CASE 1. A sixty-seven-year-old woman had a right total hip arthroplasty in January 1989 because of degenerative osteoarthrosis. The procedure was performed through a lateral approach, with use of an anterolateral trochanteric osteotomy that was reapproximated with number-2 Ethibond sutures (Ethicon, Somerville, New Jersey). A Triloc acetabular cup with a non-constrained liner that had a 10-degree elevated rim was inserted, without cement, along with a proximally porous-coated anatomic medullary locking femoral stem (all from DePuy, Warsaw, Indiana). Radiographs made in April 1993 were suggestive of failure of the polyethylene acetabular liner with superior migration of the femoral head. The hip was revised with insertion of a larger acetabular cup (Duraloc; DePuy) and a non-constrained liner as well as exchange of the femoral head. The patient subsequently had four additional episodes of posterior-superior dislocation. Each episode occurred after the limb had been placed in marked adduction. A second acetabular revision was performed in November 1993. At the time of the operation, a non-union was noted at the site of the anterolateral trochanteric osteotomy. A non-constrained polyethylene liner with a 10-degree elevated rim (DePuy) was placed in a new Duraloc acetabular component, and the twenty-eight-millimeter femoral head was exchanged.
Within three months after the second revision, three additional dislocations occurred, all of which necessitated reduction with the patient under general anesthesia. An S-ROM constrained acetabular liner (L-32 series; Joint Medical Products) with a 10-degree elevated rim was inserted in March 1994. The patient did well for nearly two years after that revision, but a fracture of the constraining ring of the acetabular liner was found during a routine follow-up examination (Fig. 2-A). An anterior dislocation occurred in February 1996, after the right limb had been hyperextended (Fig. 2-B), and the patient was taken to the operating room for open reduction. Intraoperatively, the polyethylene liner was found to be damaged at the apex of the elevated rim; the damage was thought to be secondary to impingement by the femoral neck on the rim of the liner (which, in turn, had resulted in the dislocation) or to the impact of the femoral head as it was dislocating. The constraining ring was fractured through the serial numbers that were etched on its outer surface (Fig. 2-C). Revision was performed with use of an S-ROM constrained acetabular liner with a neutral rim and a thirty-two-millimeter plus five-millimeter femoral head. At the time of the most recent follow-up, in March 1997, the patient was asymptomatic and had no signs that were suggestive of recurrent subluxation or dislocation.
CASE 2. A thirty-five-year-old woman sustained blunt trauma to the right hip in a motor-vehicle accident in November 1993. The injuries included a comminuted fracture of the posterior wall of the right acetabulum, which was treated non-operatively at another institution. Marked post-traumatic osteoarthrosis subsequently developed in the right hip. A right total hip arthroplasty was performed without cement in March 1995; the components included a porous-coated acetabular cup, a non-constrained acetabular liner with a 10-degree elevated rim, and a hydroxyapatite-coated femoral stem. Postoperatively, the patient noted a clunking sensation in the right hip with associated pain in the thigh.
The patient was seen at our institution in August 1996 because of posterior dislocation, which had occurred after marked flexion with the hip in adduction. The patient was managed with closed reduction in the emergency room and was instructed to restrict weight-bearing and to limit the range of motion for approximately three weeks. A second dislocation occurred within six weeks after the first; once again, the treatment consisted of closed reduction with the patient conscious but sedated. A revision was recommended.
The revision was performed in October 1996. The intraoperative examination revealed attenuation and partial detachment of the hip abductors and the short external rotators at the site of their insertion on the proximal part of the femur (findings that were consistent with the posterior approach that had been used for the previous operation) as well as moderate synovial hypertrophy. The acetabular and femoral components were loose, and the femoral stem was in marked anteversion (more than 30 degrees). The revision consisted of insertion of a Duraloc acetabular cup (300 series; DePuy), a non-constrained acetabular liner with a 10-degree elevated rim, a twenty-eight-millimeter plus six-millimeter ceramic femoral head, and a porous-coated S-ROM modular femoral stem.
The patient sustained two additional dislocations within two months after the revision. Both dislocations occurred when the patient, after excessive intake of alcohol, failed to adhere to the recommended restrictions of the range of motion. A second revision was therefore performed in December 1996, at which time the patient was found to have a non-union at the site of the anterolateral trochanteric osteotomy. An S-ROM acetabular component with a constrained polyethylene liner that had a neutral rim was inserted. The femoral components were retained. The patient initially did well, but she returned at six weeks because of dislocation (Fig. 3-A). Despite the fact that the constraining ring was intact, closed reduction was attempted. The attempt was not successful (Fig. 3-B), and open reduction was performed. Intraoperative examination of the polyethylene liner revealed wear of the inferomedial part of the rim, which appeared to be the result of impingement by the femoral neck. The worn area presumably was a site of contact between the components during extreme adduction and internal rotation, which was thought to have caused the femoral head to lever out of the stable acetabular component. The intact constraining ring was removed, the femoral head was reduced, and the constraining ring was reapplied to the existing acetabular liner. An effort to improve the orientation of the implant was not believed to be justified.
Dorr et al. classified instability after total hip arthroplasty according to three general categories. Type-I instability is positional; that is, it occurs after the hip is placed in a position that exceeds the mechanical stability of the prosthesis. Type-II instability is caused by soft-tissue imbalance and usually involves loss of the abductor mechanism. Type-III instability is caused by malpositioning of the femoral component or the acetabular component, or both. Specific variables within these categories, including trochanteric non-union with decreased strength of the abductor muscles, a high center of hip rotation, a previous operation, a revision, a posterior operative approach, acetabular retroversion, and a shorter length of time since the operation (with a greater risk within thirty days after the procedure), have been associated with dislocation of the hip8,10,15,16. The degree of patient compliance raises another complicating issue for the treating surgeon. Although compliance cannot be quantitated, it remains an important variable related to the development of instability after total hip arthroplasty.
Attaining the ideal orientation of the components and the proper soft-tissue tension while minimizing the possibility of soft-tissue interposition allows for an enhanced functional range of motion. In addition, the range of motion after total hip arthroplasty is directly related to the size of the femoral head and is inversely proportional to the diameter of the femoral neck11. The treatment options for instability include the use of a hip brace3 or an above-the-knee hip-spica cast15; revision of the acetabular component or the femoral component, or both3,6,8; and augmentation of the acetabular component1,12,14. The constrained acetabular liner offers an additional option for reestablishing stability in patients who have recurrent dislocation after primary or revision arthroplasty. The indications for the use of a constrained liner include marked shortening of the extremity, loss or weakness of surrounding muscles and soft tissues, multidirectional instability intraoperatively, recurrent instability after a previous revision, and conditions that make it difficult to enforce restrictions because of alterations in the patient's mental status or muscle control (for example, senile dementia, seizures, and Parkinson disease)1,2,7,11,13.
In 1986, Bryan and Reeve reported the first case, to our knowledge, in which an articulated total hip arthroplasty failed because of a fatigue fracture of the constraining ring that resulted in recurrent dislocation. The implant used in that patient was an SRN total hip prosthesis (Joint Medical Products). Chronic impingement of the constraining ring against the neck of the femoral component was postulated to have been the cause of failure of the ring.
Anderson et al. as well as Fischer and Kiley retrospectively reviewed the cases of a total of seventy-three patients in whom a constrained liner was used. Eleven of the seventy-three patients had a dislocation. Those authors described several different modes of failure, including dissociation of the polyethylene liner from the acetabular cup (a so-called pull-out mechanism) and disengagement of the femoral head from the liner with or without failure of the metal constraining ring on the neck of the liner (a so-called lever-out mechanism). A review of the cases of the patients who had recurrent dislocation suggested that the only factor that was predictive of failure of the constrained liner was an increased abduction angle of the metallic acetabular cup7. Lombardi et al., in a retrospective review of the results achieved with use of the S-ROM constrained acetabular insert, reported a rate of redislocation of 9 per cent (five of fifty-five patients). This rate compared favorably with the rate of 19 per cent (thirty-three of 176 patients) that had been observed after revisions performed with use of a standard, non-constrained liner11. Those authors noted that 600 pounds (272 kilograms) of direct outward pull was needed to dislocate a thirty-two-millimeter femoral head from a constrained S-ROM component whereas 325 pounds (147 kilograms) was needed to dislocate a twenty-eight-millimeter femoral head from such a component; however, this variable was not analyzed in their results.
The use of an acetabular liner with a 10-degree or 20-degree elevated rim may be advantageous when a patient has a non-constrained arthroplasty. However, when a patient has a constrained arthroplasty, the use of such a liner may actually contribute to the lever-out mechanism (depending on the placement of the elevated portion of the rim) by allowing the femoral neck to make contact with the polyethylene liner earlier than would be the case if the liner had a neutral rim, regardless of the size of the femoral head. The lever-out phenomenon also may occur when the femoral head has a long so-called apron or skirt, which would make contact with the polyethylene liner through a compromised range of motion. At the extreme limits of motion, contact occurs eventually, regardless of the configuration of the polyethylene.
Failure of the constraining ring on an acetabular liner, as was seen in two of our four patients, is an unusual but worrisome complication. Our review of the literature revealed only one report of such failure2. The fact that the lever-out mechanism can occur even when the constraining ring is intact further emphasizes the need for careful positioning of the components and reestablishment of the soft-tissue envelope about the hip. Whether an improvement in the design or the material composition of the constraining ring would decrease the rate of dislocation by the lever-out mechanism without increasing the likelihood of dislocation by the pull-out mechanism merits additional research. Patient compliance remains one variable over which the surgeon has little control. As with other modes of failure, dislocation of a hip in the setting of a constrained liner necessitates open reduction and, if necessary, simultaneous revision.
Twelve patients have been managed with a constrained acetabular liner at our institution to date. The fact that four of these twelve liners failed suggests that this type of implant should be used cautiously, in only selected settings. Careful attention to reestablishment of the dynamic and static soft-tissue envelope, proper orientation of the implants, and an intraoperative dynamic evaluation of the range of motion of the hip with the implants in place should enhance the benefits associated with the constrained acetabular liner. The avoidance of a polyethylene liner with an elevated rim and consideration of the use of a thirty-two-millimeter femoral head, despite concerns regarding increased wear of the polyethylene, may also contribute to stability.