Total hip arthroplasty with cement has proved to be one of the most successful elective operative procedures in orthopaedic practice. This success has resulted in use of this procedure for younger patients and patients who have more challenging diagnoses.
While Charnley and Feagin3 reported from Wrightington Hospital in 1973 that "the policy in this Centre [is] not to attempt the operational reconstruction of late cases of ... congenital dislocation of the hip," increased experience has allowed us to extend the indication for total hip arthroplasty to patients who have dysplastic and dislocated hips.
The anatomy of the acetabular region in such hips is changed because of the impaired articulation. Computerized tomography has greatly facilitated understanding of the altered anatomy, and it is usually possible to decide preoperatively if bulk autogenous graft will be necessary to augment the fixation of the acetabular component8,10,11,16. However, as a general rule, if there is sufficient contact between the femoral head and the acetabulum to produce secondary osteoarthrosis, then there is sufficient bone in the acetabulum to accommodate the socket of the Charnley low-friction prosthesis without the use of a bone graft.
The purpose of the present study was to review our results of total hip arthroplasty at a mean of approximately ten years in patients who had congenital dislocation or dysplasia of the hip.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Hospital Clinico Jose Joaquin Aguirre, University of Chile, Santiago, Chile.
‡2810 Grants River Circle, Sugar Land, Texas 77479.
§Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, England.
One hundred and ninety patients (232 hips) had primary low-friction arthroplasty at Wrightington Hospital between 1965 and 1987 as treatment for osteoarthrosis secondary to acetabular dysplasia resulting from congenital dysplasia or dislocation of the hip.
Preoperatively, all patients fulfilled the requirement of the Charnley pseudarthrosis test2—that is, it was determined that the patient's over-all function would be improved by pseudarthrosis (should the hip replacement fail).
Thirty-three patients (thirty-four hips) who had not been followed for at least three years were excluded from the study. Sixteen patients (sixteen hips) who had died from causes unrelated to the total hip replacement were included in the survivorship analysis only. The remaining 141 patients (182 hips) form the basis of this clinical and radiographic study.
On the basis of the four grades of Crowe et al., the hips were divided into two groups: Group S, which included 136 hips (ninety-seven patients) that were dysplastic or subluxated (Crowe grades I, II, and III) (Fig. 1), and Group D, which included forty-six hips (forty-four patients) that were completely dislocated (Crowe grade IV) (Fig. 2). Thirty-nine patients in Group S and two patients in Group D had bilateral arthroplasty.
Fifty-one hips (38 per cent) in Group S and eighteen hips (39 per cent) in Group D had been operated on previously, with twenty-two (16 per cent) of those in Group S and six (13 per cent) of those in Group D having had more than one previous operation.
A trochanteric osteotomy was performed to facilitate exposure in 164 hips that had severe anatomical distortion. A direct lateral approach was used for the remaining eighteen hips. An attempt was made to place the prosthetic cup in the anatomical position as this enhanced the possibility of restoring the Shenton line (a goal of the operation). If there had been shortening of the limb before the operation, then the anatomical location increased tension in the soft tissues, thereby helping to stabilize the reconstructed joint. Bulk autogenous graft was not used for any patient.
The trochanter did not always reach the trochanteric bed when an attempt was made to reattach it, and it was sometimes necessary to advance the trochanter by dividing the tendon, which lies at the posterior aspect of the muscle belly as it attaches to the trochanter, or to mobilize the gluteus medius muscle from the innominate bone.
The mean age of the eighty-six women and eleven men in Group S at the time of the operation was 44.5 years (range, 19.5 to 76.5 years), and the mean age of the thirty-nine women and five men in Group D was 41.5 years (range, 20.5 to 59.5 years). The mean duration of follow-up was 9.9 years (range, 3.1 to 22.8 years) for all 141 patients, 9.5 years for Group S, and 10.1 years for Group D.
Although the Charnley 22.25-millimeter femoral head was used in all hips, a variety of components were used to accommodate the anatomy of each individual hip (Table I). A femoral component of so-called regular size was used in only sixty-eight hips (50 per cent) in Group S and in only fourteen hips (30 per cent) in Group D. A so-called offset-bore acetabular component (the polyethylene is machined eccentrically, thereby reducing the outer diameter to thirty-five millimeters) was used in thirty-seven hips (27 per cent) in Group S and in twenty-seven hips (59 per cent) in Group D.
The system of Merle d'Aubigné and Postel was used to assess the hip. With this system, pain, function, and the range of motion are graded separately on a scale of 1 to 6 points. A higher score indicates a better result. Function of the abductors was determined by assessing the patient's ability to actively abduct the straight lower limb.
The stability of the acetabular components was assessed radiographically with the method of DeLee and Charnley and that of the femoral components, with the method of Gruen et al. A cup or stem was considered to be stable if there was no evidence of migration or radiolucent lines at the bone-cement interface. An acetabular cup was considered to be definitely loose if it had migrated more than five millimeters, and a femoral stem was considered to be definitely loose if it had subsided more than five millimeters. A cup or stem was considered to be possibly loose if there was a continuous radiolucent line of at least two millimeters in width at the bone-cement interface12.
Survivorship Analysis
Survivorship analysis was performed with use of the product-limit method1.
Functional Outcome
Group S
One hundred and twenty-eight hips (94 per cent) caused no or only slight pain (6 or 5 points, respectively, with the system of Merle d'Aubigné and Postel) postoperatively compared with nine hips (7 per cent) preoperatively (Fig. 3-A). One hundred and thirteen hips (83 per cent) had normal or nearly normal function (6 or 5 points, respectively) postoperatively compared with four hips (3 per cent) preoperatively (Fig. 3-B). One hundred and eleven hips (82 per cent) had a full or nearly full range of motion (6 or 5 points, respectively) postoperatively compared with thirteen hips (10 per cent) preoperatively (Fig. 3-C).
Preoperatively, twenty-five hips (18 per cent) were in patients who were able to perform active abduction and straight-leg raising of more than 60 degrees; forty-two hips (31 per cent), active abduction and straight-leg raising of 31 to 60 degrees; and sixty-nine hips (51 per cent), active abduction and straight-leg raising of less than 31 degrees. Postoperatively, the numbers of hips were 106 (78 per cent), twenty-four (18 per cent), and six (4 per cent), respectively.
Group D
Forty-four hips (96 per cent) caused no or only slight pain postoperatively compared with one hip (2 per cent) preoperatively (Fig. 4-A). Twenty-five hips (54 per cent) had normal or nearly normal function postoperatively compared with ten hips (22 per cent) preoperatively (Fig. 4-B). Thirty-three hips (72 per cent) had a full or nearly full range of motion postoperatively compared with nine hips (20 per cent) preoperatively (Fig. 4-C).
Preoperatively, three hips (7 per cent) were in patients who were able to perform active abduction and straight-leg raising of more than 60 degrees; fifteen hips (33 per cent), active abduction and straight-leg raising of 31 to 60 degrees; and twenty-eight hips (61 per cent), active abduction and straight-leg raising of less than 31 degrees. Postoperatively, the numbers of hips were twenty-five (54 per cent), fifteen (33 per cent), and six (13 per cent), respectively.
Complications
Five patients had deep venous thrombosis postoperatively, and they were managed with anticoagulants. No patient had a clinical pulmonary embolism. Transient postoperative sciatic-nerve palsy developed in three patients, but it resolved by the time of the most recent follow-up examination. An extraperitoneal hematoma developed in one patient from each group as a result of damage to the iliac veins; each patient had a good recovery after an operative repair. There were four postoperative dislocations (two in each group), which were treated non-operatively. Other complications included heterotopic bone formation (ten hips [five in each group]), postoperative hematoma (seven hips), and a fracture of the femoral shaft (one hip).
Radiographic Results
At the most recent follow-up examination, there was one definitely loose cup in Group D and there were none in Group S. Four cups (9 per cent) in Group D and five cups (4 per cent) in Group S were possibly loose.
One hundred and thirteen hips had no subsidence of the femoral stem, and forty-six hips had two to five millimeters of subsidence. None of these hips were clinically symptomatic. Three stems, one in Group D and two in Group S, had subsided more than five millimeters and were considered definitely loose. One hip in Group D and one hip in Group S had fracture of the cement mantle. A complete radiolucent line of less than two millimeters in thickness was found at the bone-cement interface in one hip in Group D and one hip in Group S. The remaining stems had incomplete radiolucent lines that were less than two millimeters thick, but these lines were not considered to be important.
Revision
Fourteen hips (10 per cent) in Group S and eight hips (17 per cent) in Group D had a revision operation. Both components were revised in two hips (two patients) in Group S.
The rate of revision for loosening of the acetabular cup was high: nineteen (10 per cent) of 182. Twelve (9 per cent) of the 136 cups in Group S and seven (15 per cent) of the forty-six cups in Group D were revised. Three cups that had been placed in a false acetabulum did not need to be revised. Seven of the cups that failed were so-called offset-bore components. The cups failed at a mean of 5.9 years in Group D and at a mean of 8.1 years in Group S.
Over-all, five femoral stems (3 per cent) were revised: one (2 per cent) from Group D and four (3 per cent) from Group S. Of these stems, two (from Group S) fractured and three became aseptically loose.
A Girdlestone resection arthroplasty was performed in four hips because of loss of acetabular bone stock (two hips) or infection (two hips).
Survivorship Analysis
The estimated probability of survival of the implant (cup and stem), with the 95 per cent confidence interval, at fifteen years was 74 ± 4.7 per cent for Group S and 68 ± 11.6 per cent for Group D. The difference between the curves was not significant at the 10 per cent level (p = 0.096, Mantel-Cox test) (Fig. 5).
Patients who have osteoarthrosis secondary to acetabular dysplasia or dislocation can be severely disabled. Most of these patients are women and have an early onset of symptoms. Many have young children whose rearing is complicated by the patient's pain and disability.
In Group D (complete dislocation; Crowe grade IV), 17 per cent (eight) of the hips were revised compared with 10 per cent (fourteen) of the hips in Group S (subluxation; Crowe grades I, II, and III). This difference appears to reflect the increasing difficulty of achieving fixation in a dislocated hip.
An effort should be made to obtain osseous coverage of the acetabular component, and in some hips this can be achieved only in the false acetabulum. The anterior osseous coverage could be severely reduced in the true acetabulum because of the reduced anteroposterior diameter of the dysplastic bone, and in some hips the false acetabulum is used because the bone stock may be more favorable13,17.
In the present series, the socket was placed in the false acetabulum in three hips, and function was satisfactory except for limited abduction (Figs. 6-A and 6-B). The over-all rate of definite and possible loosening of the acetabular component (revision of the cup or radiographic evidence of loosening at follow-up) was 16 per cent (twenty-nine cups). Loosening was more common in the hips that had a complete dislocation (twelve cups; 26 per cent) than in those that were dysplastic or subluxated (seventeen cups; 13 per cent). Loosening of the cup appears to depend on the amount of osseous coverage obtained at the time of the operation5,7,14 (Figs. 7-A, 7-B, and 7-C).
Despite marked anteversion and a narrow femoral canal in some patients, only 3 per cent (five) of the femoral stems were revised. In addition, three stems that had subsided more than five millimeters in association with a fracture of the cement mantle and radiolucent lines were considered to be definitely loose. Thus, the over-all rate of loosening was 4 per cent (eight stems).
In conclusion, total hip arthroplasty provides worthwhile long-term function in this severely disabled group of patients. Patients must be cautioned that the ideal goals of limb-length equalization and eradication of a limp are not likely to be realized completely and that rehabilitation may be prolonged. Whereas patients may be able to resume active walking without the use of aids by six months after a standard total hip arthroplasty performed to treat degenerative osteoarthrosis in maturity, fifteen to eighteen months may be needed before optimum rehabilitation can be achieved after total hip arthroplasty performed for osteoarthrosis secondary to acetabular dysplasia or dislocation.
In this study, we evaluated a severely disabled group of patients who fulfilled the requirements of the Charnley pseudarthrosis test2. Our indications for total hip arthroplasty in patients who have acetabular dysplasia or dislocation are changing as acceptable long-term results of total hip arthroplasty with cement have been confirmed. Contemporary operative techniques for dysplasia now often include bone-grafting.