- Email: [email protected]
A mechanical evaluation of an elastic femoral prosthetic stem
W21-929U/87 S3.00 + .OO Q 1987 Pergamon Journals Ltd.
.I. Biomeehanics Vol. 20, No. 2. pp 179-185, 1987 Printed in Gnat Britain
A MECHANICAL
EVALUATION OF AN ELASTIC FEMORAL PROSTHETIC STEM
W. R. NEWCOMBE*, R. U. REPOt, N. B. EATON* and J. OCCHIONORELLI~ *Department of Mechanical Engineering, McMaster University, Hamilton, Ontario, Canada L8N 325: TDepartment of Surgery, Section of Orthopaedic Surgery, McMaster University Medical Centre, Hamilton, Ontario, Canada LSN 325 Abstract-Studies are being conducted in our laboratory to test the concept of introducing an elastomer to attenuate and damp forces applied to the hone interface in a major weighthearing joint replacement prosthesis. An analogue of a fully constrained intramedullary stem type prosthesis has heen developed in a segmental femoral replacement prosthesis of the dog. The layer of silastic was introduced to damp forces at the hone-prosthesis interface. This paper describes the response of this elastomer prosthesis to torsional and bending loads, and defines the upper limits of elastomer strain. The low modulus silastic displayed surprisingly low strain for applied loads, particularly in bending tests, in this prosthetic configuration. The results of these mechanical studies serve as a bench mark for the eventual design and material selection of an elastomer for human prosthetic use.
were manufactured. Figure 1 shows a prosthesis in situ
INTRODUCTION
in a dog’s femur. This prosthesis has a square stem, as
Studies carried out on the dynamic compressive properties of articular cartilage (Repo and Finlay, 1977; Finlay and Repo, 1978 a, b) led to the suggestion that a material similar to articular cartilage should be included in total joint replacement design. Such an elastomer material would function to damp peak dynamic loads applied to the bone interface, and thereby protect cement and its interfaces. This concept of introducing a low modulus material into joint replacement design has been, of course, discussed, and experiments have been conducted over several decades (Swanson, 1973). In current medical practice there are at least two human clinical trials underway incorporating lower modulus materials at the bone interface for hip implants (the Isoelastic RM Prosthesis and the Proplast@ Anaform@). There is, therefore, considerable current interest in low modulus arthroplosty for major weight bearing joints. The purpose of this present article is to describe the mechanical characteristics of a low modulus femoral replacement prosthesis which has already had a canine clinical trial (Repo et al., 198 1, in press). Characterization of such a low modulus prosthesis in an engineering sense is essential for the subsequent development of a human joint replacement which incorporates an elastomer for cushioning applied loads. METHODSAND MATERIALS Two basic models of prostheses having their stems seated in silastic elastomer contained in an outer shell
Received October 1984, in revised form June 1986.
Correspondence should he addressed to Dr R. U. Repo, Department of Surgery, St. Joseph’s Hospital, 50 Charlton Avenue East, Hamilton, Ontario, Canada L8N 4A6.
shown in Fig. 2a, embedded in a silastic cushion within an outer metal casing. Figure 2b shows an alternative design with a tapered stem of square cross-section with its surrounding silastic cushion and corresponding metal casing. In both designs the elastomer was 1.5 mm thick Dow-Corning grade silastic with a durometer hardness of 60 Shore A. This stiffness of silastic was chosen for the prostheses because at this stiffness the silastic most nearly resembles the stiffness of canine articular cartilage in dynamic compression. This similarity was established by performing high strain rate compression tests of 9 mm diameter samples of silastic in a drop tower apparatus at a controlled strain rate of 1000 S( - 1). The details of the test procedure have been fully described elsewhere including the stress-strain curves derived from compression of canine articular cartilage (Repo and Finlay, 1977; Finlay and Repo, 1978a). These elastomer embedded prostheses were designed for insertion in the distal end of a dog’s femur so as to replace a short section of the femur. The objectives of the animal study (Repo et al., 1981, in press) were to determine the clinical tolerance of an uncemented femoral replacement prosthesis. The results of clinical trials have already been reported (Repo et al., 1981, in press). The scope of this present report is restricted to the results of some mechanical testing of the silastic at high strain rates in compression and testing of these prostheses in torsion and bending at a low strain rate. The original design, as shown in Fig. 2a, had a collar of elastomer under the flange which did not stand up to the compression loads exerted on the prosthesis in oioo in dogs, and the design was therefore modified to provide a larger flange and a tapered stem, Fig. 2b. These were then used td evaluate bending and torsional load responses as would occur in a hip prosthesis and constrained knee prosthesis.
179
W. R. NEWCOMBE et al.
180
L
\
Silastic Cushion-l Smm \ Thick All Around
65~5mm-_3~S,3m~6~ommJ 64
I
1’
-52SmmF
-28.2mm-
iSilastic Cushion-l Smm Thick All Around
Stem Cross Section Tapers From 5.5mm x5.5mm to 8 7 x8.7mm 04
Fig. 2. (a) Original design for a prosthesis with a square stem and elastomer collar. Also seen in Fig. 1.
(b) Modified design. Prosthesis has a tapered stem designed to prevent subsidence by the tapered fit against the bone.
To measure the bending response, three samples each of the square and tapered stems were mounted in methylmethacrylate cement (PMMA) in a metal fixture. The advantage of testing the torsional and bending characteristics with the prosthesis embedded in PMMA was to avoid the effect of bone strain and other m~urement artefacts introdu~ by a test system with the prosthesis inside bone. Since the objective of these tests was to measure the load-deflection properties of the elastomer itself, the use of PMMA mounting proved advantageous. As shown in Fig. 3a, a force was applied perpendicular to the axis of the stem at the center of the pin joint, using a special adapter. The load was applied using an Instron TT-D floor model testing machine, and it is assumed that the reaction is a distributed linearly-varying compressive force system similar to that shown in Fig. 3b. The stems were loaded vertically perpendicular to the casing and the elastomer cushion was deflected at a cross head speed of OS1 mm s- ‘. This is a low strain rate, but it has the advantage of providing information on the real load carrying ability of the elastomer. It is known that viscoelastic materials such as silastic demonstrate increasing stiffnesses with increasing strain rates. ~nsequently, low strain rate testing allows the elastomer to reach maximum strain for a given load. Because of impact conditions and dynamic clinical joint loading, it is known that greater strain rates will occur in use. This is why the drop tower tests of silastic in compression were done to provide data on the dynamic properties of silastic used in these prostheses.
PROSTHESIS
LOAD APPLICATOR
CEMENT
I
’ INSTRON
k
LOAD.
CELL
I
BASE
\
Fig. 3. [a) Schematic drawing of the arrangement for testing the elastomer coated stems in bending. (b) Assumed reactive force system for calculating maximum deflection in elastomer.
Fig. 1. This X-ray photograph shows the prosthesis inserted in a dog’s femur. All femoral loads pass through the elastomer.
181
Fig. 4. IDevice used to apply a pure torsional load to prosthesis through the Ins&on testing met
182
Mechanical evaluation of femoral prosthesis The device shown in Fig. 4 was constructed to convert a vertical load applied by the Instron testing machine to a pure torsional load on the specimens. The shaft has two bearings, and the specimen was cemented into a holder which had sufficient freedom of motion so that the stem could be centered exactly on the shaft center-line. The stems were twisted through 20” at a rate of 0.12” s-l. This torsional rate is less than that expected in clinical conditions in the human, but as already pointed out, low strain rates were used in this study to determine the maximum deflection of the elastomer coating. The deflection was recorded on the integral machine recorder in inches, and mathematically converted to degrees.
183
7----l
160.0 3 s .g Q E 9 s
120.0
80.0
40.0
0. 0.
.200
.400
A00
.a00
1.00
DEFLECTION (mm) RESULTS
Results of controlled high strain rate compression testing of the silastic used here are illustrated in Fig. 9. Comparison of the curve profile in Fig. 9 to previously published stress-strain data on canine articular cartilage in compression (Finlay and Repo, 1978a) reveals surprising similarity. At 10 % compression at a strain rate of 1000 S( - 1) the average stress in the silastic and canine cartilage samples was in the range of 4 to 5 N mm - ‘. Thus these high strain rate tests confirm that the silastic as used in these prostheses had compressive behaviour very similar to articular cartilage when tested dynamically. Typical results for the bending tests on the straight stems are shown in Fig. 5. There is quite good agreement between the three samples tested. After an initial deflection of the elastomer cushion the response is fairly linear until about 33 % strain at point A in the cushion. At this strain the material begins to slip within the shell, and the volume is displaced from under the stem. The results then become artefactual. In practice one would not wish to design for more than 25 y0 strain on the elastomer and probably much less than that to limit the stress, thus increasing the time to failure. The tests on the tapered stems show a wider variation (Fig. 6). On further inspection this was found to be caused by poor centering of the stems within the shell with the resultant unequal thickness of the elastomer cushion. More consistent results could be achieved with a better control on the manufacturing process. Although, it is realized that the sample number is too small for statistical confidence, general trends and logical inference indicate that the straight stems are stiffer in bending than the tapered stems. The deflection plotted on the graphs is measured at the point of application of the load, and the maximum deflection in the elastomer, occurring at point A in Fig. 3b, was calculated to be about 77 y0 of the measured deflection by a simple moment calculation knowing the stem dimensions. In trying to give a physical interpretation of these results, the cushion material will undergo a maximum strain of about 25 % at point A in Fig. 3b when resisting a bending moment roughly equivalent
Fig. 5. Force-deflection curves of three. rectangular stem prostheses tested in bending as per Fig. 3.
2oor-----l
DEFLECTION (mm)
Fig. 6. Force-deflection curves of three tapered stem prostheses tested in bending as per Fig. 3.
to a 1 kg mass acting on the end of a lever 10 cm in
length. Fairly consistent results were again obtained for the straight stems under torsional loading as shown in Fig. 7, whereas, the tapered stems show a wider variation, Fig. 8. The tapered stems in general have a higher resistance to torque because the average crcsssection of the metal stem is larger providing a greater area on which the elastic cushion can act. The straight stems rotated 20” under a torque of approximately 5 Nm is equivalent to the torque exerted by a 1 kg mass sitting on the end of a lever 0.5 m in length. It is also equivalent to the approximate maximum torque that can be exerted by the human forearm in supination.
W. R.
NEWCOMBE et al.
DISCUSSION
5.00
0.
10.0
15.0
20.0
2
1
ANGULAR DEFLECTION (degrees)
Fig. 7. Torque-angutar deflection curves for three different rectangular stem prostheses tested in torsion as per Fig. 4.
.O-
.o .o .o -
0.
5.00
10.0
15.0
20.0
:
ANGULAR DEFLECTION (degrees)
Fig. 8. Torque-angular deflection curves for three different tapered stem prostheses tested in torsion as per Fig. 4. Note the much greater stiffness of these stems at maximum angular deflection compared to stems in Fig. 7.
50 E 40
STRAIN
mm
mJlcu mm
50 50 50 50
3.0 3.0 3.0 3.0
523E + 00 4MEt 00 4.52E +OO 4.35E *OO
B 2 30 8 z $
20 10 0 0
10
20 STRAIN
30
40
50
%
Fig. 9. Stress-strain curves of four samples of silaatic compressed at a strain rate of 1000 S( - 1) together with data on energy absorption.
Many joint replacement designs have been proposed, and are available for most joints in the human body. Total knee and hip joints are the most heavily loaded, and these are normally cemented in place. A major problem still exists because these joints invariably loosen at the bone-cement interface, especially in the younger patient where they undergo active use. Therefore, it is necessary to investigate other approaches to artificial joint design and installation to improve or minimize failure. Firstly, one should determine why loosening occurs. It is generally understood that loss of fixation with cemented implants is multifactorial involving prosthetic, geometric, material, biologic, surgical technique and patient factors. It is suggested, however, that one of the significant factors involved in loss of cement fixation is the difference in the stiffness of the implant materials (cement and metal) as compared to trabecular bone. In an effort to equalize the modulus of stiffness of the prosthetic implant, an Isoelastic prosthesis has been developed (Morshcer et al., 198 1) and a second generation isoelastic prosthesis model is currently in clinical trial (Bombelli and Mathys, 1982). A low modulus interface is also the objective behind the Proplast@ Anaform@ hip prosthesis which has a porous, sponge-like coating created for fibrous tissue ingrowth. Swanson (1973) described the use of silastic sheets at the interface of bone and modified McIntosh tibia1 plateau prosthesis in rheumatoid patients. Silastic has been introduced at the bone interface in other hip prosthesis experiments by Leidholt and Gorman (1968), Esslinger and Rutkowski (1973) and G. F. W. Price (1978). A silastic and polyethylene shock absorbing total hip has been reported in a human clinical trial by Ficat et al. (1984). Ficat’s hip does not place the elastomer at the bone interface but into the femoral neck area, but the idea of cushioning applied loads on the prosthesis is an integral part of his hip design. The aforementioned studies represent efforts to improve function of total joint replacements because contributing factors to knee and hip joint loosening may be the shock waves that are transmitted through the skeletal structure on heel strike. Light and coworkers (1980) have shown that significant spikes or deceleration transients occur in the acceleration curve measured in the tibia when the heel strikes the ground. These occur when the person is in bare feet or is wearing hard leather heels. It is believed that these shock waves may have deleterious effects on the paraosteal tissues, and produce back pain, osteoarthritic changes and cartilage fatigue (Hawes et al., 1979). An energy absorbing material called Sorbothane@ has been developed, and claimed to be effective in reducing peak forces and the rate of loading when it is included in the heels and soles of running shoes. Surgeons who are aware of this recommend its use in shoes for the management of some musculoskeletal diseases, and it
Mechanical evaluation of femoral prosthesis
should be a requirement for patients who have been fitted with artificial hip or knee joints. The prosthesis which we have investigated here utilizes a low modulus elastomer which closely matches the stiffness of articular cartilage. The importance of the damping properties of healthy articular cartilage on subchondral bone has been emphasized in studies by Radin et al., (1978, 1970), Radin and Paul (1970), as well as those by Repo andFinlay (1977) and Finlay and Repo (1978a, b, 1979). The particular prosthesis tested in this study can be considered an experimental analogue of a fully constrained stem type prosthesis such as a constrained hinge joint. Torsion and bending are principal modes of loading as well as compression, and in this experiment the torsional and bending loads are transmitted through the elastomer. The rationale for selecting these two modes was that this particular experimental prosthesis was not well suited for loading the elastomer in compression, and furthermore, the number of variables was reduced. It is of interest to note that the results of the mechanical tests suggest that a rather significant torsional strain would occur in the elastomer in clinical use in the canine femur based on the fact that a torque of only 5 N m produced a torsional deflection of 20”. The animal clinical results (Repo et al., 1981, in press) suggested no clinical torsional instability at all since the dogs were able to run and jump with no evident disability. It is, however, prudent to suggest that a stiffer torsional resistance be considered in any eventual human implant employing an elastomer to obviate the sense of rotational instability that could result. Experimental values of torsional and bending strain found in this study in the range of applied loads serve as an important benchmark from which a human joint prosthesis could be developed. SUMMARY
AND CONCLUSIONS
A low modulus canine segmental femoral replacement prosthesis embedded in an elastomer of similar stiffness to canine articular cartilage has been tested for torsional and bending characteristics. Previous animal clinical studies suggested good clinical function of the prosthesis to substitute for a length of the femur but the results of the mechanical studies performed here indicate a rather surprisingly large angular deflection in torsion in the elastomer. The results of these mechanical tests serve as a benchmark for the eventual design and development of a human joint replacement prosthesis which incorporates an elastomer to attenuate force applied to the bone or cement interfaces. Acknowledgements-The authors wish to thank Dr. J. B. Finlay for his contribution in testing the silastic at high strain
185
rates. This work was supported by a grant from The Canadian Medical Research Council, Grant No. MA-6674.
REFERENCES
Bombelli, R. and Mathys, R. (1982) Cementless isoelastic RM total hip prosthesis. J. R. Sot. Med., 75, 588-597. Esslinger, J. 0. and Rutkowski, E. J. (1973) Studies of the skeletal attachment of experiment hip prostheses in the pygmy goat and dog. J. Biomed. mat. Res. 7 (Symp. 4). 187-193. Ficat, P., Julia, F. and Boussaton, M. (1984) Prothese totale amortie et rotatoire. Rhumatologie 36, 313320. Finlay, J. B. and Repo, R. U. (1978a) Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Biomed.
BME-25, 3439.
Finlay, J. B. and Repo, R. U. (1978b) Cartilage impact: in vitro effect of bone and cement. J. Biomechanics 11, 379-388. Finlay J. B. and Repo, R. U. (1979) Energy absorbing ability of articular cartilage during impact. Med. Biol. Engng Comput. 17,397-403.
Hawes. D., Light, L. H. and Repond, E. (1979) Modelling the distortion produced by heel strike transients in soft tissue. J. PhysioL 296, l&l 1P. Leidholt, J. D. and Gorman, H. A. (1968) Use of silastic cushions on a metallic hip prosthesis in dogs. C/in. Orthop. 61, 164-166. Light, L. H., McLellan, G. E. and Klenerman, L. (1980) Skeletal transients on heel strike in normal walking with different footware. J. Biomechanics 13, 477-480. Morshcer, E., Bombelli, R., Schenk, R. and Mathys, R. (1981) The treatment of femoral neck fractures with an isoelastic endoprosthesis implanted without bone cement. Archs Orthop. Traum. Surg. 98, 93-100.
Price, G. F. W. (1978) Interim report on a modification of methylmethacrylate fixation of prosthesis into bone. J. Bone Jt Surg. 60-B, 293. Radin, E. L. and Paul, L. L. (1970) Does cartilage compliance reduce skeletal impact loads? The relative force.attenuating properties of articular cartilage, synovial fluid, periarticular soft tissues and bone. Arthritis Rheum. 13, 139-144. Radin, E. L., Paul, L. L. and Lowry, M. (1970) A comparison of the dynamic force transmitting properties of subchondral bone and articular cartilage. J. Bone Jr Surg. 52A, 44+456. Radin, E. L., Ehrlich, M. G., Chernack, R., Abernethy, P., Paul, L. L. and Rose, R. M. (1978) Effect of repetitive impulsive loading of the knee joints of rabbits. C/in. Orthop. 131, 288-293. Repo, R. U. and Finlay, J. B. (1977) Survival of articular cartilage after controlled impact. J. Bone Jr Surg. 59-A, 1068-1076. Repo, R. U., Newcombe, W. R., Belbeck, L. and Winfield, D. (1981) Interface studies ofan elastic canine femoral replacement prosthesis. Orthop. Trans. 5,471. Repo, R. U., Newcombe, W. R. and Winfield, D. (1983) An elastic canine total hi-an alternative method of total joint attachment to bone. Orthop. Trans. 7, 307-308. Rep, R. U., Newcombe, W. R. and Occhionorelli, J. (in press) An elastic femoral replacement prosthesis: an experimental study. Acta orthop. be/g. Swanson, A. B. (1973) Low modulus force-dampening materials for knee joint prostheses. Acta orthop. beig. 39, 116-137.
.I. Biomeehanics Vol. 20, No. 2. pp 179-185, 1987 Printed in Gnat Britain
A MECHANICAL
EVALUATION OF AN ELASTIC FEMORAL PROSTHETIC STEM
W. R. NEWCOMBE*, R. U. REPOt, N. B. EATON* and J. OCCHIONORELLI~ *Department of Mechanical Engineering, McMaster University, Hamilton, Ontario, Canada L8N 325: TDepartment of Surgery, Section of Orthopaedic Surgery, McMaster University Medical Centre, Hamilton, Ontario, Canada LSN 325 Abstract-Studies are being conducted in our laboratory to test the concept of introducing an elastomer to attenuate and damp forces applied to the hone interface in a major weighthearing joint replacement prosthesis. An analogue of a fully constrained intramedullary stem type prosthesis has heen developed in a segmental femoral replacement prosthesis of the dog. The layer of silastic was introduced to damp forces at the hone-prosthesis interface. This paper describes the response of this elastomer prosthesis to torsional and bending loads, and defines the upper limits of elastomer strain. The low modulus silastic displayed surprisingly low strain for applied loads, particularly in bending tests, in this prosthetic configuration. The results of these mechanical studies serve as a bench mark for the eventual design and material selection of an elastomer for human prosthetic use.
were manufactured. Figure 1 shows a prosthesis in situ
INTRODUCTION
in a dog’s femur. This prosthesis has a square stem, as
Studies carried out on the dynamic compressive properties of articular cartilage (Repo and Finlay, 1977; Finlay and Repo, 1978 a, b) led to the suggestion that a material similar to articular cartilage should be included in total joint replacement design. Such an elastomer material would function to damp peak dynamic loads applied to the bone interface, and thereby protect cement and its interfaces. This concept of introducing a low modulus material into joint replacement design has been, of course, discussed, and experiments have been conducted over several decades (Swanson, 1973). In current medical practice there are at least two human clinical trials underway incorporating lower modulus materials at the bone interface for hip implants (the Isoelastic RM Prosthesis and the Proplast@ Anaform@). There is, therefore, considerable current interest in low modulus arthroplosty for major weight bearing joints. The purpose of this present article is to describe the mechanical characteristics of a low modulus femoral replacement prosthesis which has already had a canine clinical trial (Repo et al., 198 1, in press). Characterization of such a low modulus prosthesis in an engineering sense is essential for the subsequent development of a human joint replacement which incorporates an elastomer for cushioning applied loads. METHODSAND MATERIALS Two basic models of prostheses having their stems seated in silastic elastomer contained in an outer shell
Received October 1984, in revised form June 1986.
Correspondence should he addressed to Dr R. U. Repo, Department of Surgery, St. Joseph’s Hospital, 50 Charlton Avenue East, Hamilton, Ontario, Canada L8N 4A6.
shown in Fig. 2a, embedded in a silastic cushion within an outer metal casing. Figure 2b shows an alternative design with a tapered stem of square cross-section with its surrounding silastic cushion and corresponding metal casing. In both designs the elastomer was 1.5 mm thick Dow-Corning grade silastic with a durometer hardness of 60 Shore A. This stiffness of silastic was chosen for the prostheses because at this stiffness the silastic most nearly resembles the stiffness of canine articular cartilage in dynamic compression. This similarity was established by performing high strain rate compression tests of 9 mm diameter samples of silastic in a drop tower apparatus at a controlled strain rate of 1000 S( - 1). The details of the test procedure have been fully described elsewhere including the stress-strain curves derived from compression of canine articular cartilage (Repo and Finlay, 1977; Finlay and Repo, 1978a). These elastomer embedded prostheses were designed for insertion in the distal end of a dog’s femur so as to replace a short section of the femur. The objectives of the animal study (Repo et al., 1981, in press) were to determine the clinical tolerance of an uncemented femoral replacement prosthesis. The results of clinical trials have already been reported (Repo et al., 1981, in press). The scope of this present report is restricted to the results of some mechanical testing of the silastic at high strain rates in compression and testing of these prostheses in torsion and bending at a low strain rate. The original design, as shown in Fig. 2a, had a collar of elastomer under the flange which did not stand up to the compression loads exerted on the prosthesis in oioo in dogs, and the design was therefore modified to provide a larger flange and a tapered stem, Fig. 2b. These were then used td evaluate bending and torsional load responses as would occur in a hip prosthesis and constrained knee prosthesis.
179
W. R. NEWCOMBE et al.
180
L
\
Silastic Cushion-l Smm \ Thick All Around
65~5mm-_3~S,3m~6~ommJ 64
I
1’
-52SmmF
-28.2mm-
iSilastic Cushion-l Smm Thick All Around
Stem Cross Section Tapers From 5.5mm x5.5mm to 8 7 x8.7mm 04
Fig. 2. (a) Original design for a prosthesis with a square stem and elastomer collar. Also seen in Fig. 1.
(b) Modified design. Prosthesis has a tapered stem designed to prevent subsidence by the tapered fit against the bone.
To measure the bending response, three samples each of the square and tapered stems were mounted in methylmethacrylate cement (PMMA) in a metal fixture. The advantage of testing the torsional and bending characteristics with the prosthesis embedded in PMMA was to avoid the effect of bone strain and other m~urement artefacts introdu~ by a test system with the prosthesis inside bone. Since the objective of these tests was to measure the load-deflection properties of the elastomer itself, the use of PMMA mounting proved advantageous. As shown in Fig. 3a, a force was applied perpendicular to the axis of the stem at the center of the pin joint, using a special adapter. The load was applied using an Instron TT-D floor model testing machine, and it is assumed that the reaction is a distributed linearly-varying compressive force system similar to that shown in Fig. 3b. The stems were loaded vertically perpendicular to the casing and the elastomer cushion was deflected at a cross head speed of OS1 mm s- ‘. This is a low strain rate, but it has the advantage of providing information on the real load carrying ability of the elastomer. It is known that viscoelastic materials such as silastic demonstrate increasing stiffnesses with increasing strain rates. ~nsequently, low strain rate testing allows the elastomer to reach maximum strain for a given load. Because of impact conditions and dynamic clinical joint loading, it is known that greater strain rates will occur in use. This is why the drop tower tests of silastic in compression were done to provide data on the dynamic properties of silastic used in these prostheses.
PROSTHESIS
LOAD APPLICATOR
CEMENT
I
’ INSTRON
k
LOAD.
CELL
I
BASE
\
Fig. 3. [a) Schematic drawing of the arrangement for testing the elastomer coated stems in bending. (b) Assumed reactive force system for calculating maximum deflection in elastomer.
Fig. 1. This X-ray photograph shows the prosthesis inserted in a dog’s femur. All femoral loads pass through the elastomer.
181
Fig. 4. IDevice used to apply a pure torsional load to prosthesis through the Ins&on testing met
182
Mechanical evaluation of femoral prosthesis The device shown in Fig. 4 was constructed to convert a vertical load applied by the Instron testing machine to a pure torsional load on the specimens. The shaft has two bearings, and the specimen was cemented into a holder which had sufficient freedom of motion so that the stem could be centered exactly on the shaft center-line. The stems were twisted through 20” at a rate of 0.12” s-l. This torsional rate is less than that expected in clinical conditions in the human, but as already pointed out, low strain rates were used in this study to determine the maximum deflection of the elastomer coating. The deflection was recorded on the integral machine recorder in inches, and mathematically converted to degrees.
183
7----l
160.0 3 s .g Q E 9 s
120.0
80.0
40.0
0. 0.
.200
.400
A00
.a00
1.00
DEFLECTION (mm) RESULTS
Results of controlled high strain rate compression testing of the silastic used here are illustrated in Fig. 9. Comparison of the curve profile in Fig. 9 to previously published stress-strain data on canine articular cartilage in compression (Finlay and Repo, 1978a) reveals surprising similarity. At 10 % compression at a strain rate of 1000 S( - 1) the average stress in the silastic and canine cartilage samples was in the range of 4 to 5 N mm - ‘. Thus these high strain rate tests confirm that the silastic as used in these prostheses had compressive behaviour very similar to articular cartilage when tested dynamically. Typical results for the bending tests on the straight stems are shown in Fig. 5. There is quite good agreement between the three samples tested. After an initial deflection of the elastomer cushion the response is fairly linear until about 33 % strain at point A in the cushion. At this strain the material begins to slip within the shell, and the volume is displaced from under the stem. The results then become artefactual. In practice one would not wish to design for more than 25 y0 strain on the elastomer and probably much less than that to limit the stress, thus increasing the time to failure. The tests on the tapered stems show a wider variation (Fig. 6). On further inspection this was found to be caused by poor centering of the stems within the shell with the resultant unequal thickness of the elastomer cushion. More consistent results could be achieved with a better control on the manufacturing process. Although, it is realized that the sample number is too small for statistical confidence, general trends and logical inference indicate that the straight stems are stiffer in bending than the tapered stems. The deflection plotted on the graphs is measured at the point of application of the load, and the maximum deflection in the elastomer, occurring at point A in Fig. 3b, was calculated to be about 77 y0 of the measured deflection by a simple moment calculation knowing the stem dimensions. In trying to give a physical interpretation of these results, the cushion material will undergo a maximum strain of about 25 % at point A in Fig. 3b when resisting a bending moment roughly equivalent
Fig. 5. Force-deflection curves of three. rectangular stem prostheses tested in bending as per Fig. 3.
2oor-----l
DEFLECTION (mm)
Fig. 6. Force-deflection curves of three tapered stem prostheses tested in bending as per Fig. 3.
to a 1 kg mass acting on the end of a lever 10 cm in
length. Fairly consistent results were again obtained for the straight stems under torsional loading as shown in Fig. 7, whereas, the tapered stems show a wider variation, Fig. 8. The tapered stems in general have a higher resistance to torque because the average crcsssection of the metal stem is larger providing a greater area on which the elastic cushion can act. The straight stems rotated 20” under a torque of approximately 5 Nm is equivalent to the torque exerted by a 1 kg mass sitting on the end of a lever 0.5 m in length. It is also equivalent to the approximate maximum torque that can be exerted by the human forearm in supination.
W. R.
NEWCOMBE et al.
DISCUSSION
5.00
0.
10.0
15.0
20.0
2
1
ANGULAR DEFLECTION (degrees)
Fig. 7. Torque-angutar deflection curves for three different rectangular stem prostheses tested in torsion as per Fig. 4.
.O-
.o .o .o -
0.
5.00
10.0
15.0
20.0
:
ANGULAR DEFLECTION (degrees)
Fig. 8. Torque-angular deflection curves for three different tapered stem prostheses tested in torsion as per Fig. 4. Note the much greater stiffness of these stems at maximum angular deflection compared to stems in Fig. 7.
50 E 40
STRAIN
mm
mJlcu mm
50 50 50 50
3.0 3.0 3.0 3.0
523E + 00 4MEt 00 4.52E +OO 4.35E *OO
B 2 30 8 z $
20 10 0 0
10
20 STRAIN
30
40
50
%
Fig. 9. Stress-strain curves of four samples of silaatic compressed at a strain rate of 1000 S( - 1) together with data on energy absorption.
Many joint replacement designs have been proposed, and are available for most joints in the human body. Total knee and hip joints are the most heavily loaded, and these are normally cemented in place. A major problem still exists because these joints invariably loosen at the bone-cement interface, especially in the younger patient where they undergo active use. Therefore, it is necessary to investigate other approaches to artificial joint design and installation to improve or minimize failure. Firstly, one should determine why loosening occurs. It is generally understood that loss of fixation with cemented implants is multifactorial involving prosthetic, geometric, material, biologic, surgical technique and patient factors. It is suggested, however, that one of the significant factors involved in loss of cement fixation is the difference in the stiffness of the implant materials (cement and metal) as compared to trabecular bone. In an effort to equalize the modulus of stiffness of the prosthetic implant, an Isoelastic prosthesis has been developed (Morshcer et al., 198 1) and a second generation isoelastic prosthesis model is currently in clinical trial (Bombelli and Mathys, 1982). A low modulus interface is also the objective behind the Proplast@ Anaform@ hip prosthesis which has a porous, sponge-like coating created for fibrous tissue ingrowth. Swanson (1973) described the use of silastic sheets at the interface of bone and modified McIntosh tibia1 plateau prosthesis in rheumatoid patients. Silastic has been introduced at the bone interface in other hip prosthesis experiments by Leidholt and Gorman (1968), Esslinger and Rutkowski (1973) and G. F. W. Price (1978). A silastic and polyethylene shock absorbing total hip has been reported in a human clinical trial by Ficat et al. (1984). Ficat’s hip does not place the elastomer at the bone interface but into the femoral neck area, but the idea of cushioning applied loads on the prosthesis is an integral part of his hip design. The aforementioned studies represent efforts to improve function of total joint replacements because contributing factors to knee and hip joint loosening may be the shock waves that are transmitted through the skeletal structure on heel strike. Light and coworkers (1980) have shown that significant spikes or deceleration transients occur in the acceleration curve measured in the tibia when the heel strikes the ground. These occur when the person is in bare feet or is wearing hard leather heels. It is believed that these shock waves may have deleterious effects on the paraosteal tissues, and produce back pain, osteoarthritic changes and cartilage fatigue (Hawes et al., 1979). An energy absorbing material called Sorbothane@ has been developed, and claimed to be effective in reducing peak forces and the rate of loading when it is included in the heels and soles of running shoes. Surgeons who are aware of this recommend its use in shoes for the management of some musculoskeletal diseases, and it
Mechanical evaluation of femoral prosthesis
should be a requirement for patients who have been fitted with artificial hip or knee joints. The prosthesis which we have investigated here utilizes a low modulus elastomer which closely matches the stiffness of articular cartilage. The importance of the damping properties of healthy articular cartilage on subchondral bone has been emphasized in studies by Radin et al., (1978, 1970), Radin and Paul (1970), as well as those by Repo andFinlay (1977) and Finlay and Repo (1978a, b, 1979). The particular prosthesis tested in this study can be considered an experimental analogue of a fully constrained stem type prosthesis such as a constrained hinge joint. Torsion and bending are principal modes of loading as well as compression, and in this experiment the torsional and bending loads are transmitted through the elastomer. The rationale for selecting these two modes was that this particular experimental prosthesis was not well suited for loading the elastomer in compression, and furthermore, the number of variables was reduced. It is of interest to note that the results of the mechanical tests suggest that a rather significant torsional strain would occur in the elastomer in clinical use in the canine femur based on the fact that a torque of only 5 N m produced a torsional deflection of 20”. The animal clinical results (Repo et al., 1981, in press) suggested no clinical torsional instability at all since the dogs were able to run and jump with no evident disability. It is, however, prudent to suggest that a stiffer torsional resistance be considered in any eventual human implant employing an elastomer to obviate the sense of rotational instability that could result. Experimental values of torsional and bending strain found in this study in the range of applied loads serve as an important benchmark from which a human joint prosthesis could be developed. SUMMARY
AND CONCLUSIONS
A low modulus canine segmental femoral replacement prosthesis embedded in an elastomer of similar stiffness to canine articular cartilage has been tested for torsional and bending characteristics. Previous animal clinical studies suggested good clinical function of the prosthesis to substitute for a length of the femur but the results of the mechanical studies performed here indicate a rather surprisingly large angular deflection in torsion in the elastomer. The results of these mechanical tests serve as a benchmark for the eventual design and development of a human joint replacement prosthesis which incorporates an elastomer to attenuate force applied to the bone or cement interfaces. Acknowledgements-The authors wish to thank Dr. J. B. Finlay for his contribution in testing the silastic at high strain
185
rates. This work was supported by a grant from The Canadian Medical Research Council, Grant No. MA-6674.
REFERENCES
Bombelli, R. and Mathys, R. (1982) Cementless isoelastic RM total hip prosthesis. J. R. Sot. Med., 75, 588-597. Esslinger, J. 0. and Rutkowski, E. J. (1973) Studies of the skeletal attachment of experiment hip prostheses in the pygmy goat and dog. J. Biomed. mat. Res. 7 (Symp. 4). 187-193. Ficat, P., Julia, F. and Boussaton, M. (1984) Prothese totale amortie et rotatoire. Rhumatologie 36, 313320. Finlay, J. B. and Repo, R. U. (1978a) Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Biomed.
BME-25, 3439.
Finlay, J. B. and Repo, R. U. (1978b) Cartilage impact: in vitro effect of bone and cement. J. Biomechanics 11, 379-388. Finlay J. B. and Repo, R. U. (1979) Energy absorbing ability of articular cartilage during impact. Med. Biol. Engng Comput. 17,397-403.
Hawes. D., Light, L. H. and Repond, E. (1979) Modelling the distortion produced by heel strike transients in soft tissue. J. PhysioL 296, l&l 1P. Leidholt, J. D. and Gorman, H. A. (1968) Use of silastic cushions on a metallic hip prosthesis in dogs. C/in. Orthop. 61, 164-166. Light, L. H., McLellan, G. E. and Klenerman, L. (1980) Skeletal transients on heel strike in normal walking with different footware. J. Biomechanics 13, 477-480. Morshcer, E., Bombelli, R., Schenk, R. and Mathys, R. (1981) The treatment of femoral neck fractures with an isoelastic endoprosthesis implanted without bone cement. Archs Orthop. Traum. Surg. 98, 93-100.
Price, G. F. W. (1978) Interim report on a modification of methylmethacrylate fixation of prosthesis into bone. J. Bone Jt Surg. 60-B, 293. Radin, E. L. and Paul, L. L. (1970) Does cartilage compliance reduce skeletal impact loads? The relative force.attenuating properties of articular cartilage, synovial fluid, periarticular soft tissues and bone. Arthritis Rheum. 13, 139-144. Radin, E. L., Paul, L. L. and Lowry, M. (1970) A comparison of the dynamic force transmitting properties of subchondral bone and articular cartilage. J. Bone Jr Surg. 52A, 44+456. Radin, E. L., Ehrlich, M. G., Chernack, R., Abernethy, P., Paul, L. L. and Rose, R. M. (1978) Effect of repetitive impulsive loading of the knee joints of rabbits. C/in. Orthop. 131, 288-293. Repo, R. U. and Finlay, J. B. (1977) Survival of articular cartilage after controlled impact. J. Bone Jr Surg. 59-A, 1068-1076. Repo, R. U., Newcombe, W. R., Belbeck, L. and Winfield, D. (1981) Interface studies ofan elastic canine femoral replacement prosthesis. Orthop. Trans. 5,471. Repo, R. U., Newcombe, W. R. and Winfield, D. (1983) An elastic canine total hi-an alternative method of total joint attachment to bone. Orthop. Trans. 7, 307-308. Rep, R. U., Newcombe, W. R. and Occhionorelli, J. (in press) An elastic femoral replacement prosthesis: an experimental study. Acta orthop. be/g. Swanson, A. B. (1973) Low modulus force-dampening materials for knee joint prostheses. Acta orthop. beig. 39, 116-137.