JBJS | Long-Term Results of Total Hip Replacement in Young Patients Who Had Ankylosing Spondylitis. Eighteen to Thirty-Year Results with Survivorship Analysis*

The Journal of Bone & Joint Surgery, Volume 79, Issue 8

Articles | August 01, 1997

Long-Term Results of Total Hip Replacement in Young Patients Who Had Ankylosing Spondylitis. Eighteen to Thirty-Year Results with Survivorship Analysis*

DAVID H. SOCHART, F.R.C.S.†; MARTYN L. PORTER, F.R.C.S.‡, WIGAN UNITED KINGDOM

View Disclosures and Other Information

Investigation performed at the Centre for Hip Surgery, Wrightington Hospital, Wigan

The Journal of Bone & Joint Surgery. 1997; 79:1181-9

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

Article

Figures

References

text A A A

Abstract

We determined the long-term results of total hip replacement in a series of young patients who had ankylosing spondylitis. Between 1966 and 1978, forty-three Charnley low-friction arthroplasties were performed in twenty-four patients who had an average age of 28.8 years (range, nineteen to thirty-nine years) at the time of the operation. There were seventeen men and seven women. The average duration of follow-up for the series was 22.7 years (range, one month [a perioperative death] to 30.3 years). Four patients (seven hips) had died an average of 12.6 years (range, one month to 18.7 years) after the operation. The twenty surviving patients had clinical and radiographic follow-up until the time of writing or until both of the original components had been revised.All patients had substantial relief of pain and improvement of function and the range of motion of the joint. Twenty-one patients (88 per cent; thirty-nine hips) were completely free of pain (6 points, according to the scale of Merle d'Aubigné and Postel), and the remainder had only slight discomfort (5 points). Ten acetabular components and one femoral component were revised because of aseptic loosening, and one patient had a revision of both components because of late deep infection. Three additional femoral components were revised during a revision operation for a loose acetabular component. Although the femoral components were not loose, they had been in place for more than ten years and it was thought likely that the bearing surface was damaged. The average time to revision was 13.3 years (range, 4.0 to 20.3 years). At an average of 22.7 years, thirty-eight (88 per cent) of the original femoral components and thirty-two (74 per cent) of the original acetabular components remained in situ. The average annual rate of acetabular wear was 0.12 millimeter for the entire series. Only six hips (14 per cent) had minor heterotopic ossification, and none of the hips had clinically important ossification (class III or IV according to the system of Brooker et al.).To our knowledge, the present report describes the largest series of total hip arthroplasties, with the longest duration of follow-up, in young patients who had ankylosing spondylitis. Survivorship analysis with use of the Kaplan-Meier method revealed that the probability of survival of the femoral component (with 95 per cent confidence intervals) was 91 per cent (83 to 99 per cent) at twenty years and 83 per cent (72 to 94 per cent) at thirty years. The probability of survival of the acetabular components was 73 per cent (61 to 84 per cent) at twenty years and 70 per cent (57 to 83 per cent) at thirty years. The probability that both components would survive was 91 per cent (82 to 100 per cent) at ten years, 73 per cent (61 to 84 per cent) at twenty years, and 70 per cent (57 to 83 per cent) at thirty years. The Charnley low-friction arthroplasty provided consistently good long-term results, with a low rate of complications and revisions, in this group of young patients.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

The hip is involved in 25 to 50 per cent of patients who have ankylosing spondylitis, and 50 to 90 per cent of patients who have such involvement have it bilaterally^2,5,38,41. The typical age at the onset of ankylosing spondylitis is between fifteen and twenty-five years. However, the younger the age at the onset, the more severe the disease is likely to be and the higher the probability that a total hip replacement will be performed because of the resulting loss of motion combined with severe and often debilitating pain^{2,13,16,25,41}.

Many authors have expressed concern about performing a total hip replacement in patients who have ankylosing spondylitis because of the perceived high risks of a recurrence of ankylosis, mechanical failure, and poor function^2,5,25,37,41. In addition, young patients are expected to have increased rates of wear and loosening of the components because they place greater demands on the prosthesis than do older patients^{6,12,35,42,43}. Most series^{2,12,13,16,25,26,35,38,39} have consisted of a small number of patients who had a wide range of ages, a diverse array of implants, and a limited duration of follow-up, which has made it difficult to draw any firm conclusions regarding the true outcome of the operation. Furthermore, in some reports^26,33,34,40, patients who had ankylosing spondylitis were grouped with those who had inflammatory arthritides, such as rheumatoid arthritis, and the results were not analyzed with regard to the specific preoperative diagnosis. This was done even though the diseases have different characteristics. Ankylosing spondylitis occurs most commonly in men or boys; the onset is earlier than that of inflammatory arthritides and thus operations are performed at a younger age; and ankylosing spondylitis tends to spare the upper limbs, resulting in higher demands on hip prostheses because the over-all function is better.

In the present report, we describe what we believe to be the largest series of total hip arthroplasties, with the longest duration of follow-up, in young patients who had anklyosing spondylitis.

*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.

†7 Woodlea, Walkden Road, Worsley, Manchester M28 2QJ, United Kingdom.

‡The Center for Hip Surgery, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, United Kingdom.

*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.

†7 Woodlea, Walkden Road, Worsley, Manchester M28 2QJ, United Kingdom.

‡The Center for Hip Surgery, Wrightington Hospital, Hall Lane, Appley Bridge, Wigan, Lancashire WN6 9EP, United Kingdom.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 1-A, 1-B, and 1-C: A twenty-six-year-old woman who had a bilateral Charnley low-friction arthroplasty of the hip for the treatment of ankylosing spondylitis. Fig. 1-A: Preoperative radiograph made in 1966.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B Radiograph made one year postoperatively, in 1967.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-C Radiograph made thirty years after the operation, in 1996, showing no evidence of loosening of any component. Wear of the acetabular component averaged 0.14 millimeter per year bilaterally. The hips remained free of pain, and the patient could walk long distances. The cumulative range of motion was 210 degrees, and pain, function, and range of motion received a score of 6, 5, and 5 points, respectively^7,29.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 2-A, 2-B, and 2-C: Bar graphs showing the average preoperative and postoperative scores according to the system of Merle d'Aubigné and Postel[29]. Fig. 2-A: Pain.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-B Function.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-C Range of motion.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3 Thirty-year survivorship curves (and 95 per cent confidence intervals), calculated with use of the Kaplan-Meier method²⁴, for the acetabular and femoral components. The survival of both original components was determined by the survival of the acetabular component; these two curves therefore are identical and superimposed.

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

All patients with ankylosing spondylitis who had had a total hip replacement before the age of forty years were identified from the clinical database at the Centre for Hip Surgery at Wrightington Hospital near Wigan, United Kingdom. The medical records and the radiographs of twenty-four patients who had had a total of forty-three arthroplasties were retrieved. Most of the patients had had regular follow-up examinations at the clinic, and those who had been lost to follow-up were traced through the National Health Service Central Register at the Office for National Statistics. Four patients (seven hips) had died, but they had been followed until shortly before death. All patients who had survived were contacted and evaluated, initially with use of a written questionnaire and then with a telephone interview. The patients who lived a long distance from the clinic sent radiographs made at their local hospital, and the remainder were seen at specific research clinics for clinical and radiographic assessment (Figs. 1-A, 1-B, and 1-C).

Preoperative and postoperative pain, function, and range of motion were graded with use of the 6-point scale of Merle d'Aubigné and Postel³⁰, as modified by Charnley⁷. The modification includes an assessment of preoperative disability with the addition of a prefix (A, B, or C) to the numerical classification. The prefix A indicates involvement of only one hip; B, bilateral involvement of the hip; and C, some other factor that contributed to the failure to achieve normal locomotion, such as polyarticular arthritis, neurological dysfunction, or cardiorespiratory impairment. Four patients had had a previous operation on the hip: three had had an osteotomy, and one had had insertion of an acrylic hemiprosthesis.

The main indication for a total hip replacement in patients who have ankylosing spondylitis is the combination of loss of motion, poor posture, and pain. Arthroplasty often provides substantial relief of pain, but improvement in range of motion is less predictable^2,5,38,39,45. Most of the patients in the present study were selected for the operation on the basis of the pseudarthrosis test¹⁷; an arthroplasty was performed only if the preoperative grades for function and pain indicated that a Girdlestone excisional arthroplasty would result in improvement. The rationale was that even if the prosthesis were to fail and subsequently be removed, the patient would still be better off. As a result, although the patients who were selected for the procedure were young, they tended to be severely disabled by hip disease.

All operations were done in a Charnley-Howorth clean-air enclosure with laminar airflow (Howorth Engineering, Brinscall Blackburn, United Kingdom), first introduced in 1966, and the operating team wore body-exhaust suits after the introduction of the suits in 1970¹⁹. A lateral approach was used with a planar trochanteric osteotomy and standard Charnley reattachment with stainless-steel wires^3,7,8. All patients were managed with a Charnley prosthesis (Thackray, Leeds, United Kingdom). This implant consists of a tapered, monoblock, stainless-steel femoral stem with a 22.25-millimeter-diameter head that forms a low friction-torque articulation with a non-flanged ultra-high molecular weight polyethylene acetabular component with an integral wire marker. Both components were inserted with radiopaque self-curing polymethylmethacrylate cement that was pressurized digitally without previous occlusion of the distal portion of the femoral intramedullary canal.

An anteroposterior radiograph of the pelvis was made immediately after the operation, before discharge from the hospital, and at each subsequent visit to the clinic. Patients had a clinical and radiographic follow-up examination annually for the first ten years and on alternate years thereafter. The anteroposterior radiographs of the pelvis were made with use of a standard technique to show both hips. The beam was centered over the pubic symphysis, and the hips were in extension and neutral rotation with the patellae pointing up. The radiographs were examined for evidence of demarcation of the acetabular components in the zones described by DeLee and Charnley¹⁰. Loosening was assessed with use of the criteria of Hodgkinson et al.¹⁸. An acetabular component was considered to be loose if a continuous radiolucent line of any thickness was evident in all three zones (type-3 demarcation) or if the cup had changed position or migrated (type-4 demarcation). Migration of the cup was assessed on sequential radiographs by comparing the vertical distance from the center of the cup, measured with use of the integral wire marker, to a horizontal line joining the two anatomical teardrops. Wear of the acetabular components was measured with a graded circular gauge that used the standard 22.25-millimeter-diameter femoral head as a point of reference in order to account for the differences in magnification between radiographs^14,44.

The position of the femoral stem within the medullary canal was considered to be neutral when the long axis of the stem of the prosthesis was aligned with the neutral axis of the femoral shaft. If the axis of the stem was at an angle to the neutral axis of the femoral shaft, with the distal tip and the axis of the implant directed medially, then the prosthesis was in valgus alignment; if the axis and the distal tip of the implant pointed laterally, then it was in varus alignment. The femoral components were assessed for demarcation and endosteal lysis with use of the zonal system of Gruen et al.¹⁵ and by identification of any of the features associated with loosening as described by Loudon and Charnley²⁸ as well as Loudon and Older²⁹. Definite loosening was defined as more than five millimeters of subsidence or continuous demarcation around the stem. Probable loosening was defined as subsidence of two to five millimeters or a radiolucent line surrounding 50 per cent of the stem (three zones) or more. Heterotopic ossification was classified with use of the method described by Brooker et al.⁴, but no specific prophylactic measures were used to prevent the formation of heterotopic bone. Most patients received thromboembolic prophylaxis with a daily dose of 600 to 1200 milligrams of Plaquenil (hydroxychloroquine sulfate)^22,27 beginning on the day before the operation and continuing until the patient was fully mobile, at an average of fourteen days postoperatively.

The Kaplan-Meier method²⁴ was used to calculate the probability of retention of the original prosthesis from the time of the initial arthroplasty until one of three end points: a revision for any reason, a revision of the femoral stem, or a revision of the acetabular component. In addition, the 95 per cent confidence intervals were calculated.

Results

Introduction | Materials and Methods | Results | Discussion | References

Of the 10,469 total hip replacements that were performed from 1966 to 1978 at the Centre for Hip Surgery at Wrightington Hospital near Wigan, United Kingdom, 280 (2.7 per cent) were performed on patients who were less than forty years old and 117 (1.1 per cent) were performed on patients of all ages who had ankylosing spondylitis. Twenty-four patients who had ankylosing spondylitis and were less than forty years old had forty-three total hip replacements, which was 0.4 per cent of the total procedures and 37 per cent of the procedures performed on patients of all ages who had ankylosing spondylitis. The average age of the patients in the present study at the time of the index arthroplasty was 28.8 years (range, nineteen to thirty-nine years). There were seventeen men and seven women, and nineteen of the patients had a staged bilateral procedure. At the time of the operation, eighteen of the patients were in category C; four, category B; and two, category A according to the system of Charnley⁷.

Four patients (seven hips) had died. One had died as a result of aortic dissection secondary to aortitis during the perioperative period of the second arthroplasty of the hip, which was performed two years after the arthroplasty of the contralateral hip. The other three patients had died an average of seventeen years (range, fourteen to eighteen years) after the operation. Three of these patients (six hips) still had the original implants in situ at the time of death and had been followed on a regular basis, with no evidence of loosening of any component.

A deep infection caused by a coagulase-negative Staphylococcus, which led to septic loosening and revision nine years after the operation, developed in one patient. None of the patients had a thromboembolic complication, breakage of an implant, or dislocation. Three hips (7 per cent) had trochanteric displacement associated with broken wires. Two patients needed removal of the wires, and one had trochanteric reattachment. Two patients (two hips) had late asymptomatic breakage of wires. Only six hips (14 per cent) had heterotopic ossification; three were rated as class I according to the system of Brooker et al.⁴, and three were rated as class II. Thirty-seven hips (86 per cent) showed no evidence of heterotopic ossification, and none of the hips had class-III or IV heterotopic ossification.

The twenty patients (thirty-six hips) who were still alive had been followed for an average of 23.4 years (range, 18.0 to 30.3 years). The average duration of follow-up for the entire series, including the patients who had died, was 22.7 years (range, one month [the perioperative death] to 30.3 years).

All patients had a substantial clinical improvement in terms of pain, function, and range of motion (Figs. 2-A, 2-B, and 2-C). The average score for pain was 3.7 points (range, 2 to 6 points) preoperatively, with four hips that were pain-free because of osseous ankylosis of the hip. The average postoperative score was 5.8 points (range, 5 to 6 points). Twenty-one patients (88 per cent; thirty-nine hips) were entirely free of pain, and the remainder had only occasional discomfort. Although the score for function was designed originally for assessment of patients who had monoarticular disease, it improved from an average of 2.8 points (range, 1 to 5 points) preoperatively to 5.3 points (range, 2 to 6 points) after the operation. The score for the range of motion averaged 2.3 points (range, 1 to 4 points) before the operation, representing a cumulative range of 60 to 100 degrees, and it improved to 5 points (range, 4 to 6 points) postoperatively, an average cumulative range of 210 degrees (range, 160 to 260 degrees).

Ten acetabular components in eight patients were revised because of aseptic loosening and one component, because of a late deep infection. Radiographically, two cups had type-3 loosening (a radiolucent line in all three zones) and nine, type-4 (an average of seven millimeters [range, four to twenty millimeters] of migration)¹⁸. Five patients also had revision of the femoral component during the same operation. Only two of the femoral components were actually loose (one because of infection), and the other three were removed as a routine part of the procedure as they had been in place for more than ten years and were considered likely to have a damaged bearing surface. The average time to revision was 13.3 years (range, 4.0 to 10.1 years) for the acetabular component and 13.4 years (range, 4.0 to 22.8 years) for the femoral component.

Thirty-two (74 per cent) of the original acetabular components had not been revised by the time of the latest follow-up examination (twenty-six hips) or by the time of death (six hips). Fourteen of the twenty-six hips had no evidence of demarcation (type 0); seven had radiolucent demarcation in one zone (type 1); and one had demarcation in two zones (type 2). No hip had type-3 demarcation of the acetabular component. Four implants were definitely loose (type 4), having migrated an average of nine millimeters (range, two to twenty millimeters). However, the patients remained asymptomatic.

Rates of wear (penetration) were calculated for all of the acetabular components until the time of death, revision, or the latest follow-up evaluation. The total amount of wear averaged 2.35 millimeters (range, 0.2 to 4.94 millimeters), yielding an average annual rate of wear of 0.12 millimeter (range, 0.01 to 0.26 millimeter) for the entire series. The average rate of wear of the eleven components that had been revised for loosening was 0.18 millimeter (range, 0.07 to 0.22 millimeter) per year, which was higher than that for the group that had not been revised (0.1 millimeter per year; range, 0.01 to 0.26 millimeter).

Twenty-two femoral components were in valgus alignment (the preferred position at our institution during some of the study period), eighteen were in neutral, and three were in slight varus alignment. Thirty-seven implants (86 per cent) had not subsided, and six implants had subsided between two and five millimeters. Three implants that had subsided were associated with a fracture of the cement at the tip of the prosthesis. Two of these implants were subsequently revised: one, because of aseptic loosening, and one, because of infection.

Thirty-eight (88 per cent) of the original femoral implants remained in situ at the time of the most recent follow-up examination or the time of death. Five femoral components (12 per cent) had been revised: one was in the group of four patients who died and four were in the group of patients who were seen at the latest follow-up examination. Only two femoral components (5 per cent) actually loosened. Twenty-six (81 per cent) of the thirty-two femoral components seen at the latest follow-up examination had no evidence of loosening, and three had only one zone of demarcation. Loosening was apparent in three hips: two had three affected zones, and one had complete demarcation of all seven zones associated with subsidence of five millimeters and a fracture of the cement at the tip of the prosthesis. Only three hips (7 per cent) in the series had osteolysis. Two of those hips were revised (one, because of infection). The third hip, which had an area of lysis one centimeter long and two millimeters wide in zone 5 according to the system of Gruen et al.¹⁵, was one of the hips that had loosening of the femoral component and demarcation in zones 1, 2, and 3.

Notable changes in the medial portion of the femoral neck (the calcar) were seen in fourteen hips (33 per cent). Four of them had marked osteopenia of the calcar, and ten had a loss of height of the calcar that averaged 4.4 millimeters (range, two to fifteen millimeters). The femoral component was not revised in any of those hips, and it was loose in only one. Hypertrophy of the femoral cortex around the tip of the stem (zones 3, 4, and 5) was seen in ten (23 per cent) of the hips in the entire series and was not associated with loosening.

The probability of survival of the original acetabular component (with 95 per cent confidence intervals) was 73 per cent (61 to 84 per cent) at twenty years and 70 per cent (57 to 83 per cent) at thirty years. The probability of survival of the femoral stem was 91 per cent (83 to 99 per cent) and 83 per cent (72 to 94 per cent), respectively (Fig. 3). The probability of survival of both components at twenty and thirty years was determined by that of the acetabular component and therefore was 73 per cent (61 to 84 per cent) and 70 per cent (57 to 83 per cent), respectively.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

Many surgeons have been hesitant to perform total hip replacements on young patients because of the belief that the high functional demands placed on the implants by such patients were likely to result in increased rates of wear and loosening^{6,12,17,35,43}. In addition, patients who have ankylosing spondylitis have been thought to be among the poorest candidates for hip arthroplasty at any age because of increased rates of heterotopic ossification leading to a reduction in the postoperative range of motion, to complete ankylosis, or to mechanical failure^2,5,25,37,41. Ankylosing spondylitis usually develops between the ages of fifteen and twenty-five years, often necessitating a total hip arthroplasty within twenty years after the onset. In addition, the correction of long-standing contractures and the accurate placement of the acetabular component in the presence of pelvic obliquity are technically difficult. For all of these reasons, many surgeons consider the performance of total hip arthroplasty in these patients to be particularly demanding^5,13,25,39.

Of the 10,469 arthroplasties done at our institution from 1966 to 1978, 0.4 per cent (forty-three) were performed in patients with ankylosing spondylitis who were less than forty years old, but those arthroplasties accounted for 37 per cent of the 117 arthroplasties performed in patients of all ages who had a preoperative diagnosis of ankylosing spondylitis. This finding is consistent with those in previous reports, in which patients who had ankylosing spondylitis, especially those who had an early onset, were found to need an operation at a relatively young age^5,41. Twenty-two (92 per cent) of the twenty-four patients in the present study were in category B or C according to the system of Charnley⁷, and nineteen (79 per cent) of them had a bilateral arthroplasty. Again, this observation is similar to the findings in other reports, in which the rates of bilateral involvement of the hip in patients who had ankylosing spondylitis were 48 per cent (eleven of twenty-three)², ten of nineteen³⁸, 59 per cent (fifty-one of eighty-seven)⁵, and 62 per cent (thirty-three of fifty-three)⁴¹.

The short to intermediate-term results reported after total hip replacement for treatment of ankylosing spondylitis have been encouraging, with low rates of early complications, loosening, and revision^23,25,40,41. However, some results give cause for concern. Bisla et al.² reported radiolucent lines around the acetabular component in twenty-one (62 per cent) of thirty-four hips at an average of only 3.5 years after the operation, and Kilgus et al.²⁵ reported non-progressive radiolucent lines involving twenty-two (76 per cent) of twenty-nine conventional, cemented implants that had been followed for 6.3 years. Studies with longer-term follow-up are essential to reveal the true rate of failure²³, and most of the reports of which we are aware fall short of this goal.

Postoperatively, our patients had substantial relief of pain and improvement with respect to function and the range of motion. All of the patients had either slight or no pain in the hip (a score of 5 or 6 points), and 88 per cent (twenty-one patients; thirty-nine hips) were entirely free of pain in the hip³⁰. The range of motion, which had averaged 60 to 100 degrees before the operation, improved to a cumulative range that averaged 210 degrees (range, 160 to 260 degrees) for the series but was more than 210 degrees (a score of 5 or 6 points³⁰) for 84 per cent (thirty-six) of the hips. Improvement of function was less consistent because many patients were limited by polyarticular disease. However, 71 per cent (seventeen) of the patients were able to walk long distances without the aid of a cane. Our results are similar to those of studies in which patients were followed for as long as ten years. In those reports, the rates of good or excellent results were 70 per cent (fifty-one of seventy-three hips)³⁵, 73 per cent (sixty-three of eighty-six hips)⁴¹, 86 per cent (119 of 138 hips)⁵, 92 per cent (sixty-three of sixty-eight hips)¹⁷, and 94 per cent (thirty-two of thirty-four hips)². Our results are also similar to those of studies in which the rates of complete relief of pain associated with ankylosing spondylitis were 89 per cent (123 of 138 hips)⁵, nine of ten hips³⁵, 97 per cent (twenty-eight of twenty-nine hips)³⁸, and 100 per cent (seventeen hips¹⁸ and thirty-three hips³⁹). There were few complications in the present series. Only one patient had a late deep infection, and none of the patients had thromboembolism, breakage of an implant, or dislocation.

Thirty-two (74 per cent) of the acetabular components and thirty-eight (88 per cent) of the femoral components were in situ at the latest follow-up examination or at the time of death, at an average of 22.7 years (range, one month to 30.3 years). Eleven acetabular components (26 per cent) had been revised: ten (23 per cent), because of aseptic loosening, and one (2 per cent), because of infection. As four of the twenty-six original implants were definitely loose in the patients who had survived, the rate of aseptic loosening was 33 per cent (fourteen prostheses) for the entire series. The fact that patients who have loosening and even migration of an implant can remain asymptomatic has been documented previously and is strong justification for the long-term follow-up of all patients after an arthroplasty^7,37,39,40. The annual rate of wear of the acetabular components for the entire series was 0.12 millimeter (range, 0.01 to 0.26 millimeter) and is similar to that in other series that have had rates of 0.07 to 0.21 millimeter per year^{9,14,20,44,45}.

Five femoral components (12 per cent) were revised, but only two (5 per cent) of these actually were loose; the other three were replaced because they had been in situ for more than ten years and were considered likely to have a suboptimum bearing surface. Most tapered prostheses initially subside to a minor degree, but if end-bearing is sufficient to cause a fracture of the tip of the cement mantle, continued subsidence is possible as the prosthesis is supported only by the intact, proximal part of the mantle. Subsidence of more than five millimeters indicates that the remaining, proximal portion of the cement is unable to resist the applied load, and prosthetic loosening becomes more likely^28,29. Of the thirty-two original femoral implants in the patients who had survived, one was definitely loose and two were probably loose, yielding a rate of aseptic loosening for the entire series of 12 per cent (five hips). Thirty-seven prostheses (86 per cent) had not subsided, and six (14 per cent) had subsided between two and five millimeters. None of the prostheses had subsided more than five millimeters, but those associated with a fracture of the cement at the tip of the prosthesis were definitely loose. Osteolysis was seen in only three hips (7 per cent), all of which had a loose component. Two of those components were revised.

Cortical hypertrophy, which was seen around the tip of the implant in ten hips (23 per cent), is thought to represent the physiological adaptation of bone to the transition from an artificially rigid proximal segment of the femur to a normal distal segment of the femur^7,8. Cortical hypertrophy was not associated with loosening. It has been reported in 8.5 to 26 per cent of hips in series ranging from fifty-seven to 582 hips^7,23,42,45. Three hips (7 per cent) in our series had non-union of the greater trochanter, and all three had an additional operation: two had removal of the wire and one had reattachment. These findings are similar to those of other reports^3,21,36,39, in which the rates of non-union have ranged from 1 to 5 per cent in series ranging from seventy-seven to 932 patients. However, the routine use of trochanteric osteotomy remains controversial, and many surgeons now reserve it for complex or revision procedures.

Heterotopic ossification developed in only 14 per cent (six) of the hips and was minor (class I or II), despite the fact that no specific prophylaxis had been used. This rate was similar to the rates of 5 per cent (twenty of 379 hips)⁷, 15 per cent (347 of 2311 hips)¹¹, 21 per cent (twenty-one of 100 hips)⁴, and 30 per cent (152 of 507 hips)³² that have been reported after arthroplasty in the general population. Conversely, patients who have ankylosing spondylitis traditionally are considered to be at high risk for heterotopic ossification, and investigators have reported rates of 40 per cent (thirty-nine of ninety-eight hips)³⁷, 47 per cent (twenty-five of fifty-three hips)²⁵, 56 per cent (fifty-five of ninety-nine hips)⁴¹, 62 per cent (twenty-one of thirty-four hips)², and 76 per cent (twenty-two of twenty-nine hips)³⁸. Also, the rates of clinically important quantities of bone that were likely to interfere with function (class III and IV) have been reported to be 11 per cent (eleven of ninety-eight hips³⁷ and six of fifty-three hips²⁵), 23 per cent (six of twenty-six hips)³⁸, and 26 per cent (nine of thirty-four hips)². The risk of heterotopic ossification is higher for men as well as for hips that have had a trochanteric osteotomy or a previous infection^11,25. The reported rates of heterotopic ossification after a second procedure, when heterotopic ossification developed after the arthroplasty of the first hip, have been 65 per cent (sixty-four of ninety-eight hips)³⁷, and 100 per cent (twenty-two hips)¹¹.

The concepts of the original design of the Charnley prosthesis have remained unaltered to the present day. Adherence to the principle of low frictional torque is achieved with use of the 22.25-millimeter-diameter femoral head, which minimizes the sliding distance and results in lower rates of wear (penetration) and of loosening of the acetabular component. The Charnley low-friction prosthesis has been shown to perform well in the long term, both in young patients and in the general population^{17,23,31,33,34}. In addition, if the components have been fixed solidly at the time of the operation, the bone-cement interface is not likely to deteriorate with time or to result in late failure⁴⁵. Because of improvements in cementing technique¹, implant materials, and design that have occurred since the prosthesis was introduced, it is not unreasonable to expect additional reductions in the rates of aseptic loosening of both the acetabular and the femoral component.

More studies with at least twenty years of follow-up are necessary in order to reveal the ultimate performance of the current implants. Ideally, these studies should include large numbers of patients in well defined diagnostic groups and with consistently high rates of follow-up. As the number of patients in studies of the general population inevitably decreases because of death, the most accurate long-term data is accrued from studies of patients who are relatively young at the time of the operation. Therefore, studies involving young patients are of particular importance.

In 1979, shortly after the completion of the operations in the present series of patients, Charnley⁸ stated that experience with arthroplasty in patients who were less than forty-five years old was only beginning. The long-term results in this group of patients, which represent the outcome of the early pioneering experience with use of relatively unsophisticated techniques, have been good. Survivorship analysis showed that the probability of survival (and 95 per cent confidence intervals) of the acetabular and femoral components at twenty years was 73 per cent (61 to 84 per cent) and 91 per cent (83 to 99 per cent), respectively, which was similar to that in previous studies^23,34. The probability of survival of the acetabular and femoral components at thirty years was 70 per cent (57 to 83 per cent) and 83 per cent (72 to 94 per cent), respectively, with a 70 per cent (57 to 83 per cent) likelihood of survival of both original components. The Charnley low-friction arthroplasty with cement currently represents a reliable solution for young patients who have ankylosing spondylitis as the prosthesis has a good rate of survival and there are few complications.

NOTE: The authors thank Mrs. E. Jenkins of the Clinical Outcomes Research Group, Wrightington Hospital, for assistance in compiling and analyzing the data.

References

Introduction | Materials and Methods | Results | Discussion | References

Barrack, R. L.; Mulroy, R. D. Jr.; and Harris, W. H.: Improved cementing techniques and femoral component loosening in young patients with hip arthroplasty. A 12-year radiographic review. J. Bone and Joint Surg.,74-B(3): 385-389, 1992.74-B(3)385 1992

Bisla, R. S.; Ranawat, C. S.; and Inglis, A. E.: Total hip replacement in patients with ankylosing spondylitis with involvement of the hip. J. Bone and Joint Surg.,58-A: 233-238, March 1976.58-A233 1976

Boardman, K. P.; Bocco, F.; and Charnley, J.: An evaluation of a method of trochanteric fixation using three wires in the Charnley low friction arthroplasty. Clin. Orthop.,132: 31-38, 1978.13231 1978 [PubMed]

Brooker, A. F.; Bowerman, J. W.; Robinson, R. A.; and Riley, L. H., Jr.: Ectopic ossification following total hip replacement. Incidence and a method of classification. J. Bone and Joint Surg.,55-A: 1629-1632, Dec. 1973.55-A1629 1973

Calin, A., and Elswood, J.: The outcome of 138 total hip replacements and 12 revisions in ankylosing spondylitis: high success rate after a mean followup of 7.5 years. J. Rheumatol.,16: 955-958, 1989.16955 1989 [PubMed]

Chandler, H. P.; Reineck, F. T.; Wixson, R. L.; and McCarthy, J. C.: Total hip replacements in patients younger than thirty years old. A five-year follow-up study. J. Bone and Joint Surg.,63-A: 1426-1434, Dec. 1981.63-A1426 1981

Charnley, J.: The long-term results of low-friction arthroplasty of the hip performed as a primary intervention. J. Bone and Joint Surg.,54-B(1): 61-76, 1972.54-B(1)61 1972

Charnley, J.: Low Friction Arthroplasty of the Hip. Theory and Practice. New York, Springer, 1979.

Charnley, J., and Halley, D. K.: Rate of wear in total hip replacement. Clin. Orthop.,112: 170-179, 1975.112170 1975 [PubMed]

DeLee, J. G., and Charnley, J.: Radiological demarcation of cemented sockets in total hip replacement. Clin. Orthop.,121: 20-32, 1976.12120 1976 [PubMed]

DeLee, J.; Ferrari, A.; and Charnley, J.: Ectopic bone formation following low friction arthroplasty of the hip. Clin. Orthop.,121: 53-59, 1976.12153 1976 [PubMed]

Dorr, L. D.; Takei, G. K.; and Conaty, J. P.: Total hip arthroplasties in patients less than forty-five years old. J. Bone and Joint Surg.,65-A: 474-479, April 1983.65-A474 1983

Finsterbush, A.; Amir, D.; Vatashki, E.; and Husseini, N.: Joint surgery in severe ankylosing spondylitis. Acta Orthop. Scandinavica,59: 491-496, 1988.59491 1988

Griffith, M. J.; Seidenstein, M. K.; Williams, D.; and Charnley, J.: Socket wear in Charnley low friction arthroplasty of the hip. Clin. Orthop.,137: 37-47, 1978.13737 1978 [PubMed]

Gruen, T. A.; McNeice, G. M.; and Amstutz, H. C.: "Modes of failure" of cemented stem-type femoral components. A radiographic analysis of loosening. Clin. Orthop.,141: 17-27, 1979.14117 1979 [PubMed]

Gualtieri, G.; Gualtieri, I.; Hendriks, M.; and Gagliardi, S.: Comparison of cemented ceramic and metal-polyethylene coupling hip prostheses in ankylosing spondylitis. Clin. Orthop.,282: 81-85, 1992.28281 1992 [PubMed]

Halley, D. K., and Charnley, J.: Results of low friction arthroplasty in patients thirty years of age or younger. Clin. Orthop.,112: 180-191, 1975.112180 1975 [PubMed]

Hodgkinson, J. P.; Shelley, P.; and Wroblewski, B. M.: The correlation between the roentgenographic appearance and operative findings at the bone-cement junction of the socket in Charnley low friction arthroplasties. Clin. Orthop.,228: 105-109, 1988.228105 1988 [PubMed]

Howorth, F. H.: Air flow patterns in the operating theatre. Engin. Med.,9: 87-92, 1980.987 1980

Isaac, G.; Wroblewski, B. M.; Atkinson, J. R.; and Dowson, D.: A tribological study of retrieved hip prostheses. Clin. Orthop.,276: 115-125, 1992.276115 1992 [PubMed]

Jensen, N. F., and Harris, W. H.: A system for trochanteric osteotomy and reattachment for total hip arthroplasty with a ninety-nine percent union rate. Clin. Orthop.,208: 174-181, 1986.208174 1986 [PubMed]

Johnson, R., and Loudon, J. R.: Hydroxychloroquine sulfate prophylaxis for pulmonary embolism for patients with low-friction arthroplasty. Clin. Orthop.,211: 151-153, 1986.211151 1986 [PubMed]

Joshi, A. B.; Porter, M. L.; Trail, I. A.; Hunt, L. P.; Murphy, J. C. M.; and Hardinge, K.: Long-term results of Charnley low-friction arthroplasty in young patients. J. Bone and Joint Surg.,75-B(4): 616-623, 1993.75-B(4)616 1993

Kaplan, E. L., and Meier, P.: Nonparametric estimation from incomplete observations. J. Am. Statist. Assn.,53: 457-481, 1958.53457 1958

Kilgus, D. J.; Namba, R. S.; Gorek, J. E.; Cracchiolo, A., III; and Amstutz, H. C.: Total hip replacement for patients who have ankylosing spondylitis. The importance of the formation of heterotopic bone and of the durability of fixation of cemented components. J. Bone and Joint Surg.,72-A: 834-839, July 1990.72-A834 1990

Lakatos, J., and Csakanyi, L.: Comparison of complications of total hip arthroplasty in rheumatoid arthritis, ankylosing spondylitis, and osteoarthritis. Orthopedics,14: 55-57, 1991.1455 1991 [PubMed]

Loudon, J. R.: Hydroxychloroquine and postoperative thromboembolism after total hip replacement. Am. J. Med.,85 (Supplement 4A): 57-61, 1988.85 (Supplement 4A)57 1988

Loudon, J. R., and Charnley, J.: Subsidence of the femoral prosthesis in total hip replacement in relation to the design of the stem. J. Bone and Joint Surg.,62-B(4): 450-453, 1980.62-B(4)450 1980

Loudon, J. R., and Older, M. W. J.: Subsidence of the femoral component related to long-term outcome of hip replacement. J. Bone and Joint Surg.,71-B(4): 624-628, 1989.71-B(4)624 1989

Merle d'Aubigné, R., and Postel, M.: Functional results of hip arthroplasty with acrylic prosthesis. J. Bone and Joint Surg.,36-A: 451-475, June 1954.36-A451 1954

Neumann, L.; Freund, K. G.; and Sorensen, K. H.: Total hip arthroplasty with the Charnley prosthesis in patients fifty-five years old and less. Fifteen to twenty-one-year results. J. Bone and Joint Surg.,78-A: 73-79, Jan. 1996.78-A73 1996

Ritter, M. A., and Vaughan, R. B.: Ectopic ossification after total hip arthroplasty. Predisposing factors, frequency, and effect on results. J. Bone and Joint Surg.,59-A: 345-351, April 1977.59-A345 1977

Sarmiento, A.; Ebramzadeh, E.; Gogan, W. J.; and McKellop, H. A.: Total hip arthroplasty with cement. A long-term radiographic analysis in patients who are older than fifty and younger than fifty years. J. Bone and Joint Surg.,72-A: 1470-1476, Dec. 1990.72-A1470 1990

Schulte, K. R.; Callaghan, J. J.; Kelley, S. S.; and Johnston, R. C.: The outcome of Charnley total hip arthroplasty with cement after a minimum twenty-year follow-up. The results of one surgeon. J. Bone and Joint Surg.,75-A: 961-975, July 1993.75-A961 1993

Sharp, D. J., and Porter, K. M.: The Charnley total hip arthroplasty in patients under age 40. Clin. Orthop.,201: 51-56, 1985.20151 1985 [PubMed]

Sochart, D. H.; Paul, A. S.; and Kurdy, N. M.: A new osteotome for performing chevron trochanteric osteotomy. Acta Orthop. Scandinavica,66: 445-446, 1995.66445 1995

Sundaram, N. A., and Murphy, J. C. M.: Heterotopic bone formation following total hip arthroplasty in ankylosing spondylitis. Clin. Orthop.,207: 223-226, 1986.207223 1986 [PubMed]

Walker, L. G., and Sledge, C. B.: Total hip arthroplasty in ankylosing spondylitis. Clin. Orthop.,262: 198-204, 1991.262198 1991 [PubMed]

Welch, R. B., and Charnley, J.: Low-friction arthroplasty of the hip in rheumatoid arthritis and ankylosing spondylitis. Clin. Orthop.,72: 22-32, 1970.7222 1970 [PubMed]

White, S. H.: The fate of cemented total hip arthroplasty in young patients. Clin. Orthop.,231: 29-34, 1988.23129 1988 [PubMed]

Williams, E.; Taylor, A. R.; Arden, G. P.; and Edwards, D. H.: Arthroplasty of the hip in ankylosing spondylitis. J. Bone and Joint Surg.,59-B(4): 393-397, 1977.59-B(4)393 1977

Williams, W. W., and McCullough, C. J.: Results of cemented total hip replacement in juvenile chronic arthritis. A radiological review. J. Bone and Joint Surg.,75-B(6): 872-874, 1993.75-B(6)872 1993

Witt, J. D.; Swann, M.; and; and Ansell, B. M.: Total hip replacement for juvenile chronic arthritis. J. Bone and Joint Surg.,73-B(5): 770-773, 1991.73-B(5)770 1991

Wroblewski, B. M.: Direction and rate of socket wear in Charnley low-friction arthroplasty. J. Bone and Joint Surg.,67-B(5): 757-761, 1985.67-B(5)757 1985

Wroblewski, B. M.: 15-21-year results of the Charnley low-friction arthroplasty. Clin. Orthop.,211: 30-35, 1986.21130 1986 [PubMed]

Topics