Acetabular Component Thickness Does Not Affect Mid-term Clinical Results in Hip Resurfacing

Abstract

Background

The benefits of using thin acetabular components for hip resurfacing have been shown in terms of bone conservation, but there currently are little data available in the literature addressing the mid-term clinical results of these devices.

Questions/purposes

We aimed to determine whether thinner acetabular components altered mid-term postoperative clinical scores, complication rates, survivorship, radiographic appearance, and metal ion levels.

Methods

Two hundred eighty-one patients with unilateral disease received a 5-mm thick acetabular shell and 223 received a 3.5-mm shell. The femoral component implanted in both groups was identical. We compared clinical scores, complication rates, survivorship, radiographic results, and ion levels between these two groups.

Results

UCLA hip scores were similar (pain, p = 0.0976; walking, p = 0.9571; function, p = 0.9316; activity, p = 0.2085). Complications were higher in the 5-mm group (6.4% versus 1.8%, p = 0.0431). Both groups were similar regarding survivorship (p = 0.3181), cup radiolucency at 5 years (p = 0.107), and metal ion levels (cobalt p = 0.404, chromium p = 0.250).

Conclusions

With comparable mid-term clinical results, there is no tangible reason to abstain from using the 3.5-mm acetabular component.

Level of Evidence

Level III, retrospective comparative study. See the Guidelines for Authors for a complete description of levels of evidence.

Introduction

Hip resurfacing provides stability because of the larger femoral head [10, 17] and maintains normal biomechanics and leg length equality [14, 24], all with minimal bone loss for a subsequent hip arthroplasty [26]. Metal-on-metal bearings have enhanced the bone-conserving nature of the procedure because of the ability to manufacture thin yet strong monoblock acetabular components with ingrowth surface while accommodating a close to normal-sized femoral head. After 6 years of implantation of a standard 5-mm thick porous-coated acetabular component, thinner shells (3.5 mm) were introduced in our series. The benefits of the thinner components in terms of bone conservations have been reported [19], but the effect of acetabular cup thickness on the clinical results of hip resurfacing have not been studied.

The purpose of our study was to compare the clinical scores, complication rates, survivorship of the acetabular component, radiographic findings, and serum metal ion levels of the thin and thick acetabular designs in a large series of patients treated with metal-on-metal hip resurfacing arthroplasty.

Materials and Methods

We conducted a retrospective review of patients treated with a hip resurfacing device (Conserve^®Plus; Wright Medical Technology, Inc, Arlington, TN, USA). All patients with Charnley Class A disease (patients with unilateral hip disease) operated on between November 1996 and March 2008 were included in this study. Five hundred four patients were selected, of which 281 received a 5-mm thick acetabular shell and 223 received a hip resurfacing device with a thinner, 3.5-mm acetabular shell which was introduced in 2003. The 3.5-mm cup group had 73% men and 27% women and the 5-mm cup group had 72% men and 28% women. The proportion of patients treated for osteoarthritis secondary to developmental dysplasia of the hip was comparable between groups (Table 1). At the time of writing, 14 patients (10 in the 5-mm shell group and four in the 3.5-mm shell group) had died of causes unrelated to the surgery. The coverage of the femoral head by the socket was the same for both types of acetabular components and varied with femoral component size from 159° for a 36-mm head to 163° for a 54-mm head. Although the 5-mm thick socket had a constant thickness, the 3.5-mm socket was designed with greater thickness at the dome (4.7 mm) compared with the rim (3.5 mm) to maintain the same resistance to deformation during interference impaction as the thicker acetabular shell [19]. The gain in volume for a 3.5-mm thick cup of average size (48 mm inside diameter) is 1.2 mm³ when compared with a 5-mm thick component. Regarding the insertion technique, the 5-mm shell was attached to the inserter with three bayonet locked fixation points, whereas the 3.5-mm shell had one of the bayonet prongs rigidly fixed by the inserter (Fig. 1). The change also was made to reduce the thickness of the end piece so the surgeon could have a better view of the anterior acetabular rim during the insertion. The femoral component implanted in both groups was identical.

Table 1.

Comparative summary of patient demographics and surgical parameters relevant to the study

Variable	5-mm shell group	3.5-mm shell group	p value
Months of followup	111 (1–194)	57 (1–106)	0.0001
Male/female (%)	203 (72%)/78 (28%)	163 (73%)/60 (27%)	0.9112
Age at surgery (years)	48 (14–78)	52 (14–84)	0.0001
Patient height (m)	1.75 (1.47–1.98)	1.76 (1.47–1.98)	0.4518
Patient weight (kg)	84 (45–164)	83 (45–153)	0.6491
Body mass index	27 (17–46)	27 (17–39)	0.2260
Femoral head size (mm)	46.6 (36–54)	48.9 (38–56)	0.0001
Acetabular shell size (mm)	56.6 (46–64)	54.9 (44–62)	0.0001
Abduction angle (°)	43.1 (16–66)	45.4 (29–62)	0.0001
Anteversion angle (°)	18.2 (3–40)	17.4 (3–42)	0.2643
CPR distance (mm)	14.4 (3.2–22.8)	14.9 (3.1–24.1)	0.1434
Patients with developmental dysplasia of the hip	32 (11%)	27 (12%)	0.8145

Ranges in parentheses; CPR distance = contact patch to rim distance.

Fig. 1A–B.

Shown are (A) a 5-mm shell engaged with the original inserter featuring a three-bayonet locking mechanism, and (B) a 3.5-mm shell engaged with the more advanced inserter also using a three-bayonet system but featuring a more secure one-prong locking mechanism.

The original 5-mm thick acetabular component was approved by the FDA for use in hip resurfacing in 2009, and at our center we implanted the 3.5-mm shell under off-label use. None of the components was hydroxyapatite-coated.

One surgeon (HCA) completed all of the surgeries using a previously described posterior approach [2]. A 1-mm press-fit preparation and identical insertion techniques were used for both acetabular cup designs.

Preoperative and postoperative UCLA hip scores [6] were recorded by the senior author (HCA) for all hips at each patient visit. Complications were noted from chart review and revisions of the resurfacing device. AP radiographs were taken at each followup. Particular attention was directed to the interface between bone and the cementless acetabular component. At the time of this writing, five patients (1.8%) in the 5-mm shell group and 16 patients (7.2%) in the 3.5-mm shell group were missing radiographs beyond the 1-year mark. These patients were not included in the subsequent comparative analysis of radiographic lucencies around the cup. Radiolucencies in the DeLee and Charnley zones [12] and the presence of gaps on the immediate postoperative film were recorded. All radiographs were examined from digital films with a high level of magnification. Three of the authors (MA-H, KMT, MJL) performed the initial radiographic review. Intraobserver and interobserver reliabilities were assessed using a subset of 43 patients evaluated twice by all three experimenters. The two measurements were made with a lapse of 24 hours between the two readings while blinded to the previous reading and those of the other experimenters. Kappa coefficients were computed and yielded excellent agreement for intraexperimenter reliability on gaps (kappa = 0.84), substantial agreement for intraexperimenter reliability on cup interface radiolucencies (kappa = 0.64) and interexperimenter reliability on gaps (kappa = 0.69), and moderate agreement for interexperimenter reliability on cup interface radiolucencies (kappa = 0.47). The senior author (HCA) reviewed post hoc the radiographs flagged by the three experimenters as having radiolucent lines at the bone-cup interface and minor differences were resolved by consensus. Einzel-Bild-Roentgen-Analyse software (EBRA-CUP – Release 2003; University of Innsbruck, Innsbruck, Austria) was used to calculate acetabular component abduction and anteversion angles [8, 18]. A subgroup of 121 patients, 53 with 3.5-mm shells and 68 with 5-mm shells, had previous serum metal ion studies with a minimum followup of 1 year after surgery However, we restricted this subgroup to 87 patients (46 with 5-mm shells and 41 with 3.5-mm shells) who had optimal acetabular component positioning defined as a contact patch to rim (CPR) distance of 10 mm or more [29] and a radiographically secure (absence of three-zone radiolucency) bone-cup interface to control for possible surgical technique bias and ensure a fair comparison of serum ion concentrations between the two designs. For these patients, the mean time of followup to the blood drawing was 78 months (range, 12–158 months).

UCLA hip scores and serum cobalt and chromium concentrations were compared using the Mann-Whitney U-test. Chi-square tests were used to compare complication rates and the prevalence of radiographic features. To focus the study on the comparative performance of the two cup designs, Kaplan-Meier survivorship estimates were computed using time to revision for aseptic acetabular component failure as the end point and the survival curves were compared between groups using the log-rank test. The Cox proportional hazard ratio was used to evaluate the predictive value of postoperative radiographic gaps on the loosening rate of the cups, adjusting for component positioning (CPR distance) and size.

Institutional review board approval was granted to perform this study.

Results

Postoperative Functional Outcomes

UCLA hip scores were similar between patients with 5-mm and 3.5-mm shells (pain, p = 0.0976; walking, p = 0.9571; function, p = 0.9316; activity, p = 0.2085) (Table 2).

Table 2.

Postoperative UCLA scores compared between groups*

UCLA hip scores	5-mm shells	3.5-mm shells	p value
Pain	10 (9 to 10)	9 (9 to 10)	0.0976
Walking	10 (10 to 10)	10 (10 to 10)	0.9571
Function	10 (10 to 10)	10 (10 to 10)	0.9316
Activity	8 (7 to 9)	8 (7 to 9)	0.2085

* Median scores and (interquartile ranges) are presented.

Complications

Patients with the 5-mm thick shell had a higher complication rate (p = 0.0431). Eighteen patients in the 5-mm thick shell group (6.4%) experienced complications (six dislocations, five femoral nerve palsies, eight blood-related complications; one patient had femoral nerve palsy and thrombophlebitis). Six patients (1.8%) in the 3.5-mm group experienced complications (two dislocations, one peroneal nerve palsy, three blood-related complications). All nerve palsies resolved without treatment and all blood-related complications resolved with conventional treatments. Two of the dislocations required reoperations, whereas the others stabilized after closed reduction.

Survivorship

Both groups had similar survivorship (Fig. 2; log-rank test, p = 0.318). There were eight revision surgeries for cup failure performed in the 5-mm shell group (three in men and five in women). The mean time to revision was 115 months (range, 56–169 months) and mean femoral component size was 44.7 mm (range, 42–48 mm). In the 3.5-mm shell group, there were two such revisions. One was performed 46 months after surgery in a man with a 52-mm femoral component and the other at 44 months in a woman with a 44-mm femoral component. The 7-year Kaplan-Meier survivorship for the 5-mm thick group was 99.6% (95% CI, 97.1%–99.9%) and for the 3.5-mm group was 98.7% (95% CI, 94.7%–99.7%). The 10-year Kaplan-Meier survivorship for the 5-mm group was 97.7% (95% CI, 0.93.8%–99.2%).

Fig. 2.

Comparative Kaplan-Meier survivorship curves between the 5-mm and the 3.5-mm thick implants are shown. The time to revision for aseptic loosening of the acetabular component was used as the end point. Ninety-five percent CIs are shown at 7, 10, 13, and 15 years of followup. The sharp decrease at 169 months is attributed to one failure at a level of followup in which only five hips were followed, as indicated by the wide 95% CI.

Osseointegration

Both groups had similar rates of gap in Zone 2 (p = 0.0813). There were 34 hips (12%) showing a gap in Zone 2 on the postoperative radiograph in the 5-mm cup group while 17 hips (7%) from the 3.5-mm cup group had a similar feature (Fig. 3).

Fig. 3A–B.

(A) An immediate postoperative AP radiograph shows a 2-mm gap (arrow) in DeLee and Charnley Zone 2. (B) One year after surgery, the gap has filled and the bone-cup interface shows good osseointegration.

In addition, from the total 51 hips with postoperative gaps, only eight showed cup radiolucencies on the last followup films, and we found no association between the presence of a postoperative gap and the rate of revision for acetabular component loosening (p = 0.111) (Table 3).

Table 3.

Cox proportional hazard ratio

Variable	Hazard ratio	p value	95% CI
Femoral head size (mm)	0.9948	0.957	0.8258	1.1985
CPR distance (mm)	0.8899	0.24	0.7326	1.0809
Presence of a cup gap	3.119	0.111	0.7689	12.6516

CPR distance = contact patch to rim distance.

Although there appeared to be a difference in the rates of radiolucency in at least one of the DeLee and Charnley zones at last followup (p = 0.0001), we found similar results when accounting for followup time (p = 0.1450). Among the hips reconstructed with a 5-mm cup, 37 (13%) had a radiolucency in at least one of the DeLee and Charnley zones at the time of last followup (Fig. 4) versus 10 (5%) in the group treated with the 3.5-mm shell. However, the average followup time for patients who received the 5-mm cup was 108 months compared with 54 months for the patients who received the 3.5-mm cup (p = 0.0001). Therefore, to account for the difference in followup time, we reviewed the radiographic data of the 5-mm group within 5 years of surgery, excluding radiographs taken beyond, and noted 23 cups with radiolucencies (8%). In these conditions, the lucency rate in the 3.5-mm group was similar to that of the 5-mm group.

Fig. 4A–C.

(A) An immediate postoperative AP radiograph shows intimate contact of the socket with the bone. (B) One year after surgery, a small (< 1 mm) radiolucency is visible in DeLee and Charnley Zone 3 (arrow). (C) Twelve years later, the same radiolucency is still visible (arrow) but has not progressed. The patient’s UCLA hip scores at last followup were 10, 10, 10, and 8 for pain, walking, function, and activity, respectively.

Metal Ion Concentrations

Both groups had similar serum cobalt and chromium ion concentrations among patients with a CPR distance of 10 mm or greater (n = 46 in the 5-mm group and n = 41 in the 3.5-mm group). The median cobalt concentration was 1.1 μg/L (range, 0.3–5.6 μg/L) for the 5-mm group and 1.0 μg/L (range, 0.4–6.0 μg/L) for the 3.5-mm group (p = 0.404). The median chromium concentration was 1.4 μg/L (range, 0.1–6.3 μg/L) for the 5-mm group and 1.3 μg/L (range, 0.3–6.4 μg/L) for the 3.5-mm group (p = 0.250).

Discussion

The advantage of using a thinner acetabular shell in hip resurfacing is the ability to conserve bone stock on the acetabular or the femoral side, or both [19]. However, it is necessary to verify that mid- to long-term safety and efficacy of such devices are at least as good as those of the original components before their use can be generalized. In this study, we compared UCLA hip scores, complication rates, acetabular component survivorship, radiographic features relevant to socket fixation, and serum metal ion concentrations between groups of patients treated with the same femoral hip resurfacing device but acetabular components of different thicknesses.

The limitations of our study come from the consecutiveness of the two series that were compared. It is possible that the surgeon’s surgical technique for implantation of the acetabular component kept improving, although no intentional surgical change was made between the two implantation periods. However, these possible improvements would have been small considering that the senior author (HCA) had been performing hip resurfacing arthroplasties for more than 20 years when this series was initiated. In addition, there was a difference in length of followup between the two groups and this could have affected our comparison of clinical scores and metal ion levels. However, there is evidence in the literature that clinical scores after THA are maintained beyond 7 years [9], and metal ion levels do not increase in well-functioning hips once stable state wear rate has been reached [3].

We found no difference in UCLA hip scores between the two groups. This result was expected because the modifications made to the acetabular shell aimed at conserving bone with minimal alteration of the device geometry and maintaining coverage of the femoral head by the acetabular shell. The magnitude of the scores recorded was well within the range of UCLA hip scores reported for this procedure and the device used in this study [5].

The reduction in complication rates between the 5-mm and the 3.5-mm shell groups can hardly be attributed to the use of a different acetabular component, but rather to progress in instrumentation and technique. Approximately 1/3 of the reported complications were femoral nerve palsies associated with the use of an anterior pelvic stabilizer, which may have pressed on the patient’s femoral triangle. The use of this stabilizer was discontinued before the series of 3.5-mm shells was initiated [4]. However, it is possible that the slight reduction in dislocation rate may be a consequence of a more favorable head-to-cup-diameter ratio as suggested by Kelley et al. [15].

The survivorship of the two cup designs was similar. This is most likely the result of using the same surgical technique and the same porous coating on both acetabular components. The 7-year survivorship of both cups was high and the 10-year survivorship of the 5-mm cup confirms the potential of this technology for long-term durability. These results match those of the most successful current designs [13, 27].

We did not find any association between the presence of postoperative radiographic gaps in De Lee and Charnley Zone 2 and the occurrence of aseptic loosening of the cup. This might suggest that initial bone-cup apposition at the periphery of the component may be sufficient to ensure durable osseointegration. This result is in agreement with those of previous reports of filling of the postoperative gap within a few years after surgery and no incidence of component loosening [11, 25, 30]. The prevalence of radiolucencies around the cup at 5 years for both groups was comparable to the rate reported by Ollivere et al. [23] with the Birmingham prosthesis with the same followup time.

Serum cobalt and chromium ion concentrations did not differ between the two subgroups of patients with metal ion data and good socket positioning. This was expected because the bearing characteristics of the two acetabular components were identical and the manufacturing tolerances for roundness and clearance were the same. In addition, the median values for serum cobalt and chromium were as low as any previously reported for any type of hip resurfacing design [1, 16, 21, 22, 28].

Based on the results of our study, the clinical performance of the 5-mm and the 3.5-mm thick cups was equally good considering that the only difference observed was a lower rate of postoperative complications explained by a change in instrumentation. With similar clinical outcomes at 5-years minimum followup, comparable metal ion concentrations, and 7-year survivorship, there is no reason to abstain from using the 3.5-mm thick acetabular component. Excellent results of ultraporous trabecular-like acetabular component backing (made of titanium or tantalum) have been reported [7, 20] for conventional THAs and open the way for additional improvement of cementless fixation of hip resurfacing devices.