Predicting the Position of the Femoral Head Center

Transcription

1 The Journal of Arthroplasty Vol. 14 No Predicting the Position of the Femoral Head Center Nobuhiko Sugano, MD, Philip C. Noble, PhD, and Emir Kamaric, MS Abstract: To find an accurate method to predict the position of the normal head center in severely deformed hips without a contralateral reference, we studied the relationships between the head center and dimensions of the normal proximal femur using cadaveric specimens. From a large anatomic collection, 32 cadaveric femora with neck-shaft angles ranging from 115 to 146 were selected. The two parameters with the greatest correlation with the height of the femoral head were the height of the neck isthmus (r =.932) and the lowest point of neck saddle (r =.790). Medial head offset was most strongly correlated with the offset of the neck isthmus (r =.945) and the distance from the medullary axis to the outer borders of the medial cortex at the + 30% level (r =.861 ). As a rule of thumb, the height of the head center can be predicted as the height of the midpoint of the neck isthmus plus 10 or as the height of the neck saddle with an accuracy of and The offset of the head center can be predicted as the offset of the midpoint of the neck isthmus plus 15 or as the distance from the medullary axis to the medial cortex at the +30% level plus 15 with an accuracy of +4.6 and Key words: femoral head center, femoral head offset, total hip arthroplasty, preoperative planning. In total hip arthroplasty, normal joint mechanics are restored by recreating the mechanical balance of forces acting about the hip. Many surgeons attempt to restore the original head center or, at least, reference the position or their restoration with respect to the original head center. However, the normal, physiological location of the head center is not always evident in the arthritic joint, owing to the presence of bony deformities. If the skeletal anatomy of the affected hip is only minimally abnormal, the appropriate position for placement of the prosthesis can be directly determined using the original femoral head. In cases of unilateral joint disease with femoral head deformity, preoperative From the Joseph Barnhart Department of Orthopedic Surgery, Baylor College of Medicine, Houston, Texas. Submitted August 19, 1997; accepted April 6, Reprint requests: Nobuhiko Sugano, MD, Baylor College of Medicine, 6565 Fannin F115, Houston, TX Copyright 1999 by Churchill Livingstone / $10.00/0 planning is often performed using radiographs of the normal contralateral hip [1,2]. This planning is not possible in cases of bilateral involvement with significant deformities of both joints, as sometimes observed in patients with severely collapsed osteonecrosis, rheumatoid arthritis, or rapidly destructive arthropathy of the hip. Moreover, reference to a normal, centralateral femur is not possible in revision cases in which bilateral femoral components are present. In each of these scenarios, the position of the head center must be estimated from anatomic landmarks present within the femur. Previous authors have demonstrated that the height of the femoral head with respect to the lesser trochanter and its medial offset from the medullary axis are highly variable [3]. Few studies, however, have reported head position related to bony landmarks within individual femora. A coon assumption is that the cephalad position of the head center corresponds to the proximal tip of the greater 102

2 The Position of the Femoral Head Sugano et al. 103 trochanter [4,5]. Others have noted a wide variation in the prominence of the greater trochanter and have suggested that this landmark is of limited predictive value {3,6]. To the best of our knowledge, the medial offset of the head center has not been related to anatomic parameters that can be readily applied to severely affected hips. To find an accurate method to predict the position of the normal head center in severely deformed hips without a contralateral reference, we studied the relationships between the head center and dimensions of the normal proximal femur using cadaveric specimens. Y Materials and Methods Thirty-two cadaveric femora with neck-shaft angles ranging from 115 to 146 were selected from a large anatomic collection. Standard anteroposteriot radiographs of each specimen were prepared in a plane parallel to the axis of the femoral neck, with a tube-film distance of 72 inches. This distance generated radiographic magnification averaging 1.5 %. Any specimen with radiographic evidence of metastatic disease, previous trauma, or other significant pathology was excluded from the study. Dimensional data describing the morphology of the proximal femur were measured with reference to a standard radiographic axis system. The vertical axis was defined by the medullary axis of each femur calculated as a line passing through the midpoints of the medullary canal, 20 proximal and 20 distal to the canal isthmus. The horizontal axis was defined by a line passing through the center of the lesser trochanter, perpendicular to the medullary axis. The following landmarks were located on each radiograph (Fig. 1 ): 1. Femoral head center, determined by using a set of concentric circular templates 2. Neck isthmus, determined by finding the shortest distance between the superior and the inferior borders of the femoral neck 3. Axis of the femoral neck, defined by a line passing through the midpoint of the isthmus and the head center 4. Point of intersection of the medullary axis and the axis of the femoral neck 5. Lowest point on the upper surface of the femoral neck (the saddle of the neck) 6. Tip of the greater trochanter 7. Points of intersection of the horizontal axis and the external borders of the medial and lateral cortices, determined at 3 levels, located at distances of 0.1 G, 0.2 G, and 0.3 G above and below the horizontal axis, where each level was separated by Fig. 1. The proximal femur with reference to an x-y coordinate system. 1, The center of the femoral head; 2, the midpoint of the femoral neck isthmus; 3, the point of intersection of the medullary axis (y-axis) and the axis of the femoral neck, defined by a line passing through the midpoint of the isthmus and the head center; 4, the lowest point on the upper surface of the femoral neck (the neck saddle); 5, the tip of the greater trochanter; 6, the lower edge of the neck isthmus; 7, the upper edge of the neck isthmus; 8-21, a series of points defined by the intersection of a line perpendicular to the y-axis and the external borders of the medial and lateral cortices at 3 levels (10%, 20%, and 30% of the height of the greater trochanter) above and below the level or the lesser trochanter. 10% of the vertical distance from the tip of the greater trochanter to the center of the lesser trochanter (G) Each of these points was electronically digitized using a back-lit digitizing screen (GTCO, GTCO Corp., Columbia, MD). The following geometric parameters were calculated using the digitized data: 1) the midpoint of the isthmus of the femoral neck and 2) the neck-shaft angle, defined as the angle between the femoral neck axis and the medullary axis. To predict the position of the head center, correlations between the x, y coordinates of each of the anatomic landmarks and the height and the medial offset of the femoral head center were calculated using simple regression analysis. Four parameters, 2 having the strongest correlation with the height of the femoral head and 2 with the head offset, were used to calculate the corresponding univariate regression equations to predict the coordinates of the

3 104 The Journal ofarthroplasty Vol. 14 No. 1 January 1999 Table 1. Average Values of Anatomic Parameters Average SD Minimum Maximum Parameters () () (ram) () Neck-shaft angle Femoral head diameter Neck isthmus width Height of head center Neck isthmus height Neck saddle height Greater trochanter height (G) Medial offset of head center Neck isthmus offset Outer cortical width at 30% of G above the lesser trochanter Outer cortical width at 20% of G above the lesser trochanter Outer cortical width at 10% of G above the lesser trochanter Outer cortical width at the lesser trochanter Outer cortical width at 10% of G below the lesser trochanter Outer cortical width at 20% of G below the lesser trochanter Outer cortical width at 30% of G below the lesser trochanter i head center. Then, for each method of calculation, the average error in predicting the position of the head center was calculated as the difference between the actual and predicted values. Results The femoral specimens were obtained from 11 men and 2I women with an average age of 67.2 years (range, years). The average neck-shaft angle ( ; range, ) was quite similar to normal femoral anatomy previously reported [3,6]. Eight femora had a valgus neck (neckshaft angle >132 ), 16 femora had a neck-shaft angle in the normal range ( ), and 8 femora had a varus neck (neck shaft angle < 120 ). The average values for each of the dimensional parameters are presented in Table 1. Data describing the correlation between the height of the femoral head center and each of the anatomic parameters are presented in Table 2. The two parameters with the greatest correlation were the height of the neck isthmus (r =.932) and the lowest point of the neck saddle (r =.790) (Fig. 1, points 2 and 4). The tip of the greater trochanter was located at an average of 10.3 proximal to the femoral head center; however, the distance between these two landmarks was highly variable (r =.594, range, ). Correlations between the offset of the femoral head center and the various anatomic parameters are presented in Tables 3 and 4. The medial offset of the femoral head was most strongly correlated with the offset of the neck isthmus (r =.945) and the distance from the medullary axis to the periosteal border of the medial cortex at the +30% level (r =.861) and the +20% level (r =.821). Strong correlations were also noted between the medial head offset and the femoral head diameter (r =.704), the width of the isthmus of the femoral Table 2. Correlation of Each Anatomic Parameter With the Height of the Femoral Head Center Correlation Linear Regression Parameters (x) Coefficient (r) Formula Head diameter Neck isthmus width Neck isthmus height Neck saddle height Greater trochanter height (G) Outer cortical width at 30% of G above the lesser trochanter Outer cortical width at 20% Of G above the lesser trochanter Outer cortical width at 10% of G above the lesser trochanter Outer cortical width at the lesser trochanter Outer cortical width at 10% of G below the lesser trochanter Outer cortical width at 20% of G below the lesser trochanter Outer cortical width at 30% of G below the lesser trochanter x x x x x x x x x x x x

4 The Position of the Femoral Head Sugano et al. 105 Table 3. Correlation of the Parameters With Offset of the Femoral Head Center Correlation Linear Regression Parameters (x) Coefficient (r) Formula Head diameter Neck isthmus width Height of the femoral head center Great trochanter height (G) Neck saddle height Neck isthmus offset Outer cortical width at 30% of G above the lesser trochanter Outer cortical width at 20% of G above the lesser trochanter Outer cortical width at 10% of G above the lesser trochanter Outer cortical width at the lesser trochanter Outer cortical width at 10% of G below the lesser trochanter Outer cortical width at 20% of G below the lesser trochanter Outer cortical width at 30% of G below the lesser trochanter I x x x x x x x x x x x x x neck (r =.692), and the coordinates of the periosteal border of the medial cortex at the other levels studied (r = ). Much weaker correlations were exhibited with the coordinates of the outer borders of the lateral cortex at each level (r = ). There was also no significant correlation between the height and the medial offset of the head center (r =.064). Using linear regression formulas (Tables 1 to 3), the height and offset of the head center were predicted using each of the 2 parameters that were most strongly correlated with head position. The difference between the true and predicted locations of the head center was then calculated for each femur. The height of the head center was predicted to _+4.9 (95% confidence interval) using the height of the midpoint of the neck isthmus and using the height of the neck saddle (Figs. 2, 3). Similarly, head offset was predicted to an accuracy of _+4.1 using the medial offset of the neck isthmus and +_ 6.4 using the distance from the medullary axis to the medial cortex at the +30% level (Figs. 4, 5). When simpler formulas were used to predict the height and offset of the head center, the range of 9 5 % confidence interval of each formula (values in parentheses) was obtained as follows: 1. Height of the femoral head = l0 + the height of the midpoint of the neck isthmus (+5.1 ) 2. Height of the femoral head = height of the neck saddle (-+8.3 ) 3. Femoral head offset the offset of the midpoint of the neck isthmus (-+4.6 ) 4. Femoral head offset = 15 + the distance from the medullary axis to the medial cortex at the +30% level (-+6.3 ) Discussion In total hip replacement, it is widely believed that the center of the femoral head is level with or slightly superior to the tip of the greater trochanter [4,5]. This assumption is reflected in the design of standard equipment intended to facilitate restoration of the head center, including cutting guides that guide the surgeon in placement of the femoral neck osteotomy and some transparent templates supplied for preoperative planning using standard hip radiographs. In this study, the location of the head center with respect to the tip of the trochanter was found to be highly variable, ranging from 1.6 superior to 22.1 inferior, with an average value of 10.3 Table 4. Correlation of the Parameters With Offset of the Femoral Head Center Level of Measurement Site Along the Medullary Axis* Parameters 30% 20% 10% LT -- 10% -20% -30% Medial cortex Lateral cortex *Expressed as a percentage of the vertical distance from the tip of the greater trochanter to the center of the lesser trochanter (LT).

5 106 The Journal of Arthroplasty Vol. 14 No. 1 January I 5O 00 ~ q O ~ e., 40,o,~ 35 ~, y = x 40,/ Height of the midpoint of the femoral neck isthmus (x) Fig. 2. Scatterplot and 95% confidence interval bands of the height of the head center and the height of the midpoint of the femoral neck isthmus Offset of the midpoint of the neck isthmus (x) Fig. 4. Scatterplot and 95% confidence interval bands of the offset of the head center and the offset of the midpoint of the femoral neck isthmus. inferior. This variation is due to several factors, including the inclination of the neck, the size of the metaphysis, and the prominence of the greater trochanter. More reliable parameters for estimating the height of the head center were the height of the neck saddle and the midpoint of the neck isthmus. Many dimensional parameters were strongly correlated with the medial offset of the head center, including the offset of the midpoint of the neck isthmus and the offset of the medial outer cortex at the level of 30% of the greater trochanter height above the lesser trochanter. The offset was not strongly correlated with the height of the head. This means that in total hip replacement, the height and medial offset of the femoral head prostheses cannot always be restored using implants with one standard neck-shaft angle, even if the level of the femoral neck osteotomy is varied. This problem becomes most acute in the femur with a varus neck, in which restoration of the correct leg length necessitates a low osteotomy, which in some cases may be close to the superior border of the lesser trochanter. This procedure requires sacrifice of an excessive amount of bone stock and often leads to an inferior fit of cementless prostheses within the medullary canal. Our results demonstrate that the normal position of the center of the femoral head may be predicted more accurately using anatomic landmarks other than the greater trochanter, even in cases of rheumatoid arthritis or osteoarthritis secondary to hip dysplasia with severe bilateral deformities of the femoral heads. Clearly, the accuracy of the predicted head position depends on the amount of the femo Lf/ 60 /,..., t. 45 J J 7 7,i/1~1,,~ /:J ~,= x 40,, J Height of the neck saddle (x) Fig. 3. Scatterplot and 95% confidence interval bands of the height of the head center and the height of the neck saddle. "~ 35 O 30 / y = ~ x Fig. 5. Scanerplot and 95% confidence interval bands of the offset of the head center (y) and the distance from the medullary axis to the medial outer cortex at 30% of the greater trochanter height above the lesser trochanter (x).

6 The Position of the Femoral Head Sugano et al. 107 ral neck that is present preoperatively because the most accurate estimates of head height and offset are based on the neck isthmus. However, even in cases of more extensive bone loss, a useful estimate of head position can be obtained using the height of the neck saddle and the distance from the medullary axis to the medial cortex at the +30% level. As a rule of thumb, the height of the head center can be predicted as the height of the midpoint of the neck isthmus plus 10 or as the height of the neck saddle. The offset of the head center can be predicted as the offset of the midpoint of the neck isthmus plus 15 or as the distance from the medullary axis to the medial cortex at the +30% level plus f 5. Conclusion The best method to predict the position of the head center from the dimensions of the proximal femoral metaphysis is based on the midpoint of the neck isthmus. When this landmark cannot be used because of the presence of some pathological deformity of the neck, the physiological head position may be estimated from the height of the neck saddle and the distance from the medullary axis to the medial cortex at the + 30% level. References 1. D'Antonio JA: Preoperative templating and choosing the implant for primary THA in young patient. AAOS Instruct Course Lect 43:339, Dorr DD, Rubash HE: Primary total hip arthroplasty in the older patient: optimizing the results. AAOS Instruct Course Lect 43:347, Noble PC, Alexander JW, Lindahl L J, et al: The anatomic basis of femoral component design. Clin Orthop 235:148, Harkess JW: Arthroplasty of hip. p In Crenshow AH (eds): Campbell's operative orthopaedics, 8th ed. Mosby-Year Book, St Louis, Levy RN: The location of the level of femoral neck transection for prosthetic hip arthroplasty. Clin Orthop 171:51, Noble PC, Box GG, Kamaric E, et al: The effect of aging on the shape of the proximal femur. Clin Orthop 316:31, 1995