In recent years, three-dimensional templating has been used for preoperative THA planning [13, 14]; however, this method requires specialized software and is time-consuming. Contrary, 2D preoperative templating using plain radiographs and axial CT images is simple, inexpensive, and can be completed in a few minutes. Knowing the ilioischial line on CT can provide information on the offset for cup placement, which may be helpful for creating more accurate 2D radiographic templates. The offset position of the cementless cup is critical to avoid iliopsoas impingement [15–19], ensure adequate acetabular coverage [7, 20], and for THA stability [21–23]. When placing the cup, the anterior edge of the cup should not overhang the anterior acetabular wall, and there should be adequate coverage of the host bone above the cup to prevent cup loosening. For patients with acetabular dysplasia, where the cup center-to-edge angle is <0º, the surgeon should consider placing the cementless cup more medially or with the hip center slightly higher to stabilize the cup [23]. It is also possible to use a cemented cup and graft the bulk bone of the acetabulum if necessary [24, 25]. The cup size can be predicted by CT at the level of the acetabulum, and the distance from the ilioischial line on CT to the cup can be measured and incorporated into the radiographic template to adjust the simulation, especially for patients with acetabular dysplasia. Global offset is also vital for hip function [26]; however, the need for cup medialization may require the selection of a high offset stem.
The iliac crest line is also an important landmark for identifying hip joint morphologies, including coxa profunda and protrusio acetabuli. In their CT-based imaging study, O'Sullivan et al. [8] identified the location of the ilioischial line in 10 cadaveric pelvises to be posterior to the acetabulum and at the radiographic interface of the cortical bone of the posterior column. However, a 5° rotation of the pelvis yielded a change in the radiographic location of the tear drop used as a reference for the ilioischial line. In their investigation of the morphology of the acetabulum in coxa profunda, Fujii et al. [9] reported the ilioischial line to be located at the outermost point of the sciatic acetabular medial wall. We hypothesized that the ilioischial line would be located in the bony region where the radiographic beam overlaps to the greatest extent with the cortical bone of the acetabular medial wall on axial CT imaging. This required considering the angle of the radiographic beam relative to the ilioischial line. On a related note, the X-ray beam offset has previously been described to measure acetabular cup anteversion after THA [26–30]. Goergen and Resnick proposed a correction factor of 5° for the radiographic angle for AP pelvis radiographs centered on the symphysis pubis; this correction factor was, however, variable for different relative cup positions [29]. Using trigonometry, Widmer calculated a radiographic beam offset of 5.46° to measure the cup anteversion angle [27]. Measuring the distance from the pelvic center to the ilioischial line on AP radiographs obtained with patients in the supine position, we calculated a radiographic angle of 4°. Our hypothesis regarding the relationship between line α and ilioischial line is theoretical but consistent with previous opinions [8, 9]. Our study aimed to prove that line α is the bony region representing the ilioischial line on axial CT images by confirming that distances A and B were nearly equivalent. We showed that these two distances are strongly correlated (R2=0.87), with an average difference of 0.2 mm (calculated from the regression line y=x+0.2). The ilioischial line was located in the cortical bone area of the acetabular medial wall, which overlapped to the greatest extent with line α. The average width of the cortical bone defining the ilioischial line was about 2 cm in width, but the measurement reliability was only fair-to-moderate. The zone classification showed that the ilioischial line concentrated in zone 3 and was located posteriorly in almost 90% of cases. The more posterior the zone containing the ilioischial line, the shorter the cup distance. We did exclude cases where the cup and ilioischial line either touched or overlapped; therefore, there is a possibility for localization of the ilioischial line in zone 4 in some cases.
Measurement errors of distances A and B can be caused by the magnification rate of the X-ray beam, differences in pelvic rotation in the supine position during imaging, and metal artifacts in the cup. O'Sullivan et al. [8] reported that the relationship between the ilioischial line and tear drop could be altered with as little as 5° of pelvic rotation. The Bland–Altman plot identified only 3 outliers among the 51 hips included in our study. Therefore, we can assume that our correction of 0.9 for radiographic magnification was appropriate. Errors are also possible on the CT image, although the software does correct for these greatly [31]. A specific issue is the limit of 3 mm between slices; thus, the image used for measurement may not be exactly through the femoral head center. Additionally, the X-ray beam offset of 4° was calculated only in Japanese patients, with the possibility of differences based on race. However, we did note the high correlation between distances A and B. Therefore, it is reasonable to consider line α as representing the point of the ilioischial line at the CT slice level. The error using this method was within 2 mm, with this level of accuracy being sufficient for clinical applications, such as preoperative planning of THA cup placement.
On the pelvic CT axial image, the cortical bone region at the contact point of the acetabular medial wall, which overlaps the line tilted at 4° from the floor to the hypotenuse, is the bony component representing the ilioischial line. As such, the offset distance based on the ilioischial line of the cup can be measured on the CT axial slice. This would provide accurate preoperative templating from radiographs and CT images without the need for specialized software, which could improve THA outcomes.
Limitations
The limitations of our study are as follows. First, the number of cases was small. Second, we did not evaluate the pelvic tilt on radiographs and CT images. Third, although it is desirable to calculate and measure the radiographic offset beam based on the size of each pelvis, it is complicated in actual clinical practice; therefore, we simplified it to 4º. However, in individual cases, the radiation angle ranges from 3.3° to 4.3°, and this difference has little effect on distance B. Finally, the use of our methods for preoperative planning needs to be evaluated.