From January 1997 through December 2000, we prospectively followed a series of 113 consecutive patients at our institution who underwent primary total hip replacements (THA) using the Metasul Hip System (Zimmer, Warsaw, IN).
All patients provided informed written consent for this long-term clinical and biological prospective study. In accordance with the practice of the ethics board at our institution, they received complete information about the study, including the goals and the procedures for blood sampling.
Exclusion criteria were the presence of other orthopedic implants, a history of professional or dental exposure to titanium or cobalt, previous conservative surgical procedures on either hip, or renal disease. All enrolled patients were below 60 years of age at the time of surgery. In our practice, fewer than 20% of all patients undergoing primary THA during the study period were eligible for inclusion in this study.
Mean age at the time of arthroplasty was 54 (30–60) years. 53 females and 56 males were included. The etiologies were primary arthritis (121 cases) and osteonecrosis (13 cases). Mean body mass index was 26 (19–34). During the preoperative period, 77% of patients had an activity level graded as 4 or 5 according to the Devane activity score (Devane and Horne 1999
). Clinical data for each patient were obtained by office interview and physical examination before surgery, and at each follow-up examination. Hip function results were rated using the Harris hip score grading system.
Of the 113 patients (138 hips), 4 patients (4 hips) were lost to follow-up in the first 5 years. At their last review, they had no pain and none had loosened implants.
The remaining 109 patients (134 hips) were available for complete clinical and radiographic analysis 7–11 years post-operatively. No patients in the series died.
We used cemented Metasul cups (Zimmer, Warsaw, IN) for all replacements. This cup is composed of a single component, namely a 3-mm thick cobalt-chromium (Co-Cr) inlay embedded in a polyethylene socket. The cemented femoral stem (Alizé; Fournitures Hospitalières, Quimper, France) is collarless, oval in cross section, and straight. The implant was made of titanium alloy (Ti-6Al-4V) with a polished surface coated with titanium oxide (TiO2) obtained by anodization. 6 stem sizes were available. The modular stem was combined with a 28-mm Metasul femoral head (Zimmer) with a 12/14, 5°43’ taper.
Head-neck compatibility was assessed by a unitary control for the taper angle (precision: 1 minute) and by the diameter at the base and the diameter at the summit (precision: 1 micron) (Pexit Dorsey Gage, Cambridge, UK). Palacos Genta cement (Scherring Plough, Brussels, Belgium) was used for both components.
All procedures were performed following the standard practice at our institution, namely using an anterior approach on a Judet orthopedic table, by 4 different surgeons. To obtain a complete and thick cement mantle, the femoral canal was over-reamed by 2 mm and distally occluded by a resorbable femoral plug (Synplug; Zimmer). The cement was injected in retrograde fashion and pressurized by a hand-driven syringe.
Anteroposterior and lateral radiographs of each hip were taken before and immediately after surgery, 6 weeks after discharge from the hospital, and at 3 months, 6 months, 1 year, and then annually. Acetabular radiolucent lines and osteolysis were measured according to the zones described by DeLee and Charnley. Radiographic loosening of the cup was defined as the presence of radiolucent lines measuring 2 mm in at least two DeLee-Charnley zones, axial cup migration of > 5 mm, or > 5° of change in cup inclination on the anteroposterior radiographs of the pelvis.
On the femoral side, we assessed the presence and progression of radiolucent lines according to Gruen et al., calcar resorption or atrophy, subsidence, periprosthetic osteolysis, and cortical hypertrophy. Loosening of the stem was defined as migration exceeding 3 mm or a continuous radiolucent line wider than 2 mm. Radiographic measurement of wear was not possible as no distinction could be made between the edge of the femoral head and the metal articulation surface of the acetabular component.
Serum ion determination
Titanium, cobalt, and chromium ion measurements were performed in all patients. None of the patients had renal insufficiency or chromium and cobalt devices of any kind that could create bias in the serum metal determinations. Blood samples were taken just before implantation and 3 months, 6 months, 1 year, and every year postoperatively until the last endpoint. To avoid metal contamination, blood samples were drawn using a sampling kit dedicated to trace element determination; this kit included an S-Monovette needle (ref. 85.1162.400; Sarstedt, Marnay, France) and lithium heparin in a 7.5-mL S-Monovette for trace metal analysis (ref. 01.1604.400, Sarstedt). All metal ion measurements were performed on 2 samples. Titanium was measured in diluted blood plasma by inductively coupled optical emission spectrometry (ICP-OES) on a JY24 spectrometer (Jobin Yvon, Longjumeau, France). The detection limit for titanium was 1.4 μg/L of plasma. Concentrations under the detection level were set at half the detection limit value (0.7 μg/L) to allow conventional statistical analysis. The serum cobalt and chromium levels were also determined in the same serum samples. Until 2004, these concentrations were simultaneously determined by electrothermal atomic absorption spectrometry using a 5100 spectrometer, and thereafter by using a SIMAA 6100 spectrometer (Perkin Elmer, Courtabœuf, France). In our laboratory, the detection limits of cobalt and chromium in serum were 0.3 μg/L. Concentrations below that level were defined as 0.15 μg/L to allow statistical analysis. As an internal quality control, Seronorm levels I and II (Sero, distributed by Ingen, Rungis, France) were also determined in each analytical session.
All statistical analyses were performed using SAS 9.1 software (SAS institute, Cary, NC). Prosthesis survival analysis was calculated using the Kaplan-Meier method with exchange of the cup, the stem, or both components as endpoint.
Our study included 25 bilateral joint replacements, and the events of these cannot be considered statistically independent. The statistical methods used required independent observations, however. Thus, we included only randomized observations from 1 hip of each patient in all statistical tests.
Progressive radiolucent lines at the bone-cement interface of the socket, leading to aseptic loosening at the 7-year follow-up. A. After one year. B. After 5 years. C. At the 7-year follow-up.
Serum metal levels were expressed as medians with twenty-fifth and seventy-fifth percentiles (IQR). Normally distributed data were analyzed with t-tests or ANOVA, and non-parametric data were analyzed with the Mann-Whitney U test.
Cups were categorized as either small (44–50 mm) or large (52–60 mm). Chi-square tests were performed to compare patients with and without radiolucent lines. Fisher’s exact tests were performed to compare patients with and without cup revision. Log-rank test was performed to determine a surgeon factor and whether the rate of loosening and revision was dependent upon surgeons. Statistical significance for all tests was set at p < 0.05.