Vol 10, No 4, July/August 2002 233 Dislocation is one of the most com- mon and distressing early complica- tions of total hip arthroplasty. The reported incidence of dislocation ranges from 0% to 10% after a prima- ry arthroplasty and from 10% to 25% after a revision arthroplasty. A wide variety of predisposing causes and associated factors have been suggest- ed. 1,2 Pellicci et al 3 described the use of a posterior approach and enhanced soft-tissue repair in an attempt to decrease the early incidence of dislo- cation. Nonsurgical treatment of the initial dislocation with a cast or brace is successful in approximately two thirds of patients. However, when surgical treatment is required for recurrent dislocation, satisfactory results have been achieved in only 60% of hips using a wide variety of techniques. 1 Additionally, the chance of success is even less when a precise etiology cannot be determined. It is for these situations that constrained components have been considered. By definition, constrained total hip arthroplasty components include a mechanism that locks the prosthetic femoral head into a polyethylene acetabular component. A thorough understanding of the design features of constrained components in total hip arthroplasty, indications for their use, and results and complications is essential for the effective application of this technique. Historical Perspective The use of constrained total hip arthroplasty components has been limited. 4 Sivash first reported on his constrained prosthesis in 1963 in Moscow, at a conference on tubercu- losis of bones and joints. 5 The Sivash prosthesis was a locked one-piece prosthesis, with the cup and head- neck components fabricated as a connected whole (Fig. 1). The first components were fabricated of steel; they were later modified to include chrome-cobalt and titanium alloys. The acetabular component was a threaded hemisphere made of a tita- nium alloy and was available in 51-, 57-, and 65-mm diameter sizes. The femoral component had a chrome- cobalt head welded onto a titanium- alloy stem and was available in three sizes: 14-, 16-, and 18-mm proximal diameter. The articulating surface was polyethylene. Fixation was Dr. Lachiewicz is Professor, Department of Orthopaedics, University of North Carolina at Chapel Hill, Chapel Hill, NC. Dr. Kelley is Associate Professor, Department of Orthopaedics, University of North Carolina at Chapel Hill, Chapel Hill. One or more of the authors or the departments with which they are affiliated have received something of value from a commercial or other party related directly or indirectly to the sub- ject of this article. Reprint requests: Dr. Lachiewicz, 242 Burnett- Womack Building, CB 7055, Chapel Hill, NC 27599. Copyright 2002 by the American Academy of Orthopaedic Surgeons. Abstract The use of a constrained component may be appropriate for the surgical treatment of recurrent dislocation due to soft-tissue insufficiency following a total hip arthroplasty. Constrained components usually include a locking mechanism incorporated into the polyethylene acetabular liner to keep the prosthetic femoral head in place. Two differ- ent prosthetic designs are available and have been approved by the U.S. Food and Drug Administration. The S-ROM constrained component uses additional polyeth- ylene in the rim, which deforms to more fully capture the femoral head and then is held in place by a metal locking ring. The Howmedica Osteonics constrained compo- nent is a tripolar device; its bipolar component articulates with another polyethylene liner. These constrained components transfer hip forces that would otherwise lead to dislocation to the locking mechanism, the liner-shell interface, or the bone-prosthesis interface. These forces may eventually contribute to failure of the component due to loosening, dissociation, breakage, or recurrent dislocation. Studies of these compo- nents show a failure rate of 4% to 29% at relatively short-term follow-up. J Am Acad Orthop Surg 2002;10:233-238 The Use of Constrained Components in Total Hip Arthroplasty Paul F. Lachiewicz, MD, and Scott S. Kelley, MD Perspectives on Modern Orthopaedics either press-fit or cemented. After appropriate reaming of both surfaces, the femoral component was implant- ed, followed by impaction or cement- ing of the locked-on acetabular com- ponent. Sivash reported its use in 200 cases, with 1- to 9-year follow-up in 169 patients. 5 The most common indications were ankylosing spondy- litis in 107 patients and tuberculous arthritis in 56 patients. Although there was no detailed analysis of results, Sivash reported that the pros- thesis fractured in 13 hips. A modified Sivash prosthesis with specially designed rasps was described in 1974. 6 A case report in 1981 described the successful use of this modified prosthesis for recurrent dislocation and anecdotally suggest- ed that cerebral palsy, Parkinson’s disease, and loss of hip muscula- ture were indications for its use. 7 Although the prosthesis was used ex- tensively in Europe, 8 it was used only sporadically in the United States. Bryan and Reeve 9 described a case of a patient with recurrent dis- location who was treated with this device. Failure was eventually caused by fatigue fracture of the con- straining ring and severe polyethyl- ene wear and metal-metal abrasion. Koffman 10 reported the use of three different designs of constrained com- ponents (including the Sivash) in five hips of four patients with spastic cerebral palsy. The Sivash prosthesis was implanted in the only ambulato- ry patient and failed because of dislocation and early acetabular loosening. Current Constrained Designs Two constrained total hip arthro- plasty liner systems are presently approved by the U.S. Food and Drug Administration (FDA) and have data published on results of their use. They are the S-ROM con- strained acetabular liner (Poly-Dial; DePuy Orthopaedics Warsaw, IN) and the Howmedica Osteonics con- strained acetabular liner (Stryker Howmedica Osteonics, Rutherford, NJ). Other constrained liners have been used in FDA trials or are in development. The S-ROM constrained acetabu- lar liner has been available since 1987. It was marketed under a Premarket Notification from the FDA. To date, according to the com- pany, more than 6,000 have been implanted. This constrained acetab- ular liner was designed for use with S-ROM metal shells. The constraint is derived from the addition of extra polyethylene in the rim, which deforms to more fully capture the femoral head implant (Fig. 2). In addition, a capture ring provides increased constraint. The design of this component allows the head to dissociate from the liner before the forces dislodge the acetabular shell from the pelvic bone. Cameron 11 reported that the force required for withdrawal of this component is 60 inch-pounds and that the metal constraining ring increases the hold- ing power to 300 inch-pounds. Lombardi et al 12 reported that the metal ring provided a constraining force of more than 600 pounds with a 32-mm head and 325 pounds with a 28-mm head. However, they found that the amount of leveraged torque required to pry the femoral head out of the liner was 150 lbs/in 2 . The optimal amount of torque required for removal of the femoral head from a constrained acetabular component is not known. The S- ROM component, which is “dialed” into the acetabular shell, is currently available with an internal diameter of 28 or 32 mm and with a standard rim or a 10° elevated rim. The liner is available to fit acetabular shells with an outer diameter of 48 mm to 68 mm and is fabricated of cross- linked polyethylene, with a mini- mum thickness of 5 mm. The aver- age arc of motion (when used with Constrained Components in Total Hip Arthroplasty Journal of the American Academy of Orthopaedic Surgeons 234 Figure 1 The one-piece Sivash constrained component for total hip arthroplasty. (Reprinted with permission from Amstutz HC, Grigoris P: Metal on metal bearings in hip arthroplasty. Clin Orthop 1996;[329 suppl]:S11-S34.) Figure 2 The S-ROM constrained acetabu- lar liner with locking ring and correspond- ing uncemented metal shell. (Reprinted with permission from Kaper BP, Bernini PM: Failure of a constrained acetabular prosthesis of a total hip arthroplasty: A report of four cases. J Bone Joint Surg Am 1998;80:561-565.) an S-ROM femoral component) is reported to be 88° with a 28-mm head and 98° with a 32-mm head. 12 This arc of motion is probably less when an elevated rim liner is com- bined with a “skirted” modular femoral head component. The Howmedica Osteonics con- strained acetabular liner was intro- duced as a custom component in 1988 13 and was marketed under a Premarket Approval from the FDA, until recently converted to a class II device. This component is basically a tripolar device (Fig. 3): a polyeth- ylene inner liner is covered with a polished cobalt-chrome shell; the shell articulates with another poly- ethylene liner (the outer bearing), which is inserted into a standard noncemented acetabular shell. The inner liner accepts a 22-, 26-, or 28- mm femoral head and has a locking ring identical to the ring in a bipolar prosthesis. Some authors have suggested that the constrained acetabular liner can be cemented into a well-fixed acetab- ular shell or into an acetabulum pre- pared for cement fixation. 14 The Howmedica Osteonics constrained acetabular liner has been cemented into both an acetabular shell of another manufacturer and into an acetabulum prepared for cement fix- ation. 14 If the former technique is used, the surgeon should carefully consider preoperatively if this con- strained liner will fit. The smallest acetabular shell into which this liner could be safely cemented is probably 52 mm. According to the manufac- turer, the total range of motion is 72° when it is used with 50- to 54-mm outer acetabular shells, 82° with a 56- mm shell, and 84° with 58- to 74-mm shells. The two polyethylene articu- lating surfaces have a thickness rang- ing from 5.2 to 7.7 mm for the inner bearing and from 4.3 to 10.4 mm for the outer bearing. The polyethylene thickness varies based on femoral head size and acetabular shell diam- eter. The pullout strengths of the three segments of this tripolar liner have not been reported. Indications The use of a constrained acetabular component in total hip arthroplasty is indicated for recurrent dislocation of the hip due to soft-tissue insuffi- ciency (capsular or abductor muscu- lature) that is not amenable to repair or augmentation. If the abductor mechanism has been resected, then reconstruction with a constrained system may be required. Soft-tissue laxity (not insufficiency) due to short- ening of the prosthetic hip may be treated by lengthening the femoral neck and/or lateralizing the acetabu- lar component or liner. Component malposition, loosening, or wear should be treated by revision of one or both components. Dislocation re- sulting from impingement of bone or a “skirted” femoral head against an elevated-rim acetabular liner should not routinely be treated by a con- strained component. Bone impinge- ment can be treated by bone resection and impingement of the femoral head by revision of the head, liner, or acetabular component. Acute avulsion of a greater trochanteric osteotomy or fracture of the greater trochanter should be treated by sur- gical repair and/or advancement. However, recurrent dislocation due to a chronic nonunion of the greater trochanter, with severe and irrepara- ble loss of abductor muscle function, may be an indication for use of the constrained component. Recurrent dislocation associated with a large mismatch between the femoral head size and the outer acetabular compo- nent diameter, as reported by Kelley et al, 15 should be treated by revision to a larger head and corresponding acetabular liner, if possible. Late (>1 to 2 years postopera- tively) recurrent dislocation, which may be associated with weight loss, decrease in muscle mass, and/or chronic disease (cancer, rheumatoid arthritis) without component malpo- sition, is extremely difficult to treat. The constrained component may be a reasonable option in patients with these conditions. When late disloca- tion is associated with an acute or chronic infection, the treatment is complex, must be individualized, and may involve the use of a con- strained component. Contraindications for the use of constrained components include acute dislocation, dislocation due to component loosening or malposi- tion, insufficient acetabular bone structure, acute infection, skeletal immaturity, and neurologic spas- Paul F. Lachiewicz, MD, and Scott S. Kelley, MD Vol 10, No 4, July/August 2002 235 Outer bearing (UHMWPE) Inner bearing shell (CoCr) Acetabular shell Inner bearing (UHMWPE) Bipolar retaining ring (UHMWPE) Figure 3 A, The Howmedica Osteonics constrained acetabular liner. B, Schematic showing the tripolar nature. UHMWPE = ultra-high-molecular-weight polyethylene, CoCr = cobalt- chrome. (Adapted with permission from Stryker Howmedica Osteonics, Rutherford, NJ.) A B ticity. Neurologic spasticity may seem to be an attractive indication for the use of this component, but Root et al, 16 in reporting the results of total hip arthroplasty without con- strained components performed in patients with cerebral palsy, found that only 2 of 15 patients had a recur- rent dislocation, and both had com- ponent malposition. The prophylactic use of con- strained components in primary or revision total hip arthroplasty is con- troversial. Because good data are lacking, constrained acetabular liners should not be used routinely in these situations. Larger femoral head sizes, femoral necks with greater length and offset, and/or elevated rim liners are better choices. Results and Complications Theoretically, constrained acetabular components should transfer the forces that would otherwise lead to dislocation to the locking mecha- nism, the liner-shell interface, or the bone-prosthesis (or bone- cement) interface. If the hip center is shifted laterally, which may occur with either of the two available con- strained components, these forces may be increased. The reported results of constrained components have demonstrated four types of fail- ure: loosening of the acetabular component; 12 dissociation of the con- strained liner from the shell (with redislocation) 17,18 (Fig. 4); material failure (breakage) or disengagement of the constraining ring (with or without redislocation) 17,19 (Fig. 5); and dissociation of a modular femoral head from its neck. 20 An additional potential mode of failure is excessive wear of a thin acetabu- lar liner interface. According to information avail- able from the manufacturer, the S- ROM constrained acetabular liner has a low rate of dissociation- dislocation in the more than 6,000 implanted since 1987. However, a careful clinical review of these cases has not been performed, and there are relatively few published data on the component. Lombardi et al 12 reported a retro- spective review of 57 S-ROM con- strained acetabular liners implanted in 55 patients. Six were used in primary arthroplasties and 51 in revision arthroplasties. Of these, however, only 31 were done for dis- location, and of those, 13 patients had experienced multiple disloca- tions (average, 2.7; range, 2 to 5). Al- though the clinical follow-up period for the entire groupwas relatively short (mean, 30.2 months), two patients experienced early definite radiographic loosening of the ace- tabular component with screw breakage and migration. Five of 55 patients (9%) experienced failure due to redislocation at a mean of 2.5 months (range, 1 to 9 months) post- operatively. Three of these five patients had undergone the proce- dure because of recurrent disloca- tion, and thus the failure rate of the constrained component for this indi- cation was 23% (3 of 13). Open reduction was necessary when a dis- location of this constrained compo- nent occurred. Anderson et al 17 reported the results of S-ROM constrained acetabular liners in 21 patients, 18 of whom had experienced recurrent dislocation. At a mean follow-up of 31 months (range, 24 to 64 months), 15 patients (71%) reported no fur- ther dislocations. However, six patients (29%) reported eight redis- locations at a mean of 10 months postoperatively (range, 1 to 30 months). In four cases, the polyeth- ylene liner (still securely fixed to the femoral head) was levered out of the metal shell; in two failures, the femoral head pulled out of the liner; and in two other dislocations, the metal retaining ring disengaged from the polyethylene liner. In all six patients with redislocations, the preoperative diagnosis was recur- rent dislocation, for a failure rate of 33%. However, no loosening of the 19 porous-coated acetabular compo- nents was reported in this study. Fisher and Kiley 18 reported two cases of failure of the S-ROM com- ponent. One occurred 9 months postoperatively and was due to fail- ure of the retaining ring and poly- ethylene wear; the other occurred 5 months postoperatively, with both loosening of the metal shell and pullout of the polyethylene liner from the shell following a traumatic event. Of 51 hips in which the S- ROM constrained component was used, either for recurrent dislocation or in extensive revisions, there were 5 failures—3 redislocations and 2 dissociations (10% failure)—and all required open reduction or revision of the component. 18 Of 12 patients managed with the S-ROM component at their institu- tion, Kaper and Bernini 19 reported Constrained Components in Total Hip Arthroplasty Journal of the American Academy of Orthopaedic Surgeons 236 Figure 4 Radiograph showing the S-ROM constrained polyethylene component dislo- cated from the metal shell. Probably there is also loosening of the acetabular shell. (Reprinted with permission from Anderson MJ, Murray WR, Skinner HB: Constrained acetabular components. J Arthroplasty 1994;9:17-23.) failure in four. In two, the constrain- ing ring had fractured, and in the other two, the liner had pulled out of the metal acetabular shell. Because two of the failures involved an ele- vated-rim constrained liner, these authors suggested that the use of that liner may contribute to a lever-out mechanism. McPherson et al 21 recently de- scribed a new technique that resulted in the successful closed reduction of a dislocated S-ROM constrained liner in three medically compromised patients. With the patient under gen- eral anesthesia and using fluoroscop- ic guidance, the femoral head was perched into the opening of the acetabular component. With the leg positioned in 40° of abduction and 30° of flexion, a minimum of three people using a “bear hug” maneuver of the hip and pelvis apply a contin- ued axial compressive force for at least 90 to 120 seconds, until an audi- ble and palpable clunk of reduction has occurred. No complications were reported, but all three patients later had revision or resection arthroplas- ty. The advantage of this technique is the ability to delay revision surgery until conditions (the patient’s health and the availability of equipment and personnel) are more favorable. Because of wide variation in series size and in indications for surgery, little information beyond anecdotal case reports can be gleaned from the four series reviewing the use of S- ROM constrained components. 12,17-19 However, the reported rate of failure or redislocation is high (9% to 33%). There is even less published experience with the Howmedica Osteonics constrained acetabular liner, in part because its use was ini- tially restricted to two medical cen- ters. Goetz et al 14 reported the use of this acetabular liner for recurrent dis- location in 56 hips. Forty-six con- strained components were inserted without cement, and 10 were inserted with cement (four of these were cemented into acetabular shells of another manufacturer). The 38 patients (39 hips) still living at the time of the report had been followed for a mean of 5.3 years (range, 3 to 8 years), and the deceased 16 patients had been followed for a mean of 2.3 years (range, 1 to 81 months). One patient was lost to follow-up. Only two patients (4%) experienced failure described as “recurrent dislocation.” However, in one patient, the acetabu- lar shell (with screws) pulled out of the pelvis, and in the other, the cemented constrained component dissociated from a well-fixed shell. Seven hips (13%) required revision surgery in the follow-up period, including four for infection and one for acetabular component loosening. Radiographic analysis was per- formed for 38 hips with a minimum 2-year follow-up. There was acetabu- lar osteolysis in 2 of 27 hips (7%) treated with a new acetabular shell and a new constrained liner, both in- serted without cement. There was also definite loosening of 2 of 34 uncemented acetabular components (6%) and 2 of 33 uncemented femoral components (6%). Goetz et al 14 emphasized that, because the primary goal of these revisions was a stable hip, the patients and surgeons were willing to accept the increased risk of polyethylene wear, osteolysis, and component loos- ening. It also should be emphasized that these patients were predominant- ly elderly, debilitated women with a mean age of 71 years. There are no published reports on the use of this constrained component in younger or active patients, in whom an even higher rate of failure of fixation would be expected. Summary Constrained components should be used judiciously for the surgical treatment of recurrent dislocation of the hip. The ideal patient is an elder- ly, low-demand patient with recur- rent dislocation despite well-fixed and properly positioned compo- nents. The etiology of these dislocations is usually soft-tissue (capsule or musculature) insufficien- cy around the prosthetic hip joint. These components should be consid- ered for use only when other options are exhausted and only when bipolar arthroplasty, resection arthroplasty, or a constrained acetabular liner remains. For the two presently avail- able constrained hip components, the rates of failure, including redislo- cation, dissociation of the liner from the acetabular shell, and loosening of the acetabular shell, are reported to be from 4% to 29% at short-term follow-up. Based on the limited published data regarding these con- strained components, prophylactic use of these components is not presently recommended because of the danger of excessive wear of thin polyethylene, breakage, and acceler- ated loosening of components. Paul F. Lachiewicz, MD, and Scott S. Kelley, MD Vol 10, No 4, July/August 2002 237 Figure 5 Radiograph showing the con- straining ring displaced from the S-ROM polyethylene component. (Reprinted with permission from Anderson MJ, Murray WR, Skinner HB: Constrained acetabular components. J Arthroplasty 1994;9:17-23.) References 1. Morrey BF: Difficult complications after hip joint replacement: Dislocation. Clin Orthop 1997;344:179-187. 2. Paterno SA, Lachiewicz PF, Kelley SS: The influence of patient-related factors and the position of the acetabular com- ponent on the rate of dislocation after total hip replacement. J Bone Joint Surg Am 1997;79:1202-1210. 3. Pellicci PM, Bostrom M, Poss R: Pos- terior approach to total hip replacement using enhanced posterior soft tissue repair. Clin Orthop 1998;355:224-228. 4. Schneider PG: Total replacement arthroplasty of the hip joint, in Chapchal G (ed): Arthroplasty of the Hip. Stuttgart, Germany: G Thieme, 1973, pp 113-167. 5. Sivash KM: The development of a total metal prosthesis for the hip joint from a partial joint replacement. Re- constr Surg Traumatol 1969;11:53-62. 6. Russin LA, Russin MA: Abstract: A preliminary study of total hip joint replacement by means of the Russin- modified Sivash prostheses: 100 cases. 41st Annual Meeting Proceedings, Dallas, Texas. Chicago, IL: American Academy of Orthopaedic Surgeons, 1974, p 119. 7. Russin LA, Sonni A: Indications for the use of a constrained THR prosthe- sis. Orthop Rev 1981;10:81-84. 8. Radulovic B, Kenig I, Radovanovic M: Indications for Sivash type total hip prosthesis, in Charnley J (ed): Low Friction Arthroplasty of the Hip: Theory and Practice. Berlin, Germany: Springer- Verlag, 1979, pp 74-81. 9. Bryan WJ, Reeve RE: Dislocation and failure of an articulated total hip replacement: A case report. Orthopedics 1986;9:1113-1115. 10. Koffman M: Proximal femoral resec- tion or total hip replacement in severe- ly disabled cerebral-spastic patients. Orthop Clin North Am 1981;12:91-100. 11. Cameron HU: Use of a constrained acetabular component in revision hip surgery. Contemp Orthop 1991;23:481-484. 12. Lombardi AV Jr, Mallory TH, Kraus TJ, Vaughn BK: Preliminary report on the S-ROM constraining acetabular insert: A retrospective clinical experi- ence. Orthopedics 1991;14:297-303. 13. Goetz DD, Capello WN, Callaghan JJ, Brown TD, Johnston RC: Salvage of total hip instability with a constrained acetabular component. Clin Orthop 1998;355:171-181. 14. Goetz DD, Capello WN, Callaghan JJ, Brown TD, Johnston RC: Salvage of a recurrently dislocating total hip pros- thesis with use of a constrained acetabular component: A retrospective analysis of fifty-six cases. J Bone Joint Surg Am 1998;80:502-509. 15. Kelley SS, Lachiewicz PF, Hickman JM, Paterno SM: Relationship of femoral head and acetabular size to the prevalence of dislocation. Clin Orthop 1998;355:163-170. 16. Root L, Goss JR, Mendes J: The treat- ment of painful hip in cerebral palsy by total hip replacement or hip arthro- desis. J Bone Joint Surg Am 1986;68: 590-598. 17. Anderson MJ, Murray WR, Skinner HB: Constrained acetabular compo- nents. J Arthroplasty 1994;9:17-23. 18. Fisher DA, Kiley K: Constrained acetab- ular cup disassembly. J Arthroplasty 1994;9:325-329. 19. Kaper BP, Bernini PM: Failure of a con- strained acetabular prosthesis of a total hip arthroplasty: A report of four cases. J Bone Joint Surg Am 1998;80:561-565. 20. Namba RS, Van der Reis WL: Femoral head and neck dissociation after a total hip arthroplasty with a constrained acetabular liner. Orthopedics 2000;23: 489-491. 21. McPherson EJ, Costigan WM, Gerhardt MB, Norris LR: Closed reduction of dislocated total hip with S-ROM constrained acetabular component. J Arthroplasty 1999;14:882-885. Constrained Components in Total Hip Arthroplasty Journal of the American Academy of Orthopaedic Surgeons 238