioactive coating. The purpose of this work was to evaluate a locked pivot with a two-third proximal hydroxyapatite coating.

Material and methods: From 1996 to 2007, 158 femoral implants (Linea™) were implanted for hip arthroplasty revision. The 143 cases with more than one year follow-up were reviewed (14 deaths, 1 lost to follow-up); 83 revisions for aseptic loosening, 41 periprosthetic fractures, 19 revisions in an infected context. Patients were reviewed retrospectively by a non-operator observer. These were older patients (mean age 68 years, range 31–93) and 36 patients had complex situations requiring at least two prior prostheses; 25% had severe grade 3 or 4 (SOFCOT) bone lesions and 59% diaphyseal damage.

Results: Clinical improvement was significant, the Postel Merle d’Aubigné score improved from 7.7 + 4.3 (0–17) to 15.6 + 2.2 (8–18) at mean 50 months (14–131) (p< 0.001). At last follow-up, 14 patients (10%) reported thigh pain and only seven pivots (5%) had not achieved Engh osteointegration (five with thigh pain two without). There were no fractures. There was on non-union of the femorotomy. According to the Hoffman index, femoral bone regeneration was significant at the metaphyseal and diaphyseal levels. Eighteen of the 19 infections cured, all periprosthetic fractures healed. The implant survival was 88.9% at 91 months (65–96.7%). There were five pivot replacements for non integration and/or thigh pain which resolved in all cases; there were no other pivot replacements. The quality of the metaphyseal and diaphyseal filling was predictive of the quality of the bone fixation of the pivot (p< 0.01).

Discussion: Compared with older models of locked pivots, this implant reduced the rate of thigh pain while allowing constantly satisfactory bone reconstruction. The metaphyseal and diaphyseal filling index is the main factor predictive of clinical and radiological success by favouring osteointegration. Maximal filling, obtained with a full range of implant diameters and lengths, contributes to this good result.

Purpose: We performed mechanical trials to quantify the contribution of locking to the stability of revision femoral implants. The implant tested was a revision prosthesis with anatomic metaphyseal contact locked with three distal bolts measuring 4.5 mm.

Material and methods: Twelve implants were impacted into composite saw bones with constant and known dimensions and mechanical properties. Three displacement sensors were used to measure micromovements between the prosthesis and the bone: three specially designed force sensors were inserted into the bolt holes to measure the force distributions for each hole. Measurements were made with an Instron. Compression cycles (780 1-Hz cycles, 100daN applied to the femoral head) and torsion cycles (780 1-Hz cycles, 4.5 Nm applied to the femoral head) were used to simulate loading and weight-bearing and to estimate the evolution of the system. Trials were conducted in two different configurations: stable metaphyseal prosthesis, unstable metaphyseal prosthesis (simulating surgical resection). These two configurations were tested with a locked and with a non-locked implant.

Results: Loading distribution between the bolts was variable and depended on the insertion conditions, implant/bolt tolerance, and the quality of the supporting bone. In the “stable” metaphyseal configuration, the bolts carried a large percentage of the compression force (up to 30%) despite the support provided by the metaphysis; when exposed to torsion stress, the metaphyseal form of the prosthesis carried the charge and avoided this phenomenon. Locking had only minimal effect on micromovements, the impaction and the form of the prosthesis maintaining its stability. For the “unstable” configuration, locking created a stable situation: micromovements were limited to those observed in the stable prosthesis (< 150 μm), compatible with bone regrowth. The bolts carried most of the charge (74.8 ±20%; 56.0±41.7%) during the compression and torsion tests. Loading created major stress within the bolts whose properties (strict diameter 4.5 mm, lateral threading) should be taken into consideration to avoid risk of rupture beyond the elastic limit of the material.

Discussion: These results can be reasonably extrapolated to surgical situations leading to the following conclusions: locking is useful and reliable after surgical resection, all the holes available should be used for locking, “rational” unlocking can be useful if “physiological” metaphyseal stress is desired.