A randomized controlled trial will be conducted and concealed allocation will be used to allocate patients to either metal-on-polyethylene or metal-on-metal cementless total hip arthroplasty. The randomization procedure is based on sequentially numbered opaque sealed envelopes, produced by an external institution not involved in the selection, clinical care and evaluation of the patients. The study design, procedures and informed consent are approved by the local Medical Ethical Committee (registration number 2005–42). The trial is registered in the Netherlands Trial Registry (NTR1399). Guidelines of the Consort Statement are followed .
The study will be conducted at the Department of Orthopaedic Surgery of the Martini Hospital, which is a large teaching hospital in the city of Groningen, the Netherlands. Patients to be included suffer from non-inflammatory degenerative joint disease of the hip including osteoarthritis, avascular necrosis and traumatic arthritis, are aged between 18 and 80 and are admitted for primary cementless unilateral THA. Patients with active infection, revision arthroplasty, marked bone loss, and unwillingness or inability to follow instruction are excluded. Participation in the study is voluntary and informed consent is required. The inclusion period is planned from September 2006 to September 2009.
Patients in the metal-on-metal group will receive a metal-on-metal articulation total hip arthroplasty, a cementless plasma sprayed porous coated titanium alloy acetabular component with a cobalt-chromium liner (M2a-Magnum™, Biomet) and a cobalt-chromium femoral head with a carbon concentration between 0.20% and 0.30%. The radial clearances of the articulations vary between 17.5 and 150 micrometers. The head sizes vary from 38 to 60 mm, depending on the shell sizes which range from 44 to 66 mm. The geometry of the patient determines the largest possible shell size and head size to be implanted.
Patients in the metal-on-polyethylene group will receive a metal-on-polyethylene total hip arthroplasty, a cementless plasma sprayed porous coated titanium alloy acetabular component (Mallory-Head®, Biomet) with a polyethylene liner (ArCom™, Biomet) and a 28 mm cobalt-chromium femoral head with a carbon concentration between 0.20% and 0.30%. In both the MM and MP groups the same cementless femoral component is used: a proximally plasma sprayed porous coated titanium alloy (Ti6Al4V) stem (Mallory-Head®, Biomet).
According to the surgeon's preference, a posterolateral or anterolateral surgical approach in lateral decubitus position is used. Antibiotic prophylaxis with a first-generation cephalosporin will be given preoperatively and during the first twenty-four hours intravenously. All patients will be treated postoperatively following a standardized protocol, in terms of analgesia and mobilization. As prophylaxis against thrombosis, oral anticoagulation by coumarin-derivate is given 6 weeks postoperatively.
In this study the following outcome parameters will be assessed: bone densitometry and serum metal ion concentration (primary outcome parameters), self reported functional status, physician reported functional status, range of motion, number of dislocations, radiographic evaluation and prosthetic survival (secondary outcome parameters). Measurements will take place preoperatively, perioperatively, and postoperatively (6 weeks, 1 year, 5 years and 10 years).
Bone mineral density (BMD) measurements will be performed using a dual energy x-ray absorptiometry (DEXA) scanner (Hologic Inc., Bedford, Mass., United States) in order to calculate bone density changes around the acetabular component. Four horizontal regions of interest (ROI) are defined, as suggested by Wilkinson . In addition, an extra ROI is defined in the os ilium to serve as control. The manufacturer's metal removal software will be used. The contralateral normal hip will be scanned following a standard manufacturer's protocol to establish BMD in the femoral neck, trochanter, intertrochanteric, total hip and Ward's triangle sites.
Serum metal ion concentration
Serum ion concentrations for cobalt, chromium and titanium will be determined by venous blood sampling. Cobalt and titanium concentrations are analyzed by inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7500 series, Agilent Technologies) and chromium is measured by grafite furnace atomic absorption spectrometry with Zeeman correction (GFAAS; Varian 220Z, Varian Inc.). The patients' sera may also be used to assess cytokine levels and effects of these ions on osteoblast cells.
Surgical approach, surgical time and intra-operative blood loss are recorded. Perioperative complications will be registered, including hip dislocations.
Self-reported and physician reported functional status and range of motion
The validated Oxford self-rating questionnaire will be used to assess self reported functional status . The validated Harris Hip Score is used to assess patient and physician reported functional status, as well as range of motion [11, 12].
During every follow-up visit standard supine anteroposterior (AP) pelvic hip radiographs (with 115% magnification) will be taken. The AP radiographs at 6 weeks will serve as baseline, and will be compared to the X-rays 5 years and 10 years postoperatively. Radiographs are reviewed for presence of femoral radiolucent lines and scored according to the 7 zones described by Gruen et al. . Peri-acetabular radiolucencies are assessed according to the three zones of De Lee and Charnley . The scoring will be undertaken by an independent reviewer.
It is our hypothesis that large head metal-on-metal arthroplasties will show less bone mineral density loss and higher serum metal ion concentrations compared to the conventional 28 mm metal-on-polyethylene articulations. In order to detect a least clinical relevant difference in bone mineral density (BMD) of 0.25 g/cm2 with a standard deviation of 0.4, 41 patients are needed in each group (alpha 0.05, power 0.80). Based on previous work with cemented THA we expect a drop-out rate of 10%, but we also expect conversion to cemented cups if adequate cementless fixation fails. We therefore aim to include 50 patients in each group. Comparable studies also used 50 patients in each group [15, 16]. In order to detect a clinical difference of 2.5 μg/liter in serum metal ion concentration with a standard deviation of 1.8, 8 patients per group are needed (alpha 0.05, power 0.80). To compensate for patients withdrawn from the study, (the first) 15 patients will be included in each group. A comparable study used 10 patients in each group .
The Statistical Package for the Social Sciences version 14.0 for Windows (SPSS Inc.) will be used. Group comparisons are based on intention-to-treat analysis. Non-parametric tests are used for comparisons of means within groups (Wilcoxon's Signed Ranks Test) and between groups (Mann-Whitney Test) if our expectation of a skewed distribution of Oxford and Harris Hip scores postoperatively is confirmed. Chi-square (Fisher's Exact) tests are employed for analyses of categorical variables. Cumulative implant survival is calculated by Kaplan-Meier time series (Mantel-Cox log rank test). A two-sided p-value of < 0.05 is assumed to be significant.