For hip dysplasia, the evaluation of the anterior coverage of the hip joint is as important as that of the lateral coverage. Currently, the ACEA measured on FP radiographs is one of the most widely used parameters for evaluating anterior coverage. It has been proven to be related to clinical results.[2, 7, 25–29] Clinically, radiologists are not always familiar with FP radiographs. Sometimes it seems redundant to take an additional FP radiograph. Thus, hip surgeons often use pelvic AP radiographs to estimate the anterior coverage.
The crossover index has been proven to be related to the ACEA in patients with Femoroacetabular impingement (FAI). [30] However, to our knowledge, no study has included patients with hip dysplasia. In this study, the crossover index had no correlation with the ACEA preoperatively. However, after the deficiency was corrected in the surgery, there was a weak correlation. The result is similar to what has been described in patients with FAI.
In our study, when the crossover sign was used to detect anterior overcoverage in preoperative patients with hip dysplasia, none of the results agreed with the results from the ACEA. Postoperatively, some crossover sign results agreed with the results from ACEA, but there was an unsatisfactory specificity (75.0%) and a worse sensitivity (30.0%). Therefore, the positive crossover sign in the clinic might only be a reminder of excessive anterior coverage. It is not recommended to use only the crossover sign as the decisive parameter. The first impression from the positive crossover sign must be confirmed by other measurements, such as ACEA.
The reason why the agreement between excessive anterior coverage detected by ACEA and detected by the crossover index/the crossover sign is not high originated from their different methodologies. Essentially, the “crossover” is determined by the relationship between the anterior rim and the posterior rim of the acetabulum. A positive crossover sign means that the anterior acetabular wall is partially larger than the posterior acetabular wall. Positive results may be caused by not only excessive coverage of the acetabulum but also deficient posterior coverage. Deficient posterior walls are frequently seen in patients with hip dysplasia, [38] which greatly affects the determination of the anterior coverage. Therefore, using the crossover sign to determine the anterior coverage of the acetabulum might lead to overestimation of the anterior coverage. This also explained why only a postoperative correlation between the crossover sign and the ACEA existed. The corrected posterior wall decreases the error caused by the deficient posterior wall, and the value of the crossover sign in determining anterior coverage increases.
When we measured the AWI, we used the location of the anterior rim of the acetabulum on the femoral head (rather than the relationship between the anterior wall and the posterior wall of the acetabulum used in the crossover sign and crossover index) on the pelvic AP radiograph to determine the anterior coverage. The AWI measurement avoids the effect of a deficient posterior wall of the acetabulum. It is weakly but significantly correlated with the ACEA. When comparing the preoperative and postoperative data, it was found that the preoperative correlation was lower (Fig. 3). On the preoperative scatter plots of the AWI and ACEA, there are some points far from the regression line. When we traced these points, some common characteristics were found in the cases that were represented by these points. In all these cases, the lateral contours of the anterior acetabulum spines (AIIS) are clearly visible on pelvic AP radiographs, as demonstrated in Fig. 3A. On the FP radiograph, it could be seen that the anterior edge of the articular surface was located only slightly anterior to the top of the femoral head center. The supporting area of the articular surface was not sufficient (Fig. 3B). The prominent AIIS instead of the real anterior rim of the socket formed the seemingly normal anterior wall on the pelvic AP radiograph (Fig. 3A). In this case, the anterior coverage of the hip joint was greatly overestimated on the pelvic AP radiograph because of the effect of the AIIS. This can also explain why there is only a moderate correlation between AWI on pelvic radiographs and CT. [39] In the same patient, after the acetabulum was reoriented in PAO, the supporting area of the acetabulum rotated laterally and anteriorly. As the anterior hip coverage significantly increased, the ACEA increased significantly as well (Fig. 3D). However, the AWI did not lead to any changes on the pelvic AP radiographs (Fig. 3C).
For the above reasons, the weakness of estimating the anterior coverage on pelvic AP radiographs lies in the fact that some of the anterior rims identified on AP radiographs are not the real anterior rims of the socket. “Anterior walls” on AP radiographs are enlarged by the AIIS. This interference factor affected our determination of the anterior coverage evaluation to a certain extent.
The rule of thirds classification, which was verified by Professor Tannast, is based on the concept of AWI. Therefore, the method has problems similar to AWI. In this study, we did not replicate results from Professor Tannast. The rule of thirds was verified not by direct measurements of the ACEA or CT data but by the software Hip2Norm (University of Bern, Bern, Switzerland).[32, 37] Essentially, Hip2Norm is also based on measurements from a pelvic AP radiograph (Fig. 4). The principle is calculating the anterior coverage on the horizontal plane by the projection of the anterior and posterior walls of the pelvis on the sphere of the femoral head in the statistical model.[40, 41] Because Hip2Norm is also based on the AP radiograph, the mechanism is the same as AWI and the rule of thirds. The underlying logic of all 3 methods is the same, so a perfect correlation between them does not mean that they reproduce anterior hip coverage perfectly.
The AIIS greatly affected the identification of the anterior rim of the joint surface on pelvic AP radiographs. Only when the AIIS is not prominent does the pelvic anterior wall, which we can see on a pelvic AP radiograph, agree with the anterior wall of the socket. Without the AIIS’ interference, we can see that the rule of thirds will slightly underestimate the anterior coverage with trigonometric calculation. The results of the calculations show that when the ACEA ranges from 20° to 40°, the AWI value approximately ranges between 0.06 and 0.23 rather than ranging from 0.33–0.66, which would be recommended by the rule of thirds (Fig. 5). In this study, the AWI was 0.21 ± 0.11 (0.00 ~ 0.44) for all patients with an ACEA between 20 and 40°, which was also significantly lower than the reference value of 0.33–0.66. The anterior coverage of some hips was underestimated (preoperative 9 hips, accounting for 16.98% of the total and postoperative 34 hips, accounting for 81.13%).
When the AIIS is prominent, the anterior rim identified on a pelvic AP radiograph is located laterally and inferiorly to the anterior border of the joint surface. Because it is impossible to visually distinguish the acetabular surface from the AIIS on the AP radiograph, measuring the anterior wall on a pelvic AP radiograph will result in overestimation of the anterior coverage. For these patients, there is a possibility of obtaining deficient postoperative anterior coverage if we correct the acetabulum based only on pelvic AP radiographs. If FP radiographs can be obtained before surgery, the correction of the anterior coverage can be more accurate (Fig. 3). For patients with hip dysplasia, a certain error rate is obtained when estimating the anterior coverage on pelvic AP radiographs.
Therefore, it is risky to evaluate anterior coverage using pelvic AP radiographs instead of FP radiographs.
There are some limitations in this study: 1. Because the accuracy of anterior coverage measured from CT data in the supine position was greatly affected by pelvic tilt,[13, 17] the gold standard of anterior coverage should be anterior coverage measured on standing CT data. Because we do not perform hip CT on standing position right now, after careful consideration, we finally decided to use ACEA, a 2D parameter, as a standard of all the parameters. There are several reasons that we took this approach. (1) ACEA is taken in the patient’s standing position. It can reproduce the weight-bearing state of the hip joint in the standing position. (2) ACEA has been confirmed in clinical practice and widely accepted by most doctors in clinical practice.[2, 7, 25–29] The acetabulum is a three-dimensional structure, whereas the ACEA on an FP radiograph is only a two-dimensional parameter. We will try to solve this problem in subsequent studies.