- Email: [email protected]
Charnley low-friction arthroplasty
The Journal of Arthroplasty Vol. 17 No. 6 2002
Charnley Low-Friction Arthroplasty A Worldwide Retrospective Review at 15 to 20 Years John Older, FRCS
Abstract: This is a retrospective survivorship analysis review of 5,089 Charnley low-friction arthroplasties performed as a primary procedure at 8 hospitals around the world before December 31, 1980. Data collected were simple with a well-defined endpoint: death or revision. There was no clinical or radiographic evaluation. The mean age at operation was 63 years; 57% of hips were in women, and 43% were in men. The preoperative diagnosis was primary osteoarthrosis in 74% of hips. KaplanMeier survivorship analysis using failure as revision of any component for any cause showed probability of survival at 20 years of 83% (women, 86%, and men, 78%). In patients aged 70 to 80 at surgery, probability of survival was 92%; in patients aged ⱕ40, probability of survival was 67%. This study showed almost identical durability of the femoral versus acetabular component— 87%. Key words: Charnley, primary hip arthroplasty, retrospective survivorship analysis. Copyright 2002, Elsevier Science (USA). All rights reserved.
femoral component with a 22.225-mm diameter head. Six of the units later changed to a matte surface and round-backed prosthesis. The cups were ultra-high molecular weight polyethylene with an outer diameter of either 40 or 44 mm. CMW cement (available in the early 1970s) (Thackray, Blackpool, United Kingdom) at first was used by all surgeons. Half the surgeons later changed to Simplex (Howmedica, Limerick, Ireland) or Palacos R (Heraeus-Kulzer, Darmstadt, Germany). Everyone used finger packing—The Charnley 2-thumb technique. No guns were used. Every surgeon used a classic Charnley lateral approach with trochanteric osteotomy. The data obtained were simple: patient age, patient sex, diagnosis of the hip problem requiring arthroplasty, date of the primary operation, date of last evaluation, and death or revision with the reason for reoperation. All index LFA operations in this series were completed before December 31, 1980. A few assumptions were necessary collecting these simple data. When the data of the last follow-up evaluation or death was given in months or years, the first full date to fall in the range was used
The aim of this study was to follow-up and evaluate the long-term efficacy of the Charnley low-frictional torque arthroplasty (LFA) using data from various hospitals around the world.
Materials and Methods Data were collected from 8 centers. In most of the hospitals, the LFA was performed by 1 experienced high-volume surgeon who originally had worked at Wrightington with Charnley, who was the surgeon himself in 1 unit (Table 1). A Charnley hip prosthesis was used in all patients. Initially, all surgeons used a stainless steel, polished surface, flat-backed From the John Charnley Hip Unit, King Edward VII Hospital, Midhurst, West Sussex, United Kingdom. Submitted December 20, 2000; accepted December 14, 2001. No benefits or funds were received in support of this study. Reprint requests: John Older, FRCS, John Charnley Hip Unit, King Edward VII Hospital, Midhurst, West Sussex, GU29 OBL, UK. Copyright 2002, Elsevier Science (USA). All rights reserved. 0883-5403/02/1706-0001$35.00/0 doi:10.1054/arth.2002.31973
675
676 The Journal of Arthroplasty Vol. 17 No. 6 September 2002 Table 1. Contributors Louis Brady Ake Carlsson Eduardo Garcia-Cimbrelo Nas Eftekhar Henry Hamilton Richard Johnston John Older* Johan Walters
Table 3. Patient Status
Florida Malmo Madrid New York Ontario Iowa Midhurst Cape Town
US Sweden Spain US Canada US UK South Africa
Active Dead Lost to follow-up Revised Removed
n
%
2,563 1,808 208 478 32
50.4 35.5 4.1 9.4 0.6
*Surgeon—John Charnley
to keep all estimates conservative. If January 1992 was given as the last date of follow-up, the first of the month was used as the date. If the only information given was a year, the first day of the year was used. When a revision for infection was listed it was assumed both components were removed; in roughly half of the revisions for infections, details of the component removed were not given. It was not possible to evaluate radiographs or to make any detailed clinical evaluation, such as the Charnley modification of the d’Aubigne and Postel numerical grading system. Survivorship analysis was performed using the product limit method [1]. This method recalculates the survival rate each time a failure occurs [2].
Results Up to the end of 1980, 5,089 Charnley LFAs were performed as a primary procedure at 8 institutions included in this study (Table 2). All the data refer to hips and not to patients. Some patients had bilateral LFAs and are counted as 2 hips throughout the article. The mean patient age at operation was 63 years; 2,683 hips (57%) were in women, 2,036 hips (43%) were in men, and the gender was unknown in 371 hips. The preoperative diagnosis was nonin-
flammatory primary osteoarthrosis in 3,776 hips (74%), inflammatory arthritis in 402 (8%) hips, and other in 97 (2%). No diagnosis was available in 814 (16%) cases. At the last review, half of the original hip prostheses still were being actively used, whereas a little more than a third of the patients in the series had died. Of hips, 4% had been lost to follow-up, and 9% required revision (Table 3). There was considerable variation in the average follow-up of active patients among centers (Table 4). Of all patients, 61% had ⱖ10 years’ follow-up. This includes all patients followed, no matter what the endpoint. Of the patients who were still alive at their last evaluation and unrevised, 78% had follow-up ⬎10 years. Of the patient population, alive on their last observation and not revised, 22% had not been seen in the 10 years after their operation. Of all patients, 35% had follow-up ⬎15 years. Of the patients who still were considered active, 55% had follow-up ⬎15 years. Of active patients, 45% had not been seen for ⱖ5 years. Of all patients, 10% had at least 20 years’ follow-up, which calculated to 17% for patients still active at last evaluation (Table 5). We hope that this assessment provides the reader with a sense of the quality of follow-up in this collection of hip arthroplasties.
Survivorship Analysis This article is entirely a survivorship analysis using the Kaplan-Meier [2] technique, which has
Table 2. Retrospective Data Collection
Table 4. Average Follow-Up of Active Patients
Unit
No. of Hips
Unit
n
Years (mean)
1 2 3 4 5 6 7 8
446 628 830 790 199 1683 369 144
1 2 3 4 5 6 7 8
353 268 332 567 53 838 91 61
9.6 13.0 18.3 17.6 6.7 12.5 22.0 17.4
Charnley LFA Worldwide Review at 20 Years • John Older Table 5. Demographics Follow-Up Beyond Years 10 15 20
All (%)
Active (%)
61 35 10
78 55 17
been analyzed critically in depth by Murray, Carl, and Bulstrode [3]. The endpoint in assessing probability of survival is paramount. In this series, the endpoint is solely whether the prosthesis required revision. The Kaplan-Meier estimates of survival for the 5,089 primary total hip arthroplasties is shown in Table 6. This table illustrates the importance of having a huge database because this overcomes potential errors of Kaplan-Meier probability of survival and gives an extremely high statistical significance and ensures that bias is distributed randomly [3].
Discussion The survivorship analysis of 5,089 of the primary cemented Charnley LFAs in this series is relatively high, with a probability of survival at 20 years of 83% (Fig. 1A). Excluding revisions for infection, the survivorship is 84%. The probability of survival at 20 years in women is 86% (Fig. 1B) and in men it is 78% (Fig. 1C). There was considerable variation in the average follow-up of active patients among centers. Two centers with poor follow-up were excluded. Analyzing the data from the remaining 6 centers shows survivorship is 83%. This percentage is identical to
677
that with the 8 centers and emphasizes the importance of large numbers (Table 7). Demographics are important because male gender and young age increase the risk for revision resulting from aseptic loosening. In this study, the probability of survival in patients aged 70 to 80 years at surgery was 92% (Fig. 1D). This percentage confirms that for the elderly patient the outcome of total hip arthroplasty is good. In younger and more active patients, the results are significantly worse and not acceptable. Patients aged ⱕ40 years at operation in the series had a probability of survival of 67% (Fig. 1E). The expectations of young patients are unrealistic and conflict with the longterm survival of cemented and uncemented implants. This conflict indicates the importance of redefining indications for total hip arthroplasty and the further development of new bearing configurations in young and active individuals. Comparing the relative durability of the femoral versus the acetabular component, excluding infection or stem fractures, this study showed little difference (87.3%, stem, and 87.2%, acetabulum). The good survival of the cup reflects the different opinion between the United States and Europe with respect to using cemented or cementless cups. The difference may be related to the use of the 22.225-mm head of the femoral component and the thick-walled plastic cup used in the Charnley LFA. Charnley’s pioneering work shaped the concept of the LFA: a small-diameter head, 22.225 mm, articulating in a cup of maximal external diameter to reduce frictional torque at the cement interface and prolong acetabular fixation. Initially, in all units, the acetabular cups were standard and unflanged as evolved in 1962. The long posterior wall was introduced in 1972 to enhance the stability of the arthroplasty, especially in adduction, flexion,
Table 6. Survivorship
n
n
%
% Survival Estimate 20 years
5,089 5,089 5,089 5,089 2,036 2,682 1,258 175
510 450 369 349 256 224 52 47
10.0 8.8 8.4 7.9 12.6 8.4 4.1 26.9
82.9 84.2 87.3 87.2 77.8 85.4 92.3 67.1
Failure
All patients All patients excluding infected revisions Femoral stem Cup Male Female Aged 70–80 y at surgery Aged ⬍40 y at surgery
95% Confidence Interval for Mean 81.3–84.5 82.6–85.8 85.8–88.7 85.6–88.7 74.7–80.8 83.3–87.6 89.5–95.0 58.6–75.6
678 The Journal of Arthroplasty Vol. 17 No. 6 September 2002
Fig. 1. Charnley LFA survivorship analysis. Failure ⫽ Revision of any component for any cause. Graphs showing the probability of survival at 20 years with a 95% confidence range limit for survivorship.
Table 7. Comparison of Units With Variable Follow-Up No. of Hips 8U 6U
5,089 2,960 Probability of 20-yr survival 83%
Failures 510 (10%) 355 (12%)
and internal rotation. This wall reduced the dislocation rate. The same year, a flange was added to the cup. This flange is trimmed to customize the cup to fit against the bone of the acetabulum. The flange increases cement pressurization, improves load distribution, and protects the acetabular bearing surface from the ingress of cement particles. The unique S-shaped flange of the Ogee Cup (Thackray, Leeds, United Kingdom), concave to convex, was
Charnley LFA Worldwide Review at 20 Years • John Older
introduced in 1982 to make best use of available bone stock. Late in this series, 4 centers changed to flanged cups with a long posterior wall. Stem fractures of the femoral component were a major problem in the early flat-back series. To improve stability and reduce fractures, the design was changed. The round-back stem replaced the flatback stem in 1972. The greater surface area of metal strengthened the stem. The radiused edges reduced stress within the cement mantle. In 1975, the undercut flange was introduced to improve pressurization of cement and load distribution and to resist subsidence. Six centers changed to a matte-surface, round-backed, and flanged prosthesis when they became available. Ortron 90 (Thackray, Leeds, United Kingdom) is a cold worked stainless steel with high fatique strength and corrosion resistance. Since its introduction in 1982 to replace 316 stainless steel, no cases of stem fracture have occurred. The high strength of Ortron 90 enabled the diameter of the neck to be reduced from 12.7 mm to 10 mm. This reduction increases the range of motion before neck impingement occurs from 90° to 108°. This increased range of motion in turn enhances the life of acetabular fixation. The original femoral flat back was polished until 1969, when all implants manufactured by Thackray had a Vaquasheen finish. Vaquasheen is a nonpolished coating considered to be an ideal finish to an implant whose design philosophy is founded on stability and not subsidence. The outcome of the polished compared with matte surface femoral components is being debated. Some authorities consider that the surface of a femoral component cannot be judged alone but has to be considered together with the overall design of the component and the technique of insertion and long-term behavior between implant, cement, and bone. In this series, in 6 of the 8 units centers, there was a change from polish to matte, but so far no significant variation has been shown with respect to type of prosthesis and outcome. The definition for failure in this series is revision— exchange of 1 or both components or removal of the implant. This is a distinct endpoint, but this definition is of low sensitivity. Clinical results must not be taken as an indication of the mechanical state of the arthroplasty. All nonrevised hips are not necessarily successful. Many cup failures are symptomless for long periods. Gross loss of bone stock can occur in the absence of pain and disability. This possibility emphasizes the importance of performing regular radiographic follow-up of specific patient populations at risk so that revision surgery is performed at an early stage of the mechanical fail-
679
ure. This study recognizes its limitations; it is essentially a retrospective survivorship analysis in which the endpoint is revision. It is not a clinical or radiographic review.
Conclusion Charnley’s vocation was to give a patient a fully functional, pain-free artificial hip that lasted their life span with no complications. Charnley’s own review at 25 years showed that 92% remain good and functional until death [5]. Three quarters of the elderly patients still active retained their original Charnley hip with a probable survivorship at 25 years of 85% [5]. With ⬎1 million Charnley prostheses implanted, the clinical results now are being published from various centers around the world. The probability of the Charnley LFA surviving 20 years has been reported by Schulte and Callaghan et al [6] and Kavanagh and Wallrichs et al [7] as 84% and by Neumann, Freund, and Sorenson et al [8] as 89%. These articles are supported by the results of this study, which, using basic data and Kaplan-Meier survivorship methods, in ⬎5,000 LFAs from 8 institutions around the world shown a 20-year probable survivorship of 83%. The Charnley metallic femoral component with a small head and the polymeric cup has been a dominant factor in total hip arthroplasty for 40 years. There is now a shortfall—wear and its associated aseptic loosening, especially in young patients. The orthopaedic community is looking at alternative methods with new bearing configurations to minimize wear. Such advances could allow the possibility of operating on younger patients and eliminate the need for revision surgery in young and old patients. The tradition of metal-on-polyethylene or ceramic-on-polyethylene configuration is fighting back with a new 0 wear polyethylene, however. The long-term success of new devices must be judged against the well-proven success of the Charnley LFA.
Acknowledgment This is the first time a survivorship analysis of the longevity of a total hip arthroplasty has been made on data collected from hospitals around the world. The author thanks all the participating orthopaedic surgeons; their friendship and help has been essential for this study. Thanks also to John Blachere, biostatistician, for outstanding help with the statistical analysis.
680 The Journal of Arthroplasty Vol. 17 No. 6 September 2002
References 1. Armitage P, Berry G: Statistical methods in medical research, 2nd ed. Blackwell Scientific, Oxford, 1987 2. Kaplan EL, Meier P: Nonparametric estimation from incomplete observation. J Am Stat Assoc 53:457, 1958 3. Murray DW, Carr AJ, Bulstrode C: Survival analysis of joint replacements. J Bone Joint Surg Br 75:697, 1993 4. Dorey F, Amstutz HC: The validity of survivorship and analysis in total joint arthroplasty. J Bone Joint Surg Am 71:544, 1989 5. Older J: Charnley’s by Charnley: a minimum follow
up of twenty years. Presidential Guest Speaker. The Hip Society, Orlando, FL, 1995 6. Schulte HR, Callaghan JJ, Kelley SS, Johnston RC: The outcome of Charnley total hip arthroplasty with cement after a minimum twenty year follow up. J Bone Joint Surg Am 75:961, 1993 7. Kavanagh BF, Wallrichs S, Dewitz M, et al: Charnley low friction arthroplasty of the hip, twenty year results with cement. J Arthroplasty 9:229, 1994 8. Neumann L, Freund KG, Sorenson KH: Long term results of Charnley total hip replacement. J Bone Joint Surg Br 76:245, 1994
Charnley Low-Friction Arthroplasty A Worldwide Retrospective Review at 15 to 20 Years John Older, FRCS
Abstract: This is a retrospective survivorship analysis review of 5,089 Charnley low-friction arthroplasties performed as a primary procedure at 8 hospitals around the world before December 31, 1980. Data collected were simple with a well-defined endpoint: death or revision. There was no clinical or radiographic evaluation. The mean age at operation was 63 years; 57% of hips were in women, and 43% were in men. The preoperative diagnosis was primary osteoarthrosis in 74% of hips. KaplanMeier survivorship analysis using failure as revision of any component for any cause showed probability of survival at 20 years of 83% (women, 86%, and men, 78%). In patients aged 70 to 80 at surgery, probability of survival was 92%; in patients aged ⱕ40, probability of survival was 67%. This study showed almost identical durability of the femoral versus acetabular component— 87%. Key words: Charnley, primary hip arthroplasty, retrospective survivorship analysis. Copyright 2002, Elsevier Science (USA). All rights reserved.
femoral component with a 22.225-mm diameter head. Six of the units later changed to a matte surface and round-backed prosthesis. The cups were ultra-high molecular weight polyethylene with an outer diameter of either 40 or 44 mm. CMW cement (available in the early 1970s) (Thackray, Blackpool, United Kingdom) at first was used by all surgeons. Half the surgeons later changed to Simplex (Howmedica, Limerick, Ireland) or Palacos R (Heraeus-Kulzer, Darmstadt, Germany). Everyone used finger packing—The Charnley 2-thumb technique. No guns were used. Every surgeon used a classic Charnley lateral approach with trochanteric osteotomy. The data obtained were simple: patient age, patient sex, diagnosis of the hip problem requiring arthroplasty, date of the primary operation, date of last evaluation, and death or revision with the reason for reoperation. All index LFA operations in this series were completed before December 31, 1980. A few assumptions were necessary collecting these simple data. When the data of the last follow-up evaluation or death was given in months or years, the first full date to fall in the range was used
The aim of this study was to follow-up and evaluate the long-term efficacy of the Charnley low-frictional torque arthroplasty (LFA) using data from various hospitals around the world.
Materials and Methods Data were collected from 8 centers. In most of the hospitals, the LFA was performed by 1 experienced high-volume surgeon who originally had worked at Wrightington with Charnley, who was the surgeon himself in 1 unit (Table 1). A Charnley hip prosthesis was used in all patients. Initially, all surgeons used a stainless steel, polished surface, flat-backed From the John Charnley Hip Unit, King Edward VII Hospital, Midhurst, West Sussex, United Kingdom. Submitted December 20, 2000; accepted December 14, 2001. No benefits or funds were received in support of this study. Reprint requests: John Older, FRCS, John Charnley Hip Unit, King Edward VII Hospital, Midhurst, West Sussex, GU29 OBL, UK. Copyright 2002, Elsevier Science (USA). All rights reserved. 0883-5403/02/1706-0001$35.00/0 doi:10.1054/arth.2002.31973
675
676 The Journal of Arthroplasty Vol. 17 No. 6 September 2002 Table 1. Contributors Louis Brady Ake Carlsson Eduardo Garcia-Cimbrelo Nas Eftekhar Henry Hamilton Richard Johnston John Older* Johan Walters
Table 3. Patient Status
Florida Malmo Madrid New York Ontario Iowa Midhurst Cape Town
US Sweden Spain US Canada US UK South Africa
Active Dead Lost to follow-up Revised Removed
n
%
2,563 1,808 208 478 32
50.4 35.5 4.1 9.4 0.6
*Surgeon—John Charnley
to keep all estimates conservative. If January 1992 was given as the last date of follow-up, the first of the month was used as the date. If the only information given was a year, the first day of the year was used. When a revision for infection was listed it was assumed both components were removed; in roughly half of the revisions for infections, details of the component removed were not given. It was not possible to evaluate radiographs or to make any detailed clinical evaluation, such as the Charnley modification of the d’Aubigne and Postel numerical grading system. Survivorship analysis was performed using the product limit method [1]. This method recalculates the survival rate each time a failure occurs [2].
Results Up to the end of 1980, 5,089 Charnley LFAs were performed as a primary procedure at 8 institutions included in this study (Table 2). All the data refer to hips and not to patients. Some patients had bilateral LFAs and are counted as 2 hips throughout the article. The mean patient age at operation was 63 years; 2,683 hips (57%) were in women, 2,036 hips (43%) were in men, and the gender was unknown in 371 hips. The preoperative diagnosis was nonin-
flammatory primary osteoarthrosis in 3,776 hips (74%), inflammatory arthritis in 402 (8%) hips, and other in 97 (2%). No diagnosis was available in 814 (16%) cases. At the last review, half of the original hip prostheses still were being actively used, whereas a little more than a third of the patients in the series had died. Of hips, 4% had been lost to follow-up, and 9% required revision (Table 3). There was considerable variation in the average follow-up of active patients among centers (Table 4). Of all patients, 61% had ⱖ10 years’ follow-up. This includes all patients followed, no matter what the endpoint. Of the patients who were still alive at their last evaluation and unrevised, 78% had follow-up ⬎10 years. Of the patient population, alive on their last observation and not revised, 22% had not been seen in the 10 years after their operation. Of all patients, 35% had follow-up ⬎15 years. Of the patients who still were considered active, 55% had follow-up ⬎15 years. Of active patients, 45% had not been seen for ⱖ5 years. Of all patients, 10% had at least 20 years’ follow-up, which calculated to 17% for patients still active at last evaluation (Table 5). We hope that this assessment provides the reader with a sense of the quality of follow-up in this collection of hip arthroplasties.
Survivorship Analysis This article is entirely a survivorship analysis using the Kaplan-Meier [2] technique, which has
Table 2. Retrospective Data Collection
Table 4. Average Follow-Up of Active Patients
Unit
No. of Hips
Unit
n
Years (mean)
1 2 3 4 5 6 7 8
446 628 830 790 199 1683 369 144
1 2 3 4 5 6 7 8
353 268 332 567 53 838 91 61
9.6 13.0 18.3 17.6 6.7 12.5 22.0 17.4
Charnley LFA Worldwide Review at 20 Years • John Older Table 5. Demographics Follow-Up Beyond Years 10 15 20
All (%)
Active (%)
61 35 10
78 55 17
been analyzed critically in depth by Murray, Carl, and Bulstrode [3]. The endpoint in assessing probability of survival is paramount. In this series, the endpoint is solely whether the prosthesis required revision. The Kaplan-Meier estimates of survival for the 5,089 primary total hip arthroplasties is shown in Table 6. This table illustrates the importance of having a huge database because this overcomes potential errors of Kaplan-Meier probability of survival and gives an extremely high statistical significance and ensures that bias is distributed randomly [3].
Discussion The survivorship analysis of 5,089 of the primary cemented Charnley LFAs in this series is relatively high, with a probability of survival at 20 years of 83% (Fig. 1A). Excluding revisions for infection, the survivorship is 84%. The probability of survival at 20 years in women is 86% (Fig. 1B) and in men it is 78% (Fig. 1C). There was considerable variation in the average follow-up of active patients among centers. Two centers with poor follow-up were excluded. Analyzing the data from the remaining 6 centers shows survivorship is 83%. This percentage is identical to
677
that with the 8 centers and emphasizes the importance of large numbers (Table 7). Demographics are important because male gender and young age increase the risk for revision resulting from aseptic loosening. In this study, the probability of survival in patients aged 70 to 80 years at surgery was 92% (Fig. 1D). This percentage confirms that for the elderly patient the outcome of total hip arthroplasty is good. In younger and more active patients, the results are significantly worse and not acceptable. Patients aged ⱕ40 years at operation in the series had a probability of survival of 67% (Fig. 1E). The expectations of young patients are unrealistic and conflict with the longterm survival of cemented and uncemented implants. This conflict indicates the importance of redefining indications for total hip arthroplasty and the further development of new bearing configurations in young and active individuals. Comparing the relative durability of the femoral versus the acetabular component, excluding infection or stem fractures, this study showed little difference (87.3%, stem, and 87.2%, acetabulum). The good survival of the cup reflects the different opinion between the United States and Europe with respect to using cemented or cementless cups. The difference may be related to the use of the 22.225-mm head of the femoral component and the thick-walled plastic cup used in the Charnley LFA. Charnley’s pioneering work shaped the concept of the LFA: a small-diameter head, 22.225 mm, articulating in a cup of maximal external diameter to reduce frictional torque at the cement interface and prolong acetabular fixation. Initially, in all units, the acetabular cups were standard and unflanged as evolved in 1962. The long posterior wall was introduced in 1972 to enhance the stability of the arthroplasty, especially in adduction, flexion,
Table 6. Survivorship
n
n
%
% Survival Estimate 20 years
5,089 5,089 5,089 5,089 2,036 2,682 1,258 175
510 450 369 349 256 224 52 47
10.0 8.8 8.4 7.9 12.6 8.4 4.1 26.9
82.9 84.2 87.3 87.2 77.8 85.4 92.3 67.1
Failure
All patients All patients excluding infected revisions Femoral stem Cup Male Female Aged 70–80 y at surgery Aged ⬍40 y at surgery
95% Confidence Interval for Mean 81.3–84.5 82.6–85.8 85.8–88.7 85.6–88.7 74.7–80.8 83.3–87.6 89.5–95.0 58.6–75.6
678 The Journal of Arthroplasty Vol. 17 No. 6 September 2002
Fig. 1. Charnley LFA survivorship analysis. Failure ⫽ Revision of any component for any cause. Graphs showing the probability of survival at 20 years with a 95% confidence range limit for survivorship.
Table 7. Comparison of Units With Variable Follow-Up No. of Hips 8U 6U
5,089 2,960 Probability of 20-yr survival 83%
Failures 510 (10%) 355 (12%)
and internal rotation. This wall reduced the dislocation rate. The same year, a flange was added to the cup. This flange is trimmed to customize the cup to fit against the bone of the acetabulum. The flange increases cement pressurization, improves load distribution, and protects the acetabular bearing surface from the ingress of cement particles. The unique S-shaped flange of the Ogee Cup (Thackray, Leeds, United Kingdom), concave to convex, was
Charnley LFA Worldwide Review at 20 Years • John Older
introduced in 1982 to make best use of available bone stock. Late in this series, 4 centers changed to flanged cups with a long posterior wall. Stem fractures of the femoral component were a major problem in the early flat-back series. To improve stability and reduce fractures, the design was changed. The round-back stem replaced the flatback stem in 1972. The greater surface area of metal strengthened the stem. The radiused edges reduced stress within the cement mantle. In 1975, the undercut flange was introduced to improve pressurization of cement and load distribution and to resist subsidence. Six centers changed to a matte-surface, round-backed, and flanged prosthesis when they became available. Ortron 90 (Thackray, Leeds, United Kingdom) is a cold worked stainless steel with high fatique strength and corrosion resistance. Since its introduction in 1982 to replace 316 stainless steel, no cases of stem fracture have occurred. The high strength of Ortron 90 enabled the diameter of the neck to be reduced from 12.7 mm to 10 mm. This reduction increases the range of motion before neck impingement occurs from 90° to 108°. This increased range of motion in turn enhances the life of acetabular fixation. The original femoral flat back was polished until 1969, when all implants manufactured by Thackray had a Vaquasheen finish. Vaquasheen is a nonpolished coating considered to be an ideal finish to an implant whose design philosophy is founded on stability and not subsidence. The outcome of the polished compared with matte surface femoral components is being debated. Some authorities consider that the surface of a femoral component cannot be judged alone but has to be considered together with the overall design of the component and the technique of insertion and long-term behavior between implant, cement, and bone. In this series, in 6 of the 8 units centers, there was a change from polish to matte, but so far no significant variation has been shown with respect to type of prosthesis and outcome. The definition for failure in this series is revision— exchange of 1 or both components or removal of the implant. This is a distinct endpoint, but this definition is of low sensitivity. Clinical results must not be taken as an indication of the mechanical state of the arthroplasty. All nonrevised hips are not necessarily successful. Many cup failures are symptomless for long periods. Gross loss of bone stock can occur in the absence of pain and disability. This possibility emphasizes the importance of performing regular radiographic follow-up of specific patient populations at risk so that revision surgery is performed at an early stage of the mechanical fail-
679
ure. This study recognizes its limitations; it is essentially a retrospective survivorship analysis in which the endpoint is revision. It is not a clinical or radiographic review.
Conclusion Charnley’s vocation was to give a patient a fully functional, pain-free artificial hip that lasted their life span with no complications. Charnley’s own review at 25 years showed that 92% remain good and functional until death [5]. Three quarters of the elderly patients still active retained their original Charnley hip with a probable survivorship at 25 years of 85% [5]. With ⬎1 million Charnley prostheses implanted, the clinical results now are being published from various centers around the world. The probability of the Charnley LFA surviving 20 years has been reported by Schulte and Callaghan et al [6] and Kavanagh and Wallrichs et al [7] as 84% and by Neumann, Freund, and Sorenson et al [8] as 89%. These articles are supported by the results of this study, which, using basic data and Kaplan-Meier survivorship methods, in ⬎5,000 LFAs from 8 institutions around the world shown a 20-year probable survivorship of 83%. The Charnley metallic femoral component with a small head and the polymeric cup has been a dominant factor in total hip arthroplasty for 40 years. There is now a shortfall—wear and its associated aseptic loosening, especially in young patients. The orthopaedic community is looking at alternative methods with new bearing configurations to minimize wear. Such advances could allow the possibility of operating on younger patients and eliminate the need for revision surgery in young and old patients. The tradition of metal-on-polyethylene or ceramic-on-polyethylene configuration is fighting back with a new 0 wear polyethylene, however. The long-term success of new devices must be judged against the well-proven success of the Charnley LFA.
Acknowledgment This is the first time a survivorship analysis of the longevity of a total hip arthroplasty has been made on data collected from hospitals around the world. The author thanks all the participating orthopaedic surgeons; their friendship and help has been essential for this study. Thanks also to John Blachere, biostatistician, for outstanding help with the statistical analysis.
680 The Journal of Arthroplasty Vol. 17 No. 6 September 2002
References 1. Armitage P, Berry G: Statistical methods in medical research, 2nd ed. Blackwell Scientific, Oxford, 1987 2. Kaplan EL, Meier P: Nonparametric estimation from incomplete observation. J Am Stat Assoc 53:457, 1958 3. Murray DW, Carr AJ, Bulstrode C: Survival analysis of joint replacements. J Bone Joint Surg Br 75:697, 1993 4. Dorey F, Amstutz HC: The validity of survivorship and analysis in total joint arthroplasty. J Bone Joint Surg Am 71:544, 1989 5. Older J: Charnley’s by Charnley: a minimum follow
up of twenty years. Presidential Guest Speaker. The Hip Society, Orlando, FL, 1995 6. Schulte HR, Callaghan JJ, Kelley SS, Johnston RC: The outcome of Charnley total hip arthroplasty with cement after a minimum twenty year follow up. J Bone Joint Surg Am 75:961, 1993 7. Kavanagh BF, Wallrichs S, Dewitz M, et al: Charnley low friction arthroplasty of the hip, twenty year results with cement. J Arthroplasty 9:229, 1994 8. Neumann L, Freund KG, Sorenson KH: Long term results of Charnley total hip replacement. J Bone Joint Surg Br 76:245, 1994