Recently, metal-on-metal hip resurfacing has experienced a resurgence as an alternative to traditional total hip arthroplasty1,2. It is a procedure that offers some unique advantages over total hip arthroplasty but also is associated with some unique modes of failure. The obvious advantage lies in that fact that it is bone-sparing as the femoral neck is not transected. Herein lies its major drawback as a small number of patients who undergo hip resurfacing may incur early failure secondary to femoral neck fracture3-7.
Hip resurfacing is a technically demanding procedure. Much like traditional total hip replacement, it may be performed through a number of surgical approaches8,9. Most commonly, a posterolateral or an anterolateral exposure is employed. Each approach has its own advantages and disadvantages.
The purpose of the present report is to describe our technique involving the anterolateral approach and our early experience with metal-on-metal hip resurfacing. We include an analysis of early results, complications, and technical recommendations based on these cases.
Between July 2001 and April 2006, a total of fifty-seven hip resurfacing procedures in fifty-three patients were performed by a single surgeon (M.A.J.) through an anterolateral approach to the hip. The Cormet Hip Resurfacing system (Corin Group, Cirencester, Gloucestershire, United Kingdom) was used. The present study was part of a United States Food and Drug Administration investigational device exemption study. All patients had a minimum of two years follow-up (mean, thirty-eight months; range, twenty-nine to seventy-two months).
Forty-one hips (72%) were in men, and sixteen (28%) were in women. The patients had a mean age of fifty-two years (range, twenty-six to seventy-three years) and a mean weight of 191 lb [86.6 kg] (range, 110 to 275 lb [49.9 to 124.7 kg]) at the time of surgery. The diagnosis was primary osteoarthritis in fifty-five hips and osteonecrosis of the femoral head in two hips.
All patients were followed prospectively at six weeks, six and twelve months, and then annually with use of the Harris hip score. Radiographic evaluation was also performed at these time-intervals by an independent observer. Complications were recorded as systemic, hip-related, or device-related.
Surgical Technique
All procedures were performed in a positive-pressure surgical suite with a standard headlight worn by the primary surgeon (M.A.J.). Either general endotracheal or spinal anesthesia was administered in all cases. With this approach, a standard straight lateral incision between 10 and 14 cm is utilized (Fig. 1). After the fascia is incised, the leading edge of the gluteus medius muscle is identified and several small vessels that traverse the fat are cauterized. The anterior 25% of the gluteus medius muscle and the underlying gluteus minimus muscle are released from their tendinous attachments. The anterior, superior, and inferior part of the capsule are resected. The posterior part of the capsule is left intact. With experience, the inferior part of the capsule is best resected in two steps. An initial resection is performed with the hip located. After dislocation, the femoral head and neck are retracted in a posterior direction and the remaining inferior part of the capsule is cut after palpation and direct visualization. This step is critical as acetabular exposure can be difficult, and this release of the inferior part of the capsule is essential to allow adequate posterior retraction of the femoral head and neck.
The next step is placement of the femoral head-neck guidewire. Osteophytes are removed from the femoral neck to help to confirm anatomical landmarks (Fig. 2). Initially, a targeting jig was employed. With this technique, the distance from the tip of the greater trochanter to the insertion point on the lateral part of the femur was measured on the preoperative radiograph. This distance was then transferred to the proximal part of the femur with use of a jig. This technique, however, was found to be inaccurate. Later, a technique evolved in which the center of the femoral head and neck is visually estimated (Fig. 3). To aid in this process, the appropriate cylindrical reamer is placed over the femoral head and a guidewire is placed through its shaft and is marked (Fig. 4). The actual insertion point is a compromise between the actual center of the head and adjustments made for any head deformities that moved this point eccentrically and would lead to the creation of a femoral notch. Early in our experience, several portable radiographs were made with the guidewire in place to ensure proper placement (Fig. 5). With time, this step was eliminated.
Once the guidewire is placed, preliminary cylindrical reaming (oversized by 6 mm) is performed to allow the surgeon to assess the eventual size of the femoral component. This maneuver also allows debulking of the femoral head to provide easier access to the acetabulum. This step is important as hip resurfacing is distinctly different from total hip arthroplasty in that the femoral head and the acetabular component are not freely interchangeable. They are essentially coupled by size and, as such, the surgeon always must be sure that the acetabular size that is chosen will allow for a femoral component that fits properly and will not notch the femoral neck. Early in our experience, it was necessary to switch back and forth from the femur to the cup several times during the procedure to be certain that the femoral and acetabular sizes would match.
Acetabular exposure through the anterolateral approach can be challenging. If general anesthesia is utilized, complete muscle paralysis is helpful. A cobra-type retractor is placed over the anterior pelvic brim, and the first assistant translates the femur in a posterior direction. Sometimes a bone hook placed around the femoral neck can be helpful. In this position, the inferior part of the capsule is again examined, and usually there is additional tissue that must be incised to allow increased mobility of the femur (Fig. 6). It is emphasized that a substantial portion of the posterior part of the capsule is maintained for hip stability, and doing so also theoretically may preserve the blood supply to the femoral head. After resection of the labrum, the acetabular reaming is accomplished as in a standard total hip arthroplasty. A headlight may aid this part of the procedure.
After acetabular component placement, the femoral preparation is completed and a trial reduction is performed. It is critical to make a radiograph at this point in time in order to confirm correct acetabular positioning and seating (Fig. 7). General femoral component positioning also can be ascertained on the radiograph. This step is important as the acetabular component is placed through a small window and the insertion handle for these cups is usually somewhat bulky and can make an assessment of component status difficult. In addition, if the femoral component alignment is not acceptable, then the surgeon may elect to switch to a traditional total hip arthroplasty.
Once these steps have been successfully accomplished, it is time for femoral component cementing. In preparation for this step, it is important to create two or three small channels in the femoral head to allow for cement extravasation. If this is not done, a mantle of cement may be trapped under the cap and may prevent the component from fully seating. This trapped cement has been associated with early femoral fracture3. The cement also must be sufficiently liquid that it can flow out through these channels. This degree of fluidity demands the use of either low-viscosity cement or the use of chilled monomer with standard-viscosity cement. This step was not performed during the first forty procedures in the present series, and we believe that a femoral component that sat somewhat proud on an excessive cement mantle was a factor in at least two of the postoperative fractures. Closure is accomplished in a routine fashion.
Postoperatively, the patients were kept at 20% weight-bearing for the first six weeks and then were advanced to full weight-bearing as tolerated. For the past two years, this protocol has been changed, and all patients have been allowed to bear weight as tolerated immediately after surgery.
The mean Harris hip score improved from 52 points (range, 28 to 75 points) preoperatively to 99 points (range, 89 to 100 points) at the time of the most recent evaluation at a mean of thirty-eight months (range, twenty-nine to seventy-two months). At the time of the most recent follow-up, fifty hips (88%) had no pain, five had slight to mild pain, one had mild pain, and one had moderate pain. No patient reported marked or disabling pain. Fifty-five (96%) of the fifty-seven hips were in patients with normal function scores, and two were in patients who reported mild dysfunction.
Three hips (5%) had a failure that required revision. Two of these failures were completely displaced femoral neck fractures. One fracture occurred at one year postoperatively, and one occurred at twenty-eight months postoperatively; both hips were revised to total hip arthroplasty and were doing well at the time of the most recent follow-up visit (at two and three years after revision) with Harris hip scores in excess of 90 points. The third revision was performed because of loosening of the acetabular cup secondary to debonding of the surface coating, which occurred at twenty-three months postoperatively. The details of this case have also been submitted in a separate case report (presently unpublished). This debonding was thought to be the result of a mechanical failure of the manufacturing process. The patient had had an excellent result prior to the debonding. The cup was revised to a larger acetabular component with a conventional femoral stem, and the patient was doing well (with a Harris hip score of 95 points) at three years after revision.
Two other femoral neck fractures were successfully treated nonoperatively. Both fractures healed with protected weight-bearing and had excellent clinical results. In one case, the fracture was nondisplaced and united without complications. The components remained in their original postoperative position. At the time of the most recent follow-up (at fifty-two months), the patient continued to have an excellent result. The second fracture healed in varus, and the femoral component also shifted into varus. At forty-eight months, the patient continued to have an excellent clinical result, although the position of the femoral component created an uncertain prognosis for the overall longevity of the implant. These two cases have been detailed in a separate report that has been submitted for publication (currently unpublished).
With the exclusion of the hip that had cup failure, none of the hips demonstrated any radiolucent lines or progressive radiolucent lines about the acetabular component on radiographic evaluation. Similarly, none of the cups (excluding the one that failed) demonstrated any evidence of migration or tilt. On the femoral side, as noted above, one femoral component shifted into marked varus following a femoral neck fracture that healed in varus. None of the other hips had evidence of femoral component subsidence.
There were no deaths, pulmonary emboli, myocardial infarctions, deep infections, or hip dislocations. With the exclusion of the patient who sustained a femoral fracture that healed in a varus position, no patient had a limb-length discrepancy of >3/8 in (>0.95 cm). None of the patients used a shoe lift.
While the ultimate role for hip resurfacing remains unclear, it has been established as a potential option for younger, high-demand patients8. It can be performed through several different approaches with use of techniques that have been developed in association with the standard total hip arthroplasty experience. Although the posterolateral approach is more commonly utilized, we have found that the anterolateral approach is preferable. One reason for choosing the anterolateral approach is because of its documented association with a decreased rate of hip dislocation. This consideration is particularly important when using a resurfacing system because femoral neck impingement might occur more readily as a result of the absence of the head and neck offset that is present in association with traditional femoral components. Any contact of pelvic bone on the femoral neck not only can dislocate the head but also potentially could become the origin of a stress-riser leading to a femoral neck fracture. This lack of a femoral head and neck offset partially negates some of the intrinsic advantages of the large-diameter heads of the resurfacing systems.
To allow for adequate visualization of the acetabulum from the posterolateral approach, many have advocated the employment of extensile steps, which can include the release of the tendons of the gluteus maximus and gluteus medius. This amount of soft-tissue release places enormous value on the soft-tissue closure, and any failure of the closure could lead to posterior instability. In our series, none of the hips that were resurfaced through an anterolateral approach dislocated or subluxated.
Another potential advantage of the anterolateral approach is that there is evidence that it preserves the blood flow to the femoral head more consistently that the posterolateral approach does. Khan et al. compared the blood flow in the femoral head of patients undergoing resurfacing through both the posterolateral and anterolateral approaches9. The authors found that the blood flow, as measured on the basis of the cefuroxime concentration in the femoral head, to be significantly higher in patients who had had the anterolateral approach. In addition, cefuroxime was undetectable in association with one procedure that had been performed through the posterolateral approach. Forrest et al. compared the blood flow in the hips of patients in whom one hip had been resurfaced through an anterolateral approach and the contralateral hip was normal and had never undergone any surgery10. The authors found the blood supply of the two hips to be comparable.
Our experience with hip resurfacing reveals that it is a technically demanding procedure with a high number of early failures. The cases presented here represent the senior author's (M.A.J's.) initial experience with the procedure at a time when virtually all of the training centers were performing surgery exclusively through the posterolateral approach. Despite this, there were only three failures overall, and one was secondary to a mechanical failure of the acetabular component. To decrease the chance of failure, several factors should be considered. Proper patient selection is important as patients with abnormalities of the femoral head are not good candidates. Patients with osteoporosis, large bone cysts, or large areas of osteonecrosis are at higher risk for femoral-side failure3,4. Technically, we began by making several radiographs with the femoral head guidewire in place to be sure that its placement was accurate. With time, we were able to eliminate this step, but it prevented us from having any malpositioned femoral components early in the study. Another technical step that was found to be helpful was to make a radiograph with the cup in place and a trial cap on the femur. This step confirmed complete seating of the cup and helped us to assess limb lengths and femoral component position. It is also recommended that the inferior part of the capsule be cut in two steps to help to obtain acceptable acetabular exposure. Finally, one should understand that the femoral and acetabular components are size-linked so that it may be necessary to switch back and forth from the femur to the cup in order to determine the correct size combination.
In conclusion, we have presented our initial experience with hip resurfacing through an anterolateral approach. The early results are encouraging, and we continue to use this technique. Refinements that have been made as the study progressed have been described. Although we continue to use this surgical exposure for resurfacing for the reasons described above, we believe that successful results can also be consistently achieved with other exposures2. 