This article has Open Peer Review reports available.
Biomechanical optimization of different fixation modes for a proximal femoral L-osteotomy
© Tai et al; licensee BioMed Central Ltd. 2009
Received: 10 October 2008
Accepted: 10 September 2009
Published: 10 September 2009
Numerous proposed surgical techniques have had minimal success in managing greater trochanter overgrowth secondary to retarded growth of the femoral capital epiphysis. For reconstruction of residual hip deformities, a novel type of proximal femur L-osteotomy was performed with satisfactory results. Although the clinical outcome was good, the biomechanical characteristics of the femur after such an osteotomy have not been clearly elucidated. Therefore, this study presents a three dimensional finite element analysis designed to understand the mechanical characteristics of the femur after the L-osteotomy.
A patient with left hip dysplasia was recruited as the study model for L-osteotomy. The normal right hip was used as a reference for performing the corrective surgery. Four FEA models were constructed using different numbers of fixation screws but the same osteotomy lengths together with four FEA models with the same number of fixation screws but different osteotomy lengths. The von Mises stress distributions and femoral head displacements were analyzed and compared.
The results revealed the following: 1). The fixation devices (plate and screws) sustained most of the external loading, and the peak value of von Mises stress on the fixation screws decreased with an increasing number of screws. 2). Additional screws are more beneficial on the proximal segment than on the distal segment for improving the stability of the postoperative femur. 3). The extent of osteotomy should be limited because local stress might be concentrated in the femoral neck region with increasing length of the L-osteotomy.
Additional screw placement on the proximal segment improves stability in the postoperative femur. The cobra-type plate with additional screw holes in the proximal area might improve the effectiveness of L-osteotomies.
The hip joint is important for maintaining posture and aiding locomotion. The joint is formed by the articulation between the femoral head and the acetabulum and any alteration in the anatomy of the bony components induces abnormal mechanical forces on the joint . Patients with congenital hip dislocation, Perthes disease or septic arthritis often exhibit a deformed femoral neck and limb length discrepancy. These residual deformities compromise joint biomechanics and cause abnormal loading of the hip joint , and result in clinical symptoms: hip instability, limping gait, shortening of the extremity involved, limitation of range of motion in the hip and weakening of the hip abductors . A disease process affecting the physis of the femoral head with continued growth in the greater trochanter will result in a shortened femoral neck, coxa vara deformity and increased anteversion . The upward displacement of the greater trochanter causes poor abductor function and progressively worsens the Trendelenberg gait [5, 6].
Various surgical techniques have been proposed but have had limited success in managing femoral neck shortening and the greater trochanter overgrowth secondary to retarded growth of the femoral capital epiphysis [2, 5–8]. Although various procedures exist to treat patients with congenital hip dislocation or Perthes disease, residual deformities including leg length discrepancy, hip joint incongruity, proximal displacement of the greater trochanter, and poor joint biomechanics often persist that remain difficult to solve . For reconstruction of these residual hip deformities, Papavasiliou et al.  performed a new proximal femoral L-osteotomy in sixteen patients with residual hip deformity (coxa vara, coxa breva and high riding greater trochanter). Good results were reported in all patients after a mean of 4.3 years. The surgical procedure not only provided limb equalization but also repositioned the greater trochanter to its normal level. Additionally, the procedure ensured elongation of the femoral neck with subsequent restoration of gluteus medius length. Moreover, engagement of the femoral cortices anteriorly or posteriorly during the distraction restored the proper degree of anteversion of the femoral neck. Another advantage of the L-osteotomy is the repositioning of the femoral head deeper in the acetabulum, which improves the contact between articular surfaces. However, hip spica casts were used in the Papavasiliou study to ensure postoperative stability.
As computer technology advances, the prospects for more realistic modeling of bone diseases are encouraging. Patient-specific simulations of surgical procedures are now feasible, particularly using computed tomography and magnetic resonance imaging (MRI) techniques. Given the geometric nature of residual femur deformity, the FEA model derived from the reconstruction of 3-D CT images may be helpful for objectively analyzing the stresses in structures with complex shapes, loading and material behavior. Finite element analysis models have been applied extensively in orthopedics and have proven effective for predicting musculoskeletal mechanics in unusual circumstances [10, 11]. There is also ample precedent for the use of FEA models to elucidate mechanical behavior for a proximal femur osteotomy [12–15]. In 1987, Fyhrie and Carter  explored the role of compressive volumetric stress and strain in the early pathogenesis of femoral head necrosis. In 2002, Yang et al.  developed a three-dimensional finite element model using a surface modeling technique to assess the stress distribution at various sizes of segmental osteonecrosis. Recently, Chen et al.  designed a three-dimensional finite element analytical model for comparing postoperative stability between large cancellous screw fixation and dynamic hip screw fixation in transtrochanteric rotational osteotomy. The Chen study concluded that dynamic hip screw fixation provides better stability than large cancellous screw fixation. The same study also reported the use of computer simulation to investigate the degree to which transtrochanteric rotational osteotomy moves the region of osteonecrotic femoral head out of the weight bearing area. The results demonstrated that posterior rotational osteotomies were more effective for moving the necrotic region out of the weight bearing area during a gait cycle . Since numerous reports have shown good results in analyzing stress distribution and predicting the postoperative behavior of proximal femur osteotomies, a finite element analysis based on various fixation configurations of the L-osteotomy should be conducted due to the promising results of our previous clinical experience.
Generation of 3D solid model
This study was approved from the committee of National Science Council of the Republic of China (Contract Number: NSC92-2218-E-182A-001). A left hip dysplasia patient (male, 25 years, 71 Kg, 167 cm) was recruited as the study model for the L-osteotomy. The normal right hip was used as a reference to plan the surgery. 3-D Solid models of both the left proximal femur with hip dysplasia and the normal right femur were created using computed tomography (CT) scan images and Solidworks CAD software (SolidWorks 2004, SolidWorks Corp. Boston, MA, USA). The CT scan images of the intact femur were obtained at 1.25 mm intervals in the transverse planes starting from the femoral head using a GE Hi-speed scanner (General Electric, Milwaukee, WI, USA). The resolution for each of the CT scan image was 512 by 512 pixels, the field of view was 330 mm, and the pixel size was 0.625 mm/pixel. The cross-sectional image files of the femur were transferred to a custom-written automatic contouring program (Caotool) for the detection of the contours between the cortical bone and cancellous bone based on a dynamic density-thresholding algorithm. The parallel-stacked contours were then input into the Solidworks CAD software for the reconstruction of 3-D intact femur solid models.
Generation of finite element model
Based on the solid model of the deformed left femur, four FEA models with different screw placement configurations (P2/D2, P2/D3, P3/D2 and P3/D3; P: fixation on the proximal segment; D: fixation on the distal segment) were created with the osteotomy length kept constant at 126 mm. This length was defined by the roentgenographic measurement of the actual surgery. To investigate the influence of the osteotomy length on postoperative mechanical characteristics, an additional four FEA models with the same number of fixation screws (P3/D2) but different osteotomy lengths (116 mm, 126 mm, 136 mm and 146 mm) were also created (MENTAT 2003, MSC Software Corp., Los Angeles, USA).
Loading and boundary conditions
Convergence test and model validation
For model verification, a convergence test was used to guarantee our numerical model reached the converged results and no further mesh refinement was necessary. Based on the same solid model, four different FE models with average element lengths of 4 mm, 5 mm, 6 mm and 7 mm were created from the pre-operative femur. The convergences of the FEM models in this study were justified by the total strain energy of the structure. The total strain energy was reviewed for convergence within the four models. The tolerance level was set as the change of less than 5%.
In order to validate the FE model, the experimental results from Shih et al.  was used to compare with our FE analysis. In their experiment, the surface strains at the medial and lateral proximal femur under 2000 N vertical loading was recorded in fresh-cadaver compressive testing. To simulate this setting, an additional FE model was created from the intact normal right femur, and the surface strains at the medial and lateral proximal femur subjected to the same 2000 N vertical loading were also analyzed. The experimental and analytical results were compared for the validation of the FEA model.
Convergence test and model validation
Results of convergence test demonstrated a less than 5% changes in the total strain energy among four models. The element numbers for four different models with average element lengths of 4, 5, 6, and 7 mm were 92749, 54634, 35031, and 28087, respectively. The total strain energies for each model were 4.131, 4.055, 4.006 and 3.952 J, respectively. The percent differences of the total strain energy compared with that of the finest mesh (element number: 92749) for each of the three models were 1.836%, 3.021%, 4.324% respectively. Although the results indicated that convergence was achieved, the geometry in the medial aspect of femoral neck for FE model with 5 mm element length were somewhat distorted. Therefore, the model with an average element size of 4 mm was chosen as the base model for the creation of post-operative models.
For model validation, the previous study  had indicated that the experimental strains at the medial and lateral proximal femur were -1.098 (SD 0.134) and 0.723 (SD 0.139) microstrain, respectively, under a 2,000 N axial loading. In current study, the predicted strains at the medial and lateral proximal femur were -1.211 and 0.742 under a 2,000 N axial loading. The two results were comparable which indicated that our finite element model was reliable for further simulation and analysis.
L-osteotomy with different numbers of fixation screws but the same osteotomy length (126 mm)
L-osteotomy with the same number of fixation screws (P3/D2) but a different osteotomy length
Congenital hip dysplasia, Perthes disease and septic arthritis at an early age result in an abnormal gait with limb length shortening, often progressing to advanced osteoarthritis requiring total hip replacement later in life. Many surgical procedures exist for treating these patients, but residual deformities are often present. The L-osteotomy method described by Papavasiliou et al. addresses all these deformities by ensuring elongation of the femoral neck, biomechanical improvement of the joint, congruity of the femoral head and equalization of the leg length discrepancy . However, this surgery is a demanding technique, and proper planning for this osteotomy requires a clear understanding of postoperative mechanical performance.
In the present study, the integral femur was sectioned into two separate fragments after the L-osteotomy. The upper fragment contained the entire femoral head and medial proximal femoral shaft, and the lower fragment contained the greater trochanter and the remaining distal femoral shaft. These two separated fragments were fixed into one integral structure by the instrumentation of the fixation devices (plate and screws). It was reasonable to postulate that the fixation devices would sustain most of the external loading applied on the entire femur. As Figure 7 shows, when the L-osteotomy was performed with the same osteotomy length (126 mm), the von Mises stress of the fixation screws decreased with an increasing number of screws because the external load was shared by the additional fixation screws.
Additionally, the femoral head displacement decreased as the number of fixation screws increased (Figure 8). The experimental results further indicated that proximal placement of the fixation screws would be more effective than distal placement for improving stability of the postoperative femur in circumstances when the number of fixation screws is limited. As Figure 9 shows, in an L-osteotomy performed with the same number of fixation screws (P3/D2) but different osteotomy length, the moment acting on the entire structure increased with increasing length, and led to a higher stress concentration on the femoral neck region.
Although an improvement of the integral stability of the postoperative femur was expected by performing an L-osteotomy with more fixation screws, proximal placement of the screw (P3/D2) stabilized the femur more effectively than distal placement of the screw (P2/D3). A Cobra-type plate accommodating more proximal screws would be a good alternative for fixation of the L-Osteotomy.
Finite element analysis is a complementary tool for evaluating the feasibility, efficacy and overall biomechanical characteristics of different surgical techniques. Although clinical trials could summarize general results regarding different osteotomy methods, many fundamental issues still remain controversial and poorly understood. Frequently, scientific inquiries are simulated by the introduction of new surgical techniques. However, systematic investigations are needed to test the treatment principles. Finite element analysis is a noninvasive fracture monitoring technique which can monitor the stress and biomechanical responses of different treatments without clinical influence.
The validation of our model was done with use of the normal right femur, but not the left femur with hip dysplasia. This is because it's almost not possible to access the human cadaveric femur with residual deformities to perform an experiment with L-osteotomy. The validation of FE model in current study is thus conducted using FE model created from the intact normal right femur of the same patient, and the results were compared with those from previous experimental research . In the finite element model, the screw threads and tips are not modeled because FE models with screw threads and tips will result in a large increase of element number and computation time. The simplified FE models without taking threads and tips into consideration may have an impact on the analytic results for local area close to the screw/bone interface. However, we believe these may not cause a global effect on the resultant FE analysis.
Several possible factors affecting the FEA results of the current study must be noted. First, only a single patient with residual deformity was recruited in the present study. Therefore, the results from the FEA might not be regarded as general rules that can be applied to preoperative planning of an L-osteotomy. Importantly, although the results indicated that increased numbers of fixation screws together with a shorter osteotomy length are beneficial for the postoperative performance of the hip joint, preoperative planning should still be individualized based on the extent of the residual deformity of the femur. Second, the FE model was validated based on the intact condition without osteotomy, which may have an impact on the analytic results for the post-operative FE models with osteotomy. However, the boundary conditions including material properties, element types and element length are identical for FE models with or without osteotomy, and we believe that our results provide useful information to orthopedic surgeons performing reconstruction of residual hip deformities with proximal femur L-osteotomy. Third, the only loading condition considered was the single legged gait stance. Therefore, further investigations of the effects of other loading conditions might be necessary in the future. Fourth, the bone plate interfaces were assumed to be fully bonded without considering loosening of the fixation device. Therefore, these FEA results might only be interpreted under a well-fixed condition without implant loosening. Still, attentions need to be paid toward the actual application of this technique. Based on the limitations of this numerical investigation, the model simplifications, such as the material properties, screw geometry and load conditions might influence the accuracy of the mechanical responses and stress distributions obtained in this study. The results from this finite element simulation were based on an objective to provide a way to eliminate the problems encountered with L-osteotomy technique as a clinical treatment substitution.
In Papavasiliou's report, the L-osteotomy not only provided limb equalization but also repositioned the greater trochanter to its normal level . However, good clinical outcomes should not only rely on proper patient selection but also on a good preoperative planning and precise execution of the plan during surgery. In current study, the importance of fixation stability of L-osteotomy was clearly demonstrated by FE analysis. In addition, the significance of fixation screws number and osteotomy length contributing to the stress distribution after L-osteotomy is also highlighted. Since L-osteotomy is a technically demanding procedure and associated with high complication risks, it is suggested that the indication for L-osteotomy should be very strict and that surgical principles should be abided carefully to avoid catastrophic complications.
The fixation devices (plate and screws) sustained most of the external loading, and the peak value of von Mises stress on the fixation screws decreased with an increasing number of screws.
Placement of the screw on the proximal segment rather than on the distal segment enhanced the stability of the postoperative femur.
The extent of osteotomy should be limited because a high local stress concentration might occur in the femoral neck region as the L-osteotomy length increases.
The authors would like to thank the National Science Council of the Republic of China for the financially supporting this research under Contract No. NSC92-2218-E-182A-001. Written consent for publication was obtained from the patient.
- Delp SL, Maloney W: Effects of hip center location on the moment-generating capacity of the muscles. J Biomech. 1993, 26 (4-5): 485-499. 10.1016/0021-9290(93)90011-3.View ArticlePubMedGoogle Scholar
- Techdjian MO, Kelikian AS: Distal and lateral advancement of the greater trochanter. Congenital dislocation of the hip. Edited by: Tachdjian MO. 1982, New York: Churchill-Livingstone, 721-39.Google Scholar
- Kelikian AS, Tachdjian MO, Askew MJ, Muvali J: Greater trochanteric advancement of the proximal femur: a clinical and biomechanical study. Proceedings of the 11th Open Scientific Meeting of the Hip Society. 1983, St. Louis: CV Mosby, 77-103.Google Scholar
- Lloyd RGC, Wetherill MH, Fraser M: Trochanteric advancement for premature arrest of the femoral capital growth plate. J bone Joint Surg Br. 1985, 67 (1): 21-24.Google Scholar
- Coleman SS: Reconstructive procedures in congenital dislocation of the hip. Recent Advances in Orthopaedics. Edited by: McKibbon B. 1983, Edinburg, Churchill-Livingstone, 23-43.Google Scholar
- Tauber C, Ganel A, Horoszowski H, Farine I: Distal transfer of the greater trochanter in coax vara. Acta Orthop Scand. 1980, 51 (4): 661-666.View ArticlePubMedGoogle Scholar
- Cohen J: Congenital dislocation of the hip. Case report of an unusual complication and unusual treatment. J bone Joint Surg Am. 1971, 53 (5): 1007-1011.PubMedGoogle Scholar
- Johnston RC, Brand RA, Crowninshield RD: Reconstruction of the hip. A mathematical approach to determine optimum geometric relationships. J bone Joint Surg Am. 1976, 61 (5): 639-652.Google Scholar
- Papavasiliou VA, Kirkos JM: Reconstruction of residual deformities of the hip. Clin Orthop. 1997, 341: 123-127.View ArticlePubMedGoogle Scholar
- Cody DD, Gross GJ, Hou JH, Spencer HJ, Glodstein S, Fyhrie DP: Femoral strength is better predicted by finite element models than QCT and DXA. J Biomech. 1999, 32 (10): 1013-1020. 10.1016/S0021-9290(99)00099-8.View ArticlePubMedGoogle Scholar
- Testi D, Viceconti M, Brauffaldi F, Cappello A: Risk of fracture in elderly patients: a new predictive index based on bone mineral density and finite element analysis. Comput Meth Prog Bio. 1999, 60 (1): 23-33. 10.1016/S0169-2607(99)00007-3.View ArticleGoogle Scholar
- Fyhrie DP, Carter DR: Compressive volumetric strain as a risk factor in aseptic osteonecrosis. Orthop Trans. 1987, 12: 497-Google Scholar
- Yang JW, Koo KH, Lee MC, Yang P, Noh MD, Kim SY, Kim KI, Ha YC, Joun MS: Mechanics of femoral head osteonecrosis using three-dimensional finite element method. Arch Orthop Trauma Surg. 2002, 122 (2): 88-92. 10.1007/s004020100324.View ArticlePubMedGoogle Scholar
- Chen WP, Tai CL, Shih CH, Hsieh PH, Leou MC, Lee MS: Selection of Fixation Devices in Proximal Femur Rotational Osteotomy: Clinical Complications and Finite Element Analysis. Clin Biomech. 2004, 19 (3): 255-262. 10.1016/j.clinbiomech.2003.12.003.View ArticleGoogle Scholar
- Chen WP, Tai CL, Tan TF, Shih CH, Ho SH, Chen CY, Lee MS: The degree to which transtrochanteric rotational osteotomy in osteonecrosis of the femoral head moves the region of necrosis out of the weight bearing area as evaluated by computer simulation. Clin Biomech. 2005, 20 (1): 63-69. 10.1016/j.clinbiomech.2004.08.001.View ArticleGoogle Scholar
- Mann KA, Bartel DL, Wright TM: Coulomb frictional interfaces in modeling cemented total hip replacements: a more realistic model. J Biomech. 1995, 28 (9): 1067-1078. 10.1016/0021-9290(94)00158-Z.View ArticlePubMedGoogle Scholar
- Huang TJ, Hsu RW, Tai CL, Chen WP: A biomechanical analysis of triangulation of anterior vertebral double-screw fixation. Clin Biomech. 2003, 18 (6): S40-45. 10.1016/S0268-0033(03)00083-4.View ArticleGoogle Scholar
- Brown TD, Hild GL: Pre-collapse stress redistributions in femoral head osteonecrosis a three-dimensional finite element analysis. J Biomech Eng. 1983, 105 (2): 171-176. 10.1115/1.3138402.View ArticlePubMedGoogle Scholar
- Brown TD, Way ME, Ferguson AB: Mechanical characteristics of bone in femoral capital aseptic necrosis. Clin Orthop. 1981, 240-246. 156Google Scholar
- Shih CH, Chen WP, Tai CL, Kuo RF, Wu CC, Chen CH: New concept- biomechanical studies of a newly designed femoral prosthesis (cervical-trochanter prosthesis). Clin Biomech. 1997, 12 (8): 482-490. 10.1016/S0268-0033(97)00032-6.View ArticleGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2474/10/112/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.