Is Anterior Rotation of the Acetabulum Necessary to Normalize Joint Contact Pressure in Periacetabular Osteotomy? A Finite-element Analysis Study : Clinical Orthopaedics and Related Research®

Secondary Logo

Journal Logo

Advanced Search
CLINICAL RESEARCH

Is Anterior Rotation of the Acetabulum Necessary to Normalize Joint Contact Pressure in Periacetabular Osteotomy? A Finite-element Analysis Study

Kitamura, Kenji MD1; Fujii, Masanori MD, PhD1; Iwamoto, Miho MD, PhD1; Ikemura, Satoshi MD, PhD1; Hamai, Satoshi MD, PhD1; Motomura, Goro MD, PhD1; Nakashima, Yasuharu MD, PhD1

Author Information

1Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan

M. Fujii, Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi. Higashi-ku, Fukuoka 812-8582, Japan, Email: [email protected]

This work was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (No. JP18K09109, JP21K09281).

Each author certifies that there are no funding or commercial associations (consultancies, stock ownership, equity interest, patent/licensing arrangements, etc.) that might pose a conflict of interest in connection with the submitted article related to the author or any immediate family members.

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.

Ethical approval for this study was obtained from the Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan (approval number 30-137).

Clinical Orthopaedics and Related Research 480(1):p 67-78, January 2022. | DOI: 10.1097/CORR.0000000000001893

Abstract

Background 

Inappropriate sagittal plane correction can result in an increased risk of osteoarthritis progression after periacetabular osteotomy (PAO). Individual and postural variations in sagittal pelvic tilt, along with acetabular deformity, affect joint contact mechanics in dysplastic hips and may impact the direction and degree of acetabular correction. Finite-element analyses that account for physiologic pelvic tilt may provide valuable insight into the effect of PAO on the contact mechanics of dysplastic hips, which may lead to improved acetabular correction during PAO.

Questions/purposes 

We performed virtual PAO using finite-element models with reference to the standing pelvic position to clarify (1) whether lateral rotation of the acetabulum normalizes the joint contact pressure, (2) risk factors for abnormal contact pressure after lateral rotation of the acetabulum, and (3) whether additional anterior rotation of the acetabulum further reduces contact pressure.

Methods 

Between 2016 and 2020, 85 patients (92 hips) underwent PAO to treat hip dysplasia. Eighty-two patients with hip dysplasia (lateral center-edge angle < 20°) were included. Patients with advanced osteoarthritis, femoral head deformity, prior hip or spine surgery, or poor-quality images were excluded. Thirty-eight patients (38 hips) were eligible to participate in this study. All patients were women, with a mean age of 39 ± 10 years. Thirty-three women volunteers without a history of hip disease were reviewed as control participants. Individuals with a lateral center-edge angle < 25° or poor-quality images were excluded. Sixteen individuals (16 hips) with a mean age of 36 ± 7 years were eligible as controls. Using CT images, we developed patient-specific three-dimensional surface hip models with the standing pelvic position as a reference. The loading scenario was based on single-leg stance. Four patterns of virtual PAO were performed in the models. First, the acetabular fragment was rotated laterally in the coronal plane so that the lateral center-edge angle was 30°; then, anterior rotation in the sagittal plane was added by 0°, 5°, 10°, and 15°. We developed finite-element models for each acetabular position and performed a nonlinear contact analysis to calculate the joint contact pressure of the acetabular cartilage. The normal range of the maximum joint contact pressure was calculated to be < 4.1 MPa using a receiver operating characteristic curve. A paired t-test or Wilcoxon signed rank test with Bonferroni correction was used to compare joint contact pressures among acetabular positions. We evaluated the association of joint contact pressure with the patient-specific sagittal pelvic tilt and acetabular version and coverage using Pearson or Spearman correlation coefficients. An exploratory univariate logistic regression analysis was performed to identify which of the preoperative factors (CT measurement parameters and sagittal pelvic tilt) were associated with abnormal contact pressure after lateral rotation of the acetabulum. Variables with p values < 0.05 (anterior center-edge angle and sagittal pelvic tilt) were included in a multivariable model to identify the independent influence of each factor.

Results 

Lateral rotation of the acetabulum decreased the median maximum contact pressure compared with that before virtual PAO (3.7 MPa [range 2.2-6.7] versus 7.2 MPa [range 4.1-14 MPa], difference of medians 3.5 MPa; p < 0.001). The resulting maximum contact pressures were within the normal range (< 4.1 MPa) in 63% of the hips (24 of 38 hips). The maximum contact pressure after lateral acetabular rotation was negatively correlated with the standing pelvic tilt (anterior pelvic plane angle) (ρ = -0.52; p < 0.001) and anterior center-edge angle (ρ = -0.47; p = 0.003). After controlling for confounding variables such as the lateral center-edge angle and sagittal pelvic tilt, we found that a decreased preoperative anterior center-edge angle (per 1°; odds ratio 1.14 [95% CI 1.01-1.28]; p = 0.01) was independently associated with elevated contact pressure (≥ 4.1 MPa) after lateral rotation; a preoperative anterior center-edge angle < 32° in the standing pelvic position was associated with elevated contact pressure (sensitivity 57%, specificity 96%, area under the curve 0.77). Additional anterior rotation further decreased the joint contact pressure; the maximum contact pressures were within the normal range in 74% (28 of 38 hips), 76% (29 of 38 hips), and 84% (32 of 38 hips) of the hips when the acetabulum was rotated anteriorly by 5°, 10°, and 15°, respectively.

Conclusion 

Via virtual PAO, normal joint contact pressure was achieved in 63% of patients by normalizing the lateral acetabular coverage. However, lateral acetabular rotation was insufficient to normalize the joint contact pressure in patients with more posteriorly tilted pelvises and anterior acetabular deficiency. In patients with a preoperative anterior center-edge angle < 32° in the standing pelvic position, additional anterior rotation is expected to be a useful guide to normalize the joint contact pressure.

Clinical Relevance 

This virtual PAO study suggests that biomechanics-based planning for PAO should incorporate not only the morphology of the hip but also the physiologic pelvic tilt in the weightbearing position in order to customize acetabular reorientation for each patient.

Introduction

Periacetabular osteotomies (PAOs) are performed in young adults with symptomatic hip dysplasia to delay or prevent the development of hip osteoarthritis [5, 55]. PAO improves acetabular coverage of the femoral head and reduces abnormal joint contact pressure through multiplanar acetabular correction [1, 16, 17, 25, 29]. Although favorable intermediate-term to long-term outcomes of PAO have been reported, previous studies have shown that inadequate acetabular correction is a risk factor for an inferior outcome, along with advanced age, Tönnis grade ≥ 2, and joint incongruity [26, 33, 48, 50, 53]. Considering substantial individual variations in acetabular version and deficiency types, acetabular correction must be customized for each patient rather than applying uniform correction [10, 11, 34, 57].

Although the lateral center-edge angle and Tönnis angle are the most common parameters used to assess appropriate acetabular correction, a recent study [53] revealed that the anterior center-edge angle has a greater impact on improving the natural history of the disease after PAO. Hip dysplasia typically manifests as anterolateral acetabular deficiency [10, 11], in which shearing stress and contact stress are concentrated on the anterolateral acetabular rim [13, 22]. Therefore, adequate sagittal plane correction of the anterior under-coverage is crucial for successful hip preservation [15, 20, 40]. However, care should be taken not to overcorrect the sagittal plane because excessive anterior coverage can result in iatrogenic femoroacetabular impingement [26, 33], and anterior rotation of the acetabulum should be reserved for patients with residual anterior under-coverage after lateral acetabular rotation [20]. Currently, the characteristics of patients who undergo combined sagittal and coronal plane correction are unknown, and to our knowledge, no appropriate algorithm has been established for effectively combining sagittal and coronal acetabular correction to optimize the joint contact mechanics.

Individual-specific finite-element modeling has been shown to be potentially useful for ensuring the appropriate acetabular correction during the preoperative planning of PAO [28, 57]. However, in these previous studies, acetabular correction was performed only in the coronal plane, and the effect of sagittal plane correction on the joint contact mechanics was not evaluated. Although a few studies have considered sagittal plane correction [27, 36, 56], these studies involved finite-element analyses of dysplastic hips using the supine position or the standardized pelvic position, which was based on the anterior pelvic plane’s coordinate system. Recent studies have revealed that the sagittal pelvic tilt varies widely among candidates for hip preservation surgery and suggested that assessments in the standard pelvic position may overlook changes in acetabular coverage and joint contact stress in the weightbearing position [22, 41, 44]. Finite-element analyses that incorporate the physiologic pelvic tilt may provide additional insight into the effect of acetabular correction in the sagittal and coronal planes on the contact mechanics of dysplastic hips, which can lead to improved acetabular reorientation during PAO.

We therefore performed virtual PAO using finite-element models with reference to the standing pelvic position to clarify (1) whether lateral rotation of the acetabulum normalizes the joint contact pressure, (2) risk factors for abnormal contact pressure after lateral rotation of the acetabulum, and (3) whether additional anterior rotation of the acetabulum further reduces contact pressure.

Patients and Methods

Patients

Between September 2016 and March 2020, 85 patients (92 hips) with symptomatic hip dysplasia underwent transposition osteotomy of the acetabulum, one of the established PAOs to treat symptomatic hip dysplasia, which is characterized by a lateral approach and spherical osteotomy [8, 9]. Supine and standing AP pelvic radiographs and pelvic CT images were obtained for each patient during a preoperative examination. The inclusion criterion for this study (which was met by 82 patients [82 hips]) was the presence of hip dysplasia with a lateral center-edge angle < 20° on supine AP pelvic radiographs [51]. In patients with bilateral hip dysplasia, the operated-on side was investigated. Patients were excluded if they had advanced osteoarthritis (Tönnis grade ≥ 2 [42]) (13 patients), major femoral head deformity (one patient), history of surgery on either hip (15 patients), history of treatment for spinal disease (one patient), or images with insufficient quality for analysis (14 patients). Thus, 38 patients were eligible for this study. All patients were women and had a mean age of 39 ± 10 years (Table 1).

Table 1. - Patient demographic and radiographic data
Parameters Hip dysplasia (n = 38 hips) Control (n = 16 hips) p value
Age in years, mean ± SD 39 ± 10 36 ± 7 0.11
Sex, % female (n) 100 (38) 100 (16) > 0.99
BMI in kg/m2, mean ± SD 22 ± 3 21 ± 3 0.36
Height in cm, mean ± SD 157 ± 6 159 ± 6 0.59
Weight in kg, mean ± SD 55 ± 9 53 ± 8 0.48
Laterality, % right hip (n) 61 (23) 50 (8) 0.48
Tönnis classification, % (n) < 0.001
 Grade 0 61 (23) 100 (16)
 Grade 1 39 (15) 0 (0)
Lateral center-edge angle in degrees, mean ± SD 11 ± 1 32 ± 1 < 0.001

The AP pelvic radiographs and CT images of 33 women volunteers obtained for previous studies were reviewed as the control group [12, 22]. No participants in the control group had a history of diseases or articular symptoms in their hips, as determined by medical interviews and radiographic examinations. All participants provided written informed consent to participate in this study and were informed of radiation exposure. Seven patients with frank or borderline hip dysplasia (lateral center-edge angle < 25°) and 10 without suitable images were excluded. Thus, 16 individuals (16 randomly selected hips) with a mean age of 36 ± 7 years were included as the control group (Table 1).

CT Evaluations

Pelvic CT images were taken from the superior rim of the pelvis to the distal femur (matrix: 512 × 512, field of view: 261-670 mm, slice thickness: 1 mm or 2 mm) with the patient or volunteer in the supine position. We measured sagittal pelvic tilt with the participant in the standing position and the morphologic parameters using the three-dimensional (3D) Template software (Kyocera Medical Corporation). The x and y axes corresponded to the transverse and sagittal axes on the axial CT slice, respectively, while the z axis corresponded to the longitudinal axis of the scanner. First, the coordinate system of the CT scanner was aligned with the anterior pelvic plane’s coordinate system to standardize the position of the pelvis. Next, the sagittal pelvic tilt on the standing AP radiograph was reproduced on the digitally reconstructed radiographs by matching the vertical-to-horizontal ratio of the pelvic foramen. A previous study demonstrated that the correlation coefficient for the vertical-to-horizontal ratio between AP radiographs and CT-based images in the same pelvic tilt was 0.99 (p < 0.001) [35]. Sagittal pelvic tilt was measured as the angle formed by the anterior pelvic plane and the z axis (anterior pelvic plane angle), with positive values representing anterior tilt of the pelvis [41]. Morphologic parameters were measured on CT images with the standing pelvic position as a reference, including the anterior, lateral, and posterior center-edge angles; acetabular roof obliquity; acetabular anteversion angle; and acetabular inclination angle [32].

Virtual PAO and Finite-element Analysis

Mechanical Finder version 10 (Research Center for Computational Mechanics, Inc.) was used to create 3D surface models of the hemipelvis, proximal femur, and articular cartilage in order to describe the bony shape and density distribution visible on CT images [22, 46]. The articular cartilage of the acetabulum and femoral head was modeled with a constant thickness (1.8 mm) as a homogeneous and isotropic material [28, 57]. Virtual PAO was performed to mimic transposition osteotomy of the acetabulum [8, 9]. In our experience, in the clinical setting, the radius of the osteotomy line is 40 mm in most female patients. The pubic osteotomy site was located on the iliopubic tubercle. Therefore, virtual PAO was performed on the pelvic models with a spherical osteotomy line (radius of 40 mm) centered on the femoral head (Fig. 1) [20]. The acetabular fragment was reoriented in four patterns with reference to the standing pelvic position. First, the acetabular fragment was rotated laterally in the coronal plane to achieve a lateral center-edge angle of 30° on a standing AP pelvic radiograph (Fig. 2A). Then, the acetabular fragment was rotated anteriorly in the sagittal plane at 0°, 5°, 10°, and 15° (Fig. 2B).

F1
Fig. 1:
This 3D surface model represents a dysplastic hip with a spherical osteotomy line (radius: 40 mm) centered on the femoral head center on (A) AP and (B) lateral views.
F2
Fig. 2:
(A) The acetabular fragment was rotated laterally in the coronal plane to achieve a lateral center-edge angle of 30° to restore the normal lateral coverage of the femoral head. (B) After lateral rotation, the acetabular fragment was rotated anteriorly in the sagittal plane by 0°, 5°, 10°, and 15°.

The 3D surface models of bone and articular cartilage were meshed using a previously described method [22] (Fig. 3A). The mean numbers of finite elements and shell elements did not differ between the models of before and after virtual PAO (1,359,085 ± 128,332 versus 1,357,732 ± 128,988; p = 0.64 and 64,853 ± 5459 versus 65,215 ± 4912; p = 0.08, respectively). To allow for bone heterogeneity, the distribution of bone mineral densities (ρ in g/cm3) was estimated from the Hounsfield units of each image by assuming a linear relationship between the values of these units and bone mineral density [22, 31, 46]. Next, the elastic modulus of the finite-element model was evaluated using the average bone mineral density value of the element, as described by Keyak et al. [21]. The Poisson ratio of the bone was set at 0.3. The elastic modulus and Poisson ratio of the articular cartilage were set at 15 MPa and 0.45 MPa, respectively [28, 57].

F3
Fig. 3:
(A) This finite-element model represents the distribution of the elastic modulus (in MPa) in a dysplastic hip after virtual PAO. The meshed bone models were produced with a 2-mm tetrahedral element and a 0.4-mm triangular shell element on the surface. The meshed cartilage models of the acetabulum and femoral head were discretized using a locally refined 0.5-mm to 2.0-mm tetrahedral element in the weightbearing region of the acetabular cartilage. Three nodal shell elements, each with a thickness of 0.0005 mm, were placed on the surface of the acetabular cartilage to visualize the contact pressure on the acetabular cartilage. (B) The loading scenario was based on a single-leg stance, with the hip contact force acting on the nodal point at the center of the hip. During loading, the iliac crest and pubic area were completely fixed, and the distal femur was kept free only in the z direction while restrained in the x and y directions. Tied-contact and sliding-contact constraints were set on the cartilage-to-bone and cartilage-to-cartilage interfaces, respectively. The acetabular fragment was reconnected to the pelvis through a tied contact to simulate complete bony union. Frictional shear stress between the contacting articular surfaces was ignored.

Boundary and Loading Conditions

Nonlinear contact analyses were performed using the finite-element models of the dysplastic hips before and after four patterns of virtual PAO and the control group to calculate the joint contact area and joint contact pressure of the acetabular cartilage. In all analyses, load was applied with the participant in the standing pelvic position. The finite-element model of the femur was standardized with reference to the coordinate system described by the International Society of Biomechanics [52]. The definitions of tied-contact and sliding-contact constraints were set using reported cartilage-to-bone and cartilage-to-cartilage interfaces [4]. In the virtual PAO models, the acetabular fragment was reconnected to the pelvis through tied contact to simulate complete bony union. The iliac crest and pubic region were completely fixed, while the distal femur was restrained in the x and y directions and kept free only in the z direction. The loading scenario was based on a single-leg stance, with the hip contact force acting on the nodes of the femoral head’s center (Fig. 3B) [3]. A consistent weight of 500 N was defined for all patients in order to avoid the scaling effect of weight on the absolute contact pressure values. The total joint contact force was set at 1158 N, and the components of the x, y, and z axes were set at 150 N, 71 N, and 1146 N, respectively. The loaded nodes were allowed to move only in the direction of the applied load.

Ethical Approval

The ethical review board at our institution approved this retrospective study (approval number 30-137). All participants in both groups provided written informed consent to participate in this study and were informed of the radiation exposure required.

Statistical Analysis

A t-test or Wilcoxon rank sum test was used to compare continuous parameters between the hip dysplasia and control groups after we confirmed normal distribution and homoscedasticity (Shapiro-Wilk W test and F test). The chi-square test was used to compare categorical parameters between the two groups, while the paired t-test or Wilcoxon signed rank test with Bonferroni correction was used to compare continuous parameters before and after virtual PAO. The Dunnett or Steel test with the normal hip as a control was used for multiple comparisons, as appropriate. Statistical significance was set at p < 0.05. The correlation between two continuous parameters was evaluated using Pearson or Spearman correlation coefficient, as appropriate. The cutoff value of the normal maximum joint contact pressure was calculated to be 4.1 MPa using a receiver operating characteristic curve (sensitivity 100%, specificity 94%, area under the curve 0.99). An exploratory univariate logistic regression analysis was performed to screen for preoperative factors associated with abnormal contact pressure after lateral rotation of the acetabulum among morphologic factors (anterior, lateral, and posterior center-edge angles; acetabular roof obliquity; acetabular anteversion angle; and acetabular inclination angle) and sagittal pelvic tilt (anterior pelvic plane angle). Variables with p values < 0.05 (anterior center-edge angle and sagittal pelvic tilt) were included in a multivariable model to identify the independent influence of each factor. Receiver operating characteristic curves were plotted to calculate the sensitivity, specificity, and cutoff value of the independent factor. Statistical analyses were performed using JMP® version 15.0 (SAS Institute).

Results

Does Lateral Rotation of the Acetabulum Normalize the Joint Contact Pressure?

The mean contact area was smaller (500 ± 134 mm2 versus 919 ± 121 mm2, mean difference 420 mm2 [95% CI 342-497]; p < 0.001) and the median maximum contact pressure was higher (7.2 MPa [range 4.1-14] versus 3.5 MPa [range 2.2-4.4], difference of medians 3.7 MPa; p < 0.001) in dysplastic hips than in controls (Fig. 3). When the acetabulum was rotated laterally to a lateral center-edge angle of 30° (median lateral rotation angle 18.4° [range 12.6°-36.5°]), the mean contact area increased (898 ± 164 mm2 versus 500 ± 134 mm2, mean difference 398 mm2 [95% CI 350-447]; p < 0.001) and the median maximum contact pressure decreased (3.7 MPa [range 2.2-6.7] versus 7.2 MPa [range 4.1-14], difference of medians 3.5 MPa; p < 0.001) compared with the preoperative value, resulting in no difference in the mean contact area (898 ± 164 mm2 versus 919 ± 121 mm2, mean difference 21 mm2 [95% CI -70 to 113]; p = 0.64) or median maximum contact pressure (3.7 MPa [range 2.2-6.7] versus 3.5 MPa [range 2.2-4.4], difference of medians 0.2 MPa; p = 0.37) between patients with hip dysplasia and control participants (Table 2). Based on the cutoff value of the normal maximum contact pressure (4.1 MPa), 63% of the hips (24 of 38 hips) achieved the normal range of maximum contact pressure after virtual PAO (Table 2). CT parameters improved after virtual PAO (to a lateral center-edge angle of 30°) except the posterior center-edge angle and acetabular anteversion (Table 3). However, the anterior center-edge angle (47° ± 6° versus 52° ± 8°, mean difference 6° [95% CI 2°-10°]; p = 0.03) and posterior center-edge angle (94° ± 9° versus 106° ± 9°, mean difference 12° [95% CI 7°-17°]; p < 0.001) were still lower and the acetabular anteversion angle was higher (29° [range 18°-57°] versus 20° [range 10°-26°], difference of medians 9°; p < 0.001) in dysplastic hips after virtual PAO than in control hips (Table 3.).

Table 2. - Contact area and the maximum joint contact pressure before and after virtual PAO in patients with hip dysplasia and controls
Parameter Mean contact area in mm2, mean ± SD Maximum contact pressure in MPa, median (range) Hips with normal contact pressure (< 4.1MPa), % (n)
Control group 919 ± 121 3.5 (2.2-4.4) 94 (15)
Hip dysplasia group 500 ± 134a,b 7.2 (4.1-14)a,b 0 (0)
After virtual osteotomy
 Lateral center-edge angle of 30° 898 ± 164c 3.7 (2.2-6.7)d 63 (24)
 +5° of anterior rotation 942 ± 165 3.5 (2.3-7.1) 74 (28)
 +10° of anterior rotation 988 ± 171 3.2 (2.3-5.4) 76 (29)
 +15° of anterior rotation 1011 ± 184 3.1 (2.3-6.4) 84 (32)
ap < 0.001 versus the control group.
bp < 0.01 versus all virtual osteotomies.
cp < 0.05 between all virtual osteotomies.
dp < 0.01 versus +10° and +15° of anterior rotation.

Table 3. - Comparison of CT parameters between before and after virtual PAO (lateral center-edge angle of 30°) and controls
CT parameters Hip dysplasia Virtual osteotomy Control p valuea p valueb
Anterior center-edge angle in degrees 38 ± 9 47 ± 6 52 ± 8 < 0.001 0.03
Lateral center-edge angle in degrees 13 (0 to 21) 32 ± 2 32 ± 5 < 0.001 0.73
Posterior center-edge angle in degrees 98 ± 17 94 ± 9 106 ± 9 0.09 < 0.001
Acetabular roof obliquity in degrees 20 ± 5 0 (-12 to 14) 4 (-13 to 9) < 0.001 0.06
Acetabular inclination in degrees 48 ± 3 40 ± 3 41 ± 3 < 0.001 0.56
Acetabular anteversion in degrees 23 ± 4 29 (18 to 57) 20 (10 to 26) 0.04 < 0.001
Values are presented as the mean ± SD or the median (range). aHip dysplasia versus control. bVirtual osteotomy versus control.

Risk Factors for Abnormal Contact Pressure After Lateral Rotation of the Acetabulum

Before virtual PAO, the maximum contact pressure was negatively correlated with the anterior and lateral center-edge angles and the standing anterior pelvic plane angle and positively correlated with acetabular roof obliquity and the acetabular inclination angle (Table 4). When the acetabular fragment was rotated laterally to a center-edge angle of 30°, the maximum contact pressure was negatively correlated with the anterior center-edge angle and standing anterior pelvic plane angle (Table 4). After controlling for potential confounding variables such as the lateral center-edge angle and sagittal pelvic tilt, we found that a decreased preoperative anterior center-edge angle (per 1°; odds ratio 1.14 [95% CI 1.01-1.28]; p = 0.01) was independently associated with abnormal contact pressure (≥ 4.1 MPa) after virtual PAO (Table 5). The receiver operating characteristic curve analysis determined that a preoperative anterior center-edge angle < 32° in the standing pelvic position was associated with abnormal contact pressure after virtual PAO (sensitivity 57%, specificity 96%, area under the curve 0.77) (Fig. 4).

Table 4. - Correlations between the maximum contact pressure and CT parameters before and after virtual PAO (lateral center-edge angle of 30°)
Before virtual osteotomy After virtual osteotomy
Parameters Correlation coefficienta p value Correlation coefficienta p value
Morphologic parameters
 Anterior center-edge angle -0.56 < 0.001 -0.47 0.003
 Lateral center-edge angle -0.57 < 0.001 -0.06 0.72
 Posterior center-edge angle -0.21 0.21 -0.06 0.70
 Acetabular roof obliquity 0.60 < 0.001 -0.18 0.27
 Acetabular inclination angle 0.39 0.02 0.04 0.83
 Acetabular anteversion angle 0.17 0.30 0.23 0.16
Sagittal pelvic tilt
 Standing anterior pelvic plane angle -0.39 0.02 -0.52 < 0.001
aPearson or Spearman correlation coefficients.

Table 5. - Multivariate analysis of risk factors associated with abnormal contact pressure (≥ 4.1 MPa) after virtual PAO
Parameters Odds ratio (95% CI) p value
Preoperative anterior center-edge angle (1° decrease) 1.14 (1.01-1.28) 0.01
Standing anterior pelvic plane angle (1° decrease) 1.13 (0.97-1.33) 0.10

F4
Fig. 4:
The receiver operating characteristic curve for abnormal maximum contact pressure after virtual PAO (to a lateral center-edge angle of 30°) is shown. Based on the curve, the cutoff value of the preoperative anterior center-edge angle in the standing pelvic position was 31.8° (sensitivity 57%, specificity 96%, area under the curve 0.77).

Does Additional Anterior Rotation of the Acetabular Fragment Further Reduce Joint Contact Pressure?

The mean contact area increased and the median maximum contact pressure decreased with additional anterior rotation of the acetabular fragment (Fig. 5). The median maximum contact pressure at 10° and 15° of anterior rotation was lower than that at 0° of anterior rotation (that is, lateral rotation alone) (Table 2). As a result, the maximum contact pressures were within the normal range in 74% (28 of 38 hips) of the hips at 5° of anterior rotation, 76% (29 of 38 hips) at 10° of anterior rotation, and 84% (32 of 38 hips) at 15° of anterior rotation (Table 2). Among the four patterns of virtual PAO, the number of hips with the lowest maximum contact pressure was the highest at 15° of anterior rotation (53%; 20 of 38 hips), followed by 32% (12 of 38 hips) at 10° of anterior rotation, 16% (six of 38 hips) at 5° of anterior rotation, and no hips without anterior rotation.

F5
Fig. 5:
The distribution of joint contact pressures on the acetabular cartilage of the right hip in representative patients from the hip dysplasia group (lateral center-edge angle of 16°) before and after virtual PAO and from the control group (lateral center-edge angle of 30°). Lateral rotation of the acetabular fragment decreased the joint contact pressure, and subsequent anterior rotation further decreased this pressure, as reflected in the color distribution.

Discussion

Inappropriate sagittal plane correction results in an increased risk of osteoarthritis progression after PAO [26, 40, 53]. Individual variations in sagittal pelvic tilt, along with acetabular deformity, affect the acetabular coverage and joint contact mechanics in dysplastic hips [22, 41, 44] and may have an impact on the direction and degree of acetabular correction in patients who undergo PAO. Therefore, we performed finite-element analyses that considered physiologic pelvic tilt to determine the effect of acetabular correction in the sagittal and coronal planes in virtual PAO. We found that normal joint contact pressure was achieved in 63% of hips (24 of 38 hips) by normalizing the lateral acetabular coverage. However, lateral acetabular rotation was insufficient to normalize the joint contact pressure in patients with more posteriorly tilted pelvises and anterior acetabular deficiency, especially those with a preoperative anterior center-edge angle < 32° in the standing pelvic position. In such cases, additional anterior rotation was effective to normalize the joint contact pressure.

Limitations

This study has several limitations. First, certain restrictions were introduced in the specification of loading and boundary conditions. In this study, only one loading condition (a single-leg stance) was investigated; other conditions corresponding to daily activities and the gait cycle were not evaluated. Previous mathematical modeling studies reported that the distribution of contact stress in dysplastic hips changes during the gait cycle, and even during stair walking in patients with large acetabular anteversion [6, 18]. However, a recent study found that acetabular coverage measured in the standing position is a suitable surrogate for coverage measured during gait [45]; therefore, we posit that our observation of the single-leg stance scenario reasonably represents loading conditions during walking [7]. Future finite-element analysis studies are needed to validate our findings in other activities. Moreover, although a constant joint resultant force was adopted for all participants in the present study, the individual variety in the geometry of the hip and changes in pelvic morphology and lateral-medial movement of the femoral head after PAO may alter muscle forces and joint reaction forces and affect the contact stress distribution in the hip [16, 25, 38]. The loading conditions applied in this study were derived from in vivo data from patients who underwent THA [3] and are considered to approximate the actual loading conditions in the native hip. However, further studies are needed to elucidate how changes in the joint reaction force after PAO may affect the accuracy of the calculation of the joint contact pressure.

Second, we did not model patient-specific cartilage or the labrum because they were not clearly identifiable on plain CT images. Previous studies demonstrated the similarity of peak contact pressures between constant-thickness cartilage models and patient-specific cartilage models [28] and the validity of finite-element models without a labrum [2, 24]. However, the labrum may play a larger role in load transfer and joint stability [14], and further studies are required to determine the effect of an absent labrum. Third, we did not model impingement and it is unclear whether the anterior rotation performed in this study would lead to an increase in iatrogenic impingement. To optimize the joint contact mechanics while avoiding impingement after PAO, future studies should simulate impingement and evaluate the balance between acetabular coverage and hip ROM. Fourth, we did not evaluate asphericity of the femoral head [39, 49] or incongruity between the acetabulum and femoral head [19], which are common findings in dysplastic hips and might contribute to the joint contact pressure’s distribution after PAO. Further studies are required to evaluate the impact of these factors on the optimal acetabular reorientation during PAO.

We further acknowledge that 55% of the patients (47 of 85 patients) were not eligible for this study, and only female patients met the inclusion criteria, which could have resulted in selection bias. The demographic and radiographic parameters of the excluded patients did not differ from those of the included patients, suggesting that the risk of a potential bias was low. However, further research is needed to address the impact of differences in hip morphology between sexes on the generalizability of our observations. Additionally, although the hip dysplasia and control groups were comparable in terms of age and BMI, the age of the control volunteers was relatively young (mean 36 ± 7 years) to define a biomechanically healthy hip. Asymptomatic hips in older people (≥ 54 years) would have been more ideal as normal control hips [30]. However, we confirmed that the control volunteers had no history of hip disease, osteoarthritis, or morphologic abnormalities, and their maximum contact pressures were comparable to the normal range reported in a previous finite-element analysis study [47]; thus, we deemed them suitable as a control group for this study. Lastly, the study population consisted of non-obese patients with a mean BMI of 22 ± 3 kg/m2. However, we defined a consistent body weight for all participants in order to avoid the scaling effect of body weight on the absolute value of contact pressure and to focus on the effect of individual differences in morphology and physiologic pelvic tilt on contact pressure. Therefore, we believe that our observations are not influenced by the patients’ BMIs.

Does Lateral Rotation of the Acetabulum Normalize the Joint Contact Pressure?

Consistent with previous studies [20, 37], the current study showed that lateral rotation of the acetabulum is effective in correcting anterior and lateral coverage. Iwamoto et al. [20] reported that 79% of hips achieved normal anterior coverage after lateral rotation to a lateral center-edge angle of 30°, whereas 19% had residual anterior under-coverage and 2% had anterior over-coverage. Previous studies [28, 57] that used individual-specific finite-element models reported that the maximum contact pressure decreased 0.75-fold to 0.96-fold after mean lateral acetabular rotation of 7° to 10°. Similar to these reports, the median maximum contact pressure in this study decreased 0.5-fold after median lateral rotation of 18.6°, and the resulting maximum contact pressure was within the normal range in 63% of hips (24 of 38 hips). Therefore, because lateral rotation of the acetabulum can normalize the anterior and lateral acetabular coverage and joint contact pressure in a substantial number of patients, anterior acetabular rotation does not appear to be necessary in all patients in terms of preventing anterior femoroacetabular impingement after PAO.

Risk Factors for Abnormal Contact Pressure After Lateral Rotation of the Acetabulum

We observed that lateral rotation of the acetabulum failed to normalize the joint contact pressure in 37% of the hips (14 of 38). Elevated joint contact pressure after lateral acetabular rotation was associated with posterior pelvic tilt and a decreased anterior center-edge angle in the standing position. Previous studies reported that posterior pelvic tilt while weightbearing decreased anterior-to-superior acetabular coverage of the femoral head and increased joint contact pressure [22, 41]. Therefore, the results of our study suggest that future morphology-based and biomechanics-based planning studies for PAO should consider the impact of individual variations in physiologic pelvic tilt while weightbearing on acetabular reorientation to optimize the mechanical environment in the hip. A previous study using CT-based simulated PAO revealed that a preoperative anterior center-edge angle < 37° was associated with residual anterior deficiency after lateral rotation of the acetabulum [20]. In our study, the anterior center-edge angle was also determined to be an independent factor for elevated contact pressure after lateral rotation. In a subgroup of patients with a preoperative anterior center-edge angle < 32° in the standing pelvic position, additional acetabular correction in the sagittal plane was necessary to normalize the joint contact pressure.

Does Additional Anterior Rotation of the Acetabular Fragment Reduce Joint Contact Pressure?

During the reorientation process of PAO, the first step is to achieve sufficient lateral coverage through coronal correction and the second step is to achieve sufficient anterior coverage through anterior rotation [9, 43]. Although several finite-element analysis studies have considered multiplanar correction including the coronal and sagittal planes [27, 36, 56], no appropriate algorithm has been established for effectively combining sagittal and coronal acetabular correction to optimize the joint contact mechanics. A previous study using theoretical models demonstrated that anterolateral rotation of the acetabular fragment was more effective in reducing contact pressure than lateral rotation alone [15]. Similarly, in our study, we observed that the mean contact area further increased, and the median maximum contact pressure further decreased as the acetabular fragment was rotated anteriorly from 0° to 15° after lateral rotation. In a simulation study, Iwamoto et al. [20] reported that 10° to 15° of anterior rotation is appropriate to achieve sufficient anterior coverage while retaining posterior coverage. Unnecessary sagittal plane correction should be avoided because it may result in posterior under-coverage or anterior over-coverage, leading to a worse prognosis [23, 26, 33, 53]. The weightbearing acetabular cartilage area is limited in patients with severe dysplasia, making it difficult to achieve normal anterolateral coverage [54]. We simulated four patterns of acetabular reorientation in this study; however, the joint contact pressure could not be normalized in 8% (three of 38) of hips with a severe form of anterior acetabular deficiency.

Conclusion

Using virtual PAO, we demonstrated that normal joint contact pressure was achieved in 63% (24 of 38 hips) of patients after normalizing lateral acetabular coverage, suggesting that anterior acetabular rotation may not always be necessary, especially when iatrogenic femoroacetabular impingement after PAO is a concern. Nevertheless, lateral acetabular rotation was insufficient to normalize the joint contact pressure in patients with more posteriorly tilted pelvises and anterior acetabular deficiency. In patients with a preoperative anterior center-edge angle < 32° in the standing pelvic position, additional anterior acetabular rotation is expected to be a useful guide to normalize the joint contact pressure. The results of this virtual PAO study suggest that future biomechanics-based planning for PAO should incorporate not only the morphology of the hip but also physiologic pelvic tilt while weightbearing to customize acetabular reorientation for individual patients.

Acknowledgments

We thank Mitsugu Todo PhD (Research Institute for Applied Mechanics, Kyushu University, Fukuoka, Japan) and Takeshi Utsunomiya MD, Kyohei Shiomoto MD, Ryosuke Yamaguchi MD, Taishi Sato MD, and Shinya Kawahara MD (Department of Orthopaedic Surgery, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan) for their invaluable advice for this study.

References

1. Abraham CL, Knight SJ, Peters CL, Weiss JA, Anderson AE. Patient-specific chondrolabral contact mechanics in patients with acetabular dysplasia following treatment with peri-acetabular osteotomy. Osteoarthritis Cartilage. 2017;25:676-684.
2. Anderson AE, Ellis BJ, Maas SA, Peters CL, Weiss JA. Validation of finite element predictions of cartilage contact pressure in the human hip joint. J Biomech Eng. 2008;130:051008.
3. Bergmann G, Deuretzbacher G, Heller M, et al. Hip contact forces and gait patterns from routine activities. J Biomech. 2001;34:859-871.
4. Caligaris M, Ateshian GA. Effects of sustained interstitial fluid pressurization under migrating contact area, and boundary lubrication by synovial fluid, on cartilage friction. Osteoarthritis Cartilage. 2008;16:1220-1227.
5. Clohisy JC, Schutz AL, St John L, Schoenecker PL, Wright RW. Periacetabular osteotomy: a systematic literature review. Clin Orthop Relat Res. 2009;467:2041-2052.
6. Daniel M, Iglič A, Kralj-Iglič V. Hip contact stress during normal and staircase walking: the influence of acetabular anteversion angle and lateral coverage of the acetabulum. J Appl Biomech. 2008;24:88-93.
7. Debevec H, Pedersen DR, Iglič A, Daniel M. One-legged stance as a representative static body position for calculation of hip contact stress distribution in clinical studies. J Appl Biomech. 2010;26:522-525.
8. Fujii M, Nakashima Y, Kitamura K, et al. Preoperative rather than postoperative intra-articular cartilage degeneration affects long-term survivorship of periacetabular osteotomy. Arthroscopy. 2021. DOI: 10.1016/j.arthro.2021.01.060.
9. Fujii M, Nakashima Y, Noguchi Y, et al. Effect of intra-articular lesions on the outcome of periacetabular osteotomy in patients with symptomatic hip dysplasia. J Bone Joint Surg Br. 2011;93:1449-1456.
10. Fujii M, Nakashima Y, Sato T, Akiyama M, Iwamoto Y. Pelvic deformity influences acetabular version and coverage in hip dysplasia. Clin Orthop Relat Res. 2011;469:1735-1742.
11. Ganz R, Leunig M. Morphological variations of residual hip dysplasia in the adult. Hip Int. 2007;17:22-28.
12. Hara D, Nakashima Y, Hamai S, et al. Kinematic analysis of healthy hips during weight-bearing activities by 3D-to-2D model-to-image registration technique. Biomed Res Int. 2014;2014:457573.
13. Henak CR, Abraham CL, Anderson AE, et al. Patient-specific analysis of cartilage and labrum mechanics in human hips with acetabular dysplasia. Osteoarthritis Cartilage. 2014;22:210-217.
14. Henak CR, Ellis BJ, Harris MD, Anderson AE, Peters CL, Weiss JA. Role of the acetabular labrum in load support across the hip joint. J Biomech. 2011;44:2201-2206.
15. Hipp JA, Sugano N, Millis MB, Murphy SB. Planning acetabular redirection osteotomies based on joint contact pressures. Clin Orthop Relat Res. 1999;364:134-143.
16. Iglic A, Iglic VK, Antolic V, Srakar F, Stanic U. Effect of the periacetabular osteotomy on the stress on the human hip joint articular surface. IEEE Trans Rehabil Eng. 1993;1:207-212.
17. Ike H, Inaba Y, Kobayashi N, et al. Effects of rotational acetabular osteotomy on the mechanical stress within the hip joint in patients with developmental dysplasia of the hip: a subject-specific finite element analysis. Bone Joint J. 2015;97:492-497.
18. Ipavec M, Brand RA, Pedersen DR, Mavcic B, Kralj-Iglic V, Iglic A. Mathematical modelling of stress in the hip during gait. J Biomech. 1999;32:1229-1235.
19. Irie T, Orías AAE, Irie TY, et al. Three-dimensional curvature mismatch of the acetabular radius to the femoral head radius is increased in borderline dysplastic hips. PLoS One. 2020;15:e0231001.
20. Iwamoto M, Fujii M, Komiyama K, et al. Is lateral acetabular rotation sufficient to correct anterolateral deficiency in periacetabular reorientation osteotomy? A CT-based simulation study. J Orthop Sci. 2020;25:1008-1014.
21. Keyak JH, Rossi SA, Jones KA, Skinner HB. Prediction of femoral fracture load using automated finite element modeling. J Biomech. 1998;31:125-133.
22. Kitamura K, Fujii M, Utsunomiya T, et al. Effect of sagittal pelvic tilt on joint stress distribution in hip dysplasia: a finite element analysis. Clin Biomech (Bristol, Avon). 2020;74:34-41.
23. Kiyama T, Naito M, Shiramizu K, Shinoda T. Postoperative acetabular retroversion causes posterior osteoarthritis of the hip. Int Orthop. 2009;33:625-631.
24. Konrath GA, Hamel AJ, Olson SA, Bay B, Sharkey NA. The role of the acetabular labrum and the transverse acetabular ligament in load transmission in the hip. J Bone Joint Surg Am. 1998;80:1781-1788.
25. 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.
26. Lerch TD, Steppacher SD, Liechti EF, Tannast M, Siebenrock KA. One-third of hips after periacetabular osteotomy survive 30 years with good clinical results, no progression of arthritis, or conversion to THA. Clin Orthop Relat Res. 2017;475:1154-1168.
27. Liu L, Ecker T, Xie L, Schumann S, Siebenrock K, Zheng G. Biomechanical validation of computer assisted planning of periacetabular osteotomy: a preliminary study based on finite element analysis. Med Eng Phys. 2015;37:1169-1173.
28. Liu L, Ecker TM, Schumann S, Siebenrock KA, Zheng G. Evaluation of constant thickness cartilage models vs. patient specific cartilage models for an optimized computer-assisted planning of periacetabular osteotomy. PLoS One. 2016;11:e0146452.
29. Mavcic B, Pompe B, Antolic V, Daniel M, Iglic A, Kralj-Iglic V. Mathematical estimation of stress distribution in normal and dysplastic human hips. J Orthop Res. 2002;20:1025-1030.
30. Mavcic B, Slivnik T, Antolic V, Iglic A, Kralj-Iglic V. High contact hip stress is related to the development of hip pathology with increasing age. Clin Biomech (Bristol, Avon). 2004;19:939-943.
31. Miyamura S, Oka K, Abe S, et al. Altered bone density and stress distribution patterns in long-standing cubitus varus deformity and their effect during early osteoarthritis of the elbow. Osteoarthritis Cartilage. 2018;26:72-83.
32. Miyasaka D, Sakai Y, Ibuchi S, Suzuki H, Imai N, Endo N. Sex- and age-specific differences in femoral head coverage and acetabular morphology among healthy subjects-derivation of normal ranges and thresholds for abnormality. Skeletal Radiol. 2017;46:523-531.
33. Myers SR, Eijer H, Ganz R. Anterior femoroacetabular impingement after periacetabular osteotomy. Clin Orthop Relat Res. 1999;363:93-99.
34. Nepple JJ, Wells J, Ross JR, Bedi A, Schoenecker PL, Clohisy JC. Three patterns of acetabular deficiency are common in young adult patients with acetabular dysplasia. Clin Orthop Relat Res. 2017;475:1037-1044.
35. Nishihara S, Sugano N, Nishii T, Ohzono K, Yoshikawa H. Measurements of pelvic flexion angle using three-dimensional computed tomography. Clin Orthop Relat Res. 2003;411:140-151.
36. Park SJ, Lee SJ, Chen WM, et al. Computer-assisted optimization of the acetabular rotation in periacetabular osteotomy using patient's anatomy-specific finite element analysis. Appl Bionics Biomech. 2018;2018:9730525.
37. Rab GT. Lateral acetabular rotation improves anterior hip subluxation. Clin Orthop Relat Res. 2007;456:170-175.
38. Srakar F, Iglic A, Antolic V, Herman S. Computer simulation of periacetabular osteotomy. Acta Orthop Scand. 1992;63:411-412.
39. Steppacher SD, Tannast M, Werlen S, Siebenrock KA. Femoral morphology differs between deficient and excessive acetabular coverage. Clin Orthop Relat Res. 2008;466:782-790.
40. Stetzelberger VM, Leibold CS, Steppacher SD, Schwab JM, Siebenrock KA, Tannast M. The acetabular wall index is associated with long-term conversion to THA after PAO. Clin Orthop Relat Res. 2021;479:1052-1065.
41. Tachibana T, Fujii M, Kitamura K, Nakamura T, Nakashima Y. Does acetabular coverage vary between the supine and standing positions in patients with hip dysplasia? Clin Orthop Relat Res. 2019;477:2455-2466.
42. Tönnis D, Legal H. Congenital Dysplasia and Dislocation of the Hip in Children and Adults. Springer-Verlag: 1987;165-171.
43. Troelsen A, Elmengaard B, Soballe K. A new minimally invasive transsartorial approach for periacetabular osteotomy. J Bone Joint Surg Am. 2008;90:493-498.
44. Troelsen A, Jacobsen S, Rømer L, Søballe K. Weightbearing anteroposterior pelvic radiographs are recommended in DDH assessment. Clin Orthop Relat Res. 2008;466:813-819.
45. Uemura K, Atkins PR, Maas SA, Peters CL, Anderson AE. Three-dimensional femoral head coverage in the standing position represents that measured in vivo during gait. Clin Anat. 2018;31:1177-1183.
46. Utsunomiya T, Motomura G, Ikemura S, et al. Effects of sclerotic changes on stress concentration in early-stage osteonecrosis: a patient-specific, 3D finite element analysis. J Orthop Res. 2018;36:3169-3177.
47. Vafaeian B, Zonoobi D, Mabee M, et al. Finite element analysis of mechanical behavior of human dysplastic hip joints: a systematic review. Osteoarthritis Cartilage. 2017;25:438-447.
48. Wells J, Millis M, Kim Y-J, Bulat E, Miller P, Matheney T. Survivorship of the Bernese periacetabular osteotomy: what factors are associated with long-term failure? Clin Orthop Relat Res. 2017;475:396-405.
49. 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.
50. Wells J, Schoenecker P, Duncan S, Goss CW, Thomason K, Clohisy JC. Intermediate-term hip survivorship and patient-reported outcomes of periacetabular osteotomy: the Washington University experience. J Bone Joint Surg Am. 2018;100:218-225.
51. Wiberg G. Studies on dysplastic acetabula and congenital subluxation of the hip joint: with special reference to the complication of osteo-arthritis. Acta Chir Scand. 1939:5-135.
52. Wu G, Siegler S, Allard P, et al. ISB recommendation on definitions of joint coordinate system of various joints for the reporting of human joint motion--part I: ankle, hip, and spine. International Society of Biomechanics. J Biomech. 2002;35:543-548.
53. Wyles CC, Vargas JS, Heidenreich MJ, et al. Hitting the target: natural history of the hip based on achieving an acetabular safe zone following periacetabular osteotomy. J Bone Joint Surg Am. 2020;102:1734-1740.
54. Xuyi W, Jianping P, Junfeng Z, Chao S, Yimin C, Xiaodong C. Application of three-dimensional computerised tomography reconstruction and image processing technology in individual operation design of developmental dysplasia of the hip patients. Int Orthop. 2016;40:255-265.
55. Yasunaga Y, Yamasaki T, Ochi M. Patient selection criteria for periacetabular osteotomy or rotational acetabular osteotomy. Clin Orthop Relat Res. 2012;470:3342-3354.
56. 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;15:632-640.
57. Zou Z, Chávez-Arreola A, Mandal P, Board TN, Alonso-Rasgado T. Optimization of the position of the acetabulum in a Ganz periacetabular osteotomy by finite element analysis. J Orthop Res. 2013;31:472-479.
© 2021 by the Association of Bone and Joint Surgeons