The primary and secondary stability of a revision implant is highly compromised in cases with important loss of bone stock from the proximal femoral after severe femoral loosening. Several methods using implants with or without cement have been proposed for reconstruction after femoral bone loss. The purpose of this study was to analyze mid-term clinical and radiological outcome with the Wagner prosthesis for revision surgery.

Fifty-two cases of aseptic femoral loosening were treated with this method in 36 women and 14 men, mean age 70 years (range 32-92 years). None of the patients was lost to follow-up; five who died after 18 months were retained for analysis. Mean follow-up was 44 months (range 18-88 months). The mean preoperative Postel Merle d'Aubigné (PMA) score was 10.5 ± 0.4. These patients had major bone loss (5 grade II, 24 grade III, 23 grade IV in the SOFCOT classification). The transfemoral access was used in 17 cases and bone grafts in 32. Clinical outcome was assessed with the PMA score and leg length discrepancy was measured. Radiographically, stability was assessed by measuring stem impaction and progression of radiolucent lines.

The overall functional score was significantly improved from 10.5 ± 0.4 preoperatively to 14.6 ± 0.5 (p < 0.001). All items on the score improved but pain relief was the most notable. Improvement in the gait score was limited due to persistent limping in 39 patients. Leg length discrepancies were found in 8 patients with 6 shortenings and 2 lengthenings. The clinical situation remained stable after one year. The implant remained stable in 48 patients (92 p. 100) and stem impaction was observed in 4 before 12 months. Two of these cases required a second revision. Metaphyseal reconstruction was observed in 42 patients (81 p. 100), including 24 (46 p. 100) who exhibited homogeneous reconstruction with trabeculation. The reconstruction did not progress further after 18 months postoperatively. Complications included four dislocations and five revision procedures (three for cup loosening, two for femoral pivot instability).

The functional outcome was similar to results reported in the literature, including the high frequency of limping that was caused by various factors (valgus prosthesis neck, leg length discrepancy, muscle deficits). Leg length discrepancies resulted from defective positioning or impaction at loading. Standard radiographic series allowed an assessment of stem impaction. Our low rate could be due to delayed weight bearing. La reconstruction of bone loss did not progress after 18 months and was independent of bone grafting, route of access, and the initial degree of loosening. The high frequency of dislocations with this type of implant can be prevented by horizontalization of the acetabulum and use of anti-dislocation inserts at the first intention revision.

Our results with the Wagner prosthesis are satisfactory both clinically and radiographically. The two drawbacks of this implant (limping and dislocation), partly due to the design of the femoral stem, have led us to study a new implant that preserves good distal anchoring and optimizes the proximal biomechanics of the hip joint.

Le but de cette étude est d'évaluer à moyen terme le résultat clinique et radiologique de la prothèse de reprise fémorale de Wagner.

Cinquante-deux descellements fémoraux aseptiques ont été traités par cette méthode chez 36 femmes et 14 hommes, d'âge moyen 70 ans (32-92). Les pertes osseuses étaient sévères (5 stade II, 24 stade III, 23 stade IV selon la classification de la SOFCOT). Dans 17 cas la voie d'abord était transfémorale ; 32 cas ont été greffés. Les résultats cliniques ont été appréciés selon la classification de Merle d'Aubigné. Radiologiquement la migration et la reconstruction osseuse ont été analysés. Aucun patient n'a été perdu de vue, cinq patients décédés à plus de 18 mois ont été inclus.

Au recul moyen de 44 mois (18-88), aucun patient n'a présenté d'infection, 5 patients ont été repris dont deux pour un problème d'enfoncement fémoral. Le score fonctionnel moyen PMA était de 14,63 (10,52 en préopératoire), l'amélioration prédominant sur la douleur, 75 p. 100 des patients présentaient une boiterie. Radiologiquement 92 p. 100 des implants étaient stables, et 81 p. 100 des cas montraient une reconstruction osseuse autour de la prothèse. L'existence de greffe ou d'une corticotomie n'a pas eu d'influence sur cette reconstruction. L'évolution de ces paramètres était stabilisée : à 1 an pour les éléments cliniques, à 18 mois pour les images radiologiques. Huit inégalités de membres (6 raccourcissements, 2 allongements) étaient recensées, ainsi que quatre cas de luxation.

Cette technique, à bon ancrage distal, donne des résultats satisfaisants lors de pertes osseuses sévères fémorales. Mais la fréquence de la boiterie et des luxations attribuées à la forme proximale de la prothèse doit faire évoluer l'implant vers un dessin plus proche de la biomécanique de la hanche.