Revision of total hip arthroplasty by long locking stem with fully hydroxyapatite-coated modular metaphysis (Reef™): A continuous series of 78 cases at a minimum 2-year follow-up.

INTRODUCTION: Since the Wagner and Vives stems were introduced in the 1980s, the range of uncemented implants for femoral revision has grown with increasing demand. Cementless modular implants with a wide range of lengths and diameters improved adaptation to the bone and secondary metaphyseal integration. Data are sparse for the Reef™ locking stem (Depuy-Synthes, Saint-Priest, France) and we therefore conducted a retrospective study at more than 2 years' follow-up, in a continuous series of total hip arthroplasty (THR) revision procedures using this long fully-hydroxyapatite-coated modular locking stem, assessing (1) implant survival, (2) clinical and functional results, and (3) radiographic results.

HYPOTHESIS: The study hypothesis was that all-cause≥2-year survival is>90%.

PATIENTS AND METHODS: Between 2007 and 2018, 413 femoral stem exchanges were performed in our center, a Reef™ model being needed in 93 cases. Etiologies comprised: Vancouver B2 or B3 periprosthetic fracture (57%, n=53), aseptic loosening (33%, n=30), septic loosening (10%, n=9) and 1 case of aseptic non-union under the stem (1%, n=1). Clinical and functional assessment used Harris Hip Score (HHS), Postel Merle d'Aubigné score (PMA) and Oxford-12 Hip Score (OHS). Radiographic assessment used immediate postoperative views and last follow-up radiographs, assessing osseointegration, filling index and stress shielding. Implant survival was estimated following Kaplan-Meier with 95% confidence intervals.

RESULTS: Mean follow-up was 6±3 years (range, 2-12). Last follow-up included 78 cases for analysis (12 deaths [13%] and 3 patients lost to follow-up before 2 years [3%]). Ten-year Reef™ implant survival was 98.7% (95% CI: 96.3-100) free of aseptic loosening (n=1) and 94% (95% CI: 87-100) for all causes (n=3). The failures related to aseptic loosening involved femoral stem fracture away from the modular part; the 2 cases of septic loosening did not undergo reimplantation for medical reasons. Preoperatively (scheduled surgeries, excluding periprosthetic fractures: n=40), mean HHS was 43±12 and 77±12 at last follow-up (81±13 in scheduled surgery [n=40] and 73±11 in emergency surgery for periprosthetic fracture [n=53]), mean PMA score was 13±2 and mean OHS was 26±9. The infection rate was 11% (n=9) including 7.6% new infection (n=6). The dislocation rate was 6% (n=5). The radiographic stress shielding rate was 11% (n=9) and significantly associated with low metaphyseal filling index (76±7% vs. 83±8% [p=0.009]).

DISCUSSION: Like other recent data, the present good survival results support the trend for uncemented stems in femoral revision, as encouraged by the 1999 symposium of the French Society of Orthopedic Surgery and Traumatology (SoFCOT). Primary diaphyseal fixation enables secondary metaphyseal integration on an implant with modularity ensuring good bone adaptation. Close contact between the bone and hydroxyapatite surfaces is a key-point in the surgical strategy.

LEVEL OF EVIDENCE: IV, retrospective study.