- Email: [email protected]
The sphericity of the bearing surface in total hip arthroplasty
The Journal of Arthroplasty Vol. 16 No. 8 2001
The Sphericity of the Bearing Surface in Total Hip Arthroplasty Hiroshi Ito, MD,* Akio Minami, MD,* Takeo Matsuno, MD,† Hiromasa Tanino, MD,† Toshio Yuhta, PhD,‡ and Ikuya Nishimura, PhD‡
Abstract: This study evaluated the sphericity of bearing surfaces in total hip arthroplasty. The out-of-roundness of metal femoral heads, the inner surface of polyethylene liners, and commercially available ball bearings was measured. The hip prostheses were obtained directly from the manufacturers. The sphericity of the bearing surfaces was significantly inferior to that of the ball bearings. The sphericity of the femoral head on the sagittal plane was inferior to that on the transverse plane. Several significant differences were found among different manufacturers. The sphericity of the femoral head on the sagittal plane and that of polyethylene significantly improved in 1999 and 2000 compared with those in 1995. Further improvement is desirable, however, because good sphericity is expected to prolong the functional performance of the prosthesis after total hip arthroplasty. Key words: total hip arthroplasty, bearing surface, sphericity, polyethylene wear.
The current limitations of total hip arthroplasty (THA) are related to the wear of the components [1]. Wear of ultra-high molecular weight polyethylene (UHMWPE) liner bearing surfaces is the dominant issue. Improvement in the wear properties of the bearing surface in THA should reduce the generation of wear particles and improve longevity. Several studies have shown that a UHMWPE liner of an artificial hip joint socket wears at a linear wear rate of approximately 0.1 to 0.2 mm/y, resulting in about 1 to 2 mm of linear wear during a 10-year period [1–3]. If the UHMWPE liner is thin, the progressive wear can re-
sult in friction between the metal back and femoral head, causing metallosis and further increasing the particle load. Wear particles stimulate a foreign body reaction, resulting in osteolysis and loosening [4,5]. The morphology of the bearing surface in THA is an important factor in UHMWPE wear. It is well known in the mechanical engineering fields that the greater sphericity a ball bearing has, the lower and more uniform the frictional force and the higher rolling quality. This situation in turn leads to an improved functional performance [6]. The outof-roundness of the bearing ring of the best ball bearing currently available is 0.001 to 0.005 m [7]. The purpose of this study was to analyze the sphericity of prosthetic femoral heads and the inner surface of UHMWPE liners to determine whether further improvements of the surface contour are required. All the prosthetic metal femoral heads and the UHMWPE liners evaluated in this study were obtained directly from the manufacturers.
From the *Department of Orthopaedic Surgery, Hokkaido University, School of Medicine, Sapporo; †Department of Orthopaedic Surgery, Asahikawa Medical College, Asahikawa; and ‡Division of Biomedical Systems Engineering, Hokkaido University, School of Engineering, Sapporo, Japan. Submitted December 29, 2000; accepted July 5, 2001. No benefits or funds were received in support of this study. Reprint requests: Hiroshi Ito, MD, Department of Orthopaedic Surgery, Hokkaido University, School of Medicine, Kita-ku Kita-15 Nishi-7, Sapporo 060-8638, Japan. E-mail: itobiro@med. hokudai.ac.jp Copyright © 2001 by Churchill Livingstone威 0883-5403/01/1608-0012$35.00/0 doi:10.1054/arth.2001.27673
Materials and Methods Out-of-roundness was assessed as an indicator representing sphericity by obtaining circular traces
1024
Bearing Surfaces in THA • Ito et al.
of the implants [8 –10]. To obtain circular traces, a cylindrical stylus was used, as shown in Fig. 1. The traces were taken on the great circles of the sphere, and the out-of-roundness was obtained from the circular traces. Out-of-roundness was defined as the scaled difference in the radii of 2 concentric spheres about a specified center (radial difference of 2 circles drawn such that they enclose the traces with minimal radial difference) [8 –10]. The 2 great circles corresponding to the 2 circular traces on sagittal and transverse planes to the bore axis of the prosthetic femoral head and a great circle corresponding to the 1 circular trace on the transverse plane of the UHMWPE liner were measured.
Fig. 1. Measurement of the out-of-roundness using a stylus instrument. (A) Out-of-roundness on the sagittal plane of the prosthetic femoral head. (B) Out-of-roundness on the transverse plane of the prosthetic femoral head. (C) Out-of-roundness on the transverse plane of the ultra-high molecular weight polyethylene liner.
1025
The out-of-roundness of metal femoral heads, the inner surface of UHMWPE liners, and industrial ball bearings was measured by RONDCOM 52B-710 (Tokyo Seimitsu, Tokyo, Japan) in the 1995 group and by Talyrond 265 (Taylor Hobson, Leicester, UK) in 1999 and 2000 group. These testing machines generally were accepted for reliable stylus contact instruments in the mechanical engineering fields. Peak-to-peak values, which were deviations between the maximal peak height and the maximal valley depth, were evaluated. A total of 50 prosthetic metal femoral heads and 22 UHMWPE liners obtained directly from the manufacturers were evaluated in 1995. Two Kyocera system heads (26-mm diameter; Kyocera, Kyoto, Japan), 3 Harris Precoat Plus heads (22-mm diameter; Zimmer, Warsaw, IN), 3 Elite heads (22-mm diameter; DePuy, Warsaw, IN), 10 PFC heads (22-mm diameter; DePuy Johnson and Johnson, Warsaw, IN), 3 Richard Modular Hip system heads (RMHS) (22-mm diameter; Smith and Nephew, Memphis, TN), and 29 Omnifit heads (19 in 22-mm diameter, 10 in 26-mm diameter; Howmedica Osteonics, Allendale, NJ) were measured, and 4 Harris Galante Porous II liners (HGP IL) (22-mm diameter; Zimmer), 2 Hylamer liners (28-mm diameter; DePuy), 10 PFC liners (22-mm diameter), 3 RMHS liners (22-mm diameter), and 3 Omnifit liners (22-mm diameter) were measured. Out-of-roundness of ball bearings available in the mechanical engineering fields were measured for the control study. A total of 43 prosthetic metal femoral heads and 40 UHMWPE liners obtained directly from the manufacturers were evaluated in 1999 and 2000. Three Kyocera system heads (26-mm diameter), 10 Harris Precoat Plus heads (5 in 22-mm diameter, 5 in 26-mm diameter), 10 Elite heads (5 in 22-mm diameter, 5 in 26-mm diameter), 10 S-ROM heads (5 in 22-mm diameter, 5 in 26-mm diameter; DePuy Johnson and Johnson), and 10 Omnifit heads (5 in 22-mm diameter, 5 in 26-mm diameter) were measured, and 10 Trilogy liners (5 in 22-mm diameter, 5 in 26-mm diameter; Zimmer), 10 Enduron liners (5 in 22-mm diameter, 5 in 26-mm diameter; DePuy), 10 S-ROM liners (5 in 22-mm diameter, 5 in 26-mm diameter), and 10 Omnifit liners (5 in 22-mm diameter, 5 in 26-mm diameter) were measured. Statistical analyses of the out-of-roundness were performed using the Kruskal-Wallis test for nonparametric data, then repeated measures analysis of variance with Scheffe´ posthoc test was applied. Probability values P⬍.05 were considered significant.
1026 The Journal of Arthroplasty Vol. 16 No. 8 December 2001 Table 1. Out-of-Roundness of Ball Bearings and Prosthetic Femoral Heads Measured in 1995 Out-of-roundness Sagittal (m) Transverse (m)
Ball bearing (n ⫽ 3)
Kyocera (n ⫽ 2)
Harris (n ⫽ 3)
Elite (n ⫽ 3)
PFC (n ⫽ 10)
RMHS (n ⫽ 3)
Omnifit (n ⫽ 29)
0.3 ⫾ 0.0 0.3 ⫾ 0.0
0.4 ⫾ 0.1 0.4 ⫾ 0.1
5.2 ⫾ 5.4 0.9 ⫾ 0.2
3.9 ⫾ 3.8 0.6 ⫾ 0.2
3.2 ⫾ 1.4 1.2 ⫾ 0.5
2.0 ⫾ 0.6 2.3 ⫾ 1.3
12.1 ⫾ 4.6 0.8 ⫾ 0.3
NOTE. Values are mean ⫾ SD.
Results The out-of-roundness of all prosthetic femoral heads and UHMWPE liners are summarized in Tables 1 to 4. The data obtained from the femoral heads (Figs. 2 and 3) and the UHMWPE liners (Fig. 4) were inferior to those of industrial ball bearings (P⬍.0001 and P⬍.0001). The out-of-roundness of the UHMWPE liners was significantly inferior to that of the femoral heads (P⬍.0001). There were no significant differences among the groups of femoral heads or liners with different diameters provided by the same manufacturer (eg, 22-mm diameter heads vs 26-mm diameter heads of S-ROM). The out-of-roundness of the femoral head on the sagittal plane was significantly inferior to that on the transverse plane in the 1995 group and the 1999 and 2000 group (P⬍.0001). Using repeated measures analysis of variance with Scheffe´ posthoc test for analysis of the femoral heads comparing the sagittal and transverse planes provided by the same manufacturer, significant differences were found in Omnifit (P⬍.0001) in the 1995 group and Omnifit (P⬍.0001) and S-ROM (P⬍.05) in the 1999 and 2000 group. In an analysis of the sagittal plane in the 1995 group including ball bearings, significant differences were found only between Omnifit and ball bearing (P⬍.003), Kyocera (P⬍.03), PFC (P⬍.0001), and RMHS (P⬍.02). In an analysis of the transverse plane in the 1995 group including ball bearings, significant differences were found between RMHS and ball bearing (P⬍.001), Kyocera (P⬍.003), Harris (P⬍.03), Elite (P⬍.003), PFC (P⬍.03), and Omnifit (P⬍.003). In an analysis of the sagittal plane in the 1999 and 2000
group including ball bearings (data in 1995), significant differences were found only between Omnifit and ball bearing (P⬍.004), Kyocera (P⬍.007), and Elite (P⬍.007). There were no significant differences in an analysis of the transverse plane in the 1999 and 2000 group including ball bearings (data in 1995). The out-of-roundness of the femoral head on the sagittal plane was worst for Omnifit in the 1995 group and in the 1999 and 2000 group. Overall the out-of-roundness of the femoral head on the sagittal plane had improved significantly in 1999 and 2000 compared with that in 1995 (P⬍.0001); however, the out-ofroundness on the transverse plane was not improved significantly. In an analysis of UHMWPE liners in the 1995 group, significant differences were found only between PFC and HGP II (P⬍.02), RMHS (P⬍.03), and Omnifit (P⬍.03). In an analysis of UHMWPE liners in the 1999 and 2000 group, significant differences were found between Enduron and Trilogy (P⬍.002), Enduron and Omnifit (P⬍.0001), S-ROM and Trilogy (P⬍.0001), and S-ROM and Omnifit (P⬍.0001). In contrast with the femoral head, the out-of-roundness of the UHMWPE liners showed the best score for Omnifit in the 1999 and 2000 group. The out-of-roundness of the UHMWPE liners improved significantly in 1999 and 2000 compared with that in 1995 (P⬍.0001).
Discussion Currently, metal femoral heads and UHMWPE liners are the commonest bearing surfaces used in
Table 2. Out-of-Roundness of Prosthetic Femoral Heads Measured in 1999 and 2000 Out-of-Roundness Sagittal (m) Transverse (m)
Kyocera (n ⫽ 3)
Harris (n ⫽ 10)
Elite (n ⫽ 10)
S-ROM (n ⫽ 10)
Omnifit (n ⫽ 10)
0.67 ⫾ 0.10 0.20 ⫾ 0.07
3.65 ⫾ 2.76 0.81 ⫾ 0.29
2.71 ⫾ 1.42 0.56 ⫾ 0.29
3.97 ⫾ 2.26 0.90 ⫾ 0.55
6.93 ⫾ 2.46 1.10 ⫾ 0.70
NOTE. Values are mean ⫾ SD.
Bearing Surfaces in THA • Ito et al.
1027
Table 3. Out-of-Roundness of Polyethylene Liners Measured in 1995 Out-of-Roundness Transverse (m)
Harris (n ⫽ 4)
Hylamer (n ⫽ 2)
PFC (n ⫽ 10)
RMHS (n ⫽ 3)
Omnifit (n ⫽ 3)
21.3 ⫾ 2.9
33.5 ⫾ 13.4
41.8 ⫾ 8.5
21.2 ⫾ 1.9
21.3 ⫾ 10.6
NOTE. Values are mean ⫾ SD.
THA. The shape and contour of the bearing surfaces are important factors in UHMWPE wear. Complete fluid lubrication generally cannot occur in the friction of 2 solid materials. Solid-solid contact inevitably occurs because of roughness of the frictional surface and insufficient accuracy of surface contour congruency between the prosthetic femoral head and the socket, usually resulting in boundary lubrication. Wear particles are produced at the solidsolid contact sites. These wear particles between the frictional surfaces abrade the surfaces by acting as a cutting tool and promote wear, causing high frictional forces. It is important to prevent this abrasive wear. Several studies concerning the morphology and the surface roughness of hip prostheses have been reported [3,11]. Amstutz [11] previously reported on the roughness of the prosthetic metal femoral heads using the stylus method. He noted that the surface finish of metal heads had been improved by Charnley and that the average roughness was ⬍2 microinches with several peak-to-valley asperities of 6 microinches. Isaac et al [3] reported that the UHMWPE wear rate of the bearing surface is related proportionally to the roughness of the femoral head. Dowson et al [12] indicated that even single scratches may produce a dramatic increase in wear rate in laboratory tests. In contrast, few studies concerning sphericity of the bearing surface have been performed. The sphericity generally is assessed by the sphericity error, which is the sum of the maximal and minimal deviations from the least-squares sphere [10]. The characteristics of the circularity profile are assessed in terms of the roundness error, the maximal peak height, the maximal valley depth, and the mean
line average height. These parameters provide information about the nature of the surface and give a quantitative basis for specifying the profile. Oonishi [9] reported that metal heads with poor out-ofroundness (9.0 –11.0 m) also showed poor UHMWPE wear behavior compared with ball bearings with good out-of-roundness (0.5 m). He noted that the protruding part of metal heads reduced roundness and decreased wear performance; however, he did not verify that how degree in out-ofroundness was acceptable in wear tests. The present results indicated that the sphericity of the prosthetic femoral head was inferior to that of commercially available ball bearings. We evaluated 20 other Omnifit heads (10 in 28-mm diameter, 10 in 32-mm diameter) in 1995. The data were comparable, but not superior to those in 22-mm and 26-mm diameter Omnifit heads. The out-ofroundness of the femoral head on the sagittal plane also was significantly inferior to that on the transverse plane. We could not obtain 28-mm and 32-mm diameter heads from other manufacturers. These data of 28-mm and 32-mm Omnifit heads were not included in this study. We focused on 22-mm and 26-mm heads and UHMWPE liners in the present study. The sphericity on the sagittal plane was inferior to that on the transverse plane. The implant manufacturers were contacted regarding the out-of-roundness in the implants in the sagittal versus the transverse plane and indicated that the manufacturing process was proprietary. We suppose that the manufacturing process was one of the responsible factors. Surface processing possibly employed a rotating system of the femoral head with the bore axis when processing a smooth surface of the femoral head. Out-of-round-
Table 4. Out-of-Roundness of Polyethylene Liners Measured in 1999 and 2000 Out-of-Roundness Transverse (m) NOTE. Values are mean ⫾ SD.
Trilogy (n ⫽ 10)
Enduron (n ⫽ 10)
S-ROM (n ⫽ 10)
Omnifit (n ⫽ 10)
12.45 ⫾ 3.97
23.38 ⫾ 9.10
29.66 ⫾ 4.95
7.07 ⫾ 2.67
1028 The Journal of Arthroplasty Vol. 16 No. 8 December 2001
Fig. 2. An Omnifit femoral head measured in 1995. Out-of-roundness on the sagittal plane was inferior to that on the transverse plane. (A) Out-of-roundness on the sagittal plane was 17.7 m. (B) Out-ofroundness on the transverse plane was 0.36 m.
ness of the transverse plane might be regulated with this process to a considerable extent compared with that of the sagittal plane. As noted, the present findings are likely to affect in vivo wear performance. We previously reported [13] that UHMWPE wear with the prosthetic femoral head and socket in poor outof-roundness coupling (9.5 m and 36 m) was 148% greater compared with those in good out-ofroundness coupling (0.5 m and 0.6 m) at 1 million cycle experiments using a hip simulator. These data indicated that good sphericity is important at least in the initial wear phase. Some prosthetic femoral heads indicated ⬎9.5 m out-of-roundness in the present study, and these femoral heads with poor sphericity might be unfavorable to wear if implanted. We propose that the femoral heads with considerably poor sphericity should not be available on the market. It is not clear how degree of out-of-roundness significantly influences UHMWPE wear in clinical practice; however, our findings suggest poor sphericity is not favorable. Further studies on the relationship be-
Fig. 3. A Kyocera femoral head measured in 1999. The best results were obtained among all measured femoral heads. (A) Out-ofroundness on the sagittal plane was 0.55 m. (B) Out-of-roundness on the transverse plane was 0.24 m.
tween the sphericity of the bearing surface and wear are necessary using hip joint simulators. UHMWPE can become elastically deformed more easily than metal, which allows relative maintenance of good congruence between the metal head and the UHMWPE liner. The UHMWPE liner with repeated elastic and plastic deformation probably deteriorates over the course of millions of gait cycles. UHMWPE liners with good sphericity may be difficult to manufacture because of their ease in elastic deformation. We reported [13] that good sphericity could be obtained, however, by processing a 22-mm diameter UHMWPE liner to a 26-mm diameter liner with a cutting machine. We believe that sphericity of UHMWPE liners also should be improved to reduce initial UHMWPE wear. The measurement method using stylus instruments sometimes is criticized [10]. The profile obtained from the confocal scanning optical microscope consists of surface roughness, waviness, and form error components. The profile is a real repre-
Bearing Surfaces in THA • Ito et al.
1029
heads and UHMWPE liners may be a problem for companies. The bearing surface with poor sphericity may increase UHMWPE wear, however, compared with that with good sphericity, and improvement of the sphericity is desirable because this would be expected to prolong the functional performance of the prosthesis after THA. The present study definitely indicated that the bearing surface with poor sphericity should not be available on the market.
Acknowledgment We thank William J. Maloney, MD, Washington University School of Medicine, for his advice and critical comments on this study.
References Fig. 4. A S-ROM ultra-high molecular weight polyethylene liner measured in 2000. The out-of-roundness was 36.40 m.
sentation of the actual surface. With a stylus contact instrument, the obtained profile is not the actual surface profile and is a filtered one because of the steel ball used for measuring the surface. The roundness error value obtained by the confocal scanning optical microscope is higher than that obtained by the stylus instrument; however, the obtained profile by the stylus instrument may provide acceptable useful values [10]. In regard to the different testing devices, it is not clear whether the data produced from each machine is comparable. It generally is accepted in the mechanical engineering fields that Talyrond 265, used in the 1999 and 2000 group, is the most reliable testing machine for a stylus contact instrument; however, this machine was not available to us in 1995. RONDCOM 52B710, used in the 1995 group, generally is accepted for one of the most reliable testing machines. We believe that the present data from each machine is appropriately comparable. The present results of significant differences in the sphericity of the bearing surface among several companies indicates that manufacturing techniques and tolerances held differed among these companies. The current inferior results of bearing surfaces compared with those of commercially available ball bearings indicate that the sphericity of bearing surfaces can be improved by appropriate changes in manufacturing technique. The increased cost for achieving morphologic good sphericity in manufacturing metal femoral
1. Schmalzried TP: Current concept review: wear in total hip and knee replacements. J Bone Joint Surg Am 81:115, 1999 2. Callaghan JJ, Albright JC, Goetz DD, et al: Charnley total hip arthroplasty with cement: minimum twenty-five-year follow-up. J Bone Joint Surg Am 82:487, 2000 3. Isaac GH, Wroblewski BM, Atkinson JR, Dowson D: A tribological study of retrieved hip prosthesis. Clin Orthop 276:115, 1992 4. Maloney WJ, Smith RL, Schmalzried TP, et al: Isolation and characterization of wear particles generated in patients who have had failure of a hip arthroplasty without cement. J Bone Joint Surg Am 77:1301, 1995 5. Urban RM, Jacobs JJ, Tomlinson MJ, et al: Dissemination of wear particles to the liver, spleen, and abdominal lymph nodes of patients with hip or knee replacement. J Bone Joint Surg Am 82:457, 2000 6. Hishiki M, Hori T, Kawabe H: Sound life of steel ball (1st report): effect of shape of steel ball on sound life. Bull Jpn Soc Proc Eng 14:255, 1980 7. Gotoh K: Steel ball [in Japanese]. Special Steel 7:30, 1998 8. Murthy TSR, Raghunatha RB, Abdin SZ: Evaluation of spherical surfaces. Wear 57:167, 1979 9. Oonishi H: Studies on the shape and contour of the metal: prosthetic head in total hip prosthesis. Orthop Surg Trauma Proc 12th SICOT, p. 107, 1972 10. Udupa G, Singaperumal M, Sirohi RS, Kothiyal MP: Assessment of surface geometry using confocal scanning optical microscope. Mechatronics 8:187, 1998 11. Amstutz HC: Complications of total hip replacement. Clin Orthop 72:123, 1970 12. Dowson D, Taheri S, Wallbridge NC: The role of counterface imperfections in the wear of polyethylene. Wear 119:277, 1987 13. Ito H, Kaneda K, Yuhta T, et al: Reduction of polyethylene wear by concave dimples on the frictional surface in artificial hip joints. J Arthroplasty 15:332, 2000
The Sphericity of the Bearing Surface in Total Hip Arthroplasty Hiroshi Ito, MD,* Akio Minami, MD,* Takeo Matsuno, MD,† Hiromasa Tanino, MD,† Toshio Yuhta, PhD,‡ and Ikuya Nishimura, PhD‡
Abstract: This study evaluated the sphericity of bearing surfaces in total hip arthroplasty. The out-of-roundness of metal femoral heads, the inner surface of polyethylene liners, and commercially available ball bearings was measured. The hip prostheses were obtained directly from the manufacturers. The sphericity of the bearing surfaces was significantly inferior to that of the ball bearings. The sphericity of the femoral head on the sagittal plane was inferior to that on the transverse plane. Several significant differences were found among different manufacturers. The sphericity of the femoral head on the sagittal plane and that of polyethylene significantly improved in 1999 and 2000 compared with those in 1995. Further improvement is desirable, however, because good sphericity is expected to prolong the functional performance of the prosthesis after total hip arthroplasty. Key words: total hip arthroplasty, bearing surface, sphericity, polyethylene wear.
The current limitations of total hip arthroplasty (THA) are related to the wear of the components [1]. Wear of ultra-high molecular weight polyethylene (UHMWPE) liner bearing surfaces is the dominant issue. Improvement in the wear properties of the bearing surface in THA should reduce the generation of wear particles and improve longevity. Several studies have shown that a UHMWPE liner of an artificial hip joint socket wears at a linear wear rate of approximately 0.1 to 0.2 mm/y, resulting in about 1 to 2 mm of linear wear during a 10-year period [1–3]. If the UHMWPE liner is thin, the progressive wear can re-
sult in friction between the metal back and femoral head, causing metallosis and further increasing the particle load. Wear particles stimulate a foreign body reaction, resulting in osteolysis and loosening [4,5]. The morphology of the bearing surface in THA is an important factor in UHMWPE wear. It is well known in the mechanical engineering fields that the greater sphericity a ball bearing has, the lower and more uniform the frictional force and the higher rolling quality. This situation in turn leads to an improved functional performance [6]. The outof-roundness of the bearing ring of the best ball bearing currently available is 0.001 to 0.005 m [7]. The purpose of this study was to analyze the sphericity of prosthetic femoral heads and the inner surface of UHMWPE liners to determine whether further improvements of the surface contour are required. All the prosthetic metal femoral heads and the UHMWPE liners evaluated in this study were obtained directly from the manufacturers.
From the *Department of Orthopaedic Surgery, Hokkaido University, School of Medicine, Sapporo; †Department of Orthopaedic Surgery, Asahikawa Medical College, Asahikawa; and ‡Division of Biomedical Systems Engineering, Hokkaido University, School of Engineering, Sapporo, Japan. Submitted December 29, 2000; accepted July 5, 2001. No benefits or funds were received in support of this study. Reprint requests: Hiroshi Ito, MD, Department of Orthopaedic Surgery, Hokkaido University, School of Medicine, Kita-ku Kita-15 Nishi-7, Sapporo 060-8638, Japan. E-mail: itobiro@med. hokudai.ac.jp Copyright © 2001 by Churchill Livingstone威 0883-5403/01/1608-0012$35.00/0 doi:10.1054/arth.2001.27673
Materials and Methods Out-of-roundness was assessed as an indicator representing sphericity by obtaining circular traces
1024
Bearing Surfaces in THA • Ito et al.
of the implants [8 –10]. To obtain circular traces, a cylindrical stylus was used, as shown in Fig. 1. The traces were taken on the great circles of the sphere, and the out-of-roundness was obtained from the circular traces. Out-of-roundness was defined as the scaled difference in the radii of 2 concentric spheres about a specified center (radial difference of 2 circles drawn such that they enclose the traces with minimal radial difference) [8 –10]. The 2 great circles corresponding to the 2 circular traces on sagittal and transverse planes to the bore axis of the prosthetic femoral head and a great circle corresponding to the 1 circular trace on the transverse plane of the UHMWPE liner were measured.
Fig. 1. Measurement of the out-of-roundness using a stylus instrument. (A) Out-of-roundness on the sagittal plane of the prosthetic femoral head. (B) Out-of-roundness on the transverse plane of the prosthetic femoral head. (C) Out-of-roundness on the transverse plane of the ultra-high molecular weight polyethylene liner.
1025
The out-of-roundness of metal femoral heads, the inner surface of UHMWPE liners, and industrial ball bearings was measured by RONDCOM 52B-710 (Tokyo Seimitsu, Tokyo, Japan) in the 1995 group and by Talyrond 265 (Taylor Hobson, Leicester, UK) in 1999 and 2000 group. These testing machines generally were accepted for reliable stylus contact instruments in the mechanical engineering fields. Peak-to-peak values, which were deviations between the maximal peak height and the maximal valley depth, were evaluated. A total of 50 prosthetic metal femoral heads and 22 UHMWPE liners obtained directly from the manufacturers were evaluated in 1995. Two Kyocera system heads (26-mm diameter; Kyocera, Kyoto, Japan), 3 Harris Precoat Plus heads (22-mm diameter; Zimmer, Warsaw, IN), 3 Elite heads (22-mm diameter; DePuy, Warsaw, IN), 10 PFC heads (22-mm diameter; DePuy Johnson and Johnson, Warsaw, IN), 3 Richard Modular Hip system heads (RMHS) (22-mm diameter; Smith and Nephew, Memphis, TN), and 29 Omnifit heads (19 in 22-mm diameter, 10 in 26-mm diameter; Howmedica Osteonics, Allendale, NJ) were measured, and 4 Harris Galante Porous II liners (HGP IL) (22-mm diameter; Zimmer), 2 Hylamer liners (28-mm diameter; DePuy), 10 PFC liners (22-mm diameter), 3 RMHS liners (22-mm diameter), and 3 Omnifit liners (22-mm diameter) were measured. Out-of-roundness of ball bearings available in the mechanical engineering fields were measured for the control study. A total of 43 prosthetic metal femoral heads and 40 UHMWPE liners obtained directly from the manufacturers were evaluated in 1999 and 2000. Three Kyocera system heads (26-mm diameter), 10 Harris Precoat Plus heads (5 in 22-mm diameter, 5 in 26-mm diameter), 10 Elite heads (5 in 22-mm diameter, 5 in 26-mm diameter), 10 S-ROM heads (5 in 22-mm diameter, 5 in 26-mm diameter; DePuy Johnson and Johnson), and 10 Omnifit heads (5 in 22-mm diameter, 5 in 26-mm diameter) were measured, and 10 Trilogy liners (5 in 22-mm diameter, 5 in 26-mm diameter; Zimmer), 10 Enduron liners (5 in 22-mm diameter, 5 in 26-mm diameter; DePuy), 10 S-ROM liners (5 in 22-mm diameter, 5 in 26-mm diameter), and 10 Omnifit liners (5 in 22-mm diameter, 5 in 26-mm diameter) were measured. Statistical analyses of the out-of-roundness were performed using the Kruskal-Wallis test for nonparametric data, then repeated measures analysis of variance with Scheffe´ posthoc test was applied. Probability values P⬍.05 were considered significant.
1026 The Journal of Arthroplasty Vol. 16 No. 8 December 2001 Table 1. Out-of-Roundness of Ball Bearings and Prosthetic Femoral Heads Measured in 1995 Out-of-roundness Sagittal (m) Transverse (m)
Ball bearing (n ⫽ 3)
Kyocera (n ⫽ 2)
Harris (n ⫽ 3)
Elite (n ⫽ 3)
PFC (n ⫽ 10)
RMHS (n ⫽ 3)
Omnifit (n ⫽ 29)
0.3 ⫾ 0.0 0.3 ⫾ 0.0
0.4 ⫾ 0.1 0.4 ⫾ 0.1
5.2 ⫾ 5.4 0.9 ⫾ 0.2
3.9 ⫾ 3.8 0.6 ⫾ 0.2
3.2 ⫾ 1.4 1.2 ⫾ 0.5
2.0 ⫾ 0.6 2.3 ⫾ 1.3
12.1 ⫾ 4.6 0.8 ⫾ 0.3
NOTE. Values are mean ⫾ SD.
Results The out-of-roundness of all prosthetic femoral heads and UHMWPE liners are summarized in Tables 1 to 4. The data obtained from the femoral heads (Figs. 2 and 3) and the UHMWPE liners (Fig. 4) were inferior to those of industrial ball bearings (P⬍.0001 and P⬍.0001). The out-of-roundness of the UHMWPE liners was significantly inferior to that of the femoral heads (P⬍.0001). There were no significant differences among the groups of femoral heads or liners with different diameters provided by the same manufacturer (eg, 22-mm diameter heads vs 26-mm diameter heads of S-ROM). The out-of-roundness of the femoral head on the sagittal plane was significantly inferior to that on the transverse plane in the 1995 group and the 1999 and 2000 group (P⬍.0001). Using repeated measures analysis of variance with Scheffe´ posthoc test for analysis of the femoral heads comparing the sagittal and transverse planes provided by the same manufacturer, significant differences were found in Omnifit (P⬍.0001) in the 1995 group and Omnifit (P⬍.0001) and S-ROM (P⬍.05) in the 1999 and 2000 group. In an analysis of the sagittal plane in the 1995 group including ball bearings, significant differences were found only between Omnifit and ball bearing (P⬍.003), Kyocera (P⬍.03), PFC (P⬍.0001), and RMHS (P⬍.02). In an analysis of the transverse plane in the 1995 group including ball bearings, significant differences were found between RMHS and ball bearing (P⬍.001), Kyocera (P⬍.003), Harris (P⬍.03), Elite (P⬍.003), PFC (P⬍.03), and Omnifit (P⬍.003). In an analysis of the sagittal plane in the 1999 and 2000
group including ball bearings (data in 1995), significant differences were found only between Omnifit and ball bearing (P⬍.004), Kyocera (P⬍.007), and Elite (P⬍.007). There were no significant differences in an analysis of the transverse plane in the 1999 and 2000 group including ball bearings (data in 1995). The out-of-roundness of the femoral head on the sagittal plane was worst for Omnifit in the 1995 group and in the 1999 and 2000 group. Overall the out-of-roundness of the femoral head on the sagittal plane had improved significantly in 1999 and 2000 compared with that in 1995 (P⬍.0001); however, the out-ofroundness on the transverse plane was not improved significantly. In an analysis of UHMWPE liners in the 1995 group, significant differences were found only between PFC and HGP II (P⬍.02), RMHS (P⬍.03), and Omnifit (P⬍.03). In an analysis of UHMWPE liners in the 1999 and 2000 group, significant differences were found between Enduron and Trilogy (P⬍.002), Enduron and Omnifit (P⬍.0001), S-ROM and Trilogy (P⬍.0001), and S-ROM and Omnifit (P⬍.0001). In contrast with the femoral head, the out-of-roundness of the UHMWPE liners showed the best score for Omnifit in the 1999 and 2000 group. The out-of-roundness of the UHMWPE liners improved significantly in 1999 and 2000 compared with that in 1995 (P⬍.0001).
Discussion Currently, metal femoral heads and UHMWPE liners are the commonest bearing surfaces used in
Table 2. Out-of-Roundness of Prosthetic Femoral Heads Measured in 1999 and 2000 Out-of-Roundness Sagittal (m) Transverse (m)
Kyocera (n ⫽ 3)
Harris (n ⫽ 10)
Elite (n ⫽ 10)
S-ROM (n ⫽ 10)
Omnifit (n ⫽ 10)
0.67 ⫾ 0.10 0.20 ⫾ 0.07
3.65 ⫾ 2.76 0.81 ⫾ 0.29
2.71 ⫾ 1.42 0.56 ⫾ 0.29
3.97 ⫾ 2.26 0.90 ⫾ 0.55
6.93 ⫾ 2.46 1.10 ⫾ 0.70
NOTE. Values are mean ⫾ SD.
Bearing Surfaces in THA • Ito et al.
1027
Table 3. Out-of-Roundness of Polyethylene Liners Measured in 1995 Out-of-Roundness Transverse (m)
Harris (n ⫽ 4)
Hylamer (n ⫽ 2)
PFC (n ⫽ 10)
RMHS (n ⫽ 3)
Omnifit (n ⫽ 3)
21.3 ⫾ 2.9
33.5 ⫾ 13.4
41.8 ⫾ 8.5
21.2 ⫾ 1.9
21.3 ⫾ 10.6
NOTE. Values are mean ⫾ SD.
THA. The shape and contour of the bearing surfaces are important factors in UHMWPE wear. Complete fluid lubrication generally cannot occur in the friction of 2 solid materials. Solid-solid contact inevitably occurs because of roughness of the frictional surface and insufficient accuracy of surface contour congruency between the prosthetic femoral head and the socket, usually resulting in boundary lubrication. Wear particles are produced at the solidsolid contact sites. These wear particles between the frictional surfaces abrade the surfaces by acting as a cutting tool and promote wear, causing high frictional forces. It is important to prevent this abrasive wear. Several studies concerning the morphology and the surface roughness of hip prostheses have been reported [3,11]. Amstutz [11] previously reported on the roughness of the prosthetic metal femoral heads using the stylus method. He noted that the surface finish of metal heads had been improved by Charnley and that the average roughness was ⬍2 microinches with several peak-to-valley asperities of 6 microinches. Isaac et al [3] reported that the UHMWPE wear rate of the bearing surface is related proportionally to the roughness of the femoral head. Dowson et al [12] indicated that even single scratches may produce a dramatic increase in wear rate in laboratory tests. In contrast, few studies concerning sphericity of the bearing surface have been performed. The sphericity generally is assessed by the sphericity error, which is the sum of the maximal and minimal deviations from the least-squares sphere [10]. The characteristics of the circularity profile are assessed in terms of the roundness error, the maximal peak height, the maximal valley depth, and the mean
line average height. These parameters provide information about the nature of the surface and give a quantitative basis for specifying the profile. Oonishi [9] reported that metal heads with poor out-ofroundness (9.0 –11.0 m) also showed poor UHMWPE wear behavior compared with ball bearings with good out-of-roundness (0.5 m). He noted that the protruding part of metal heads reduced roundness and decreased wear performance; however, he did not verify that how degree in out-ofroundness was acceptable in wear tests. The present results indicated that the sphericity of the prosthetic femoral head was inferior to that of commercially available ball bearings. We evaluated 20 other Omnifit heads (10 in 28-mm diameter, 10 in 32-mm diameter) in 1995. The data were comparable, but not superior to those in 22-mm and 26-mm diameter Omnifit heads. The out-ofroundness of the femoral head on the sagittal plane also was significantly inferior to that on the transverse plane. We could not obtain 28-mm and 32-mm diameter heads from other manufacturers. These data of 28-mm and 32-mm Omnifit heads were not included in this study. We focused on 22-mm and 26-mm heads and UHMWPE liners in the present study. The sphericity on the sagittal plane was inferior to that on the transverse plane. The implant manufacturers were contacted regarding the out-of-roundness in the implants in the sagittal versus the transverse plane and indicated that the manufacturing process was proprietary. We suppose that the manufacturing process was one of the responsible factors. Surface processing possibly employed a rotating system of the femoral head with the bore axis when processing a smooth surface of the femoral head. Out-of-round-
Table 4. Out-of-Roundness of Polyethylene Liners Measured in 1999 and 2000 Out-of-Roundness Transverse (m) NOTE. Values are mean ⫾ SD.
Trilogy (n ⫽ 10)
Enduron (n ⫽ 10)
S-ROM (n ⫽ 10)
Omnifit (n ⫽ 10)
12.45 ⫾ 3.97
23.38 ⫾ 9.10
29.66 ⫾ 4.95
7.07 ⫾ 2.67
1028 The Journal of Arthroplasty Vol. 16 No. 8 December 2001
Fig. 2. An Omnifit femoral head measured in 1995. Out-of-roundness on the sagittal plane was inferior to that on the transverse plane. (A) Out-of-roundness on the sagittal plane was 17.7 m. (B) Out-ofroundness on the transverse plane was 0.36 m.
ness of the transverse plane might be regulated with this process to a considerable extent compared with that of the sagittal plane. As noted, the present findings are likely to affect in vivo wear performance. We previously reported [13] that UHMWPE wear with the prosthetic femoral head and socket in poor outof-roundness coupling (9.5 m and 36 m) was 148% greater compared with those in good out-ofroundness coupling (0.5 m and 0.6 m) at 1 million cycle experiments using a hip simulator. These data indicated that good sphericity is important at least in the initial wear phase. Some prosthetic femoral heads indicated ⬎9.5 m out-of-roundness in the present study, and these femoral heads with poor sphericity might be unfavorable to wear if implanted. We propose that the femoral heads with considerably poor sphericity should not be available on the market. It is not clear how degree of out-of-roundness significantly influences UHMWPE wear in clinical practice; however, our findings suggest poor sphericity is not favorable. Further studies on the relationship be-
Fig. 3. A Kyocera femoral head measured in 1999. The best results were obtained among all measured femoral heads. (A) Out-ofroundness on the sagittal plane was 0.55 m. (B) Out-of-roundness on the transverse plane was 0.24 m.
tween the sphericity of the bearing surface and wear are necessary using hip joint simulators. UHMWPE can become elastically deformed more easily than metal, which allows relative maintenance of good congruence between the metal head and the UHMWPE liner. The UHMWPE liner with repeated elastic and plastic deformation probably deteriorates over the course of millions of gait cycles. UHMWPE liners with good sphericity may be difficult to manufacture because of their ease in elastic deformation. We reported [13] that good sphericity could be obtained, however, by processing a 22-mm diameter UHMWPE liner to a 26-mm diameter liner with a cutting machine. We believe that sphericity of UHMWPE liners also should be improved to reduce initial UHMWPE wear. The measurement method using stylus instruments sometimes is criticized [10]. The profile obtained from the confocal scanning optical microscope consists of surface roughness, waviness, and form error components. The profile is a real repre-
Bearing Surfaces in THA • Ito et al.
1029
heads and UHMWPE liners may be a problem for companies. The bearing surface with poor sphericity may increase UHMWPE wear, however, compared with that with good sphericity, and improvement of the sphericity is desirable because this would be expected to prolong the functional performance of the prosthesis after THA. The present study definitely indicated that the bearing surface with poor sphericity should not be available on the market.
Acknowledgment We thank William J. Maloney, MD, Washington University School of Medicine, for his advice and critical comments on this study.
References Fig. 4. A S-ROM ultra-high molecular weight polyethylene liner measured in 2000. The out-of-roundness was 36.40 m.
sentation of the actual surface. With a stylus contact instrument, the obtained profile is not the actual surface profile and is a filtered one because of the steel ball used for measuring the surface. The roundness error value obtained by the confocal scanning optical microscope is higher than that obtained by the stylus instrument; however, the obtained profile by the stylus instrument may provide acceptable useful values [10]. In regard to the different testing devices, it is not clear whether the data produced from each machine is comparable. It generally is accepted in the mechanical engineering fields that Talyrond 265, used in the 1999 and 2000 group, is the most reliable testing machine for a stylus contact instrument; however, this machine was not available to us in 1995. RONDCOM 52B710, used in the 1995 group, generally is accepted for one of the most reliable testing machines. We believe that the present data from each machine is appropriately comparable. The present results of significant differences in the sphericity of the bearing surface among several companies indicates that manufacturing techniques and tolerances held differed among these companies. The current inferior results of bearing surfaces compared with those of commercially available ball bearings indicate that the sphericity of bearing surfaces can be improved by appropriate changes in manufacturing technique. The increased cost for achieving morphologic good sphericity in manufacturing metal femoral
1. Schmalzried TP: Current concept review: wear in total hip and knee replacements. J Bone Joint Surg Am 81:115, 1999 2. Callaghan JJ, Albright JC, Goetz DD, et al: Charnley total hip arthroplasty with cement: minimum twenty-five-year follow-up. J Bone Joint Surg Am 82:487, 2000 3. Isaac GH, Wroblewski BM, Atkinson JR, Dowson D: A tribological study of retrieved hip prosthesis. Clin Orthop 276:115, 1992 4. Maloney WJ, Smith RL, Schmalzried TP, et al: Isolation and characterization of wear particles generated in patients who have had failure of a hip arthroplasty without cement. J Bone Joint Surg Am 77:1301, 1995 5. Urban RM, Jacobs JJ, Tomlinson MJ, et al: Dissemination of wear particles to the liver, spleen, and abdominal lymph nodes of patients with hip or knee replacement. J Bone Joint Surg Am 82:457, 2000 6. Hishiki M, Hori T, Kawabe H: Sound life of steel ball (1st report): effect of shape of steel ball on sound life. Bull Jpn Soc Proc Eng 14:255, 1980 7. Gotoh K: Steel ball [in Japanese]. Special Steel 7:30, 1998 8. Murthy TSR, Raghunatha RB, Abdin SZ: Evaluation of spherical surfaces. Wear 57:167, 1979 9. Oonishi H: Studies on the shape and contour of the metal: prosthetic head in total hip prosthesis. Orthop Surg Trauma Proc 12th SICOT, p. 107, 1972 10. Udupa G, Singaperumal M, Sirohi RS, Kothiyal MP: Assessment of surface geometry using confocal scanning optical microscope. Mechatronics 8:187, 1998 11. Amstutz HC: Complications of total hip replacement. Clin Orthop 72:123, 1970 12. Dowson D, Taheri S, Wallbridge NC: The role of counterface imperfections in the wear of polyethylene. Wear 119:277, 1987 13. Ito H, Kaneda K, Yuhta T, et al: Reduction of polyethylene wear by concave dimples on the frictional surface in artificial hip joints. J Arthroplasty 15:332, 2000