Strain Analysis of the Proximal Femur After Total Hip Replacement

http://www.aboutjoints.com

Strain Analysis of the Proximal Femur After Total Hip Replacement

In: Quantitative Characterization and Performance of Porous Implants
for Hard Tissue Application

by Hank C. K. Wuh,1 Lynne C. Jones,2 and David S. Hungerford3

1Medical student, Johns Hopkins University School of Medicine, Baltimore, MD 21239.

2Research associate, Orthopaedic Surgery, Johns Hopkins University School of Medicine, Baltimore, MD 21239.

3Associate professor, Orthopaedic Surgery, Johns Hopkins University School of Medicine,
and Chief, Division of Arthritis Surgery, Good Samaritan Hospital, Baltimore, MD 21239.

Reprinted with permission from STP 953 Quantitative Characterization and Performance of Porous Implants for Hard Tissue Applications, copyright ASTM International, 100 Bar Harbor Drive, West Conshohocken, PA 19428

REFERENCE: Wuh, H. C. K., Jones, L. C., and Hungerford, D. S., "Strain Analysis of the Proximal Femur After Total Hip Replacement," Quantitative Characterization and Performance of Porous Implants for Hard Tissue Applications, ASTM STP 953, J. E. Lemons, Ed., American Society for Testing and Materials, Philadelphia, 1987, pp. 249—263.

ABSTRACT: Strain analysis of human cadaver femora after cemented total hip arthroplasty (THA) has demonstrated a reduction in stress transfer along the proximal femur. The principal objective of this study was to determine the effect of the cementless application of a press-fit, porous-coated prosthesis on the strain experienced by the proximal femur. Using the photoelastic coating technique (PECT), five human cadaver specimens were subjected to strain analysis before and after cementless arthroplasty with a PCA total hip prosthesis. After total hip replacement, the strain magnitudes were reduced for all points along the medial border when the femur was subjected to loading conditions. A reduction of the level of strain ex­perienced by the calcar ranged from 34.7 to 43.7% under loads ranging from 750 to 2000 N— a considerably smaller reduction than that reported by previous investigators. The region of the greater trochanter was the only area of the lateral surface to demonstrate an increase in strain magnitude after THA; the other, more distal points laterally experienced a reduced level of strain. Increases in strain magnitude, although not statistically significant, were detected along the anterior aspect of the femora. Significant decreases in strain were observed at the two more distal points posteriorly, with no significant change proximally. As this investigation is an evaluation only of the immediate effect of the design of the prosthesis in achieving a press fit, and the specimens are without the benefit of bony ingrowth, additional studies are necessary to determine the effect of biologic ingrowth on the distribution of strain within the proximal femur.

Key Words: porous implants, stress transfer, total hip replacement, porous-coated prostheses, strain analysis, photoelastic coating technique, stress shielding

One of the most important limitations to the long-term success of total hip replacement (THR) is aseptic loosening of the prosthesis. For this reason, a great deal of interest has developed in the application of porous-surfaced implants for cementless THR. The early reported clinical results for cementless fixation have been favorable, both with the macro-porous designs widely used in Europe [1-5] and with the microporous designs extensively studied by a number of investigators in the United States [6-10].

A major concern in the clinical usage of porous-surfaced implants is proximal stress transfer from the femoral prosthesis to the surrounding cortical bone. Sir John Charnley [11] in 1965 and McKee [12] in 1970 have postulated that the cement between the bone and prosthesis distributes the load over a wider area and thus greatly reduces the stress translated to the cortical bone. The design of porous-surfaced implants for biological fixation in the absence of cement may therefore significantly affect bone stresses and subsequent bone remodeling.

The potential for the development of secondary osteoporosis due to significant stress shielding and the resulting loosening of the bone/porous implant interface is a well-recognized phenomenon [13-19]. Engh and Bobyn [20] have evaluated adaptive femoral bone modeling following the biologic fixation of porous-surfaced (modified Moore) prostheses through radiographic analysis. However, direct in vitro stress analysis of the proximal femur following arthroplasty with a porous-coated prosthesis has not been reported. Such analysis may yield important information regarding the initial stress transfer from the porous prosthesis to the surrounding cortical bone and serve to predict subsequent adaptive bone remodeling.

As the clinical experience with cementless THR is limited, there is a lack of specimens retrieved from deceased patients. Therefore, stress analyses of femora implanted with porous prostheses stabilized with biologic ingrowth have not been reported. In vitro studies using fresh cadaver femora should simulate the response to loading of the operated femur after initial implantation. Although the findings would be limited to this application, the information generated should reflect the mechanical response of bone to implantation of a press-fit prosthesis.

The purpose of this study was to evaluate the stress transfer in the proximal femur after the cementless implantation of a porous-surfaced femoral component. Traditionally, four methods have been used to evaluate stress in the loaded proximal femur: (1) theoretical mathematical analysis (including finite element analysis) [21-27], (2) the stress coat (brittle coating) technique [28,29], (3) photoelastic models [30,31], and (4) strain gages [32-37]. However, each of these methods have limitations in their applications to the study of bone, including directional and positional constraints, assumptions of homogeneity, and sensitivity. One technique that overcomes these limitations is the photoelastic coating technique (PECT). Its use toward the study of freshly retrieved bony tissues has only recently been reported by our laboratory [38,39] and will be detailed in the section on Materials and Methods. To carry out our study, physiologic loading conditions simulating a single-limb stance were applied to five freshly retrieved human cadaver femurs. Strain measurements were made using the PECT, first with the femoral head intact and then after cementless implantation of a PCA femoral component.

Materials and Methods

Five freshly retrieved, intact, human cadaver femurs were frozen immediately upon removal at 20°C and thawed to room temperature just prior to testing. It has been established that freezing has no harmful effect on the physical properties of femoral cortical bone [40]. All five femurs were roentgenographically normal. The soft tissues and the periosteal layer were thoroughly debrided. The surface was gently smoothed with fine finishing sandpaper and then wiped with gauze soaked in 30% ethyl alcohol to degrease the surface and to remove any loose particles.

Photoelastic Coating Technique

The photoelastic coating technique (PECT) involves basically four steps: (1) the casting and contouring of photoelastic plastic sheets to the testing area in this case, the proximal femur; (2) sealing the bone surface with a precoating of bonding cement; (3) bonding of the contoured photoelastic plastic to the proximal femur; and (4) measurement of the strain under load, using a reflection polariscope and an attached digital compensator (see Table 1).

TABLE 1 - The photoelastic coating technique
Step No.
Procedure
Experimental Day
1
casting and contouring
1
2
sealing
2
3
bonding
2
4
measurement
3

Casting and Contouring

A room-temperature-curing, two-component resin/hardener system, PL-1 (Measurements Group, Raleigh, NC), was used for making contourable photoelastic plastic sheets. It has a Poisson’s ratio of 0.36, a modulus of elasticity of 0.42 x 106, and a K factor (strain optical constant) of 0.10. The PL-1 was prepared according to instructions [41]. The photoelastic epoxy resin was applied to the proximal femur using two PL-1 photoelastic sheets with a gap of approximately 1 to 2 mm between the two sheets along the anterior and posterior surfaces of the femur. [The seam did not cross the points of measurement.] Care was taken not to stretch the plastic and to ensure that it was in complete contact with the surface of the proximal femur. Once contoured to the shape of the proximal femur, the PL-1 was allowed 18 h to complete its polymerization cycle. At the end of the cycle, the PL-1 was solid and was of the same size and detailed contour as the proximal femur. It was then carefully removed from the proximal femur, and measurements of the coating thickness were made along approximately 1.5 to 2.0 cm intervals using a Starrett micrometer. The thickness of the coating used during this study averaged 1.87 mm.

Sealing

A two-component resin/hardener curing adhesive, PC-1 (Measurements Group, Raleigh, NC), was carefully applied to the proximal femur in a very thin layer to plug the pores of the cortical wall. After the PC-1 had dried for 2 h, the surface of the proximal femur was prepared with fine finishing sandpaper to remove the excess PC-1, thus allowing it to remain only in the pores. This step prevented moisture from escaping during the subsequent bonding procedures.

Bonding

The surface of the sealed proximal femur was cleaned with gauze soaked in 30% ethyl alcohol. The PC-1 was prepared according to instructions and was brushed over the surface of the proximal femur in a thin uniform layer [41]. The contoured PL-1 photoelastic plastic was then carefully placed over the adhesive. Care was taken to ensure total contact between the plastic and the adhesive without underlying air pockets being present. The adhesive was allowed to cure for 12 h, after which the femur was ready for strain measurement.

Measurement

Although the technique of photoelastic coating is capable of measuring the stress distribution over the entire surface of the proximal femur, particular reference points along the femur were noted for comparison with results of other stress analysis studies previously reported. The points were marked with a wax pencil over the photoelastic coating at the calcar, the subtrochanteric region, and the area approximating the distal portion of the stem along the medial, lateral, anterior, and posterior borders. The reference points were at similar locations for each femur tested, for they were determined by the specific ratio of the distance from the femoral head to the reference point over the length of the femur. The ratio was established by the first femur tested.

When load is applied to the femoral head and the femur is viewed under polarized light, the bonded photoelastic coating experiences strain corresponding to that experienced by the underlying bone, and a characteristic colored fringe pattern can be detected. Lines of similar strain, isoclinic lines, can then be used to determine the direction of the principal strains. A digital compensator attached to a reflection polariscope, the source of polarized light, can then be used to measure the magnitude of the strain on the basis of the colored patterns, called isochromatics or fringes. The value obtained is the fringe order N and is directly proportional to the difference between the principal strains of the proximal femur established by the formula.

Єx - Єy =N λ
              

where

Єx - Єy =
 difference in principal strains,
N =
 fringe order = δ/λ ,
δ =
light retardation, in.,
λ =
the wavelength of light (22.77 x 10-6 in.),
t =
 plastic coating thickness, and
K =
sensitivity of the plastic coating, the K factor (0.10).


The thickness of the PL-1 photoelastic coating t has been measured for all specimens tested and found to be a relatively constant value. K and λ are constants. Therefore, the fringe order N is directly proportional to the difference in principal strains experienced by the proximal femur. We have chosen to report our results as changes in fringe order. However, to convert from N units to the difference in principal strains in microstrain units for this study, the fringe order value can be multiplied by 1.54 x 10-3. As the purpose of this study was to seek comparisons rather than absolute values, the data have also been presented as percentage changes (Fig. 1). In a previous study, our group found a highly significant correlation between changes in the difference in principal strains (ε1 - ε2) and the changes in the magnitude of axial strain (r = 0.996, P < 0.001) [39]. We have, therefore, chosen to use the photoelastic technique as an indirect method of detecting changes in the principal axial strain experienced by the femur. The implications of this are outlined in the section titled Discussion.

Testing Conditions

Each intact femur was fixed distally in a specially designed jig for stabilization. Using a goniometer, the femoral shaft was positioned at a 9° angle from the vertical to simulate the proper anatomical axjs. While viewing the medial aspect of the femur, the femoral head was placed directly above the medial epicondyle. Each femur was mounted in the testing rig of a MMED servohydraulic, computer-linked materials tester (Matco, La Canada, CA). Load was applied to the vertical axis in 100 to 200-N increments with strain measurements taken at 750, 1500, and 2000 N. The loading procedures and measurements were repeated for each femur. When strain measurements in the intact femurs were completed, each femur was prepared for the cementless implantation of a PCA femoral component. Photographs were taken during each point of measurement.

FIG. 1 - Summary of averaged data illustrating the change in the magnitude of strain in the proximal femur. The ratios are those of the strain values in the femur after cememtless PCA femoral arthroplasty to the strain values in the intact femur. Data are presented for reference points along the medial and lateral borders under loading conditions 750, 1500, and 2000 N. These cumulative data are based on the percentage changes for each individual femur, which served as its own control. The statistical significance of changes between the intact and implanted femora were determined using two-way analysis of variance for related measurements, with linear contrast methods used for the comparison of groups. NS indicates no significance.

PCA Implantation

Roentgenograms of each femur were examined for proper sizing of the porous femoral implant. Using the PCA total hip instrumentation system, cementless implantation of a PCA femoral component of the appropriate size, in the neutral position, was performed on each femur. Roentgenograms were then taken to confirm proper positioning. The photoelastic plastic coating around the proximal femur was entirely undisturbed. The femur was then tested under conditions identical to those used for the intact femur.

Statistical Analysis

Each femur was tested twice under each loading condition. As the results for each test were highly reproducible, these values were averaged. Changes in strain magnitude subsequent to the total hip replecement were analyzed using two-way analysis of variance (ANOVA) for related measurements. Comparisons between groups were made using linear contrast methods.

Results

Strain Distribution of the Intact Femur

Figure 2 compares the strain distribution in the proximal femur before (n = 5) and after (n = 4) cementless femoral arthroplasty for reference points along the medial and lateral borders. Along the medial wall, two of the fine intact femurs tested had a strain magnitude at the calcar greater than that at the subtrochanteric region; three of the five femurs held the reverse relationship. Both conditions have been reported in the past [23,24,34-37]. The magnitude of the distal point was the lowest value on the medial aspect in four of the five femora tested. Laterally, the subtrochanteric region was fairly consistent in having the greatest magnitude of strain, whereas the trochanteric region displayed the lowest strain values.

Figure 3 presents data for reference points along the anterior and posterior borders under the same conditions as for Fig. 2. Along the anterior margin, three of the five specimens demonstrated maximal strain magnitudes at the most proximal point, while two specimens had maximal values at the distal point. For the posterior border, the largest values were seen at the point most distal.

Fig. 2 - Bar graph illustrating the strain distribution in the intact femur and the strain distribution after cementless implantation of the femoral component. Values at various reference points along the medial and lateral borders are presented. Each femur served as its own control. The load was 2000 N. Note that, laterally, the subtrochanteric region has the greatest strain magnitude and than the trochanteric region is under tension. Medially,, no consistent relationship exists between the strain magnitude at the calcar and that at the subtrochanteric region in the intact femur.

FIG. 3 - Bar graph illustrating the strain distribution in the intact femur and the strain distribution after cementless implantation of the femoral prosthesis. Values at various reference points along the anterior and posterior borders are presented. Each femur served as its own control. The load was 2000 N.

The increase in strain magnitude in response to increasing load along the medial and lateral borders of the intact femur is illustrated in Fig. 4. To demonstrate the response to increasing loads at the calcar, Fig. 5 presents data for each femur tested.

FIG. 4 - Graph of averaged strain values for all intact femurs tested. The strain values at the various reference points along the medial and lateral borders appear to increase proportionately with the increase in load, with no abrupt changes noted up to 2000 N.

FIG. 5 - Graph illustrating the response of the calcar to an increase in load before cementless PCA arthroplasty. The strain magnitude at the calcar increases in proportion to the increase in load for the intact femur.

Strain Distribution After Cementless Femoral Arthroplasty

As demonstrated for the intact femur, the points along the medial margin demonstrated the largest measurements for strain magnitude. In contrast, while the strain values of the anterior border were the smallest found for the intact femur, the lateral margin displayed the lowest values following arthroplasty. Furthermore, along the medial and anterior borders, there was a shift in the point of greatest magnitude distally. The relationships along the lateral and posterior borders were unchanged.

As is the case with previous reports [36,37], there is a great deal of individual variation in strain magnitude and distribution from one femur to the next. For this reason, the changes in strain magnitude subsequent to cementless total hip arthroplasty (THA) are presented as a percentage of the values obtained for the intact femur (Fig. 1). The magnitude of strains exhibited by the loaded proximal femora were shown to decrease after cementless hip arthroplasty for most of the points tested. All points along the medial margins experienced lower strain magnitudes; the strains for the more distal points posteriorly and the subtrochantenc and distal points along the lateral border were also reduced. Three of the four implanted femora demonstrated an increase in strain subsequent to surgery at the greater trochanteric region. However, the level of change was not statistically significant.

Figure 6 shows that the strain response to increased loading remains approximately linear after cementless femoral arthroplasty for points along the medial and lateral margins. However, the change in response (the slope of the line) to the increased load is decreased. This is particularly apparent at the medial calcar. Figure 7 illustrates the strain response to increasing loads at the calcar after cementless THR. Again, although strain appears to increase proportionately with load, the level of increase, that is, the slope of the line, is diminished.

FIG. 6 - Graph of averaged strain values for all femurs tested after cementless implantation of the femoral component. Though the strain response to increasing load remains approximately linear, the rate of increase, or stress transfer, is diminished. The decrease is particularly noticeable at the calcar.

FIG. 7- Graph showing the level of strain experienced at the calcar after total hip arthroplasty, which increased as the femur was subjected to increasing load. However, the level of the increase in strain was of a lower magnitude than that seen for the intact femur.

Discussion

A major concern over biological fixation is the proximal stress transfer from prosthesis to bone and the potential for both pathological stress transfer and significant stress shielding. A particular area of controversy in cemented THR has been bone resorption in the calcar region and its relationship to prosthesis loosening. It is known that the calcar frequently resorbs some years after femoral component implantation in up to 70% of the cases [42-46]. Although significant stress reduction has generally been considered the mechanism of resorption at the calcar [34,36,37,47] a possible decrease in the blood supply [42,45], thermal necrosis [48], and foreign-body reaction to nonpolymerized monomers of bone cement [49] have been postulated as causes. Opinions differ as to whether this type of resorption is causally related to the loosening of prostheses [43,45,47,50].

As reported in previous laboratory studies, there was a significant decrease in the strain magnitude at the calcar following THR. Oh [36] and Oh and Harris [37] have reported greater than 90% reduction in strain values at the calcar after cemented THR (load = 2012 N). McBeath reported reductions greater than 75% (load = 445 N) [34], and Crowninshield reported up to 91% reduction for a collarless steel prosthesis and 50 to 60% reduction for a prosthesis with a collar design (load = 1000 N) [47]. Data from our study reveal a strain magnitude reduction of 34.7% at 750 N, 38.5% at 1500 N, and 43.7% at 2000 N at the calcar after cementless femoral arthroplasty with a press-fit, porous-surfaced prosthesis. These reductions are statistically significant (Fig. 1). Our findings indicate a substantially smaller reduction in strain than all previous reports for cemented THR. As our experimental technique measures the difference between the principal strains (εx  - εy) the significance of our results must be further discussed. Based on the principles of photoelasticity, a decrease of the difference in principal strains in a region known to be subjected to compressive strains is a reflection of either (1) a decrease in the hoop strains, (2) a decrease in the level of axial strain, or (3) a decrease in the hoop strains concomitant with a decrease in axial strain. As the prosthesis used in this study was a press-fit design, we suggest that our findings for the calcar reflect a decrease in the axial strain [34,39]. However, additional studies using rosette strain gages are needed to validate our results further.

Radiographic evidence of endosteal bone formation near the distal portion of the stem, indicating significant stress transfer from implant to bone, has been reported for biologic fixation [20]. In our study, we found an increase in strain at the anterior margin and strain decreases at the other margins. The increase at the anterior point distally was not statistically significant. However, the decreases in strain at the medial (13.5%) and posterior (22.4%) points distally were significant (P < 0.01 at 2000 N). Oh [36] and Oh and Harris [37] have described a gradient of strain distribution along the proximal femur with the greatest strain values at the calcar, which decrease as one moves distally in the intact femur; after THR, the distribution is reversed. In our study, the region of the calcar demonstrated the largest strain values in three of the five intact specimens. Along the medial margin, the magnitude of strain decreased as the distal point was approached. After THR, this relationship was reversed in two of the three specimens. However, these relationships have not been a universal finding, according to other previous studies [23,24].

The cadaver specimens tested for this in vitro study were obviously without the benefit of bony ingrowth. Therefore, the relatively smaller magnitudes in strain change after THR are a reflection of the press-fit design and not necessarily a result of the porous-coated surface. Theoretically, while cemented THR achieves maximum stability at time of surgery, cementless THR with porous-surfaced implants will reach optimum stability over time. If strain values were obtained after the process of bony ingrowth had been in progress, one would suspect that the strain values would be different and that the proximal femur would share a greater portion of the applied load. This may further limit the significant decrease in the strain magnitude at the calcar and along the medial borders. Future studies evaluating specimens retrieved from deceased patients into which porous-coated prostheses have been implanted would give further insight into these processes.

Acknowledgments

We wish to express our gratitude to the Anatomy Board of Maryland for making available to us the cadaver specimens used in this study.

References

  1. Judet, R., Siquier, M., and Brumpt, B., Clinical Orthopaedics and Related Research, Vol. 137, 1978, pp. 76-84.

  2. Judet, R., “Rough-Surfaced Total Hip Prosthesis Without Cement,” Orthopaedic Transactions, Vol. 5, 1981, pp. 382-383.

  3. Lord, G. and Bancel, P., “The Madreporic Cementless Total Hip Arthroplasty: New Experimental Data and a Seven-Year Clinical Follow-Up Study,” Clinical Orthopaedics and Related Research, Vol. 176, 1983, pp. 67-76.

  4. Lord, G. A., “Bone Ingrowth and Uncemented Total Hip Replacement With Madreporic Stemmed Prostheses,” Orthopaedic Transactions, Vol., 1981, pp. 383-384.

  5. Lord, G. A., Hardy, J. R., and Kummer, F. J., Clinical Orthopaedics and Related Research, Vol. 141, 1979, pp. 2-16.

  6. Cameron, H. U., Clinical Orthopaedics and Related Research, Vol. 165, 1982, pp. 188-190.

  7. Engh, C. A., Clinical Orthopaedics and Related Research, Vol. 176, 1983, pp. 52-66.

  8. Engh, C. A., Bobyn, J. D., and Gorski, J. M., Orthopaedics, Vol. 7, 1984, pp. 285-298.

  9. Hungerford, D.S., Hedley, A., Habermann, E., Borden, L., and Kenna, R. V., “Clinical Results with the PCA Hip,” Total Hip Arthroplasiy: A New Approach, University Park Press, Baltimore, 1984, Chapter 8, pp. 170-180.

  10. Ryan, G. M. and Brosher, A. P., Orthopaedics, Vol. 3, 1980, p. 660.

  11. Charnley, J., Journal of Bone and Joint Surgery, Vol. 47B, 1965, pp. 354-363.

  12. McKee, 0. K., Clinical Orthopaedics and Related Research, Vol. 72, 1970, pp. 88-103.

  13. Bobyn, J.D., Cameron, H.U., Abdulla, D., Pilliar, R.M., and Weatherly, G.C., Clinical Orthopaedics and Related Research, Vol. 166, 1982, pp. 301-312.

  14. Galante, J. 0., “Overview of Current Attempts to Eliminate Methylmethacrylate,” The Hip: Proceedings, Eleventh Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1983, pp. 181-189.

  15. Hedley, A. K., Clarke, I.C., and Kozinn, S.C., Clinical Orthopaedics and Related Research, Vol. 163, 1982, pp. 300-311.

  16. Morscher, E.W., “European Experience with Cementless Hip Replacements,” The Hip: Pro­ceedings, Eleventh Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1983, pp. 190-203.

  17. Pilliar, R.M., Journal of Orthopaedic Research, Vol. 1, 1983, pp. 189-234.

  18. Pilliar, R.M., Cameron, H.U., Birmington, A.G., Szinek, J., and Macnah, I., Journal of Biomedical Materials Research, Vol. 13, 1979, pp. 799-810.

  19. Tonino, A. J., Davidson, C.L., Kloffer, P.J., and Linclan, C.A., Journal of Bone and Joint Surgery, Vol. 58B, pp. 107-113.

  20. Engh, C. A. and Bobyn, J.D., “Evaluation of Adaptive Femoral Bone Modeling,” The Hip: Proceedings,Twelfth Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1984, pp. 110—132.

  21. Cook, S. D., Skinner, H. B., Weinstein, A. M., and Haddard, R. J., “Stress Distribution in the Proximal Femur After Surface Replacement: Effects of Prostheses and Surgical Techniques,” Biomaterials, Medical Devices, and Artificial Organs, Vol. 10, 1982, pp. 85-102.

  22. Crowninshield, R. D., Brand, R. A., and Johnston, R. C., Journal of Bone and Joint Surgery, Vol. 62A, 1980, pp. 68-78.

  23. Koch, J. C., American Journal of Anatomy, Vol. 21, 1972, pp. 177-298.

  24. Rybicki, E.F., Siruonen, F.A., and Weis, E.B., Jr., Journal of Biomechanics, Vol. 5, 1972, pp. 203—215.

  25. Skinner, H.B., Cook, S.O., Weinstein, A.M., and Haddad, R.J., Clinical Orthopaedics and Related Research, Vol. 166, 1982, pp. 277-283.

  26. Svensson, N.C., Valliaffan, S., and Wood, R.D., Journal of Biomechanics, Vol. 10, 1977, pp. 581-588.

  27. Toridis, T.G., Journal of Biomechanics, Vol. 2, 1969, pp. 163-174.

  28. Eva6s, F.G. and Lisner, H.R., “‘Stress Coat’ Deformation Studies of the Femur Under Static Vertical Loading,” Anatomy Research, Vol. 100, 1948, pp. 159-190.

  29. Kalen, R., “Strain and Stresses in the Upper Femur Studied by the Stress Coat Method,” Acta Orthopaedica Scandinavica, Vol. 31, 1961, pp. 103-113.

  30. Milch, H., Journal of Bone and Joint Surgery, Vol. 22, 1940, pp. 621-626.

  31. Pammels, F., “Uber die Bedentung der Bauprinzipien des Sturtzund Benwegrengsapparetes für die Beansprushung der Rohrenknochen,” Acta Anatomica, Vol. 12, 1951, pp. 207-227.

  32. Caler, W. E., Carter, D. R., and Harris, W. H., Journal of Biomechanics, Vol. 14, 1981, pp. 503-507.

  33. Lanyon, L.E., Paul, I.L., and Rubin, C.T., Journal of Bone and Joint Surgery, Vol. 63A, 1981, pp. 989-1001.

  34. McBeath, A.A., Schopler, S.A., and Narechairia, R.G., Clinical Orthopaedics and Related Research, Vol. 150, 1980, pp. 301-305.

  35. Mizrahi, J., Livingstone, R.P., and Rofan, I.M., Journal of Biomechanics, Vol. 12, 1979, pp. 491-500.

  36. Oh, I., “Effect of Total Hip Replacement on the Distribution of Stress in the Proximal Femur: An In Vitro Study Comparing Stress Distribution in the Intact Femur with That After Insertion of Different Femoral Components,” The Hip: Proceedings, Fifth Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1977, pp. 1-12.

  37. Oh, I. and Harris, W. H., Journal of Bone and Joint Surgery, Vol. 60A, No. 1, 1978, pp. 75-85.

  38. Jones, L.C. and Hungerford, D.S., “Measurements of Strain in the Fresh Human Femur Using the Photoelastic Coating Technique,” Transactions of the Society for Biomaterials, Vol. 8, 1985, p. 199.

  39. Jones, L.C. and Hungerford, D.S., “The Photoelastic Coating Technique—It’s Validation and Use,” Transactions of the Orthopaedic Research Society, Vol. 12, 1987, p. 406.

  40. Sedlin, E. D. and Hirsch, C., Clinical Orthopaedics and Related Research, Vol. 37, 1966, pp. 29-48.

  41. Instruction Bulletins No. IB-233, IB-221, IB-228, and IB-223-A, Measurement Group, Inc., Photoelastic Division, Raleigh, NC, 1982.

  42. Blacker, G. and Charnley, J., Clinical Orthopaedics and Related Research, Vol. 137, 1978, pp. 15-23.

  43. Coventry, M. B. and Stauffer, R. N., “Long-Term Results of Total Hip Arthroplasty,” The Hip: Proceedings,Tenth Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1982, pp. 34-41.

  44. Charnley, J. and Curic, Z., Clinical Orthopaedics and Related Research, Vol. 95, 1973, pp. 9-25.

  45. Charnley, J., Clinical Orthopaedics and Related Research, Vol. 111, 1975, pp. 105-120.

  46. Nicholson, O. R., Clinical Orthopaedics and Related Research, Vol. 95, 1973, pp. 217-223.

  47. Crowninshield, R.D., Brand, R.A., Johnston, R.C., and Pedersen, D.R., “An Analysis of Femoral Prosthesis Design: The Effects on Proximal Femur Loading,” The Hip: Proceedings, Ninth Open Scientific Meeting of The Hip Society, Mosby, St. Louis, 1981, pp. 111-122.

  48. Andersson, G. B. J., Freeman, M. A. R., and Swanson, S. A. V., Journal of Bone and Joint Surgery, Vol. 54B, 1972, pp. 590-599.

  49. Willert, H.G., Ludwig, J., and Semlitsch, M., Journal of Bone and Joint Surgery, Vol. 56A, 1974, pp. 1368-1382.

  50. Moreland, J.R., Gruen, T.A., Mai, L., and Amstutz, H.C., “Aseptic Loosening of Total Hip Replacements: Incidence and Significance,” The Hip: Proceedings, Eighth Open Scientific Meeting of The Hip Society, Mosby, 1980, pp. 281-291.