Abstract
Background
Periacetabular osteotomy (PAO) increases acetabular coverage of the femoral head and medializes the hip’s center, restoring normal joint biomechanics. Past studies have reported data regarding the degree of medialization achieved by PAO, but measurement of medialization has never been validated through a comparison of imaging modalities or measurement techniques. The ilioischial line appears to be altered by PAO and may be better visualized at the level of the inferior one-third of the femoral head, thus, an alternative method of measuring medialization that begins at the inferior one-third of the femoral head may be beneficial.
Questions/purposes
(1) What is the true amount and variability of medialization of the hip’s center that is achieved with PAO? (2) Which radiographic factors (such as lateral center-edge angle [LCEA] and acetabular inclination [AI]) correlate with the degree of medialization achieved? (3) Does measurement of medialization on plain radiographs at the center of the femoral head (traditional method) or inferior one-third of the femoral head (alternative method) better correlate with true medialization? (4) Are intraoperative fluoroscopy images different than postoperative radiographs for measuring hip medialization?
Methods
We performed a retrospective study using a previously established cohort of patients who underwent low-dose CT after PAO. Inclusion criteria for this study included PAO as indicated for symptomatic acetabular dysplasia, preoperative CT scan, and follow-up between 9 months and 5 years. A total of 333 patients who underwent PAO from February 2009 to July 2018 met these criteria. Additionally, only patients who were between 16 and 50 years old at the time of surgery were included. Exclusion criteria included prior ipsilateral surgery, femoroacetabular impingement (FAI), pregnancy, neuromuscular disorder, Perthes-like deformity, inadequate preoperative CT, and inability to participate. Thirty-nine hips in 39 patients were included in the final study group; 87% (34 of 39) were in female patients and 13% (5 of 39 hips) were in male patients. The median (range) age at the time of surgery was 27 years (16 to 49). Low-dose CT images were obtained preoperatively and at the time of enrollment postoperatively; we also obtained preoperative and postoperative radiographs and intraoperative fluoroscopic images. The LCEA and AI were assessed on plain radiographs. Hip medialization was assessed on all imaging modalities by an independent, blinded assessor. On plain radiographs, the traditional and alternative methods of measuring hip medialization were used. Subgroups of good and fair radiographs, which were determined by the amount of pelvic rotation that was visible, were used for subgroup analyses. To answer our first question, medialization of all hips was assessed via measurements made on three-dimensional (3-D) CT hip reconstruction models. For our second question, Pearson correlation coefficients, one-way ANOVA, and the Student t-test were calculated to assess the correlation between radiographic parameters (such as LCEA and AI) and the amount of medialization achieved. For our third question, statistical analyses were performed that included a linear regression analysis to determine the correlation between the two radiographic methods of measuring medialization and the true medialization on CT using Pearson correlation coefficients, as well as 95% confidence intervals and standard error of the estimate. For our fourth question, Pearson correlation coefficients were calculated to determine whether using intraoperative fluoroscopy to make medialization measurements differs from measurements made on radiographs.
Results
The true amount of medialization of the hip center achieved by PAO in our study as assessed by reference-standard CT measurements was 4 ± 3 mm; 46% (18 of 39 hips) were medialized 0 to 5 mm, 36% (14 hips) were medialized 5 to 10 mm, and 5% (2 hips) were medialized greater than 10 mm. Thirteen percent (5 hips) were lateralized (medialized < 0 mm). There were small differences in medialization between LCEA subgroups (6 ± 3 mm for an LCEA of ≤ 15°, 4 ± 4 mm for an LCEA between 15° and 20°, and 2 ± 3 mm for an LCEA of 20° to 25° [p = 0.04]). Hips with AI ≥ 15° (6 ± 3 mm) achieved greater amounts of medialization than did hips with AI of < 15° (2 ± 3 mm; p < 0.001). Measurement of medialization on plain radiographs at the center of the femoral head (traditional method) had a weaker correlation than using the inferior one-third of the femoral head (alternative method) when compared with CT scan measurements, which were used as the reference standard. The traditional method was not correlated across all radiographs or only good radiographs (r = 0.16 [95% CI -0.17 to 0.45]; p = 0.34 and r = 0.26 [95% CI -0.06 to 0.53]; p = 0.30), whereas the alternative method had strong and very strong correlations when assessed across all radiographs and only good radiographs, respectively (r = 0.71 [95% CI 0.51 to 0.84]; p < 0.001 and r = 0.80 [95% CI 0.64 to 0.89]; p < 0.001). Measurements of hip medialization made on intraoperative fluoroscopic images were not found to be different than measurements made on postoperative radiographs (r = 0.85; p < 0.001 across all hips and r = 0.90; p < 0.001 across only good radiographs).
Conclusion
Using measurements made on preoperative and postoperative CT, the current study demonstrates a mean true medialization achieved by PAO of 4 mm but with substantial variability. The traditional method of measuring medialization at the center of the femoral head may not be accurate; the alternate method of measuring medialization at the lower one-third of the femoral head is a superior way of assessing the hip center’s location. We suggest transitioning to using this alternative method to obtain the best clinical and research data, with the realization that both methods using plain radiography appear to underestimate the true amount of medialization achieved with PAO. Lastly, this study provides evidence that the hip center’s location and medialization can be accurately assessed intraoperatively using fluoroscopy.
Level of Evidence
Level III, diagnostic study.
Introduction
Acetabular dysplasia is a common and well-established cause of hip pain and dysfunction resulting from insufficient coverage of the femoral head by the acetabulum [12], and it is often accompanied by lateralization of the hip’s center. The Bernese periacetabular osteotomy (PAO) is the most common treatment for symptomatic acetabular dysplasia as it normalizes acetabular hip coverage, thus stabilizing the hip and restoring normal joint biomechanics [3, 13, 16, 20, 29]. PAO also medializes the hip’s center, to varying degrees, which improves the hips’ biomechanical position. Medialization is commonly assessed intraoperatively using fluoroscopy, and postoperatively, it is measured on AP radiographs of the pelvis. The importance of hip medialization is well established. Studies have shown that medialization of the hip’s center with THA decreases joint reaction force by 13% to 25% and abductor muscle force by 39% to 46%, resulting in better overall hip mechanics and function [1, 5].
Medialization of the hip’s center with PAO has not been investigated adequately. The distance from the femoral head to the ilioischial line has been used to mark the hip position at the center of the femoral head when comparing preoperative and postoperative or intraoperative measurements [8, 9]. However, plain radiographic assessment of lateralization has never been validated with reference to three-dimensional (3-D) imaging. Furthermore, posterior-column osteotomy and associated healing can make this traditional method difficult to use in some patients because of changes in the ilioischial line. In these patients, measuring the hip’s lateralization more inferiorly along the femoral head (one-third of the diameter of the femoral head) often makes it easier to reference the ilioischial line (Fig. 1).
Fig. 1.
A-B Coned AP plain radiographs of the pelvis were used to measure hip position. (A) On this preoperative radiograph, Line 1 represents the traditional measurement method starting from the center of the femoral head and Line 2 represents the alternative method starting at one-third of the diameter of the femoral head and extending to the ilioischial line. (B) On this postoperative radiograph, the ilioischial line is not visible at the level of the middle of the femoral head but is easily visible and possible to measure to when drawing a line starting at one-third of the femoral head diameter.
There is some existing evidence that has sought to assess the amount of medialization achieved by PAO. Using an assessment of plain radiographs, Clohisy et al. [8] reported the amount of hip medialization achieved by the Bernese PAO in patients with acetabular dysplasia, noting that the mean medialization was 9.8 mm in a population with moderate-severe dysplasia. A more recent study assessing survivorship after Bernese PAO demonstrated that surviving hips achieved twice the amount of medialization as those hips that subsequently underwent THA (1.1 mm versus 0.5 mm), suggesting that medialization of the femoral head during PAO alters hip biomechanics in such a way as to improve function and long-term outcomes [31]. Another study showed that medialization using PAO increases the moment arm of the hip abductor muscles and decreases hip loading [11]. These studies show the importance of medialization achieved by PAO, but the relative validity of available approaches to measuring medialization have not, to our knowledge, been studied formally.
We therefore asked: (1) What is the true amount and variability of medialization of the hip’s center that is achieved with PAO? (2) Which radiographic factors (such as lateral center-edge angle [LCEA] and acetabular inclination [AI]) correlate with the degree of medialization achieved? (3) Does measurement of medialization on plain radiographs at the center of the femoral head (traditional method) or inferior one-third of the femoral head (alternative method) better correlate with true medialization? (4) Are intraoperative fluoroscopy images different than postoperative radiographs for measuring hip medialization?
Patients and Methods
Patients
We performed a retrospective study using a previously established cohort of patients undergoing low-dose CT after PAO. All surgeries were performed by one surgeon (JCC), and all patients had undergone low-dose CT during a preoperative evaluation. The inclusion criteria for this study were: PAO as indicated for symptomatic acetabular dysplasia, preoperative CT scan for surgical planning, and follow-up between 9 months and 5 years. A total of 333 patients who underwent PAO from February 2009 to July 2018 met these inclusion criteria (Fig. 2). Additionally, only patients between the ages of 16 and 50 at the time of surgery were included in our study, leading to 44% (145 of 333) of patients being excluded. Patients who returned to our clinic for routine follow-up between 9 months and 5 years were offered enrollment in the current study and underwent a low-dose CT scan in lieu of a routine clinical radiographic series. Exclusion criteria included prior ipsilateral surgery, femoroacetabular impingement (FAI), pregnancy, neuromuscular disorder, Perthes-like deformity, inadequate preoperative CT, and being unable to participate (declined research, clinic appointment cancelled, CT scanner malfunction). This resulted in 79% (149 of 188) patients being excluded, leaving 39 patients eligible to be included in the study. For patients who underwent bilateral PAOs, only one hip was included in our analysis (the hip with higher quality preoperative and postoperative imaging). The independent t-test and chi-square test were used to evaluate whether any differences were present between our final cohort and all patients who underwent PAO who were initially eligible with respect to age, preoperative LCEA, AI, and sex distribution; none were detected at the 0.05 significance level.
Fig. 2.
Flow diagram showing the inclusion criteria for our current study; FAI = femoroacetabular impingement.
Thirty-nine hips in 39 patients were included in the final cohort (Table 1). Eighty-seven percent of hips (34 of 39) were in female patients and 13% (5 of 39) were in male patients. The median (range) age at the time of surgery was 27 years (16 to 49). The mean preoperative LCEA was 16° ± 6° and the mean preoperative AI was 14° ± 5°.
Table 1.
Patient demographic and radiographic data for the entire cohort
| Parameter | n = 39 hips |
| Age in years, median (range) | 27 (16-49) |
| Females, % (n) | 87 (34) |
| BMI in kg/m2, mean ± SD | 24 ± 3 |
| Right hips, % (n) | 64 (25) |
| LCEA in °, mean ± SD | |
| Preoperative | 16 ± 6 |
| Postoperative | 28 ± 5 |
| Acetabular inclination in °, mean ± SD | |
| Preoperative | 14 ± 5 |
| Postoperative | 1 ± 4 |
Radiographic Evaluation
Plain radiographic measurements were compared with the reference standard of 3-D CT measurements. All measurements were made by an independent blinded assessor (LMF) who made measurements on each imaging modality; each imaging modality was measured independent of the others. Preoperative and postoperative standing AP pelvis radiographs were obtained using a standardized technique [9]. Postoperative radiographs were taken at routine follow-up visits between 1 and 3 months postoperatively (mean 1.1 ± 0.6 months).
Given the potential effect of alterations in pelvic rotation on the appearance of the ilioischial line, we assessed pelvic rotation on preoperative and postoperative radiographs. “Good” and “fair” pelvic rotation were defined as a distance below or above 10 mm, respectively, between vertical lines through the symphysis and sacrococcygeal joint (Fig. 3) [18, 21]. Patients with good pelvic rotation on both preoperative and postoperative views were classified into the good subgroup. Patients with fair pelvic rotation on either preoperative and/or postoperative radiographs were classified into the fair subgroup. Given that pelvic radiographs sometimes show some alteration of pelvic rotation in clinical practice, we included fair radiographs in the study, but we assessed the effect of suboptimal rotation in subgroup analyses. Preoperatively, 35 radiographs met the criteria for the good subgroup, with four radiographs meeting the criteria for the fair subgroup. Postoperatively, 19 radiographs met the criteria for the good subgroup and 20 met the criteria for the fair subgroup. Eighteen radiographs met the criteria for the good subgroup both preoperatively and postoperatively; thus, there were 21 radiographs in the fair subgroup.
Fig. 3.
A-B The quality of a radiograph was determined by drawing vertical lines through the centers of the sacrococcygeal joint (Line 1) and the pubic symphysis (Line 2) and measuring the horizontal distance between them (Line 3). (A) This is an example of a “poor”-quality radiograph because the distance is greater than 10 mm. (B) This is an example of a “good” radiograph because the distance between the two vertical lines is less than 10 mm.
Primary and Secondary Study Outcomes
Our primary study goal was to determine the amount of hip medialization by comparison of preoperative and postoperative CT scans. We used a specific low-dose CT protocol developed at this institution for use in young adults with hip conditions. The CT scan in this protocol has radiation exposure (approximately 1.3 msSv) that is equivalent to that of about three to five AP pelvis radiographs [23]. CT images were analyzed with a computer modeling program (Dyonics Plan, Smith and Nephew Endoscopy, Andover, MA, USA). This allowed for neutral pelvis positioning with respect to pelvic tilt and obliquity [26]. CT measurements were considered the reference standard for measurements of true medialization. To determine medialization on CT, we measured the distance from the center of the femoral head to a vertical line representing the midline [2, 19] (Fig. 4) on preoperative and postoperative CT images. The postoperative distance was then subtracted from the preoperative distance to determine the net medialization of the hip. Measurements were made to the midline on CT (as opposed to the ilioischial line on radiographs).
Fig. 4.
Medialization measurement made on CT. A line is created connecting the center of the femoral head and a vertical line representing the midline.
Additionally, we determined whether the degree of medialization on CT correlated with plain radiographic measures of dysplasia severity. The LCEA [27] and AI (Tönnis angle) [24] were measured on preoperative radiographs to characterize the severity of acetabular dysplasia. Past studies have shown that these measurements have high interobserver and intraobserver reliability [15, 17]. Hips were then grouped by LCEA (≤ 15°, 15-20°, and 20-25°) and AI (< 15° or ≥ 15°) for subgroup analyses [6, 7, 16].
Our third study goal was to validate a plain radiographic measurement of the change in hip medialization. Hip medialization was measured on AP radiographs using two methods. First, we used a traditional method, which determines medialization by measuring the distance between the medial aspect of the femoral head and the ilioischial line at the level of the center of the femoral head on preoperative and postoperative radiographs (Fig. 1). The distance measured on the postoperative radiograph is then subtracted from the preoperative distance to determine the net hip medialization (positive indicating medialization; negative indicating lateralization) [8, 9]. Past studies have reported high reliability with this measurement, with intraclass correlation coefficients of 0.94 and 0.87 for intraobserver and interobserver reliability, respectively [10, 16]. Second, we measured medialization with an alternate method in which we started the measurement from the medial aspect of the femoral head at one-third of the diameter of the femoral head from inferior to superior (Fig. 5).
Fig. 5.
The starting point for the alternate medialization measurement was the medial aspect of the femoral head, one-third of the diameter of the femoral head from bottom to top.
Our final study goal was to determine whether intraoperative fluoroscopic images could be used as an acceptable alternative to postoperative radiographs for assessment of hip medialization. Intraoperative fluoroscopic images were available for 95% (37 of 39) of hips after PAO. A standardized technique was used for pelvic rotation and tilt with fluoroscopy. Pelvic alignment was considered acceptable when the obturator foramina were symmetric, the coccyx was in line with the pubic symphysis, and the pubic symphysis was vertical. When correct alignment was achieved, a fluoroscope was moved over the operated-on hip and an AP image was taken. Medialization was measured using the alternate measurement method.
Statistical Analysis
Our statistical analyses included a linear regression analysis to determine the correlation between the two radiographic methods of measuring medialization and the true medialization on CT using Pearson correlation coefficients, as well as 95% confidence intervals and standard error of the estimate. We also used Pearson correlation coefficients to determine the correlation of the LCEA and AI with CT medialization, as well as the correlation between measurements made on postoperative radiographs and those made on intraoperative fluoroscopic images. The correlation was rated as very weak (< 0.2), weak (0.20 to 0.39), moderate (0.40 to 0.59), strong (0.60 to 0.79), or very strong (0.80 to 1). We used a t-test to assess differences in medialization between two subgroups, and we used a one-way ANOVA to assess differences among three subgroups. A p value of less than 0.05 indicated significance.
Results
Medialization Achieved by Surgery
As assessed using the reference standard CT measurement, the mean medialization achieved with PAO was 4 ± 3 mm. Forty-six percent (18 of 39) of hips were medialized 0 to 5 mm, 36% (14) were medialized 5 to 10 mm, and 5% (2) were medialized greater than 10 mm. Thirteen percent (5) were lateralized (medialized < 0 mm).
Factors that Correlate with Medialization
AI was weakly correlated with the degree of medialization achieved, whereas LCEA was not correlated (r = 0.36; p = 0.03 for AI and r = -0.27; p = 0.10 for LCEA). We found small differences between LCEA subgroups (mean medialization on CT was 6 ± 3 mm for an LCEA of ≤ 15°, 4 ± 4 mm for an LCEA between 15° and 20°, and 2 ± 3 mm for an LCEA of 20° to 25°; p = 0.04). Greater medialization was present in hips with an acetabular inclination ≥ 15° than in those with an acetabular inclination of < 15° (6 ± 3 mm for AI ≥ 15° and 2 ± 3 mm for AI < 15°, mean difference 4 mm [95% CI 2 to 6]; p < 0.001) (Table 2).
Table 2.
Correlation coefficients between traditional and alternative methods of measuring medialization on radiographs and CT medialization measurements
| Radiographic measurement | Pearson correlation coefficient (95% CI) | p value | Standard error of the estimate in mm |
| All hips (n = 39) | |||
| Traditional method | 0.16 (-0.17 to 0.45) | 0.34 | 3 |
| Alternative method | 0.71 (0.51-0.84) | <0.001 | 2 |
| Good radiographs only (n = 18) | |||
| Traditional method | 0.26 (-0.06 to 0.53) | 0.30 | 3 |
| Alternative method | 0.80 (0.64-0.89) | <0.001 | 2 |
Which Approach to Measuring Medialization is Better?
The alternative method of measuring medialization was found to be more closely correlated with the reference standard CT measurements than was the traditional method of measuring medialization (Fig. 6). The mean medialization was 4 ± 3 mm for CT, compared with 1 ± 3 mm for the traditional measurement and 2 ± 4 mm for our alternative method of measurement (Table 3). The alternative method had a strong correlation (r = 0.71 [95% CI 0.51 to 0.84]; p < 0.001) when all hips were included and a very strong correlation (r = 0.80 [95% CI 0.64 to 0.89]; p < 0.001) when only radiographs with good preoperative and postoperative rotation were included. The traditional method had no correlation with CT both when all hips were included (r = 0.16 [95% CI -0.17 to 0.45]; p = 0.34) and when only good radiographs were included (r = 0.26 [95% CI -0.06 to 0.53]; p = 0.30). With respect to the alternative measurement, the measurements of 54% (21 of 39) of hips were within 2 mm of the CT measurement, and 82% (32 of 39) of hips were within 5 mm when all radiographs were assessed. Additionally, among the 18 hips with good radiographs, 13 had measurements within 2 mm of the CT measurement and 17 had measurements within 5 mm. Lastly, 74% (29 of 39 hips) of medialization measurements made on CT were greater than those made with the alternative measurement method on plain radiographs, indicating the tendency for plain radiographs to underestimate medialization.
Fig. 6.
A-B A hip was medialized with PAO, as shown on (A) preoperative and (B) postoperative radiographs. Medialization of this hip was 9 mm via the alternate method of measurement (measurement is made starting at one-third of the diameter of the femoral head from inferior to superior) on a plain radiograph and 8 mm with the reference standard CT method of measurement.
Table 3.
Selected publications reporting medialization measurements after PAO
| Study | Net medialization in mm | Preoperative medialization distance in mm | Postoperative medialization distance in mm | Preoperative LCEA in ° | Preoperative acetabular inclination in ° |
| Current study | Alternative: 2 ± 4 Traditional: 1 ± 3 CT: 4 ± 3 |
16 ± 4 | 13 ± 4 | 16 ± 6 | 14 ± 5 |
| Beaulé et al. [4] | 1a (-5 to 9) | 17 (-24 to 28) | 15 (0-55) | ||
| Clohisy et al. [8] | 10 | 17 | 8 | 0 | 25 |
| Kralj et al. [14] | -5 | 15 (7-26) | |||
| Siebenrock et al. [22] | 6 | 16 (6-30) | 10 (-9 to 24) | 6 (-24 to 25) | 26 (12-50) |
| Troelsen et al. [25] | 0 | 2 (1-3) | 2 (0-3) | 11a (-29 to 30) | 20a (4-53) |
| Zhu et al. [30] | 7 | 15 (9-21) | 9 (5-14) | 6 (-5 to 15) | |
| Ziran et al. [31] | 1 ± 4b 1 ± 3c |
9 ± 5b 5 ± 2c |
8 ± 4b 4 ± 2c |
0 ± 12 | 26 ± 8 |
Denotes median value. Values are otherwise reported as mean (range) or mean ± SD.
Denotes surviving hips in this study.
Denotes failing hips in this study. LCEA = lateral center-edge angle.
Correlation of Intraoperative Fluoroscopy with Plain Radiographs
Intraoperative fluoroscopy measurements (available in 37 of 39 hips) correlated well with postoperative radiographic measurements. The mean medialization distance on postoperative radiographs was 13 ± 4 mm, and the mean distance on intraoperative fluoroscopy was 11 ± 4 mm. Intraoperative fluoroscopic measurements were very strongly correlated with postoperative radiographic measurements across all 37 hips (r = 0.85; p < 0.001) and among the 18 hips with good postoperative radiographs (r = 0.90; p < 0.001). Among all hips, 46% (17 of 37) had measurements within 2 mm of the plain radiographic measurement and 84% (31 of 37) of hips had measurements within 5 mm. Additionally, among only hips with good radiographs, 56% (10 of 18) of hips had measurements within 2 mm of the plain radiographic measurement and 89% (16 of 18) had measurements within 5 mm. Lastly, 84% of measurements (31 of 37) made on plain radiographs were greater than those made with intraoperative fluoroscopy.
Discussion
The Bernese PAO is the most common treatment for symptomatic acetabular dysplasia. It can normalize acetabular coverage of the hip and medialize the hip’s center, stabilizing the hip and restoring normal joint biomechanics. Although prior studies [4, 8, 14, 22, 25, 30, 31] have reported the amount of medialization achieved via PAO as measured on plain radiographs, to our knowledge, this measurement method has not been validated with reference to 3-D imaging and its comparability with intraoperative fluoroscopy. Furthermore, an alternative method of measuring medialization on plain radiographs (starting the measurement at the lower one-third of the femoral head’s diameter) may provide more reliable and more accurate measurements because it avoids issues with landmark identification that might occur with the traditional measurement method. We found that measurement of medialization on plain radiographs strongly correlated with measurements of medialization on 3-D imaging and intraoperative fluoroscopy. Additionally, we found that the alternative method of measuring medialization on plain radiographs was superior to the traditional measurement method.
Limitations
This study has several limitations. First, although we used a standardized technique for obtaining pelvis radiographs, it was possible to experience unwanted tilt and rotation, which can introduce error and variability into measurements. For all radiographs, we measured the amount of rotation using a standardized method and created subgroups based on this measurement. Subgroup analyses were performed to account for variation in rotation on radiographs, and thus we limited the amount of influence these factors could have on our results. Although radiographs were performed standing, the CT was performed in the supine position. However, measurements of the distance between the femoral head center and midline would not be expected to differ on CT in a standing or supine position, thus allowing it to be compared with standing radiographs. Second, different landmarks were used when making measurements on radiographs and fluoroscopy and those on CT images. All changes were measured within a modality (radiograph or CT) but not across modalities. Although this could represent a source of bias within our study, the use of consistent landmarks within each imaging modality, as well as preoperatively and postoperatively, lends itself to consistent results. Third, our final study group represents a small proportion of all patients who underwent PAO at our institution and could be a source of selection bias. However, all patients with eligible follow-up during the study period were offered enrollment. When comparing the current study group to those not included from the larger cohort, we noted no differences.
In addition, the amount of medialization in our study may not be generalizable to all surgeons as it is indeed technique-dependent, and we included only one experienced surgeon. There is substantial variability in the reported amount of medialization achieved with PAO [4, 8, 14, 22, 25, 30, 31], but results of the current study question the accuracy of previous measurements made with traditional techniques. Finally, we used postoperative radiographs from 1 to 4 months after PAO. Even though it is theoretically possible that the position of the hip’s center could change in the interval until the CT was performed, this is unlikely. At 1 month postoperatively, substantial healing of the PAO usually is present and generally allows weightbearing. We have not experienced changes in position of the acetabular reduction after this point. Additionally, the hip center could be affected by the presence of osteoarthritis, but no patients encountered this in the current study.
Medialization Achieved by Surgery
There is much variability in the amount of medialization achieved with PAO for acetabular dysplasia based on a radiographic assessment [4, 8, 14, 22, 25, 30, 31]. Reported values range from -5 mm [14] (5 mm of lateralization of the hip’s center) to 9.8 mm of medialization [8] (Table 3). Our finding of a mean of 4 mm of medialization on reference standard CT measurements falls between these reported results. Variation in the preoperative position of the femoral head is notable and may affect these results. Clohisy et al. [8] and Siebenrock et al. [22] reported the largest medialization values, but they also had the largest preoperative medialization measurements (lateralized hip center). These are the oldest studies reporting medialization data and may represent different cohorts of patients with more severe deformities than those in more recent studies. Although many studies have reported medialization data as measured on plain radiographs [4, 8, 14, 22, 25, 30, 31], the current study was, to the best of our knowledge, the first to report medialization data as assessed by CT data for the Bernese PAO.
Factors that Correlate with Medialization
In the current study, there was a larger amount of medialization in patients with LCEA ≤ 15° compared with those with an LCEA between 20° to 25°, suggesting hips with more severe dysplasia, as assessed by the LCEA, experience greater hip center medialization. Other studies have reported comparable results, with mean study preoperative LCEAs of -0°, 6°, and 6° resulting in mean medializations of 10 mm, 6 mm, and 7 mm, respectively [8, 22, 30]. Conversely, other studies with mean preoperative LCEA measurements of 17° and 15° reported 1 mm of medialization and 5 mm of lateralization, respectively [4, 14]. This makes sense because hips with more severe dysplasia will require greater amounts of medialization to obtain adequate coverage of the femoral head by the acetabulum. Our study also found that hips with preoperative acetabular inclination ≥ 15° experienced greater amounts of medialization than did those with acetabular inclination less than 15°. Other studies have conflicting results. Among studies reporting that preoperative acetabular inclination was as large as 26°, two studies found large degrees of medialization [8, 22], but others reported much more modest degrees of medialization [25, 31].
Which Approach to Measuring Medialization is Better?
Although the medialization measurement has been used in studies on PAO [4, 8, 14, 22, 25, 30, 31], to our knowledge, it has not been fully validated. We found that the traditional method of measuring medialization poorly correlated with medialization measurements made with reference standard CT data and may not be a reliable marker of the hip’s center on plain radiographs. Furthermore, the alternative method of measuring medialization strongly correlated with CT measurement data when all radiographs were used, and it correlated very strongly when only radiographs showing acceptable amounts of rotation were used. Both the traditional and alternative methods using plain radiographs underestimated the true medialization by 3 mm and 2 mm, respectively. Despite this, we showed that the alternative method of measuring medialization is valid for assessing medialization of the hip’s center and is superior to the method that has been traditionally used.
Correlation of Intraoperative Fluoroscopy with Plain Radiographs
Plain radiographs are useful for assessing the location of the hip’s center preoperatively and postoperatively, but this assessment must also be made accurately intraoperatively when medializing the PAO fragment. We examined the correlation of intraoperative fluoroscopic measures of medialization to postoperative measurements made on plain radiographs. We found that measurements made on intraoperative fluoroscopy correlated very strongly compared with all radiographs as well as only radiographs showing acceptable amounts of rotation. These findings suggest that intraoperative fluoroscopy is a valid medium to assess hip medialization. This parallels the result of other studies proving that other radiographic measures of the hip (such as the LCEA, anterior center-edge angle, and AI) can be made accurately on intraoperative fluoroscopy [15, 28].
Conclusion
Our work provides new data regarding the degree of medialization obtained with PAO as assessed with reference standard CT measurements. Additionally, we provided evidence that the traditional method of measuring medialization may not be accurate, and our alternative method is a superior way of assessing the hip center’s location. We suggest transitioning to this alternative method to obtain the best clinical and research data, with the realization that both methods using plain radiography appear to underestimate the true amount of medialization achieved with PAO. Studies on this topic may underestimate the true degree of medialization. Lastly, our work suggests that fluoroscopy is an acceptable and valid method of assessing hip medialization intraoperatively.
Footnotes
The institution of one or more of the authors (JCC, JJN) has receiving funding during the study period from Zimmer Biomet. The institution of one or more of the authors (JCC) has received funding during the study period from the Curing Hip Disease Fund and the Jacqueline and W. Randolph Baker Fund.
One of the authors certifies that he (JCC), or a member of his immediate family, has received or may receive payments or benefits, during the study period, in an amount of USD 10,000 to 100,000 from Microport Orthopaedics and in an amount of USD 10,000 to 100,000 from Zimmer Biomet.
One of the authors certifies that he (JNN), or a member of his immediate family, has received or may receive payments or benefits, during the study period, in an amount of less than USD 10,000 from Responsive Arthroscopy and in an amount of USD 10,000 to 100,000 from Smith and Nephew.
All ICMJE Conflict of Interest Forms for authors and Clinical Orthopaedics and Related Research® editors and board members are on file with the publication and can be viewed on request.
Clinical Orthopaedics and Related Research® neither advocates nor endorses the use of any treatment, drug, or device. Readers are encouraged to always seek additional information, including FDA approval status, of any drug or device before clinical use.
Each author certifies that his or her institution approved the human protocol for this investigation and that all investigations were conducted in conformity with ethical principles of research.
References
- 1.Abolghasemian M, Samiezadeh S, Jafari D, Bougherara H, Gross AE, Ghazavi MT. Displacement of the hip center of rotation after arthroplasty of Crowe III and IV dysplasia: a radiological and biomechanical study. J Arthroplasty. 2013;28:1031-1035. [DOI] [PubMed] [Google Scholar]
- 2.Alazzawi S, Field MH, Bardakos NV, Freeman MAR, Field RE. The position of the centre of the femoral head relative to the midline of the pelvis: a consistent landmark in total knee replacement surgery. Knee. 2012;19:827-831. [DOI] [PubMed] [Google Scholar]
- 3.Armiger RS, Armand M, Tallroth K, Lepistö J, Mears SC. Three-dimensional mechanical evaluation of joint contact pressure in 12 periacetabular osteotomy patients with 10-year follow-up. Acta Orthop . 2009;80:155-161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Beaulé PE, Dowding C, Parker G, Ryu JJ. What factors predict improvements in outcomes scores and reoperations after the bernese periacetabular osteotomy? Clin Orthop Relat Res. 2015;473:615-622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bonnin MP, Archbold PHA, Basiglini L, Selmi TA, Beverland DE. Should the acetabular cup be medialised in total hip arthroplasty? Hip Int . 2011;21:428-435. [DOI] [PubMed] [Google Scholar]
- 6.Byrd JWT, Jones KS. Hip arthroscopy in the presence of dysplasia. Arthroscopy . 2003;19:1055-1060. [DOI] [PubMed] [Google Scholar]
- 7.Clohisy JC, Ackerman J, Baca G, et al. Patient-reported outcomes of periacetabular osteotomy from the prospective ANCHOR cohort study. J Bone Joint Surg Am. 2017;99:33-41 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Clohisy JC, Barrett SE, Gordon JE, Delgado ED, Schoenecker PL. Medial translation of the hip joint center associated with the Bernese periacetabular osteotomy. Iowa Orthop J . 2004;24:43-48. [PMC free article] [PubMed] [Google Scholar]
- 9.Clohisy JC, Carlisle JC, Beaulé PE, et al. A systematic approach to the plain radiographic evaluation of the young adult hip. J Bone Joint Surg Am . 2008;90:47-66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Clohisy JC, Carlisle JC, Trousdale R, et al. Radiographic evaluation of the hip has limited reliability. Clin Orthop Relat Res. 2009;467:666-675. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gaffney BMM, Clohisy JC, Van Dillen LR, Harris MD. The association between periacetabular osteotomy reorientation and hip joint reaction forces in two subgroups of acetabular dysplasia. J Biomech . 2020;98:109464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Gala L, Clohisy JC, Beaulé PE. Hip dysplasia in the young adult. J Bone Joint Surg Am. 2016;98:63-73. [DOI] [PubMed] [Google Scholar]
- 13.Ganz R, Klaue K, Vinh TS, Mast JW. A new periacetabular osteotomy for the treatment of hip dysplasias: technique and preliminary results. 1988. Clin Orthop Relat Res 2004;418:3-8. [PubMed] [Google Scholar]
- 14.Kralj M, Mavčič B, Antolič V, Iglič A, Kralj-Iglič V. The Bernese periacetabular osteotomy: clinical, radiographic and mechanical 7–15-year follow-up of 26 hips. Acta Orthop . 2005;76:833-840. [DOI] [PubMed] [Google Scholar]
- 15.Lehmann CL, Nepple JJ, Baca G, Schoenecker PL, Clohisy JC. Do fluoroscopy and postoperative radiographs correlate for periacetabular osteotomy corrections? Clin Orthop Relat Res . 2012;470:3508-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nepple JJ, Clohisy JC. The dysplastic and unstable hip: a responsible balance of arthroscopic and open approaches. Sports Med Arthrosc . 2015;23:180-186. [DOI] [PubMed] [Google Scholar]
- 17.Nepple JJ, Martell JM, Kim Y-J, et al. Interobserver and intraobserver reliability of the radiographic analysis of femoroacetabular impingement and dysplasia using computer-assisted measurements. Am J Sports Med . 2014;42:2393-2401. [DOI] [PubMed] [Google Scholar]
- 18.Putnam SM, Clohisy JC, Nepple JJ. Do changes in pelvic rotation and tilt affect measurement of the anterior center edge angle on false profile radiographs? A cadaveric study. Clin Orthop Relat Res . 2019;477:1066-1072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Rabari YB, Sanap A, Prasad DV, Thadeshwar KH. The distance of the centre of femoral head relative to the midline of the pelvis: a prospective x-ray study of 500 adults. Int J Res Orthop. 2017;3:565-568. [Google Scholar]
- 20.Schmitz MR, Murtha AS, Clohisy JC. Developmental dysplasia of the hip in adolescents and young adults. J Am Acad Orthop Surg . 2020;28:91-101. [DOI] [PubMed] [Google Scholar]
- 21.Siebenrock KA, Kalbermatten DF, Ganz R. Effect of pelvic tilt on acetabular retroversion: a study of pelves from cadavers. Clin Orthop Relat Res . 2003;407:241-248. [DOI] [PubMed] [Google Scholar]
- 22.Siebenrock KA, Scholl E, Lottenbach M, Ganz R. Bernese periacetabular osteotomy. Clin Orthop Relat Res . 1999;363:9-20. [PubMed] [Google Scholar]
- 23.Su AW, Hillen TJ, Eutsler EP, et al. Low-dose computed tomography reduces radiation exposure by 90% compared with traditional computed tomography among patients undergoing hip-preservation surgery. Arthroscopy . 2019;35:1385-1392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Tönnis D. Congenital Dysplasia and Dislocation of the Hip in Children and Adults. Springer Berlin Heidelberg; 1987:100-142. [Google Scholar]
- 25.Troelsen A, Elmengaard B, Søballe K. Medium-term outcome of periacetabular osteotomy and predictors of conversion to total hip replacement. J Bone Joint Surg Am. 2009;91:2169-2179. [DOI] [PubMed] [Google Scholar]
- 26.Wells J, Nepple JJ, Crook K, et al. Femoral morphology in the dysplastic hip: three-dimensional characterizations with CT. Clin Orthop Relat Res . 2017;475:1045-1054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wiberg G. Studies on dysplastic acetabulum and congenital subluxation of the hip joint: with special reference to the complication of osteoarthritis. Acta Chir Scand Suppl . 1939;58:7-135. [Google Scholar]
- 28.Wylie JD, Ferrer MG, McClincy MP, et al. What is the reliability and accuracy of intraoperative fluoroscopy in evaluating anterior, lateral, and posterior coverage during periacetabular osteotomy? Clin Orthop Relat Res. 2019;477:1138-1144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zhao X, Chosa E, Totoribe K, Deng G. Effect of periacetabular osteotomy for acetabular dysplasia clarified by three-dimensional finite element analysis. J Orthop Sci . 2010;15:632-40. [DOI] [PubMed] [Google Scholar]
- 30.Zhu J, Chen X, Cui Y, Shen C, Cai G. Mid-term results of Bernese periacetabular osteotomy for developmental dysplasia of hip in middle aged patients. Int Orthop . 2013;37:589-594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ziran N, Varcadipane J, Kadri O, et al. Ten- and 20-year survivorship of the hip after periacetabular osteotomy for acetabular dysplasia. J Am Acad Orthop Surg . 2019;27:247-255. [DOI] [PubMed] [Google Scholar]