The Comparison of Polyethylene Wear in Machined vs. Molded Polyethylene

A. Brent Bankston, M.D.+
E. Michael Keating, M.D.++
Chitranjan Ranawat, M.D.+++
Philip M. Faris, M.D.++
Merrill A. Ritter, M.D.++

+ The Orthopaedic Clinic
720 Connells Park Lane
Baton Rouge, LA 70800

++ Center for Hip and Knee Surgery
1199 Hadley Road
Mooresville, IN 46158

+++ Hospital for Special Surgery
535 E. 70th St.
New York, NY 10021
Mailing Address:
Merrill A. Ritter, M.D.

Center for Hip and Knee Surgery
1199 Hadley Road
Mooresville, IN 46158
(317) 831-2273

Abstract

Polyethylene wear debris has been identified as a cause of osteolysis and granuloma formation in total hip arthroplasty. Future improved longevity of these prostheses depends on controlling wear and wear debris products. Ultra high molecular weight polyethylene (UHMWPE) is currently the best polymer available for use in joint replacement prostheses. The production, fabrication, and sterilization of the implant can alter the molecular weight of UHMWPE and, thus, alter its resistance to wear. This study was designed to evaluate two manufacturing techniques used in production of polyethylene components. Polyethylene currently used in orthopaedic implants is either machined from polyethylene bar stock or compression molded from polyethylene resin. In this study we evaluate changes in acetabular component thickness in two well-matched groups of patients with either machined or molded polyethylene acetabular components.

Two hundred and thirty-six total hip prosthesis underwent radiographic evaluation. Radiographs from immediately postoperative and at the time of most recent follow-up were measured for change in acetabular component thickness. In 74 of these, the polyethylene was machined and in 162 the polyethylene was molded. The two groups were similar in that both the acetabular and femoral components were cemented, all acetabular components were non metal-backed, and the bearing surfaces in both groups was polished, cobalt chrome. For statistical comparison, patient matching was performed leaving 54 patients in each study group (26 female, 28 male, with an average follow-up of 6.7 years, age of 66 years, and weight of 161.0 pounds). Evaluation for differences in weight of rate of change of acetabular component thickness were compared between these patient matched groups. Radiographs were also evaluated for osteolysis, femoral subsidance, and progressive radiolucencies.

Results revealed wear rates of 0.05 mm per year for the molded polyethylene and 0.11 mm per year for the machined polyethylene in these two groups of total hip arthroplasty patients. These differences in wear rates were statistically significant (P=0.0001). The differences in volumetric wear debris production were also significantly less with the molded polyethylene as compared to the machined polyethylene.

The result of this study raises further questions regarding the consequences of polyethylene fabrication on its ultimate molecular weight and thus its wear resistance. Machined polyethylene components are machined from polyethylene bar stock which is produced by a semi-continuous ram extrusion technique. Compression molded polyethylene is the product of a one-step process. This study demonstrates the greater wear potential with machined polyethylene in total hip arthroplasty. Further laboratory evaluation of the results of this study and of UHMWPE fabricated by different methods will be necessary to clarify these clinical findings.

Introduction

Ultra high molecular weight polyethylene (UHMWPE) is currently the best polymer available for use in joint replacement prostheses. It has the lowest coefficient of friction, the best wear resistance, and the highest toughness. Numerous authors have suggested a relationship between implant loosening and polyethylene debris.3, 6, 11, 14, 15, 20, 21, 22, 23, 30, 31, 32, 40, 42, 43, 46 Osteolysis and granuloma formation at interfaces between total joint replacements prostheses and bone are of great concern to today’s orthopaedic surgeon. Although implant loosening secondary to polyethylene debris has in the past been a relatively rare problem, there is indication that as total joint replacement longevity increases the incidence of granulomatous destruction secondary to polyethylene debris will be much more prevalent.

Numerous factors have been identified which affect polyethylene debris generation.32 These include conformity of the articulating surface,3 the modulus of elasticity of the polyethylene,35 polyethylene thickness,3 metal backing,2 femoral head diameter,25 method of sterilization,24, 27, 34 and oxygen degradation.14, 25 To our knowledge, no clinical or laboratory study has examined the methods of implant fabrication and its effect on polyethylene wear. All polyethylene components currently used as orthopaedic implants are either machined from extruded bar stock polyethylene or compression molded from polyethylene resin (Figure 1). This study will examine clinical wear rates in a well-matched group of total hip arthroplasties in which machined or molded polyethylene acetabular components were used.

Materials and Methods

The study group consisted of 236 total hip replacements implanted by two surgeons at two separate institutions between 1978 and 1982. The prostheses studied were 162 TR-28 (Zimmer, Warsaw, IN) implanted between 1978 and 1980 and 74 Triad (Johnson & Johnson, Raynham, MA) implanted between 1982 and 1984. Both were one-piece prostheses with 28 mm femoral heads with varying neck length and stem sizes. The TR-28 prosthesis was a cobalt chrome prosthesis and the Triad had a titanium stem with a cobalt chrome head permanently fixed to the neck of the prosthesis. The acetabular components in both series were non-metal backed polyethylene differing only in their method of fabrication. The TR-28 acetabular component was compression molded by the implant manufacturer from PE resin produced by the Hercules company (currently Hi-mont, Wilmington, Delaware). The Triad acetabular component was machined from PE bar stock by the Westlake Corporation (Lenni, PA). Westlake produced the bar stock from PE resin supplied by Hoechst Celanese (Germany). The orthopaedic suppliers (Zimmer, Johnson & Johnson) used similar sterilization techniques consisting of irradiation with 2.5 - 3.2 MRads.

Both surgeons utilized a lateral approach with a trochanteric osteotomy for exposure and cemented both the acetabulum and the femur. There was a difference in cementing technique between the two groups. The TR-28 group was cemented using early cement techniques consisting of finger packing of simplex cement and the Triad hips were cemented using modern cementing techniques consisting of femoral canal plugging, irrigation of both the acetabulum and the femur, thorough drying and cement pressurization. The TR-28 group had an average follow-up of 8.7 years and there were 89 males and 62 females with 10 patients having bilateral procedures. The average age was 66 years with an average weight of 159 lbs. In the Triad group the average follow-up was 5.6 years with 27 females and 33 males with 14 patients undergoing bilateral procedures. The average age was 60 years with an average weight of 160 lbs.

Patient matching was performed where patients from each group were individually matched for age within 10 years, weight within 30 lbs., and follow-up within 1 year. Patients were also matched for gender. This left 54 patients in each group (26 females and 28 males). The average was 67 years (TR-28) and 65 years (Triad) with weight of 162 lbs. (TR-28) and 160.2 lbs. (Triad), follow-up 6.9 years (TR-28), 6.4 years (Triad).

The average polyethylene thickness was 12.7 mm in the molded group and 11.2 mm in the machined group. Both groups had a minimum polyethylene thickness of greater than 7.0 mm.

Measurement Technique

Standard AP radiographs taken on non-portable equipment using the same technique were performed on all patients. The most current radiograph and the initial two-month postoperative radiograph were compared for linear polyethylene wear by the measurement technique described by Livermore et al.25 (Figures 2 & 3) This technique consists of first determining the center of the femoral head with a transparency of concentric circles. A vertical line is then drawn tangent to the ischial tuberosities through the center of the femoral head (Figure 4). The direction of wear is determined with a compass and referenced relative to the tangent line. Wear medial to this line is recorded as positive. Another line is then drawn from the center of the femoral head in the wear direction. Measurement of polyethylene thickness was made along the second line from the metal-polyethylene articulation to the cement-prosthesis interface. The femoral head diameter of all prostheses was 28 mm and the radiographic diameter of the femoral head was measured to correct for magnification (Fig. 1). Measurements were made to the nearest 0.05mm using a Kanon (Model #C150) manual caliper. A computer software program (Advanced System Consultants, Indianapolis, IN) was developed to correct for magnification and to convert the radiographic measurements into actual linear and volumetric wear and to calculate wear rates. Since acetabular wear is cylindrical, volumetric wear can be calculated from the formula V-R2 W where V is the volume of wear debris, R is the femoral head radius, and W is the amount of linear wear. The wear rates were determined by dividing the measured linear wear and the calculated volumetric wear by the follow-up in years.

Radiographic Observations

Radiographs were also evaluated for progressive acetabular radiolucencies according to the criteria of DeLee and Charnley.13 Femoral radiolucencies were similarly recorded by the method described by Gruen.19 Progressive femoral osteolysis was recorded as present or absent and subsidence was graded as none, 0-5mm, 5-10mm or greater than 10mm.

Results

Results of our wear measurements for all patients included in this study showed a wear rate of 0.05 (TR-28) and 0.12 (Triad). The differences in these linear wear rates were highly significant (p=0.0001). The patient matched analysis of linear wear showed similar results with wear rates of 0.05 (TR-28) and and 0.11 (Triad) (Table I). This difference was also highly significant (p-0.0001). The angle of wear was measured as two degrees (TR-28) and four degrees (Triad) (Table IV). There was no difference in these measurements (p=0.7).

Radiographic analysis of acetabular radiolucencies showed no differences in the incidence in at least one progressive acetabular radiolucency or the incidence of complete progressive acetabular radiolucencies (p=0.275, p=0.75) (Table II). There was no significant difference in the incidence of femoral osteolysis between groups (p=0.65). There was an increased incidence of femoral subsidence in the TR-28 group (p=0.031) compared to the Triad in which modern cement technique was used (Table III).

Discussion

Polyethylene debris is generated by friction between the metal on polymer articulation in total joint replacements. Assuming that this articulation has in the past provided the lowest coefficient of friction that could be obtained in an artificial joint, our attention should be directed toward improving the metallic femoral or the polyethylene acetabular bearing surfaces to decrease wear production which can develop at the extrusion port. The powder zone can lead to dead zones in the polyethylene which can be identified by sampling of the polyethylene bars. These can greatly alter the molecular weight of the polyethylene and thus the wear resistance of the finished implant.

The molding process involves compression molding of polyethylene resin directly into the component size. This process is more expensive as implants can not be produced as rapidly as by the machining process. Polyethylene was introduced as the acetabular bearing surface in total hip arthroplasty by Charnley due to the rapid failure of polytetrafluorethylene (PTFE).6, 34 The high density polyethylene (HDPE) used by Charnley had a molecular weight of approximately 500,000. Polyethylene is polymerized from resin by heat and pressurization and due to refinements in this technique the ultra high molecular weight polyethylene (UHMWPE) used today has a MW of 3,000,000. Polyethylene wear has been directly related to its molecular weight and numerous factors have been identified which can affect the molecular weight of the polyethylene and, thus, its wear characteristics. These include irradiation for sterilization,24, 27, 34 the quality of the initial polyethylene resin, and atmospheric exposure which can lead to oxidative degradation.16, 37 Other factors affecting the rate of polyethylene wear include polyethylene thickness, joint conformity,3 and the use of metal backing.2, 4, 26, 33, 39 Wright has shown that the quality of the polyethylene can vary widely depending on its grade, the fabrication methods, and the supplier of the polyethylene.

There are various manufacturing processes which can be involved in the fabrication of polyethylene orthopaedic implants. These include ram extrusion, pressure crystallization, heat pressurization, compression molding, and uniform pressurization. The problems with heat pressing of implants to improve the surface finish have been documented in the recent literature with delamination a few millimeters under the implant surface and these methods have been discarded. Pressure crystallization is a new technique of polyethylene fabrication in which the polyethylene produced is more rigid than standard polyethylene. This technique has not been proven to reduce wear and may increase contact stresses within the polyethylene and may possibly reduce the resistance of the polyethylene to fracture. The majority of implants used today are either machined from bar stock or compression molded directly from polyethylene resin.

Machined polyethylene implants are produced by a two-step process. The initial step is the production of the polyethylene bar stock followed by the machining process which involves cutting of the bar stock to the implant size with a precision cutting lathe. This process typically involves two manufacturers. The polyethylene manufacturer who performs the bar extrusion process producing the polyethylene bar stock, and the implant manufacturer who is responsible for the machining. The step that lends itself to the development of inconsistencies in the polyethylene is the actual production of the bar stock. The polyethylene bars are produced by a semicontinuous technique called ram extrusion. In this process the polyethylene resin is heated and pressurized as in the molding process. However, instead of heating and pressurizing a single bar of polyethylene the polyethylene is extruded out of a pressurization chamber forming bars. Because this is not a continuous process there are often inconsistencies in the PE examined in a single bar. This process is being developed to bypass the shortcomings of ram extrusion and to allow the flexibility of machinery. The powder zone can lead to dead zones in the polyethylene which can be identified by sampling of the polyethylene bars. These can greatly alter the molecular weight of the polyethylene and thus, the wear resistance of the finished implant.

The molding process involves compression molding of polyethylene resin directly into the component size. This is a one-step process performed by the orthopaedic manufacturer in which a consistent amount of heat and pressurization is applied to the resin for formation of the implant. This process is more expensive as implants can not be produced as a rapidly as by the machining process. Uniform pressurization is a new technique currently being developed in which bar stock is produced by a one-step compression molding process. This process is being developed to bypass the shortcomings of ram extrusion and to allow the flexibility of machining. The machining process also typically involves two manufacturers. The polyethylene manufacturer who performs the bar extrusion process and the implant manufacturer who is responsible for the machining.

This is the first paper examining clinical wear of polyethylene which has been fabricated by two different methods. The prosthetic groups in this study were chosen for comparison because of their similarities in design and surgical technique. The Triad and TR-28 prostheses had similar head-neck offsets and identical cobalt chrome bearing surfaces. In order to eliminate variables related to age, patient weight, length of follow-up and gender the study group was individually matched to eliminate these variables. Previous studies of clinical wear have reported wear rates varying from 0.07 to 0.21 mm per year.9, 14, 18, 25, 45, 46 However, none of these studies stated whether the polyethylene examined was fabricated by machining or by the molding process. Our results show significantly increased wear in polyethylene which had been machined compared to polyethylene which had been compression molded. This report raises serious questions regarding polyethylene fabrication and standardization of these processes.

The concern about wear is not actually with the thought of the polyethylene component wearing out, but more with the production of wear debris. Bartel has shown that an acetabular component should be at least 6 mm thick.44 This reduces stress within the component and thus reduces wear rates. Assuming a linear wear rate of 0.1 mm per year it would take almost 60 years for a 6.0 mm polyethylene component to the polyethylene “wear out” if the minimal recommended thickness is chosen. However, assuming the same wear rate of 0.1mm per year, and using a known polyethylene particle size of 10 _m x 1 _m x 1 _m then there are 3.8 billion polyethylene particles produced each year. Undoubtedly, the body’s host defense mechanisms can eliminate a large percentage of these particles and at what point the reaccumulation leads to osteolysis and implant loosening has yet to be demonstrated. In this study wear is defined as a change in polyethylene thickness of the acetabular component as measured radiographically. Rose has shown that both creep and plastic deformation of polyethylene as well as abrasive wear contribute to radiographic changes in polyethylene thickness with the wear accounting for about 30% of measured wear.35 Therefore, any report of wear rate is actually a report of polyethylene deformation. Assuming that only 30% of the perceived change in polyethylene thickness is due to wear, there would still represent approximately 1 billion polyethylene particles produced each year with a wear rate of 0.1mm per year. Therefore, polyethylene debris is produced at a significant rate and this rate may threaten prosthesis longevity.

Machined polyethylene produces significantly more wear debris than comparable compression molded polyethylene. Possible reasons for the difference in wear rates between machined and molded PE can be speculated. Walker noted that there were microscopic differences in the bearing surface of the PE in machined versus molded components.41 He noted that the machining process resulted in microscopic grooves and shreds of polyethylene on the component surface. These surface irregularities could be a source of third body wear particles which could increase the rate of wear. However, Walker also noted that laboratory wear analysis showed no difference in the steady state wear rate between machined and molded polyethylene. It was postulated that there was an initial rapid wear period in machined PE which later reached a steady state and wore at a rate similar to that of molded PE. The molding or extrusion temperature has a significant effect on the final MW of the PE. The wear resistance of the implant is closely linked to the MW of the PE and minor differences in the combination of pressure and temperature used in the extrusion process compared to the molding process could lead to structural differences.

Because ram extrusion of polyethylene bar stock is not a continuous process different consistencies of the polyethylene are known to develop. Because the implant manufacturer has no control over the extrusion process variations in bar stock can go undetected. This study demonstrates how differences in manufacturing techniques can alter the resulting wear resistance of polyethylene in total joint arthroplasty. Further investigation into these and other possible variables involved in polyethylene fabrication need to be further evaluated.

Table I
Results
1Linear 1Volumetric
Follow-up Age Weight M/F Wear Rate Wear Rate Wear
(yrs) (yrs.) (lbs) (mm/yr) (mm/yr) Angle
Molded PE (162) 87 66 159.0 62/39 0.05 33.30 2°
Machined PE (74) 56 60 160.0 33/27 0.12 70.94 4°
1Machined > Molded (p=0.001)

Table II
Acetabulum - Radiographic Analysis
Progressive Radiolucencies
Patient Matched
1 Zone Complete
Molded 25.3% 13.0%
Machined 13.9% 5.4%

Table III
Femur - Radiographic Analysis
Patient Matched
Subsidence Osteolysis
TR-28 13.6% 6.3%
Triad 4.1% 5.4%
1Triad < TR-28 (p=0.03)

Table IV
1Linear Wear 1Volumetric Wear
F/U Age Weight Rate Rate
(yrs) (yrs) (lbs.) M/F (mm/yr) (mm/yr)
Molded (54) 6.9 67 162.0 26/28 0.05 31.45
Machined (54) 6.4 65 160.2 26/28 0.11 65.58
1Molded < Machined (p=0.001) Legend

Table I: Demographics and wear results of all patients evaluated.
Table II: Progressive acetabular radiolucencies in at least one of three zones and in all three zones in the patient matched groups.
Table III: The incidence of femoral subsidence greater than 5mm and osteolysis in the patient matched groups.
Table IV: Demographics and wear results in the patient matched groups.

Figure 1: Graphic of methods of PE fabrication for orthopaedic implants.
Figure 2 & 3: Two month post-op and 10 year follow-up on a patient with a TR-28 prosthesis.
Figure 4: Demonstration of lines drawn for measurement technique described by Livermore.25Bibliography

1. Bankston, A.B.; Faris, P.M.; Keating, E.M.; Ritter, M.A.: Polyethylene wear with bearing surfaces of SS, CoCr and Ti (submitted for publication), 1992.
2. Bartel, D.D.; Wright, T.M.; Edwards, D.L.: The effect of metal backing on stresses in polyethylene acetabular components. The Hip: Proceedings of the Hip Society. Page 229, C.V. Mosby, St. Louis, 1983.
3. Bartel, D.L.; Bicknell, V.L.; Wright, T.M.: The effect of conformity and thickness and material on stresses in ultra high molecular weight components for total joint replacement. J. Bone Joint Surg., 68A:1041, 1986.
4. Bayley, J.C.; Scott, R.D.; Edwards, F.C.; Holmes, G.B.: Metal backed patellar component failure following total knee replacement. J. bone Joint Surg., 70A:668, 1988.
5. Carter, D.R.: Finite element analysis of a metal backed acetabular component. The Hip: Proceedings of the Eleventh Open Scientific Meeting of The Hip Society, pp. 216, St. Louis, C.V. Mosby, 1983.
6. Charnley, J.: Tissue reactions to polytetrafluoroethylene. Lancet 2:1379, 1963.
7. Charnley, J.: Low friction arthroplasty of the hip. Springer Verlag, New York, 1979.
8. Charnley, J.; Halley, D.K.: Rate of wear in total hip replacements. Clin. Orthop., 112:170, 1975.
9. Cupic, Z.: Long-term follow-up of Charnley arthroplasty of the hip. Clin. Orthop., 141:28, 1979.
10. Dal, D.M.; Grobbelaar, C.J.; Learmonth, I.D.; et al.: Charnley low friction arthroplasty. Clin. Orthop., 211:85, 1986.
11. Dannenmaier, W.C.; Haynes, D.E.; Nelson, C.L.: Granulomatous reaction and cystic bony destruction associated with high wear rate in a total knee prosthesis. Clin. Orthop. 198:224, 1985.
12. Davidson, J.A.: Surface finish of metal and ceramic total hip bearing surfaces and the effect on long-term UHMWPE wear. The Hip Society, Washington, D.C., 1992.
13. DeLee, J.; Charnley, J.: Radiologic demarcation of cemented sockets in THR. Clin. Orthop. 121:20, 1976.
14. Dowling, J.M.; Atkinson, J.R.; Dawson, D.; Charnley, J.: The characteristics of acetabular cups worn in the human body. J. Bone Joint Surg., 60B:375, 1978.
15. Elbert, K.E.; Kurth, M.; Bartel, D.L.; et al: In vivo changes in material properties of PE and their effects on stresses associated with surface damage of PE components. Trans. Ortho. Res. Soc.; 13:53, 1984.
16. Galante, J.O.; Rostoker, W.: Wear in total hip prostheses. Acta. Orthop. Scnd.; Suppl. 145, 1973.
17. Griffith, M.J.; Seidenstein, M.K.; Williams, D.; et al: Socket wear in Charnley low friction arthroplasty of the hip. Clin. Orthop., 137:37, 1978.
18. Gruen, T.A.; McNiece, G.M.; Amstutz, H.C.: Modes of failure of cemented stem type normal components: A radiographic analysis of loosening. Clin. Orthop., 141:17, 1979.
19. Harris, W.H.; Schuyler, A.L.; Scholler, J.M.; Freyberg, R.A.; Scott, R.: Extensive localized bone absorption in the femur following total hip replacement. J. Bone Joint Surg., 58A:612, 1976.
20. Howie, D.W.; Vernon-Roberts, B.; Oakeshott, R.; Manthey, B.: A rat model of resorption of bone at the cement bone interface and the presence of polyethylene wear particles. J. Bone Joint Surg., 70A:257, 1988.
21. Johanson, N.A.; Bullough, P.G.; Wilson, P.D.; et al: The microscopic appearance of the bone cement interface in the failed total hip arthroplasty. Clin. Orthop., 216:123, 1987.
22. Johanson, N.A.; Callaghan, J.J.; Salvatti, N.A.: Fourteen-year follow-up study of a patient with massive calcar absorption. A case report. Clin Orthop., 213:189, 1986.
23. Kameli A.; Feingold, L.: A model for radiation-induced changes in the ultra high molecular weight polyethylene. J. Polym. Sci-polym. Phys., 23:2407, 1985.
24. Livermore, S.H.; Ilstrup, D.; Murray, B.: Effect of femoral head size on wear of the polyethylene acetabular component. J. Bone Joint Surg., 72A(4):518, 1990.
25. Lombardi, A.V.; Engh, J.A.; Vols, R.; et al: Fracture dislocation of the polyethylene in metal backed patellar components in total knee arthroplasty. J. Bone Joint Surg., 70A:675, 1988.
26. Lue, C.T.; Ellis, E.J.; Curgnola, A.: Effects of gamma radiation on UHMWPE. Proceedings of the 39th SPE - ANTEC Conference, p. 246, Boston, May 4-7, 1981.
27. McCoy, T.M.; Salvatti, E.A.; Ranawat, C.S.: A fifteen-year follow-up study of 100 Charnley low friction arthroplasties. Clin. Orthop. N. Amer., 19:467, 1988.
28. Miller, D.A.; Ainsworth, R.D.; Dumbleton, J.A.; Page, D.; Miller, E.A.; Shen, C.: A comparative evaluation of the wear of UHMWPE abraded with Ti, 6A1-40 wear. 28:207, 1974.
29. Mirra, J.M.; Marder, R.A.; Amstutz, H.C.: The pathology of failed total joint arthroplasty. Clin. Orthop., 170:175, 1982.
30. Murray, D.W.; Rushton, N.: Macrophages stimulate bone resportion when the phagosatize particles. J. Bone Joint Surg., 72B:988, 1990.
31. Rimnac, C.M.; Wilson, P.D.; Fuchs, M.D.; Wright, T.M.: Acetabular cup wear in total hip arthroplasty. Clin. Orthop. N. Amer., 19(3):631, 1988.
32. Ritter, M.A.; Keating, E.M.; Faris, P.M.; Brugo, G.: Metal-backed acetabular cups in total hip arthroplasty. J. Bone Joint Surg.; 72A:672, 1990.
33. Roe, R.J.; Grood, E.S.; Shastri, R.; et al.: Effect of radiation sterilization and aging on ultra high molecular weight polyethylene. J. Biomed. Mat. Res., 15:209, 1981.
34. Rose, R.M.; Nusbaum, H.J.; Schneider, N.; et al.: On the true wear rate of UHMWPE in total hip prosthesis. J. bone Joint Surg.; 62A:537, 1980.
35. Salvatti, E.A.; Wilson, F.D.; Jolley, M.N.; et al.: A ten-year follow-up study of the first 100 consecutive Charnley total hip arthroplasty. J. Bone Joint Surg., 67B:757, 1985.
36. Shastri, R.; Grood, E.S.; Roe, R.J.; Moyes, F.R.: Effect of radiation sterilization on and aging or UHMWPE. Trans. Ortho. Res. Soc.; 6:115, 1981.
37. Steine, H.L.: Ultra high molecular weight polyethylene. Engineered Materials Handbook, 2:167, Engineering Plastics. ASM Press, Metals Park, Ohio, 1988.
38. Stulberg, S.D.; Stulberg, B.N.; Namati, Y.; Tsao, A.: Failure mechanisms of metal backed patellar components. Clin. Orthop., 236:88, 1988.
39. Tanser, N.; Maloney, W.J.; Jasty, M.; Harris, W.: The progression of femoral cortical osteolysis in association with total hip arthroplasty without cement. J. Bone Joint Surg., 74A:404, 1992.
40. Walker, P.S.: Friction and wear of artificial joints in human joints and their artificial replacements. Ed. Walker, P.S., Charles Thomas, 1977.
41. Willert, H.G.; Bertram, H.; Buchorn, G.M.: Osteolysis in alloarthroplasty of the hip. The role of ultra high molecular weight polyethylene wear particles. Clin. Orthop., 258:95, 1990.
42. Willert, H.G.; Simlich, M.: Reactions of the articular capsule to wear products of artificial joint prostheses. J. Biomed. Mat. Res., 11:157, 1977.
43. Wright, P.S.: Friction and wear of artificial joints. Human Joints and Their Artificial Replacements, Charles C. Thomas, Springfield, IL, 1977.
44. Wroblewski, B.M.: Direction and rate of socket wear in Charnley low friction arthroplasty. J. Bone Joint Surg., 67B:757, 1985.
45. Wroblewski, B.M.: Wear of high density polyethylene on bone and cartilage. J. Bone Joint Surg., 61B:498, 1979.