Journal of Orthopaedic Research 852345226 Raven Press, Ltd., New York 0 1990 Orthopaedic Research Society
A Reduced-Modulus Acrylic Bone Cement: Preliminary Results Alan S. Litsky, *Robert M. Rose, ?Clinton T. Rubin, and $Elliott L. Thrasher Orthopaedic BioMaterials Laboratory, Ohio State University, Columbus, Ohio; *Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts; fMusculo-Skeletal Research Laboratory, State University of New York at Stony Brook, Stony Brook, New York; and $Department of Orthopaedic Surgery, Mt. Auburn Hospital, Cambridge, Massachusetts, U.S.A.
Summary: Excessive local contact stress is implicated as an important factor
in the initiation of the loosening process after total joint arthroplasties. A reduced-modulus acrylic bone cement, which decreases the bonecement interface stresses, was developed to test this hypothesis. The formulation consists of butylmethacrylate beads, having a glass transition temperature of 27" C, in a methylmethacrylate matrix. This cement, polybutylmethylmethacrylate (PBMMA), has an elastic modulus one-eighth that of standard PMMA bone cement, 0.27 vs. 2.1 GPa, at body temperature. In vivo use in a pilot study using the sheep total hip arthroplasty model shows a reduction in the rate of loosening of femoral components when compared both radiographically and mechanically with PMMA contrors. Key Words: Bone cement-Bonecement interface-Contact stress-Loosening.
a reduced-modulus bone cement will result in a decreased loosening rate because the contact stresses are reduced. This local biomechanical model is consistent with the reduction in loosening rates observed with improved surgical techniques. Vigorous cleaning of the bony bed, pressurization of the cement, and porosity reduction techniques improve the continuity of the bone-cement interface by reducing the population of voids and interfacial debris. Thus, these techniques reduce the overall level of contact stresses.
Clinical loosening appears to be a sequential process beginning with resorption and membrane formation at the bone-cement interface. Our hypothesis is that these first two steps of the loosening cascade are due to excessive localized stresses at the bone-cement interface. Contact stresses at discontinuous interfaces between materials of different moduli have been calculated (43) and shown to vary inversely with the moduli of each material. A soft-tissue membrane will diminish the contact stresses in the adjacent bone. A reduced-modulus cement, acting as a shear spring (l), will perform the same function without necessitating the resorption of bone from the interface (10). It follows that
Reduced-ModulusBone Cement Received December 28, 1988; accepted November 1 , 1989. Address correspondence and reprint requests to Dr. A. S. Litsky at Orthopaedic BioMaterials Laboratory, Ohio State University, S-2035 Davis Medical Research Center, 480 West Ninth Avenue, Columbus, OH 43210, U.S.A.
The best cement formulation to test the contact stress hypothesis is one identical to polymethylmethacrylate (PMMA) bone cement in every respect 623
A . S . LITSKY ET AL.
except for a lower elastic modulus at body temperature. Our approach to this ideal was to incorporate polybutylmethacrylate beads in a methylmethacrylate matrix. Polybutylmethacrylate has a glass transition temperature (27” C) below body temperature and has a known low toxicity (8). The prepolymerized beads play no active role in the curing of the bone cement. The chemically active ingredients of the reduced-modulus cement formulation are identical to those found in currently used commercial bone cements. In Vivo Evaluation
The only true test of the contact stress hypothesis can be found in a living, load-bearing system where bone can express its preference for a firm, brittle cement or a reduced-modulus, ductile cement. The sheep hip arthroplasty model has proven its validity as a reliable model for early loosening in prior studies (9). Radiolucencies and increased interface compliance have been shown to occur consistently within 1 year. Experimental Design A left total hip arthroplasty was performed on four Hampshire cross ewes using standard surgical techniques (including lavaging and plugging the canal, and pressurizing the cement) with either PMMA or polybutylmethylmethacrylate (PBMMA) bone cement used to fix the stemmed femoral prosthesis. A large canine prosthesis (Richards Mfg., Memphis, TN, U.S.A.) was used. After recovering from the procedure, the sheep were allowed to ambulate ad libitum in stalls. After 1 year, a right femoral stem was inserted using the same cement as had been employed on the left side. The animal was sacrificed after the second component was inserted. Aerobic and anaerobic cultures were taken from the left hip capsule at the time of the second surgery. Both femurs were removed for detailed radiography and torsional compliance measurements. Each animal served as its own control. Radiography and Radiographic Measurements Anteroposterior and lateral radiographs were performed under general anesthesia immediately postoperativeiy and at 3-month intervais.
J Orthop Res, Vol. 8, No. 4, 1990
Torsional Compliance Measurements The torsional compliance test was a low-torque, nondestructive, dynamic test of the bone-cement interface that measured the relative rigidity of fixation of the “experimental” femoral component compared to the contralateral “control” component, inserted just prior to sacrifice. The interanima1 variability in femoral morphology requires that each animal be used as its own control. The low torques used produced a linear torque-displacement curve, the slope of which is a measure of interface compliance. The increase in compliance between the fresh and the remodeled sides was used as a measure of the loosening, as in previous studies (9). Following torsional testing, the femurs were cut transversely with a diamond blade at 1 cm intervals beginning just beneath the neck of the prosthesis. Each section was examined grossly for evidence of cement fractures.
RESULTS Material Properties The most important difference in material properties between polybutylmethylmethacrylate (PBMMA) and polymethylmethacrylate (PMMA) is a reduction in the modulus by a factor of 8 when tested at body temperature (PBMMA, 0.27 -+ 0.012 GPa; PMMA, 2.12 ? 0.17 GPa). Another significant distinguishing feature of PBMMA is its ductility. At body temperature, PBMMA tears in a ductile fashion, absorbing much more energy prior to failure than PMMA, which fails by brittle fracture.
In Vivo Cultures taken at the time of the second procedure did not grow any aerobic or anaerobic organisms, eliminating subclinical infection as a cause of the observed loosening. Radiography None of the animals that completed the study showed any radiographic evidence of subsidence or statistically significant changes in diaphyseal cortex thickness. There was, at 1 year follow-up, dramatic radiolucency at the bone-cement interface in all zones in the PMMA sheep. No such change was
REDUCED-MODULUS BONE CEMENT evident in the PBMMA animals. Additionally, calcar resorption was clearly seen on l year films of both PMMA sheep but not on the follow-up films of the PBMMA sheep. Sample radiographs are shown in Fig. 1. The small number of animals involved in the study prevents any meaningful statistical analysis of the radiolucency. Torsional Compliance Four animals completed 1 year follow-up: two PMMA sheep and two PBMMA sheep. The torsional compliance data are listed in Table 1. Consistent results, within 1%, were obtained on five tests of each femur. The increase in compliance averaged 40.5% in the PMMA sheep and only 11.6% in the PBMMA sheep. No cement fracture was seen on the cut sections.
The local mechanical environment plays a key role in the tissue response to fractures and implants. This was suggested by Charnley (2) and has been more scientifically confirmed by the recent work of Uhthoff (13), Uhthoff and Germain (14), DiGioia (3), and Ling (7). The physiological responses that can reduce contact stresses are resorption of bone and/or the formation of a soft-tissue layer of sufficient thickness; both responses are part of the loosening process. A reduced-modulus cement can achieve the same result. The interposition of a compliant layer between an implant and bone has been discussed by several authors (7,12) and several efforts have been made to introduce a soft material into the bone-implant interface (6,ll). Torsional compliance testing was found by Radin et al. (9) to be the most reproducible means to quan-
FIG. 1. Isolated femurs from sheep 1 year after implantation: (A) PMMA-calcar resorption and radiolucency at the bone-cement interface; (8) PBMMA-preservation of the calcar and absence of interface radiolucency.
J Orthop Res, Vol. 8, No. 4, 1990
A . S . LITSKY ET AL. TABLE 1. Torsional compliance data of sheep hip arthroplasties implanted with PBMMA or PMMA bone cement Compliance" (rad/Nm)
Compliance change (left - right)
Left (1 year)
352 160 371 379
PMMA PMMA PBMMA PBMMA
0.0216 2 0.0007 0.027 f 0.001 0.0198 f 0.0010 0.023 2 0.002
0.0148 f 0.0003 0.020 ? 0.001 0.0183 f 0.0003 0.020 f 0.002
0.0068 0.007 0.0015 0.003
45.9% 35.0% 8.2% 15.0%
Comparison of the right side tested immediately postoperatively with the left side tested after 1 year of adaptive remodeling. 2 SD (n = 5).
titate the compliance changes that occur at the bone-cement interface. That study also established the correlation between the appearance of a radiolucent line at the interface, the histological deterioration of the bone-cement interdigitation, and the increase in torsional compliance. The use of a femoral stem that is undersized by current standards for the femoral canal created a situation that may have accelerated the observed loosening. This was consistent in both groups. The low torques employed are well below the elastic limits of the model, as evidenced by reproducible results on repetitive tests. The contribution of the cortical bone, the cement, or the prosthesis to the measured displacement is minimal. Prosthesiscement interface compliance should not change until after cement damage is evident. None was seen in this study. Therefore, the change in compliance should accurately reflect a loosening of the b o n e cement interface. A lower elastic modulus under physiological conditions is the major distinction between PBMMA and PMMA. The increased ductility may also be beneficial. The creep of a viscoelastic bone cement is an obvious cause for concern, but since PBMMA incorporates ductile beads in a glassy, brittle matrix, the problem should be minimized. The viscoelastic properties of PBMMA, including creep, are currently under investigation. Despite the limited number of sheep in this pilot study, marked differences were seen in both the radiographic and the mechanical results at 1 year follow-up. The sheep with PBMMA cushioning the stresses at the b o n e cement interface did not show evidence of calcar resorption or bone-cement interface radiolucency, whereas the sheep with standard PMMA had both of these signs of radiographic loosening. The torsional compliance changes support the radiographic findings: the increase in compliance seen with PMMA was 40.5% while that seen with PBMMA was only 11.6%.
J Orthop Res, Vol. 8, N o . 4, 1990
Acknowledgment: This work was partially supported by grants from the Orthopaedic Research and Education Foundation and the Whitaker Health Science Foundation. Dr. Litsky's graduate studies were supported by the Fannie and John Hertz Foundation. We would also like to thank Howmedica for supplying the Simplex P used in this work and the Richards Manufacturing Company for supplying the prostheses.
REFERENCES 1. Albrektsson T: Implant fixation by direct bone anchorage. In: The Bone-Zmplant Interface, Workshop Report, ed by JL Lewis, JO Galante, Park Ridge, IL, American Academy of Orthopaedic Surgery, 1985, pp 157-171 2. Charnley J: Acrylic Cement in Orthopedic Surgery, Baltimore, Williams and Wilkins, 1970 3. DiGioia AM 111: The role of interfragmentary strain in fracture healing, Honors thesis, Harvard Medical School, 1986 4. Erdogan F, Civelek MB: Contact problem for an elastic reinforcement bonded to an elastic plate. J Appl Mech 41: 1014-1018, 1974 5. Erdogan F, Gupta GD: The problem of an elastic stiffener bonded to a half plane. J Appl Mech 38:937-941, 1971 6. Garg A, Walker PS: The effect of the interface on the bone stresses beneath tibial components. J Biomech 9:957-%8, 1986 7. Ling RSM: Observations on the fixation of implants to the bony skeleton. Clin Orthop 210 80-96, 1986 8. Paulet G, Vidal M: Toxicity of some acrylic and methacrylic esters of acrylamide and polyacrylamides. Arch Ma1 Prof Med Tray Secur SOC36:58-60, 1975 9. Radin EL, Rubin CT, Thrasher EL, Lanyon LE, Crugnola AM, Schiller AS, Paul IL, Rose RM: Changes in the b o n e cement interface after total hip replacement. An in vivo animal study. J Bone Joint Surg [Am] 64:1188-1200, 1982 10. Rose RM, Litsky AS: Biomechanical considerations in the loosening of hip replacement prostheses. In: Current Perspectives on Implantable Devices, Vol. 1, ed by DF Williams, London, JAI Press, 1989, pp 1 4 5 11. Skalak R: Biomechanical considerations in osseointegrated prostheses. J Prosthet Dent 49843-848, 1983 12. Tiberwal SB, Grant KA, Goodfellow JW: The radiolucent line beneath the tibial component of the Oxford meniscal knee. J Bone Joint Surg [Br] 66523-528, 1984 13. Uhthoff HK: Mechanical factors influencing the holding power of screws in compact bone. J Bone Joint Surg [Bri 5k633-639, 1973 14. Uhthoff HK, Germain J-P: The reversal of tissue differentiation around screws. Clin Orthop 123:248-252, 1977