JBJS | Factors Affecting Aseptic Failure of Fixation after Primary Charnley Total Hip Arthroplasty. Multivariate Survival Analysis*

The Journal of Bone & Joint Surgery, Volume 79, Issue 11

Articles | November 01, 1997

Factors Affecting Aseptic Failure of Fixation after Primary Charnley Total Hip Arthroplasty. Multivariate Survival Analysis*

SENEKI KOBAYASHI, M.D., PH.D.†; KUNIO TAKAOKA, M.D., PH.D.†; NAOTO SAITO, M.D., PH.D.†; KENJI HISA, M.D.†, MATSUMOTO, JAPAN

View Disclosures and Other Information

Investigation performed at the Department of Orthopaedic Surgery, Shinshu University School of Medicine, Matsumoto

The Journal of Bone & Joint Surgery. 1997; 79:1618-1627

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

text A A A

Abstract

Multivariate survival analysis with use of the Cox proportional-hazards model was applied to a consecutive series of 293 primary Charnley total hip arthroplasties performed on 246 patients. The purpose of the analysis was to identify risk factors for, and to quantitate their effects on, aseptic failure of fixation. The duration of follow-up ranged from one month to twenty-three years (average, thirteen years). The end point of survival was defined as radiographic evidence of failure of fixation or as a revision operation.Failure of fixation of the acetabular component was defined as complete demarcation or migration. Failure of the femoral component was defined as progression of at least one of five postoperative signs (subsidence, demarcation of the cement, separation of the component from the cement, fracture of the cement, and endosteal cavitation) or as the occurrence of at least two of these signs. Twenty-four specific items of data for each acetabular component and thirty specific items for each femoral component formed the sets of variables for the analysis.With use of radiographic evidence of failure as the end point, the sixteen-year rates of survival (with 95 per cent confidence interval) were 83.6 ± 5.6 per cent for the acetabular components and 90.9 ± 4.1 per cent for the femoral components. With use of revision as the end point, the sixteen-year rates of survival were 92.3 ± 4.0 per cent and 95.6 ± 3.2 per cent, respectively. The most important risk factor affecting radiographic loosening of the acetabular component was rapid wear of the polyethylene (0.2 millimeter or more annually), followed by the classification of the osteoarthrosis (as hypertrophic, normotrophic, or atrophic) according to the extent of osteophyte formation. The sockets in the hips that had hypertrophic osteoarthrosis survived longer than those in the other two groups. Survival of the acetabular component as determined on the basis of revision was affected only by rapid wear of the polyethylene. Survival of the femoral component, with either radiographic failure of fixation or revision as the end point, was affected by an unfavorable geometry of the medullary canal (a so-called stovepipe canal or a large canal).Patients who have rapid wear of the polyethylene, little osteophyte formation, or an unfavorable geometry of the canal should be followed carefully. These risk factors warrant additional evaluation.

Figures in this Article

Introduction

The principal long-term complication following total hip arthroplasty performed with or without cement is aseptic failure of fixation, which may also include osteolysis. Univariate survival analyses with the actuarial and Kaplan-Meier²³ methods have been used to identify risk factors affecting aseptic loosening of total hip prostheses^{9,10,25,32,35}. These methods account for patients who have not been followed during the period of observation, and they enable the comparison of groups that have had different durations of follow-up; however, they are of limited use when the investigator must examine many variables, as they cannot be used to determine major risk factors among several variables or to ascertain interfactorial influences. Because of its multifactorial nature, the problem of failure of fixation is not amenable to univariate survival analysis. Multivariate survival analysis with the Cox proportional-hazards model⁷ makes feasible the identification of risk factors with use of all available data, and interactions among the risk factors can also be taken into account. In addition, the impact of risk factors on survival can be assessed. In order to identify the important risk factors and to quantitate their effects on failure of fixation without infection, we applied this analytical model to a consecutive series of 293 Charnley total hip arthroplasties performed in 246 patients.

*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.

†Department of Orthopaedic Surgery, Shinshu University School of Medicine, Asahi 3-1-1, Matsumoto 390, Japan. Please address requests for reprints to Dr. Kobayashi.

*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.

†Department of Orthopaedic Surgery, Shinshu University School of Medicine, Asahi 3-1-1, Matsumoto 390, Japan. Please address requests for reprints to Dr. Kobayashi.

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

Fig. 1 Diagram showing the distribution of the femoral components according to the radiographic signs that were present at the latest follow-up examination. * = radiographic failure of fixation.

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

Fig. 2 Graph showing survival curves for the 293 acetabular components, with radiographic failure of fixation (A) and revision (B) as the end points. The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of eighty-eight hips (A) and ninety-eight hips (B); the cumulative rates of survival were 83.6 ± 5.6 per cent and 92.3 ± 4.0 per cent, respectively.

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

Fig. 3 Graph showing survival curves for the 293 femoral components, with radiographic failure of fixation (A) and revision (B) as the end points. The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of ninety-nine hips (A) and 105 hips (B); the cumulative rates of survival were 90.9 ± 4.1 per cent and 95.6 ± 3.2 per cent, respectively.

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

Fig. 4 Graph showing the radiographic survival of the acetabular components in relation to the so-called biological classification of the osteoarthrosis². Fifty-two hips had hypertrophic osteoarthrosis (H); 157, normotrophic osteoarthrosis (N); thirty, atrophic osteoarthrosis (A); and fifty-four, no osteoarthrosis (broken line). The 95 per cent confidence intervals are shown on only one side of each curve. (No confidence interval is depicted for the hips that did not have osteoarthrosis.) At sixteen years, the cohorts consisted of fifteen hips that had hypertrophic osteoarthrosis, fifty-four hips that had normotrophic osteoarthrosis, seven hips that had atrophic osteoarthrosis, and twelve hips that did not have osteoarthrosis; the cumulative rates of survival (average and 95 per cent confidence interval) were 95.8 ± 4.4, 84.2 ± 6.3, 71.3 ± 21.3, and 74.9 ± 19.1 per cent, respectively.

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

Figs. 5-A, 5-B, and 5-C: Anteroposterior radiographs of the right hip of a woman who had a total hip arthroplasty because of hypertrophic osteoarthrosis when she was sixty-two years old. Fig. 5-A: Preoperatively, there is marked formation of osteophytes.

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

Fig. 5-B: Radiograph made at the time of discharge, showing a non-flanged socket that was cemented after removal of eburnated bone from the acetabular roof.

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

Fig. 5-C: Twenty years postoperatively, the well fixed cemented socket has no demarcation. The average rate of wear of the polyethylene was 0.02 millimeter per year.

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

Figs. 6-A, 6-B, and 6-C: Anteroposterior radiographs of the right hip of a woman who had a total hip arthroplasty because of atrophic osteoarthrosis when she was sixty-two years old. Fig. 6-A: Preoperatively, the femoral head has decreased in size and appears elliptical. Osteophytes are scarce. The denser areas on both the femoral and the acetabular side may represent bone collapse.

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

Fig. 6-B: Radiograph made at the time of discharge, showing a flanged socket that was cemented after anchor holes were made in preserved eburnated bone.

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

Fig. 6-C: Thirteen years postoperatively, the socket has migrated despite very little wear of the polyethylene (average rate, 0.01 millimeter per year).

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

Fig. 7 Graph comparing the radiographic survival between the thirty-one sockets that had rapid wear of the polyethylene (R) and the 262 that did not have rapid wear (S). The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of twelve hips (R) and seventy-six hips (S); the cumulative rates of survival (average and 95 per cent confidence interval) were 56.8 ± 17.0 per cent and 87.0 ± 8.8 per cent, respectively.

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

Fig. 8 Graph comparing the radiographic survival of the femoral components according to the canal-flare index³⁴ (CFI). One hundred and thirteen hips had a canal-flare index of less than 3.0 (a so-called stovepipe canal), and 180 had an index of 3.0 or more. The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of twenty-seven hips that had an index of less than 3.0 and seventy-two that had an index of 3.0 or more; the cumulative rates of survival (average and 95 per cent confidence interval) were 81.6 ± 9.4 per cent and 96.1 ± 3.5 per cent, respectively.

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

Fig. 9 Graph comparing the radiographic survival of the femoral components according to the average width of the canal (CW). One hundred and twenty-two hips had an average canal width of more than seventeen millimeters, and 171 had an average width of seventeen millimeters or less. The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of twenty-four hips that had an average canal width of more than seventeen millimeters and seventy-five that had an average canal width of seventeen millimeters or less. At sixteen years, the cumulative rates of survival (average and 95 per cent confidence interval) were 78.7 ± 10.1 per cent and 97.4 ± 2.5 per cent, respectively.

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

Figs. 10-A, 10-B, and 10-C: Anteroposterior radiographs of the left hip of a woman who had a total hip arthroplasty because of normotrophic osteoarthrosis when she was sixty-two years old. Fig. 10-A: Preoperative radiograph showing the proximal aspect of the femur with a so-called stovepipe canal (canal-flare index³⁴=2.7).

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

Fig. 10-B: Radiograph made at the time of discharge. The arthroplasty was performed with a modern technique of cementing, including use of an intramedullary plug. The average width of the canal was 18.3 millimeters.

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

Fig. 10-C: Eleven years postoperatively, there is failure of fixation of the femoral component.

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

Fig. 11 Graph comparing the rates of radiographic survival, with revision as the end point, between the thirty-one acetabular components that had rapid wear of the polyethylene (R) and the 262 that did not have rapid wear (S). The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of seventeen hips (R) and eighty-one hips (S); the cumulative rates of survival (average and 95 per cent confidence interval) were 79.0 ± 15.2 per cent and 94.9 ± 3.7 per cent, respectively.

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

Fig. 12 Graph comparing the rates of radiographic survival, with revision as the end point, for the femoral components according to the canal-flare index³⁴ (CFI). One hundred and thirteen hips had an index of less than 3.0 (a so-called stovepipe canal), and 180 had an index of 3.0 or more. The 95 per cent confidence intervals are shown on only one side of each curve. At sixteen years, the cohorts consisted of twenty-nine hips that had a canal-flare index of less than 3.0 and seventy-six hips that had an index of 3.0 or more; the cumulative rates of survival (average and 95 per cent confidence interval) were 90.0 ± 5.7 per cent and 98.8 ± 1.7 per cent, respectively.

View Larger

TABLE I DISTRIBUTION OF THE RADIOGRAPHIC SIGNS AT THE LATEST FOLLOW-UP EVALUATION, ACCORDING TO WHETHER OR NOT THE FEMORAL COMPONENT HAD BEEN REVISED

*The values are given as the ratio of hips without the sign:hips with the sign.†The difference was significant (p < 0.05), according to the chi-square test.

Radiographic Sign	Not Revised* (N = 27)	Revised* (N = 8)	P Value
Subsidence = 2 mm	11:16	1:7	0.292
Demarcation	11:16	1:7	0.292
Separation from cement	22:5	3:5	0.048†
Fracture of cement	18:9	1:7	0.022†
Cavitation	19:8	2:6	0.059
Radiographic failure of fixation	14:13	0:8	0.027†

Add to My JBJS

TABLE I DISTRIBUTION OF THE RADIOGRAPHIC SIGNS AT THE LATEST FOLLOW-UP EVALUATION, ACCORDING TO WHETHER OR NOT THE FEMORAL COMPONENT HAD BEEN REVISED

*The values are given as the ratio of hips without the sign:hips with the sign.†The difference was significant (p < 0.05), according to the chi-square test.

Radiographic Sign	Not Revised* (N = 27)	Revised* (N = 8)	P Value
Subsidence = 2 mm	11:16	1:7	0.292
Demarcation	11:16	1:7	0.292
Separation from cement	22:5	3:5	0.048†
Fracture of cement	18:9	1:7	0.022†
Cavitation	19:8	2:6	0.059
Radiographic failure of fixation	14:13	0:8	0.027†

View Larger

TABLE II VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE ACETABULAR COMPONENT

*The values are given as the average and the standard deviation.

Variable
Patient-related variables
Age at op.* (yrs.)	59.4 ± 7.8
Gender (female:male) (no. of hips)	259:34
Etiology of hip disease (no. of hips)
Osteoarthrosis secondary to congenital dislocation	158
Primary osteoarthrosis	64
Rheumatoid arthritis	30
Fracture of femoral neck	14
Avascular necrosis	10
Miscellaneous	17
Charnley functional category⁴(no. of hips)
Category A: unilateral disease	90
Category B: bilateral disease	162
Category C: disease with other functional disabilities	41
Level of activity¹⁵(no. of hips)
Level I: sedentary	60
Level II: non-strenuous activity	195
Level III: moderately strenuous activity	37
Level IV: very active	1
Height* (cm)	147.8 ± 7.9
Weight* (kg)	50.3 ± 8.8
Operative technique (no. of hips)
Removal of eburnated bone from acetabular roof	105
Retention of eburnated bone of acetabular roof	188
Design of acetabular component (no. of hips)
Non-flanged	117
Flanged	176
Radiographic findings
Preoperative radiographs
Biological classification of osteoarthrosis²(no. of hips)
Hypertrophic	52
Normotrophic	157
Atrophic	30
No osteoarthrosis	54
Acetabular angle of Sharp³⁹* (degrees)	47.3 ± 7.7
Dislocation of hip¹¹(no. of hips)
None	75
Stage A: dysplasia	74
Stage B: subluxation	69
Stage C: dislocation	75
Protrusio acetabuli⁴⁰(no. of hips)
None	283
Grade I: mild	3
Grade II: moderate	7
Radiographs made at time of discharge (no. of hips)
Osseous containment²⁷
No	8
Yes	285
Reaming beyond ilioischial line
No	248
Yes	45
Position of socket²⁸(no. of hips)
Position I: in true acetabulum	271
Position II: at roof level of true acetabulum	19
Position III: proximal to roof of true acetabulum	3
Height of socket* (mm)	18.9 ± 5.7
Horizontal location of socket* (mm)	26.0 ± 4.3
Inclination of socket* (degrees)	45.7 ± 6.2
Deficiency of cement mantle⁸(no. of hips)
Zone 1
No	248
Yes	45
Zone 2
No	86
Yes	207
Zone 3
No	229
Yes	64
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of femoral component (no. of hips)
No	272
Yes	21

Add to My JBJS

TABLE II VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE ACETABULAR COMPONENT

*The values are given as the average and the standard deviation.

Variable
Patient-related variables
Age at op.* (yrs.)	59.4 ± 7.8
Gender (female:male) (no. of hips)	259:34
Etiology of hip disease (no. of hips)
Osteoarthrosis secondary to congenital dislocation	158
Primary osteoarthrosis	64
Rheumatoid arthritis	30
Fracture of femoral neck	14
Avascular necrosis	10
Miscellaneous	17
Charnley functional category⁴(no. of hips)
Category A: unilateral disease	90
Category B: bilateral disease	162
Category C: disease with other functional disabilities	41
Level of activity¹⁵(no. of hips)
Level I: sedentary	60
Level II: non-strenuous activity	195
Level III: moderately strenuous activity	37
Level IV: very active	1
Height* (cm)	147.8 ± 7.9
Weight* (kg)	50.3 ± 8.8
Operative technique (no. of hips)
Removal of eburnated bone from acetabular roof	105
Retention of eburnated bone of acetabular roof	188
Design of acetabular component (no. of hips)
Non-flanged	117
Flanged	176
Radiographic findings
Preoperative radiographs
Biological classification of osteoarthrosis²(no. of hips)
Hypertrophic	52
Normotrophic	157
Atrophic	30
No osteoarthrosis	54
Acetabular angle of Sharp³⁹* (degrees)	47.3 ± 7.7
Dislocation of hip¹¹(no. of hips)
None	75
Stage A: dysplasia	74
Stage B: subluxation	69
Stage C: dislocation	75
Protrusio acetabuli⁴⁰(no. of hips)
None	283
Grade I: mild	3
Grade II: moderate	7
Radiographs made at time of discharge (no. of hips)
Osseous containment²⁷
No	8
Yes	285
Reaming beyond ilioischial line
No	248
Yes	45
Position of socket²⁸(no. of hips)
Position I: in true acetabulum	271
Position II: at roof level of true acetabulum	19
Position III: proximal to roof of true acetabulum	3
Height of socket* (mm)	18.9 ± 5.7
Horizontal location of socket* (mm)	26.0 ± 4.3
Inclination of socket* (degrees)	45.7 ± 6.2
Deficiency of cement mantle⁸(no. of hips)
Zone 1
No	248
Yes	45
Zone 2
No	86
Yes	207
Zone 3
No	229
Yes	64
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of femoral component (no. of hips)
No	272
Yes	21

View Larger

TABLE III VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE FEMORAL COMPONENT

*The values are given as the average and the standard deviation.†Ratio of intracortical width at a point twenty millimeters proximal to the lesser trochanter to that at the isthmus of the canal.‡Percentage of cortical thickness in total diameter of the femoral shaft at the isthmus.§Extent beyond the tip of the stem to the level at which cement filled more than 80 per cent of the canal.

Variable
Patient-related variables	Same as in Table II
Operative technique (no. of hips)
Modern cementing technique
Yes	111
No	182
Design of femoral component (no. of hips)
Non-flanged	173
Flanged	120
Radiographic findings
Preoperative radiographs
Canal-flare index³⁴*†	3.2 ± 0.74
Femoral score¹*‡	47.0 ± 10.2
Radiographs made at time of discharge
Orientation of stem (no. of hips)
Non-varus	254
Varus	39
Length of medial aspect of femoral neck* (mm)	6.3 ± 4.8
Thickness of cement mantle* (mm)
At medial aspect of femoral neck	4.4 ± 2.7
At proximal-lateral corner of stem	1.9 ± 1.6
At proximal-medial corner of stem	4.0 ± 2.1
At distal-lateral corner of stem	2.7 ± 2.1
At distal-medial corner of stem	0.8 ± 1.1
Extent of cement beyond tip of stem*§ (mm)	25.1 ± 18.5
Deficiency of cement mantle¹⁴(no. of hips)
Zone 2
No	134
Yes	159
Zone 3
No	178
Yes	115
Zone 4
No	283
Yes	10
Zone 5
No	51
Yes	242
Zone 6
No	183
Yes	110
Zone 7
No	279
Yes	14
Width of canal* (mm)	16.8 ± 2.2
Ratio of implant to canal*	0.94 ± 0.05
Ratio of stem to canal*	0.72 ± 0.08
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of socket (no. of hips)
No	257
Yes	36

Add to My JBJS

TABLE III VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE FEMORAL COMPONENT

Variable
Patient-related variables	Same as in Table II
Operative technique (no. of hips)
Modern cementing technique
Yes	111
No	182
Design of femoral component (no. of hips)
Non-flanged	173
Flanged	120
Radiographic findings
Preoperative radiographs
Canal-flare index³⁴*†	3.2 ± 0.74
Femoral score¹*‡	47.0 ± 10.2
Radiographs made at time of discharge
Orientation of stem (no. of hips)
Non-varus	254
Varus	39
Length of medial aspect of femoral neck* (mm)	6.3 ± 4.8
Thickness of cement mantle* (mm)
At medial aspect of femoral neck	4.4 ± 2.7
At proximal-lateral corner of stem	1.9 ± 1.6
At proximal-medial corner of stem	4.0 ± 2.1
At distal-lateral corner of stem	2.7 ± 2.1
At distal-medial corner of stem	0.8 ± 1.1
Extent of cement beyond tip of stem*§ (mm)	25.1 ± 18.5
Deficiency of cement mantle¹⁴(no. of hips)
Zone 2
No	134
Yes	159
Zone 3
No	178
Yes	115
Zone 4
No	283
Yes	10
Zone 5
No	51
Yes	242
Zone 6
No	183
Yes	110
Zone 7
No	279
Yes	14
Width of canal* (mm)	16.8 ± 2.2
Ratio of implant to canal*	0.94 ± 0.05
Ratio of stem to canal*	0.72 ± 0.08
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of socket (no. of hips)
No	257
Yes	36

Materials and Methods

Patients

Two hundred and ninety-six primary total hip arthroplasties were performed with use of the original Charnley technique⁴, between October 1972 and June 1984, in 249 patients who had advanced hip disease. All of the arthroplasties were performed by a senior hip surgeon at a university hospital in Japan^24,25. The prostheses were implanted through a lateral approach, and the arthroplasties included a trochanteric osteotomy. All of the acetabular components were made from the same type of polyethylene, and all of the femoral components were made from stainless steel, by the same manufacturer (C. F. Thackray, Leeds, United Kingdom). The same viscous bone cement (CMW 1; CMW Laboratories of Dentsply, Exeter, United Kingdom) was used in all procedures. The femoral head had a diameter of 22.25 millimeters. Clinical and radiographic assessment was carried out annually. Anteroposterior radiographs of both hips were made with a standard method⁴.

Three hips in three patients were excluded from the study because of a deep infection (the focus of this study was aseptic loosening). The remaining 293 hips (246 patients) were followed for one month to twenty-three years (average, thirteen years) and formed the study group. Two hundred and thirty-two acetabular components (79 per cent) and 233 femoral components (80 per cent) were followed for more than ten years (average, fifteen years).

Definition of Radiographic Evidence of Failure of Fixation and Rates of Survival

Radiographic failure of fixation of the acetabular component was defined as complete demarcation or migration, as described by Hodgkinson et al.¹⁹. Radiographic failure of fixation of the femoral component was defined on the basis of five postoperative signs: subsidence, demarcation of the cement, separation of the component from the cement, fracture of the cement, and endosteal cavitation. Demarcation of the cement was defined as radiolucency surrounded by lines of increased density. So-called linear osteolysis and focal osteolysis were included in the categories of demarcation and endosteal cavitation, respectively. In a previous study of the long-term effects of these five signs, carried out at two centers by one of us (S. K.) and colleagues, failure of fixation of the femoral component was defined as progression of at least one sign or the occurrence of at least two signs, either non-progressive or progressive²⁶. Subsidence of the femoral component was measured radiographically as described by Loudon and Charnley³⁰, with a distance of two millimeters or more considered meaningful because of the limitations of the method. Subsidence of five millimeters or more was regarded as progressive. Progression of the other signs was determined on the basis of visual assessment of serial radiographs on a large viewing box. Progression of one sign or the presence of two non-progressive signs was enough to identify failure of femoral fixation. In order to verify the credibility of the radiographic criteria, the femoral components that were associated with at least one of the five signs were examined. The distribution of these components according to the signs was depicted (Fig. 1), and the prevalences of the signs were compared, with use of the chi-square test, between the components that had and those that had not been revised (Table I). The actuarial method was used to determine the cumulative rates of survival of the acetabular and the femoral components, with radiographic evidence of failure of fixation or revision designated as the end point.

Variables

Twenty-four specific items of data for each acetabular component and thirty items for each femoral component formed the sets of variables for the subsequent factorial analyses (Tables II and III). The patient-related, technical, and implant-related variables were determined from the clinical charts. All of the radiographic assessments were performed by an experienced observer (S. K.)^24-27. All of the radiographic measurements were corrected for magnification.

Although some of the variables are self-explanatory, others require additional description. With regard to the operative technique, the acetabular bone bed was prepared with use of one of two methods. Before 1979, the socket was cemented after removal of eburnated bone from the acetabular roof; however, since that time, eburnated bone has been preserved and multiple small anchor holes have been made for fixation. A non-flanged socket was used before 1979; however, since that time, a flanged socket has been used. On the femoral side, the medullary canal was minimally prepared for fixation of the femoral component with cement before 1981; however, since that time, so-called modern techniques of cementing, including brushing, use of an intramedullary plug, and use of a vent tube, have been employed. A femoral component without a flange was used before 1979, and a component with a dorsal flange has been used since that time.

On the basis of the preoperative radiographs, osteoarthrosis was classified according to the biological reaction of the joint to the disease—that is, in terms of the extent of osteophyte formation². In hips that had hypertrophic osteoarthrosis, large osteophytes were found both on the femoral head and on the margins and floor of the acetabulum. In hips that had normotrophic osteoarthrosis, moderate-sized osteophytes were seen. In hips that had atrophic osteoarthrosis, the size of the femoral head was decreased and osteophytes were scarce.

Several variables were assessed on the radiographs that were made at the time of discharge. The height and the horizontal location of the socket were determined on the basis of the perpendicular vertical distance and the horizontal distance, respectively, between the mid-point of both ends of the wire-marker and the most distal point on the pelvic teardrop²⁷. A finite element study revealed that a decrease in the thickness of the cement from three millimeters to one millimeter caused a distinct increase in the stresses in the socket, cement, and surrounding bone³. Deficiency of the cement mantle (a thickness of less than three millimeters) of the acetabular component was determined to be present or absent in the three zones described by DeLee and Charnley⁸.

With regard to the femoral component, the author of another finite element study recommended that the cement layer be at least two millimeters thick on the proximal part of the stem and that the layer on the distal part be thicker in order to minimize stresses in the cement and at the cement interfaces²¹. Deficiency of the cement mantle (a thickness of less than two millimeters [grade 1 according to the criteria proposed by Sarmiento and Gruen³⁷]) was determined to be present or absent in six of the seven zones described by Gruen et al.¹⁴. (Zone 1, in which the cement deliberately was not packed because of trochanteric reattachment, was excluded.) The width of the canal and the ratio of the implant (the cement and the stem) to the canal were measured at four levels of the stem, and the ratio of the stem to the canal was measured at three levels of the stem²⁶. The measurements at these different levels then were averaged for each hip. As the measurements at the different levels were statistically correlated with their averages (correlation coefficient, 0.68 to 0.93; p < 10^-5), the averages were considered to represent the measurements and thus were examined in the factorial analyses.

Linear wear of the polyethylene was measured with use of the method described by Charnley and Halley⁵, and the radiographs made at the time of discharge were compared with those made at the latest follow-up examination. If there was radiographic loosening of the implant, the radiograph made just before that loosening was seen was used instead of the latest follow-up radiograph. An average annual rate of wear of 0.2 millimeter or more was regarded as rapid wear²⁷.

Factorial Analyses

Factors affecting the survival of each component of the hip prosthesis were analyzed with use of the Cox proportional-hazards model. The end point of survival was defined as radiographic evidence of failure of fixation or as revision done because of aseptic loosening. In the first stage of the factorial analysis, the Cox model was applied to each of the variables in order to screen them separately for a possible risk effect on the survival of the component. In the second stage, the selected possible risk factors were examined together with use of the Cox model to determine their contribution to survival. Testing of interactions between variables for significance was part of the model-fitting process. On the basis of the final model, a p value was calculated for each parameter. A lower p value indicated a higher risk of failure of fixation.

In order to gain a better understanding of the effects of the identified risk factors, cumulative survival curves were plotted and the differences between them were tested with use of the actuarial method. Survival curves were not plotted for comparisons of subgroups beyond the point at which the number of hips was too small for valid statistical estimation¹⁰. All statistical analyses were performed with use of Beccel Mark II computer software (Beccel, Tokyo, Japan). A p value of less than 0.05 was considered significant.

Results

Thirty-five hips had at least one of the radiographic signs that were used to judge failure of the femoral fixation (Fig. 1). Fourteen of these hips had only one sign and it was not progressive, so the femoral component was not considered to have failed. In six of the fourteen hips, the femoral component had subsided less than five millimeters. The subsidence of four of them was noted at 1.5 to ten years (average, 5.8 years) and did not progress for 6.1 to 20.6 years (average, 13.4 years). The subsidence of the remaining two components was noted at the latest follow-up evaluation, at 5.7 and 20.1 years. Five hips had demarcation at the distal tip of the cement, but it remained unchanged for 9.1 to 16.7 years. In three hips, there was a transverse fracture of the cement on the lateral side of the distal tip of the stem, but the displacement of the cement fragments remained less than one millimeter for 10.0 to 14.5 years.

The prevalences of the radiographic signs were compared between the twenty-seven femoral components that were not revised despite the presence of at least one sign and the eight components that were revised (Table I). Separation of the component from the cement, fracture of the cement, and radiographic failure of fixation were found to be significantly associated with revision (p < 0.05); however, three femoral components that did not separate from the cement and one that was not associated with a fracture of the cement were revised. No femoral component that did not have radiographic failure of fixation was revised.

With use of radiographic failure of fixation and revision as the end points, the cumulative rates of survival (with 95 per cent confidence interval) of the acetabular components at sixteen years were 83.6 ± 5.6 per cent and 92.3 ± 4.0 per cent, respectively (Fig. 2). With use of the same end points, the cumulative rates of survival of the femoral components at sixteen years were 90.9 ± 4.1 per cent and 95.6 ± 3.2 per cent, respectively (Fig. 3).

When each of the twenty-four variables related to the acetabular components was screened separately with use of the Cox model, only two were found to affect radiographic loosening: the so-called biological classification of the osteoarthrosis (p = 0.0145) and rapid wear of the polyethylene (p = 6.13 x 10^-7); with the numbers available, no association could be detected between these two variables. These two factors were analyzed together, and their statistical significance was confirmed with use of the Cox model (p = 0.0439 for classification of the osteoarthrosis and p = 2.21 x 10^-6 for rapid wear). Eleven years or more after the operation, the acetabular components in the hips that had hypertrophic osteoarthrosis had a higher cumulative rate of survival than those in the hips that had atrophic osteoarthrosis; twelve years or more postoperatively, the acetabular components in the hips that had hypertrophic osteoarthrosis had a higher cumulative rate of survival than either those in the hips that had atrophic osteoarthrosis or those in the hips that had normotrophic osteoarthrosis (Figs. 4, 5-A, 5-B, 5-C, 6-A, 6-B, and 6-C). Compared with the other sockets, the thirty-one sockets that had rapid wear of the polyethylene had a lower cumulative rate of survival four years or more after the operation (Fig. 7). With use of the t test or the chi-square test, the prevalences of twenty-three variables (all except rapid wear of the polyethylene) were compared between the sockets that had rapid wear and those that did not; with the numbers available, none of these variables was found to be associated with rapid wear.

With regard to the radiographic survival of the femoral component, application of the Cox model to each of the thirty variables led to the identification of seven possible risk factors: gender (p = 0.0115), canal-flare index³⁴ (p = 0.00024), femoral score¹ (p = 0.0068), extent of the cement beyond the tip of the stem (p = 0.0036), average width of the canal (p = 8.43 = 10^-5), average implant-to-canal ratio (p = 0.0237), and average stem-to-canal ratio (p = 0.0082). When these factors were analyzed together with use of the Cox model, only two—a low canal-flare index (p = 0.020) and a large width of the canal (p = 0.023)—were found to be risk factors for failure of fixation. There was a correlation between these two factors (correlation coefficient, -0.541; p < 10^-5). A femoral canal with a canal-flare index of less than 3.0 has been called a stovepipe canal³⁴. Six years or more after the operation, the femoral components in the stovepipe canals had a lower rate of survival than the other femoral components (Fig. 8). Eleven years or more after the operation, the femoral components in the canals that had an average width of more than 17.0 millimeters had a lower rate of survival compared with the other components (Figs. 9, 10-A, 10-B, and 10-C).

With use of revision as the end point, application of the Cox model to each of the twenty-four variables related to the acetabular components led to the identification of only rapid wear of the polyethylene as a risk factor for survival (p = 0.0228). However, the actuarial method did not show a significant difference in the rate of survival between the sockets that had rapid wear and those that did not (Fig. 11). With regard to survival of the femoral component, with revision as the end point, application of the Cox model to each of the thirty variables led to the identification of only two risk factors: the canal-flare index³⁴ (p = 0.0289) and the average width of the canal (p = 0.0116). Because there was a strong correlation between these two variables (correlation coefficient, -0.541; p < 10^-5), they were not tested together with use of the Cox model. Ten years or more after the operation, the femoral components in the stovepipe canals (Fig. 12) and those in the canals that had an average width of more than 17.0 millimeters had a lower rate of survival compared with the other femoral components.

Discussion

In a study of patients who were managed at two centers, one of us (S. K.) and colleagues defined failure of fixation of the femoral component as progression of at least one of the five signs described earlier or as the presence of at least two signs²⁶. In the present study, an isolated occurrence of one of the five signs without progression was not associated with failure, and the criteria were considered to be reasonable.

Cox proposed a multivariate proportional-hazards model for use in survivorship studies⁷. This model has characteristics of both multiple-regression analysis and the life-table method, and it takes into account varying durations of follow-up and possible interfactorial influences. In a retrospective, non-randomized study of total hip prostheses, survival is influenced by the net effect of many variables. When the survival curves of different categorical groups are compared with use of the Cox model, the arthroplasties are assigned so that their other characteristics are equally balanced between groups. Therefore, the results reflect real differences in the effectiveness of the variable without it being confounded with other factors. The Cox model also can be used to quantify the effects of factors on the rates of survival.

The Cox proportional-hazards model has been used in previous studies to investigate risk factors for aseptic loosening after total hip arthroplasty^{17,18,20,22,38}. However, none of those analyses included radiographic parameters; only patient-related, technical, and implant-related variables were evaluated. In the present study, in addition to these variables, fifteen radiographic parameters for the acetabular component and twenty-one parameters for the femoral component were considered. All of the risk factors that were identified—that is, the classification of the osteoarthrosis and rapid wear of the polyethylene (risk factors for loosening of the acetabular component) and a stovepipe canal and a large medullary canal (risk factors for loosening of the femoral component)—were radiographic. All of these risk factors except for rapid wear of the polyethylene can be evaluated on a preoperative radiograph. Rapid wear can be detected within five years after the operation. Therefore, patients for whom at least one risk factor is identified should be carefully followed.

Bombelli described the so-called biological classification of osteoarthrosis in terms of the radiographic extent of osteophyte formation². In a selected series of sixty-three hips that were followed for an average of 7.5 years after primary total hip arthroplasty, and with a technically flawed arthroplasty as a criterion for exclusion, Saito et al. found a relationship between the biological classification of osteoarthrosis and the prevalence of radiographic loosening of the socket³⁶. The present study, which involved a larger number of sockets and a longer duration of follow-up of a consecutive series, clearly demonstrates the superior radiographic durability of the acetabular component in hips that have hypertrophic osteoarthrosis.

Rapid wear of the polyethylene was the most important risk factor limiting the longevity of the socket in the present series, as it was in previous studies^12,27,41. It has been suggested that a young age^12,13,16,27, male gender^13,27,31, a greater body weight²⁹, and the size of the femoral head^4,6,29 are associated with the wear characteristics of polyethylene. The Japanese patients in the present series were older and weighed less, on the average, and female gender was more predominant than in American patients in a previous study by one of us (S. K.) and colleagues²⁷. Probably because of the relatively small numbers of young, male, and heavy patients, a significant relationship between those variables and rapid wear of the polyethylene was not detected in the present series.

In a previous study of Charnley femoral components by one of us (S. K.) and colleagues, seven risk factors were noted to affect radiographic failure of fixation²⁶. However, only univariate analyses were used in that study, and it could not be determined which of the risk factors were truly important. This problem was largely resolved in the current study by use of the Cox model. In the present study, seven variables (six of which had been identified in the previous study) were selected in the initial screening stage; however, only two were validated as risk factors for radiographic failure of fixation with use of the multivariate analysis in the second stage. Those two variables also affected the rates of survival with revision as the end point.

A so-called stovepipe canal that has a low canal-flare index³⁴ and a large medullary canal seem to be mechanically unfavorable with regard to the bone exerting resistance to stresses transmitted through the implant. Modern techniques of cementing appear to improve the quality of fixation, as poor packing of the cement, especially in a large medullary canal, is avoided; excellent long-term radiographic results have been reported³³. The present series did not demonstrate improved survival with use of newer techniques of cementing (which included brushing, use of an intramedullary plug, and use of a vent tube before insertion of the cement). Although the newer techniques were not used in 182 (62 per cent) of the 293 hips, the small femoral canal (average width, 16.8 millimeters) may have contributed to the satisfactory filling with cement. However, even when these newer techniques are used, a canal with an unfavorable geometry may still constitute a risk for aseptic loosening, as shown by our findings. Use of the newer design of the femoral component with a dorsal flange did not improve survival either.

In summary, factors affecting aseptic failure of fixation after primary Charnley total hip arthroplasty were identified in the current study. Rapid wear of the polyethylene was the most important risk factor affecting loosening of the socket, followed by a small extent of osteophyte formation. Survival of the femoral component was affected by a medullary canal with an unfavorable geometry (a so-called stovepipe canal or a large canal). To ensure longer durability of the implant, these problems should be evaluated further and attempts should be made to solve them.

Note: The authors thank Kazuo Terayama, M.D., Ph.D., for his contribution of cases.

References

Barnett, E., and Nordin, B. E. C.: The radiological diagnosis of osteoporosis: a new approach. Clin. Radiol.,11: 166-174, 1960.11166 1960 [PubMed]

Bombelli, R.: Osteoarthritis of the Hip: Classification and Pathogenesis: The Role of Osteotomy as a Consequent Therapy. Ed. 2, pp. 98-108. New York, Springer, 1983.

Carter, D. R.; Vasu, R.; and Harris, W. H.: Stress distributions in the acetabular region—II. Effects of cement thickness and metal backing of the total hip acetabular component. J. Biomech.,15: 165-170, 1982.15165 1982 [PubMed]

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]

Charnley, J.; Kamangar, A.; and Longfield, M. D.: The optimum size of prosthetic heads in relation to the wear of plastic sockets in total replacement of the hip. Med. and Biol. Eng.,7: 31-39, 1969.731 1969

Cox, D. R.: Regression models and life-tables. J. Roy. Statist. Soc., Series B,34: 187-220, 1972.34187 1972

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

Dobbs, H. S.: Survivorship of total hip replacements. J. Bone and Joint Surg.,62-B(2): 168-173, 1980.62-B(2)168 1980

Dorey, F., and Amstutz, H. C.: Survivorship analysis in the evaluation of joint replacement. J. Arthroplasty,1: 63-69, 1986.163 1986 [PubMed]

Eftekhar, N. S.: Principles of Total Hip Arthroplasty. St. Louis, C. V. Mosby, 1978.

García-Cimbrelo, E., and Munuera, L.: Early and late loosening of the acetabular cup after low-friction arthroplasty. J. Bone and Joint Surg.,74-A: 1119-1129, Sept. 1992.74-A1119 1992

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]

Gustilo, R. B., and Burnham, W. H.: Long-term results of total hip arthroplasty in young patients. In The Hip. Proceedings of the Tenth Open Scientific Meeting of The Hip Society, pp. 27-33. St. Louis, C. V. Mosby, 1982.

Hartofilakidis, G.; Stamos, K.; and Ioannidis, T. T.: Fifteen years' experience with Charnley low-friction arthroplasty. Clin. Orthop.,246: 48-56, 1989.24648 1989 [PubMed]

Havelin, L. I.; Espehaug, B.; Vollset, S. E.; and EngesæOter, L. B.: Early failures among 14,009 cemented and 1,326 uncemented prostheses for primary coxarthrosis. The Norwegian Arthroplasty Register, 1987-1992. Acta Orthop. Scandinavica,65: 1-6, 1994.651 1994

Havelin, L. I.; Espehaug, B.; Vollset, S. E.; and Engesæter, L. B.: The effect of the type of cement on early revision of Charnley total hip prostheses. A review of eight thousand five hundred and seventy-nine primary arthroplasties from the Norwegian Arthroplasty Register. J. Bone and Joint Surg.,77-A: 1543-1550, Oct. 1995.77-A1543 1995

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

Hozack, W. J.; Rothman, R. H.; Booth, R. E., Jr.; Balderston, R. A.; Cohn, J. C.; and Pickens, G. T.: Survivorship analysis of 1,041 Charnley total hip arthroplasties. J. Arthroplasty,5: 41-47, 1990.541 1990 [PubMed]

Huiskes, R.: Some fundamental aspects of human joint replacement. Analyses of stresses and heat conduction in bone-prosthesis structures. Acta Orthop. Scandinavica, Supplementum 185, 1980.

Johnsson, R.; Franzén, H.; and Nilsson, L. T.: Combined survivorship and multivariate analyses of revisions in 799 hip prostheses. A 10- to 20-year review of mechanical loosening. J. Bone and Joint Surg.,76-B(3): 439-443, 1994.76-B(3)439 1994

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

Kobayashi, S., and Terayama, K.: Factors influencing survivorship of the femoral component after primary low-friction hip arthroplasty. J. Arthroplasty,7 Supplement): 327-338, 1992.7 Supplement)327 1992

Kobayashi, S., and Terayama, K.: Factors influencing survival of the socket after primary low-friction arthroplasty of the hip. Arch. Orthop. and Trauma Surg.,112: 56-60, 1993.11256 1993

Kobayashi, S.; Eftekhar, N. S.; and Terayama, K.: Predisposing factors in fixation failure of femoral prostheses following primary Charnley low friction arthroplasty. A 10- to 20-year followup study. Clin. Orthop.,306: 73-83, 1994.30673 1994 [PubMed]

Kobayashi, S.; Eftekhar, N. S.; Terayama, K.; and Iorio, R.: Risk factors affecting radiological failure of the socket in primary Charnley low friction arthroplasty. A 10- to 20-year followup study. Clin. Orthop.,306: 84-96, 1994.30684 1994 [PubMed]

Linde, F., and Jensen, J.: Socket loosening in arthroplasty for congenital dislocation of the hip. Acta Orthop. Scandinavica,59: 254-257, 1988.59254 1988

Livermore, J.; Ilstrup, D.; and Morrey, B.: Effect of femoral head size on wear of the polyethylene acetabular component. J. Bone and Joint Surg.,72-A: 518-528, April 1990.72-A518 1990

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

McCoy, T. H.; Salvati, E. A.; Ranawat, C. S.; and Wilson, P. D., Jr.: A fifteen-year follow-up study of one hundred Charnley low-friction arthroplasties. Orthop. Clin. North America,19: 467-476, 1988.19467 1988

Malchau, H.; Herberts, P.; and Ahnfelt, L.: Prognosis of total hip replacement in Sweden. Follow-up of 92,675 operations performed 1978-1990. Acta Orthop. Scandinavica,64: 497-506, 1993.64497 1993

Mulroy, R. D., Jr., and Harris, W. H.: The effect of improved cementing techniques on component loosening in total hip replacement. An 11-year radiographic review. J. Bone and Joint Surg.,72-B(5): 757-760, 1990.72-B(5)757 1990

Noble, P. C.; Alexander, J. W.; Lindahl, L. J.; Yew, D. T.; Granberry, W. M.; and Tullos, H. S.: The anatomic basis of femoral component design. Clin. Orthop.,235: 148-165, 1988.235148 1988 [PubMed]

Ritter, M. A.: The cemented acetabular component of a total hip replacement. All polyethylene versus metal backing. Clin. Orthop.,311: 69-75, 1995.31169 1995 [PubMed]

Saito, M.; Saito, S.; Ohzono, K.; and Ono, K.: The osteoblastic response to osteoarthritis of the hip. Its influence on the long-term results of arthroplasty. J. Bone and Joint Surg.,69-B(5): 746-751, 1987.69-B(5)746 1987

Sarmiento, A., and Gruen, T. A.: Radiographic analysis of a low-modulus titanium-alloy femoral total hip component. Two to six-year follow-up. J. Bone and Joint Surg.,67-A: 48-56, Jan. 1985.67-A48 1985

Schurman, D. J.; Bloch, D. A.; Segal, M. R.; and Tanner, C. M.: Conventional cemented total hip arthroplasty. Assessment of clinical factors associated with revision for mechanical failure. Clin. Orthop.,240: 173-180, 1989.240173 1989 [PubMed]

Sharp, I. K.: Acetabular dysplasia. The acetabular angle. J. Bone and Joint Surg.,43-B(2): 268-272, 1961.43-B(2)268 1961

Sotelo-Garza, A., and Charnley, J.: The results of Charnley arthroplasty of the hip performed for protrusio acetabuli. Clin. Orthop.,132: 12-18, 1978.13212 1978 [PubMed]

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

Topics

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

Fig. 1 Diagram showing the distribution of the femoral components according to the radiographic signs that were present at the latest follow-up examination. * = radiographic failure of fixation.

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

Fig. 5-B: Radiograph made at the time of discharge, showing a non-flanged socket that was cemented after removal of eburnated bone from the acetabular roof.

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

Fig. 5-C: Twenty years postoperatively, the well fixed cemented socket has no demarcation. The average rate of wear of the polyethylene was 0.02 millimeter per year.

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

Fig. 6-B: Radiograph made at the time of discharge, showing a flanged socket that was cemented after anchor holes were made in preserved eburnated bone.

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

Fig. 6-C: Thirteen years postoperatively, the socket has migrated despite very little wear of the polyethylene (average rate, 0.01 millimeter per year).

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

Fig. 10-C: Eleven years postoperatively, there is failure of fixation of the femoral component.

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

View Larger

TABLE I DISTRIBUTION OF THE RADIOGRAPHIC SIGNS AT THE LATEST FOLLOW-UP EVALUATION, ACCORDING TO WHETHER OR NOT THE FEMORAL COMPONENT HAD BEEN REVISED

*The values are given as the ratio of hips without the sign:hips with the sign.†The difference was significant (p < 0.05), according to the chi-square test.

Radiographic Sign	Not Revised* (N = 27)	Revised* (N = 8)	P Value
Subsidence = 2 mm	11:16	1:7	0.292
Demarcation	11:16	1:7	0.292
Separation from cement	22:5	3:5	0.048†
Fracture of cement	18:9	1:7	0.022†
Cavitation	19:8	2:6	0.059
Radiographic failure of fixation	14:13	0:8	0.027†

Add to My JBJS

TABLE I DISTRIBUTION OF THE RADIOGRAPHIC SIGNS AT THE LATEST FOLLOW-UP EVALUATION, ACCORDING TO WHETHER OR NOT THE FEMORAL COMPONENT HAD BEEN REVISED

*The values are given as the ratio of hips without the sign:hips with the sign.†The difference was significant (p < 0.05), according to the chi-square test.

Radiographic Sign	Not Revised* (N = 27)	Revised* (N = 8)	P Value
Subsidence = 2 mm	11:16	1:7	0.292
Demarcation	11:16	1:7	0.292
Separation from cement	22:5	3:5	0.048†
Fracture of cement	18:9	1:7	0.022†
Cavitation	19:8	2:6	0.059
Radiographic failure of fixation	14:13	0:8	0.027†

View Larger

TABLE II VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE ACETABULAR COMPONENT

*The values are given as the average and the standard deviation.

Variable
Patient-related variables
Age at op.* (yrs.)	59.4 ± 7.8
Gender (female:male) (no. of hips)	259:34
Etiology of hip disease (no. of hips)
Osteoarthrosis secondary to congenital dislocation	158
Primary osteoarthrosis	64
Rheumatoid arthritis	30
Fracture of femoral neck	14
Avascular necrosis	10
Miscellaneous	17
Charnley functional category⁴(no. of hips)
Category A: unilateral disease	90
Category B: bilateral disease	162
Category C: disease with other functional disabilities	41
Level of activity¹⁵(no. of hips)
Level I: sedentary	60
Level II: non-strenuous activity	195
Level III: moderately strenuous activity	37
Level IV: very active	1
Height* (cm)	147.8 ± 7.9
Weight* (kg)	50.3 ± 8.8
Operative technique (no. of hips)
Removal of eburnated bone from acetabular roof	105
Retention of eburnated bone of acetabular roof	188
Design of acetabular component (no. of hips)
Non-flanged	117
Flanged	176
Radiographic findings
Preoperative radiographs
Biological classification of osteoarthrosis²(no. of hips)
Hypertrophic	52
Normotrophic	157
Atrophic	30
No osteoarthrosis	54
Acetabular angle of Sharp³⁹* (degrees)	47.3 ± 7.7
Dislocation of hip¹¹(no. of hips)
None	75
Stage A: dysplasia	74
Stage B: subluxation	69
Stage C: dislocation	75
Protrusio acetabuli⁴⁰(no. of hips)
None	283
Grade I: mild	3
Grade II: moderate	7
Radiographs made at time of discharge (no. of hips)
Osseous containment²⁷
No	8
Yes	285
Reaming beyond ilioischial line
No	248
Yes	45
Position of socket²⁸(no. of hips)
Position I: in true acetabulum	271
Position II: at roof level of true acetabulum	19
Position III: proximal to roof of true acetabulum	3
Height of socket* (mm)	18.9 ± 5.7
Horizontal location of socket* (mm)	26.0 ± 4.3
Inclination of socket* (degrees)	45.7 ± 6.2
Deficiency of cement mantle⁸(no. of hips)
Zone 1
No	248
Yes	45
Zone 2
No	86
Yes	207
Zone 3
No	229
Yes	64
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of femoral component (no. of hips)
No	272
Yes	21

Add to My JBJS

TABLE II VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE ACETABULAR COMPONENT

*The values are given as the average and the standard deviation.

Variable
Patient-related variables
Age at op.* (yrs.)	59.4 ± 7.8
Gender (female:male) (no. of hips)	259:34
Etiology of hip disease (no. of hips)
Osteoarthrosis secondary to congenital dislocation	158
Primary osteoarthrosis	64
Rheumatoid arthritis	30
Fracture of femoral neck	14
Avascular necrosis	10
Miscellaneous	17
Charnley functional category⁴(no. of hips)
Category A: unilateral disease	90
Category B: bilateral disease	162
Category C: disease with other functional disabilities	41
Level of activity¹⁵(no. of hips)
Level I: sedentary	60
Level II: non-strenuous activity	195
Level III: moderately strenuous activity	37
Level IV: very active	1
Height* (cm)	147.8 ± 7.9
Weight* (kg)	50.3 ± 8.8
Operative technique (no. of hips)
Removal of eburnated bone from acetabular roof	105
Retention of eburnated bone of acetabular roof	188
Design of acetabular component (no. of hips)
Non-flanged	117
Flanged	176
Radiographic findings
Preoperative radiographs
Biological classification of osteoarthrosis²(no. of hips)
Hypertrophic	52
Normotrophic	157
Atrophic	30
No osteoarthrosis	54
Acetabular angle of Sharp³⁹* (degrees)	47.3 ± 7.7
Dislocation of hip¹¹(no. of hips)
None	75
Stage A: dysplasia	74
Stage B: subluxation	69
Stage C: dislocation	75
Protrusio acetabuli⁴⁰(no. of hips)
None	283
Grade I: mild	3
Grade II: moderate	7
Radiographs made at time of discharge (no. of hips)
Osseous containment²⁷
No	8
Yes	285
Reaming beyond ilioischial line
No	248
Yes	45
Position of socket²⁸(no. of hips)
Position I: in true acetabulum	271
Position II: at roof level of true acetabulum	19
Position III: proximal to roof of true acetabulum	3
Height of socket* (mm)	18.9 ± 5.7
Horizontal location of socket* (mm)	26.0 ± 4.3
Inclination of socket* (degrees)	45.7 ± 6.2
Deficiency of cement mantle⁸(no. of hips)
Zone 1
No	248
Yes	45
Zone 2
No	86
Yes	207
Zone 3
No	229
Yes	64
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of femoral component (no. of hips)
No	272
Yes	21

View Larger

TABLE III VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE FEMORAL COMPONENT

Variable
Patient-related variables	Same as in Table II
Operative technique (no. of hips)
Modern cementing technique
Yes	111
No	182
Design of femoral component (no. of hips)
Non-flanged	173
Flanged	120
Radiographic findings
Preoperative radiographs
Canal-flare index³⁴*†	3.2 ± 0.74
Femoral score¹*‡	47.0 ± 10.2
Radiographs made at time of discharge
Orientation of stem (no. of hips)
Non-varus	254
Varus	39
Length of medial aspect of femoral neck* (mm)	6.3 ± 4.8
Thickness of cement mantle* (mm)
At medial aspect of femoral neck	4.4 ± 2.7
At proximal-lateral corner of stem	1.9 ± 1.6
At proximal-medial corner of stem	4.0 ± 2.1
At distal-lateral corner of stem	2.7 ± 2.1
At distal-medial corner of stem	0.8 ± 1.1
Extent of cement beyond tip of stem*§ (mm)	25.1 ± 18.5
Deficiency of cement mantle¹⁴(no. of hips)
Zone 2
No	134
Yes	159
Zone 3
No	178
Yes	115
Zone 4
No	283
Yes	10
Zone 5
No	51
Yes	242
Zone 6
No	183
Yes	110
Zone 7
No	279
Yes	14
Width of canal* (mm)	16.8 ± 2.2
Ratio of implant to canal*	0.94 ± 0.05
Ratio of stem to canal*	0.72 ± 0.08
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of socket (no. of hips)
No	257
Yes	36

Add to My JBJS

TABLE III VARIABLES THAT WERE SCREENED SEPARATELY, WITH USE OF THE COX PROPORTIONAL-HAZARDS MODEL, FOR POSSIBLE RISK EFFECTS WITH REGARD TO THE SURVIVAL OF THE FEMORAL COMPONENT

Variable
Patient-related variables	Same as in Table II
Operative technique (no. of hips)
Modern cementing technique
Yes	111
No	182
Design of femoral component (no. of hips)
Non-flanged	173
Flanged	120
Radiographic findings
Preoperative radiographs
Canal-flare index³⁴*†	3.2 ± 0.74
Femoral score¹*‡	47.0 ± 10.2
Radiographs made at time of discharge
Orientation of stem (no. of hips)
Non-varus	254
Varus	39
Length of medial aspect of femoral neck* (mm)	6.3 ± 4.8
Thickness of cement mantle* (mm)
At medial aspect of femoral neck	4.4 ± 2.7
At proximal-lateral corner of stem	1.9 ± 1.6
At proximal-medial corner of stem	4.0 ± 2.1
At distal-lateral corner of stem	2.7 ± 2.1
At distal-medial corner of stem	0.8 ± 1.1
Extent of cement beyond tip of stem*§ (mm)	25.1 ± 18.5
Deficiency of cement mantle¹⁴(no. of hips)
Zone 2
No	134
Yes	159
Zone 3
No	178
Yes	115
Zone 4
No	283
Yes	10
Zone 5
No	51
Yes	242
Zone 6
No	183
Yes	110
Zone 7
No	279
Yes	14
Width of canal* (mm)	16.8 ± 2.2
Ratio of implant to canal*	0.94 ± 0.05
Ratio of stem to canal*	0.72 ± 0.08
Rapid wear of polyethylene (no. of hips)
No	262
Yes	31
Radiographic loosening of socket (no. of hips)
No	257
Yes	36