Long-Term Results of the Birmingham Hip Resurfacing Implant in the United States

Abstract

Background:

We previously reported the 5-year results of the Birmingham Hip Resurfacing (BHR) implant. This study evaluates BHR survivorship as well as radiographic and clinical outcomes at long-term follow-up.

Methods:

A total of 224 patients with contemporary indications, including 179 patients from the original study, were included in this analysis. Survivorship was calculated. Metal ion levels and radiographs were updated. Patient-reported outcomes (PROs) were compared with those for matched patients who had undergone total hip arthroplasty (THA). The mean follow-up was 14 years.

Results:

Survivorship free from any revision and from aseptic revision was 96.0% and 97.4% at 15 years, respectively. Two patients had undergone revision since the original study. The median serum cobalt and chromium levels were 1.4 and 1.5 ppb, respectively. The PROs were similar to those for the THA cohort. Equal proportions of patients remained active; however, the BHR group trended toward more remaining highly active (p = 0.12).

Conclusions:

Although activity was similar to THA at long-term follow-up, the BHR implant remains an excellent option for the treatment of osteoarthritis in younger male patients.

Level of Evidence:

Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Metal-on-metal (MoM) surface replacement arthroplasty (SRA) gained popularity nearly 2 decades ago. In May 2006, the Birmingham Hip Resurfacing (BHR) implant (Smith & Nephew) became the first SRA implant approved by the U.S. Food and Drug Administration (FDA) for use in the United States. However, MoM SRA along with MoM total hip arthroplasty (THA) saw a continuous decline several years later due to concerns related to adverse local tissue reactions¹.

Active, younger patients with osteoarthritis are the typical patients selected for SRA due to its advantages of preserving bone stock, low dislocation rates, restoration of native length and offset, and return to high levels of activity². Despite these advantages, concerns remain in association with SRA. SRA is a more technically demanding procedure requiring an extensile approach and accurate implant positioning for improved survival^3-6. Additionally, there is a higher risk of adverse local tissue reaction due to the MoM bearing surface.

Currently, SRA is still performed at some centers in the United States. The intermediate-term survivorship of the BHR in the U.S. has been shown to be excellent at 10 years, ranging from 96% to 99%^6,7. Patient selection strongly contributes to long-term survivorship, as male patients with larger femoral head sizes and a preoperative diagnosis of osteoarthritis have lower revision rates^6-8. Our institution previously reported the 5-year results of our consecutive series of BHRs that had been performed from June 2006 to December 2011⁸. Since the initial BHRs performed in the U.S., our indications have narrowed to include males <60 years of age with a primary diagnosis of osteoarthritis in addition to femoral anatomy that would accept a femoral component of ≥48 mm⁹. The purpose of this follow-up study is to report long-term BHR survivorship and outcomes for males <60 years of age from our original series. Additionally, we compared the postoperative activity level of patients managed with BHR to a matched cohort of patients managed with THA, to determine if a difference in activity level persists at long-term follow-up.

Materials and Methods

After approval by the institutional review board, we retrospectively reviewed the records for the patients from the previously published consecutive series of BHRs⁸. The original series included the first 360 BHRs performed at our institution from June 2006 to December 2011. Because that was the initial series, women and femoral component sizes of <48 mm were included. The present follow-up study includes patients who fit the current indications for BHR: males <60 years of age with a preoperative diagnosis of osteoarthritis and proximal femoral bone able to accept a component size of ≥48 mm. We also included 2 additional years of patients who met these criteria. Overall, 250 patients fit these criteria from June 2006 to December 2013. One hundred and eighty-seven patients were from the original study. Of these 187 patients, 4 patients died and 4 withdrew from the study. Of the remaining 242 patients, 224 (93%) were available for analysis at a mean follow-up of 13.9 years (range, 10.2 to 17.3 years). The mean age (and standard deviation) was 50.4 ± 5.5 years (range, 35 to 59 years), and the mean body mass index (BMI) was 27.5 ± 3.5 kg/m² (range, 21.6 to 34.4 kg/m²). The procedures were performed by 3 fellowship-trained total joint arthroplasty surgeons (including R.L.B. and J.C.C.) through an extensile posterior approach.

Preoperative demographic data, the modified Harris hip score (mHHS), and the University of California Los Angeles (UCLA) activity score were obtained from the original study^10,11. Patients were contacted via telephone and were asked to return for clinical follow-up to obtain patient-reported outcome data, including the mHHS, UCLA score, updated radiographs, serum metal ion levels, and an additional activity and satisfaction questionnaire (see Appendix). The serum metal ions that were evaluated were cobalt and chromium. Revisions, reoperations, and complications were recorded from a review of electronic medical records, in-person patient visits, and returned questionnaires. If a patient was not able to attend an in-person visit, the questionnaires were completed over the telephone or through mail, and radiographs and metal ion laboratory tests were scheduled at their convenience. Patients were awarded $100 for completion of the questionnaires, radiographs, and metal ion tests. Follow-up was defined as the date of completion of the most recent patient-reported outcomes.

The matched cohort of patients from the original study who had been managed with THA were contacted for patient-reported outcome data, including the activity and satisfaction questionnaires, for the purpose of comparing their activity level with that of patients managed with BHR at long-term follow-up. The matched cohort was described in the original publication and included males diagnosed with hip osteoarthritis who were <60 years of age at the time of THA⁸. Patients without preoperative UCLA scores were excluded. The mean age was 50.6 years (range, 26 to 59 years) in the THA group, compared with 50.4 years (range, 35 to 59 years) in the SRA group (p = 0.689). The mean BMI was 31.1 kg/m² in the THA group compared with 27.5 kg/m² in the SRA group (p = 0.001). All THA implants were cementless with ceramic on highly cross-linked polyethylene bearings. The same surgeons performing the BHRs performed the THAs through a posterior approach. Changes in UCLA scores were compared between the 2 groups. An additional questionnaire was used to determine why the activity level changed over time. Serum cobalt and chromium levels at the time of the latest follow-up were analyzed. Radiographic analysis utilized the same methodology from the previous study to assess for radiolucencies, component loosening, femoral neck narrowing, and heterotopic ossification⁸.

Statistical Analysis

Descriptive statistics, including independent t test and chi-square analyses, were calculated for continuous and categorical variables, respectively. Kaplan-Meier curves were used to estimate survivorship free of revision at the most recent follow-up. Statistical analyses were completed with SPSS (version 29; IBM).

Results

Survivorship

The survivorship free of all-cause revision was 97.2% (95% confidence interval [CI], 93.8% to 98.7%) at 10 years and 96.0% (95% CI, 92.2% to 98.0%) at 15 years (Fig. 1-A). Survival free of aseptic revision was 98.6% (95% CI, 95.7% to 99.5%) at 10 years and 97.4% (95% CI, 93.9% to 98.9%) at 15 years (Fig. 1-B). Over the entire study period, 8 patients who were included in this follow-up study required revision surgery. Table I describes the reasons for revisions and the types of revisions performed. The most common indications for revision were infection (n = 3), pseudotumor and/or elevated metal ions (n = 2), and aseptic loosening of the femoral component (n = 2). Six patients who were included in the initial publication required revision within 6 years after the index procedure. The 2 late revisions both occurred at 12 years and were performed for elevated metal ions and for aseptic loosening of the femoral component (Fig. 2). The mean time to revision, including the prior 6 revisions, was 6.0 ± 4.1 years (range, 0.7 to 12.1 years). Four of the 5 aseptic revisions were conversions to dual-mobility THA (Smith & Nephew) with retention of the acetabular cup. The 2 hips that were revised for elevated metal ions had abduction and anteversion values of 52° and 24° and of 47° and 25°, respectively. The 3 infections were treated with explantation of the femoral and acetabular components followed by the placement of an articulating antibiotic spacer. The patient who had revision because of unexplained pain had well-fixed and well-placed implants without elevation of metal ions or pseudotumor.

Fig. 1-A.

Survivorship free from all-cause revision was 97.2% (95% CI, 93.8% to 98.7%) at 10 years and 96.0% (95% CI, 92.2% to 98.0%) at 15 years.

Fig. 1-B.

Survivorship free from aseptic-free revision was 98.6% (95% CI, 95.7% to 99.5%) at 10 years and 97.4% (95% CI, 93.9% to 98.9%) at 15 years.

Fig. 2.

Postoperative radiographs at 2, 7, and 12 years demonstrating late femoral component failure and the subsequent revision construct showing femoral component conversion to total hip arthroplasty and a dual-mobility bearing.

TABLE I.

Patient and Surgical Characteristics of BHR Revisions

Patient	Age at Surgery (yr)	Head Size (mm)	Time to Revision (yr)	Reason for Revision	Type of Revision
1	42.1	48	3.1	Pseudotumor, elevated metal ions	Femoral component revision, conversion to dual-mobility
2	36.8	50	3.7	Infection	2-stage exchange
3	46.2	52	0.7	Infection	2-stage exchange
4	55.7	54	4.8	Infection	2-stage exchange
5	53	48	5.6	Unexplained pain	Revision of both components to THA
6	52.3	50	5.7	Femoral component loosening/collapse	Femoral component revision, conversion to dual-mobility
7	53.3	50	12.0	Femoral component loosening/collapse	Femoral component revision, conversion to dual-mobility
8	56.2	50	12.1	Elevated metal ions	Femoral component revision, conversion to dual-mobility

Patient-Reported Outcomes

The mean mHHS was 92.65 ± 11.0 in the BHR group and 93.56 ± 9.4 in the THA group (p = 0.44) (Table II). The mean change in UCLA score compared with the preoperative score was 0.03 ± 2.5 for the BHR group and 0.17 ± 2.4 for the THA group; the difference between the groups was not significant (p = 0.31). When the analysis was restricted to patients with preoperative UCLA scores of 7 to 10, reflecting a very active lifestyle, both groups had similar decreases at the time of the latest follow-up (p = 0.80). Furthermore, the reasons for the change in activity levels were similar between the 2 groups, with “advanced age or no interest” being the most common reason in both groups (45.8% in the BHR group and 47.3% in the THA group) followed by “other orthopaedic problems” (35.4% in the BHR group and 37.6% in the THA group) (Table III). These findings were similar when the preoperative highly active patients were analyzed as a group (Table III). Satisfaction with the overall function of the hip was similar between all patients in the BHR group and the matched patients in the THA group (p = 0.8). In the group with higher preoperative activity levels, the satisfaction with overall function was also similar, although this difference approached significance (p = 0.12) (Table II). The rating of the hip as a percentage of a normal hip was also similar at the time of the latest follow-up among all patients (89.5 ± 11.24 for the BHR group and 88.8 ± 14.5 for the THA group; p = 0.67) (Table II).

TABLE II.

Patient-Reported Outcomes

	Preoperative UCLA 1-10			Preoperative UCLA 7-10
	BHR (N = 224)*	THA (N = 119)*	P Value	BHR (N = 139)*	THA (N = 74)*	P Value
Percentage of normal hip	89.5 ± 11.24	88.8 ± 14.5	0.67	89.8 ± 10.16	87.5 ± 16.6	0.25
Modified Harris Hip score	92.65 ± 11.0	93.56 ± 9.4	0.44	92.7 ± 11.5	94.39 ± 8.04	0.28
Satisfaction with overall function			0.8			0.12
Extremely	68.7%	59.3%		67.9%	62.0%
Very	20.1%	22.0%		23.9%	22.0%
Quite	6.1%	14.3%		4.5%	12%
Somewhat	1.4%	4.4%		0.7%	4.0%
Not satisfied	3.7%	0%		3.0%	0%

^*The values are given as the mean ± deviation or as the percentage of patient responses.

TABLE III.

Activity Levels for BHR and THA Groups

	Preoperative UCLA 1-10			Preoperative UCLA 7-10
	BHR (N = 224)*	THA (N = 119)*	P Value	BHR (N = 139)*	THA (N = 74)*	P Value
Postoperative UCLA score	7.7 ± 2.1	7.2 ± 2.1	0.02 †	7.8 ± 2.1	7.9 ± 1.6	0.82
Change in UCLA score	0.03 ± 2.5	0.17 ± 2.4	0.31	−0.94 ± 2.07	−0.86 ± 1.8	0.80
Percentage of patients meeting high-activity UCLA score (9-10)	60.5%	52.2%	0.45
Why activity level changed			0.584			0.77
Symptoms in operatively treated hip	15.3%	9.7%		14.1%	9.1%
Other orthopaedic problem	35.4%	37.6%		35.3%	32.7%
Other medical problem	3.5%	5.4%		4.7%	5.5%
Advanced age or no interest	45.8%	47.3%		45.9%	52.7%

^*The values are given as the mean ± deviation or as the percentage of patient responses.

^†Significant.

Metal Ion Levels

Of the 224 patients in the BHR group, 115 were available for updated metal ion levels. There was no significant difference in BMI (p = 0.921) between those with metal ion levels and those without, but more older patients had metal ions available (p = 0.016). Serum cobalt ranged from 0.0 to 9.8 ppb (mean, 1.8 ± 1.5 ppb; median, 1.4 ppb). Serum chromium ranged from 0.4 to 12.9 ppb (mean, 2.2 ± 2.0 ppb; median, 1.5 ppb). When patients with bilateral BHR were excluded, the remaining 93 patients had serum cobalt values ranging from 0.0 to 9.8 ppb (mean, 1.7 ± 1.48 ppb; median, 1.10 ppb) and serum chromium values ranging from 0.4 to 8.0 ppb (mean, 1.9 ± 1.46 ppb; median, 1.40 ppb).

Radiographic Results

Of the 224 patients who were included in this follow-up study, 127 (57%) had anteroposterior pelvic and cross-table lateral radiographs available for review. There were no differences between those with or without radiographs in terms of age (p = 0.161) or BMI (p = 0.213). Of the available patients, 13 (10.2%) had any radiolucency around the acetabular component and 4 (3.1%) had any radiolucency around the femoral component (Table IV). Five patients had neck narrowing, but none of those patients had progression to component failure. One of those 5 patients had remodeling-type neck narrowing of 10% that stabilized after 3 years. The other 4 patients had impingement-type neck narrowing (Fig. 3). All unrevised components were radiographically well-fixed at the most recent follow-up.

Fig. 3.

Anteroposterior pelvic radiographs demonstrating impingement-type neck narrowing of the left hip at 5-year intervals. Progressive posterior pelvic tilting can be appreciated at 10 and 15 years postoperatively.

TABLE IV.

Implant Radiolucencies

Component	No. of Patients
Acetabular component
Charnley Zone I	4
Charnley Zone II	3
Charnley Zone III	4
Charnley Zones I and II	1
Charnley Zones II and III	1
Femoral component
Hing Zone 1	1
Hing Zone 2	1
Hing Zones 1 and 3	2

Discussion

This long-term follow-up study from our original U.S. series, beginning in 2006 after the BHR implant was first approved by the FDA, supports the efficacy and longevity of this implant in appropriately selected patients. With our current indications, we previously reported survivorship free from all-cause revision of 98.1% at 5 years and 96.8% at 10 years. The survivorship free from aseptic revision was 99.5% at 5 years and 98.2% at 10 years. In this updated series, the 15-year survivorship free from any revision and from aseptic revision was excellent, at 96.0% and 97.4%, respectively. This study also demonstrated that the activity level beyond 10 years after the index BHR was not different from that in males of similar ages and preoperative activity levels who underwent THA.

The survivorship in this long-term follow-up study is similar to that in other U.S. studies evaluating outcomes of BHRs at 10 years. Samuel et al.⁶ evaluated the outcomes after BHR in a single-surgeon series of 350 hips and reported a 10-year survivorship of 99% in males who had a mean age of 52 years at the time of surgery. Su et al.⁷ evaluated 10-year outcomes after BHR in a multicenter study of 218 hips and reported survivorship of 96% in males under 65 years of age at the time of surgery. Both of those studies included female patients because they were performed at a time when female sex was not a contraindication for BHR. In our original series, the 10-year survivorship free from aseptic revision was 98.2% in males <60 years old⁸. The present long-term report of 224 BHRs demonstrated continued excellent survivorship beyond 15 years.

One of the advantages of the BHR is its ability to accommodate younger patients, allowing them to maintain high activity levels postoperatively. Previous comparative studies and systematic reviews have demonstrated advantages in postoperative activity levels^2,12-14. In a systematic review, Hellman et al. concluded that SRA allowed for better activity scores and return to high-level activities compared with THA². A 2-year comparison study by MacKenzie et al.¹⁴ and a 10-year comparison study by Haddad et al.¹² both demonstrated higher proportions of patients with an SRA who were participating in higher-level activities at the time of the latest follow-up^12,14. Our prior study also demonstrated a higher proportion of patients with BHR participating in high-level activities compared with THA at 5 and 10 years postoperatively⁸. In the current long-term follow-up study, we demonstrated a gradual decline in activity levels beyond 10 years in both groups. This finding underscores the importance of patient factors beyond the type of implant. Patients in both the BHR and THA groups reported loss of interest or advanced age to be the most common reason for a decline in activity level. This response was also the most common one among patients with higher preoperative activity levels (UCLA scores of 7 to 10).

Metal debris and metal ions are a concern with MoM bearing surfaces. In the present study, the metal ions at long-term follow-up remained generally stable at acceptable levels, which is similar to the findings in other reports. Holland et al. reported similar metal ion levels at a mean of 9.5 years of follow-up in a study from the United Kingdom (median chromium concentration, 1.74 ppb; median cobalt concentration, 1.67 ppb)¹⁵. One of the exceptions in the current follow-up study involved a patient who had late increasing metal ions and pseudotumor formation beyond 10 years postoperatively. Radiographic comparison in that case demonstrated progressive posterior pelvic tilt, which functionally increases the abduction angle and anteversion of the acetabular component, thereby increasing the likelihood of edge-loading and associated metal ion release¹⁶. Tamura et al. assessed pelvic sagittal inclination 10 years after THA and observed posterior pelvic tilting over time¹⁷. In that study, no late complications were determined to be related to the pelvic position.

The present study is not without limitations. The survey-based patient-reported outcome data are limited by response bias, including recall bias regarding the reasons why the patients became less active after surgery. Furthermore, postoperative activity was influenced by postoperative recommendations to avoid high-impact activities following THA. Additionally, this study was limited by lack of radiographic follow-up for all patients. If all patients had anteroposterior and cross-table lateral radiographs available for review, this study could have included a larger patient cohort. Although this study is generalizable to the patient population meeting modern indications for SRA, a larger study population would have allowed the results to be more reliably generalizable to this cohort.

Overall, on the basis of implant survival and level of satisfaction, this long-term follow-up study supports the use of BHR in select patients. Long-term monitoring is suggested for patients with pain or elevated metal ions. Factors unrelated to implant choice were associated with activity level decline over a decade postoperatively. Patients should be counseled that although activity levels have been shown to be higher in the first decade after SRA, activity gradually equalizes to that after THA due to the effects of age.

Appendix

Supporting material provided by the authors is posted with the online version of this article as a data supplement at jbjs.org (http://links.lww.com/JBJS/I414).