Young osteoarthritic male patients have been considered the ideal candidates for Metal-on-Metal (MoM) hip resurfacing arthroplasty (HRA), based on generally good long term results. In contrast, hip resurfacing in young female patients has become controversial. Recently, one implant manufacturer withdrew 46mm and smaller components, citing poorer than expected 10 year outcomes in females with smaller HRAs. Whether this difference is related to gender or to component size is still debated. Possible reasons for higher failure rates reported in females include higher rates of hip dysplasia, poorer bone quality and the risk of higher wear in some smaller sized implants with low cup coverage angles. We reviewed HRA revision specimens with the aim of comparing mode of failure, time to revision, femoral cement characteristics and acetabular bone attachment in specimens larger and smaller than 46mm and from male versus female patients. The study included all of the MoM HRA devices in our collection. Of the 284 hip resurfacing devices with complete clinical information, 131 were from male and 153 from female patients. Femoral sizes ranged from 36 – 58mm, median and mode 46mm; median size in females was 44 and 50mm in males. Time to failure ranged from 1 to 178 months, median 24 mos. Seven designs were represented but the majority were Conserve Plus (n=105 WMT, USA) and BHR (n=78 Smith & Nephew, USA) which differ in cementing technique. 131 femoral components were sectioned and the width of the cement mantle and the amount of cement in the head were measured. Where available, the amount of bone attached to the cup porous surface (n=91), tissue ALVAL scores (n=75) and bearing wear depth (n=138) were included in the multivariate analysis. As a function of gender, there were no significant differences in time to revision, cement measurements or ALVAL scores. Wear depth was significantly higher in females (femoral 41um vs 21um; cup 50um vs 16um, p=0.05). As a function of size (46 and less = small), the <46mm group had a slightly shorter time to revision, 30 vs 38 months, p=0.04). Bone ingrowth ranged from 0 to 60% (Figure 1) and significantly less bone attachment was noted in both the smaller and larger components (p = 0.001). Other characteristics were similar in both groups. When wear-related failure modes (cup malposition, lysis, high ions) were compared, no differences between male and female or large vs small were found. The amount of cement in the femoral heads covered a wide range but femoral loosening or fracture rates were not different as a function of size or gender.Methods
Results
Bone ingrowth fixation of large diameter, beaded cobalt chromium cups is generally considered to be reliable but this is typically judged radiographically. To date, implant retrieval data of attached bone has been limited. This study evaluated correlations between the pre-revision radiographic appearance and the measured amount of bone attachment on one design of porous coated cup. Twenty-six monoblock, CoCr Birmingham Hip Resurfacing (BHR, Smith and Nephew, TN, USA) cups with macroscopic beads and hydroxyapatite coating were studied. Seventeen were revised for acetabular malposition with the remainder revised for femoral loosening (4), pain (1), infection (1), dislocation (1) or lysis (2). Median time to revision was 35 months (10 – 70 months). Ten patients were female; the median age of all patients was 54 years. The pre-revision radiographs were visually ranked for cup-bone integration as follows: 0 = none, 1 = < 50%, 2 = 50 – 75%, 3 = 76 – 95%, 4 = > 96% integration. Rankings were made for the superior and inferior aspects, without knowledge of the appearance of bone on the retrievals. The revised cups were photographed at an angle so the dome and the cup periphery were visualized. The area of bone in four equal segments in each of the superior and inferior aspects was measured with image analysis software. A probe was used to differentiate bone from soft tissue. Only bone that covered the beads was counted. Correlation coefficients were calculated for the radiographic and image analysis data. Radiographically, most cups were assessed as having more than 50% of bone attachment and 7 cups were ranked as having almost total integration with bone. Only 2 cups were assessed radiographically as fully loose. Measured total bone attachment ranged from none to 55%. Superior and inferior percent ingrowth were highly correlated (corr=0.68, p<0.001) but there was no correlation between percent bone and x-ray rank (inferior corr=0.01, p=0.96; superior corr=0.23, p=0.26). There was no correlation between cup malpositioning as a reason for revision and x-ray integration ranking (superior p=0.34; inferior p=0.80).Methods
Results
Unlike conventional radiographic methods, the newly introduced EOS system provides simultaneously-synchronized anteroposterior (AP) and true-lateral (LAT) x-ray images. EOS offers considerable potential for calculating parameters such as true femoral and acetabular angular positioning, impingement sites, and also for measuring wear in polyethylene cups. In this study we used THA wear-simulation fixtures to assess 3D-wear in polyethylene cups using EOS algorithms. A validated phantom apparatus was used to simulate values of three-dimensional wear, controlled in the 3 directions (antero-posterior, medio-lateral, cranio-caudal) using micrometers. (Figure 1) 24 simulations of wear with controlled amplitudes and directions were imaged using the biplane EOS slot-scanning system. Wear amplitudes were between 0 and 3464 μm. Using dedicated software, wear was measured by a 2D/3D matching of 3D spheres onto the 2D frontal and lateral radiographs, allowing the determination of the 3D coordinates of both the cup and femoral head centers and thus the calculation of a 3D wear vector. (Figure 2) Measured wear vector were compared to real wear vectors in terms of amplitude and direction.3D wear vectors were measured twice by 3 independent observers (for a total of 144 measurements) in order to evaluate intra- and inter-observer reliability. There was a strong correlation between the measured wear amplitude and the real wear amplitude (Pearson's r = 0,99). Mean error when comparing wear measurement amplitude with real wear amplitude was 356 μm (SD = 127 μm). None of the 144 measurements presented an error over 1 mm. The accuracy of wear direction evaluation was highly correlated with wear amplitude (Spearman's rho = 0,98), the measurement of 3D wear direction presenting an accuracy better than 15° for wear amplitudes over 1,5 mm. Intra-observer errors for wear amplitude were between 138 μm and 221 μm depending on the observer. Inter-observer error for wear amplitude was 333 μm.Material and methods
Results
It has been suggested that metal ion levels are indicative of 182 patients (73 females and 109 males)with a unilateral Conserve Plus (WMT, TN USA) MMHRA and had who had provided blood for metal ion analysis data from December 2000 to June 2011 were retrospectively studied. Only measurements made more than 12 months after surgery were included in order to exclude hips that had yet to reach steady-state wear. For patients with multiple draws, the most recent qualifying draw was used. Activity level was assessed by the UCLA activity score. The mean age was 51.5 years (20.0 to 77.5 years). The mean follow-up time for the last blood draw was 70 months (range, 12 to 165). Serum cobalt (CoS) and chromium (CrS) levels were analyzed using inductively coupled plasma mass spectrometry in a specialized trace element lab. Using acetabular abduction and anteversion measured by EBRA, component size, and reported coverage angle of the acetabular component, the CPR distance was calculated as previously described. Multiple logistic regression was performed to identify significant relationships between high metal ion levels (7 μg/L or greater) and gender, activity and CPR distance. The median CoS level for the entire cohort was 1.13 μg/L (range, 0.15 to 175.30), and the median CrS level was 1.49 μg/L (range, 0.06 to 88.70). The average CPR distance was 13.8 mm (range, 3.2 to 22.1). There was a significant association between low CPR values and CoS and CrS. There was a 37-fold increase in the risk of CoS >7μg/L (p=0.005) and 11-fold increase in the risk of CrS > 7μg/L (p=0.003) when CPR distance was 10 mm or less. No associations were shown for gender and UCLA activity scores. CPR distance was found to be a reliable predictor of ion levels > 7μg/L and appears to be a useful indicator to evaluate the multi-factorial process of edge-loading and wear. Patients with a low CPR distance should be monitored for increased metal ion levels.
