A Comparison of Radiographic Outcomes after Total Hip Arthroplasty between the Direct Lateral Approach and Posterior Lateral Approach with EOS 2D/3D X‐Ray Imaging System

Abstract

Objective

The classical approaches for total hip arthroplasty (THA) are the direct lateral approach (DLA) and posterior lateral approach (PLA). There are few studies comparing implant orientation with these two approaches, and the impact of surgical approaches on implant orientation remains controversial. With the rise of the EOS imaging system, we aimed to use it to identify the differences between and factors associated with implant orientation after THA using DLA and PLA.

Methods

In our department from January 2019 to December 2021, 321 primary unilateral THAs that used PLA and DLA were enrolled. A total of 201 patients who received PLA and 120 patients who received DLA were included in this study. Two blinded observers measured each case using EOS imaging data. Postoperative imaging metrics and other relevant influencing factors of the two surgical approaches were compared. Postoperative imaging metrics, including the anteversion and inclination of the cup, anteversion of the stem, and combined anteversion were measured based on EOS. Other relevant influencing factors included age, approach, gender, laterality, BMI, anterior pelvic plane inclination, femoral head diameter, femoral offset, lateral pelvic tilt, pelvic incidence, pelvis axial rotation, sacral slope, sagittal pelvic tilt, and surgery time. Multiple linear regression analyses were performed to identify the predictors of acceptability for each imaging data point.

Results

No dislocation was found in the 321 patients who underwent primary THA during this period. The mean anteversion and combined anteversion of the cups using the DLA were 21.33° ± 17.31° (−51.7°–60.8°) and 33.71° ± 20.85° (−38.8°–77.6°) and PLA were 25.34° ± 12.76° (−5.5°–57.0°) and 42.37° ± 18.85° (−8.7°–84.7°), respectively. The DLA group had smaller anteversion (p = 0.038) and combined anteversion (p < 0.001). We found that surgical approach (p < 0.05), anterior pelvic plane inclination (p < 0.001), gender (p < 0.001), and femoral head diameter (p < 0.001) were important factors affecting acetabular cup anteversion (R ² = 0.375) and combined anteversion (R ² = 0.525).

Conclusions

In total hip arthroplasty, different prosthesis installation directions should be made according to different surgical approaches. Compared with the direct lateral approach, the acetabular anteversion can be intentionally enlarged when using the posterolateral approach. Surgical approach, anterior pelvic plane inclination (APPI), gender, and femoral head diameter were significant predictors of prosthesis orientation. The anterior pelvic plane inclination may be a useful standard for assessing the position of the prosthesis using EOS.

Table 1. Clinical description of the patients (PLA n = 201; DAL n = 120) (no missing data for the variables presented). The bolded values indicate a P‐value <0.05. BMI, body mass index; PLA, the posterior lateral approach; DLA, the direct lateral approach. Table 2. Counts and percentages of variables comparing the direct lateral approach and the posterior lateral approach. The bolded values indicate a P‐value <0.05. PLA, the posterior lateral approach; DLA, the direct lateral approach. Table 3. Multiple linear regression analysis with implant orientation and 14 factors. The dependent variables are implant orientations. The independent variables are influencing factors. The bolded values indicate a P‐value <0.05; (A), anterior pelvic plane; (P), patient's plane. Table 4. Multiple linear regression analysis with implant orientation and 13 factors in the direct lateral approach. The dependent variables are implant orientations. The independent variables are influencing factors. The bolded values indicate a P‐value <0.05; (A), anterior pelvic plane; (P), patient's plane. Table 5. Multiple linear regression analysis with implant orientation and 13 factors in the posterior lateral approach. The dependent variables are implant orientations. The independent variables are influencing factors. The bolded values indicate a P‐value <0.05; (A), anterior pelvic plane; (P), patient's plane. Table 6. Evaluation of intra‐rater and inter‐rater reliability in DDH cases. (A), anterior pelvic plane; (P), patient's plane. Table 7. Evaluation of inter‐rater reliability in all cases. (A), anterior pelvic plane; (P), patient's plane.

Introduction

Total hip arthroplasty (THA) is an effective treatment for coxarthrosis pain, ¹ restoring hip joint function, and improving the quality of life in patients. ² Currently, the classical approaches for THA are the posterior lateral approach (PLA) and direct lateral approach (DLA). ³ Potential disadvantages of DLA include splitting of the abductor muscles, which may cause gluteal insufficiency and postoperative pain, and potential disadvantages of the PLA include a higher dislocation rate, which is one of the most common complications following THA. ⁴ To avoid dislocation, the orientation of implant should be planned reasonably. Implant orientation, ⁵ including the anteversion and inclination of the cup and anteversion of the stem prosthesis, is a predictive risk factor for dislocation and is associated with clinical outcomes after THA. ⁶

In the 1970s, Lewinnek first defined the safe zone of the acetabulum cup. It was defined as 40° ± 10° of inclination and 15° ± 10° of anteversion, ⁷ was conceived to have a lower rate of dislocation, ⁸ but the accuracy and reproducibility are uncertain. In order to study the acetabular cup and femoral stem as a whole, Dorr et al. proposed the concept of combined anteversion. ⁹ Combined anteversion, ¹⁰ as the sum of the acetabular and femoral stem inversions, was conceived to predict and evaluate the effect of THA ¹¹ and concluded to be 37.6° ± 7° (range, 19°–50°). ⁹ Although the range of combined anteversion still requires further verification, many studies have achieved good results. ¹¹ We should select appropriate equipment to measure the above indicators.Standard radiography and computed tomography (CT) are usually used to complete imaging measurements ¹² to evaluate the orientation of the prosthesis and bony structures, ¹³ but they still have drawbacks. Standard radiography cannot be reconstructed three‐dimensionally. CT cannot measure hip joint parameters in standing position. In this study, we used a more advanced radiography device, the biplanar imaging system EOS (EOS Imaging, Paris, France), which has many advantages, including three‐dimensional (3D) reconstruction, ¹⁴ capability to avoid errors caused by bony deformities, ¹⁵ and no magnification errors for the bilateral full‐length limb measurements in the weight‐bearing standing position. ¹⁶ Nowadays, EOS is used increasingly frequently in joint surgery because of its 3D reconstruction capabilities and accuracy. ¹⁷ EOS has good reproducibility. ¹⁸ Peter et al.'s ¹⁴ study demonstrated that EOS is a satisfactory device for lower limb examinations.

There are few studies comparing implant orientation with these two approaches, and the impact of surgical approaches on implant orientation remains controversial. Therefore (i) the primary goal was to compare the differences in postoperative imaging between the two surgical approaches utilizing the EOS imaging system, and (ii) the secondary aims of this study were to investigate the safe zone of prosthesis positioning and the factors influencing prosthesis position.

Material and Methods

This retrospective study was approved by the First Affiliated Hospital of Soochow University Review Board (IRB2022371). Between January 1, 2019, and December 31, 2021, 352 patients underwent primary THA at our institution (Fig. 1). To eliminate the impact of human error, the information of 31 patients was removed from the overall data, and 321 cases were eventually included. The human error here refers to inaccurate positioning of the reference point on the 3D image due to prosthetic occlusion. A total of 201 patients who received PLA and 120 patients who received DLA were included in this study. All operations in this study were performed by the same chief physician, who had more than 20 years of professional experience in total hip arthroplasty. A Zimmer Biomet femoral stem and acetabular component were used for all operations (Zimmer Biomet Corporation, USA).

Fig. 1.

Subject flow diagram. THA, total hip arthroplasty.

Inclusion and Exclusion Criteria

The inclusion criteria were as follows: (i) patients with end‐stage hip diseases who underwent primary THA in our department from January 2019 to December 2021; (ii) underwent PLA or DLA; and (iii) complete image data from postoperative EOS imaging. The average follow‐up time was 12.3 ± 5.9 months (range, 3–21 months).

The exclusion criteria were as follows: (i) patients with disease that prevent accurate imaging measurements due to prosthetic occlusion during radiography; (ii) history of previous surgeries on the affected lower extremity; (iii) bilateral hip replacement and either knee replacement (including TKA and UKA); and (iv) severe deformities, such as DDH cases with Crowe type III and IV, etc. (Figure 1).

Demographics

Preoperative general patient data and postoperative 3D reconstructed hip data were collected. Preoperative patient data included age, gender, height, weight, body mass index (BMI), and the direction of the affected limb. The 3D reconstructed hip model included pelvic incidence, sacral slope, sagittal pelvic tilt, lateral pelvic tilt, pelvic axial rotation, anterior pelvic plane inclination, acetabular cup inclination, acetabular cup anteversion, femoral head diameter, femoral offset, neck length, neck shaft angle, femoral torsion, stem anteversion, femur length, femoral mechanical angle, and hip knee shaft (HKS) angles. The patients were diagnosed with femoral head necrosis (n = 98, 30.5%), femoral neck fracture (n = 61, 19.0%), hip osteoarthritis (n = 92, 28.6%), developmental hip dysplasia (DDH) (n = 62, 19.3%), traumatic arthritis (n = 3, 0.009%), rheumatoid arthritis (n = 2, 0.006%), and coxotuberculosis (n = 3, 0.009%).

Surgical Approach

The patients were classified into two groups based on the surgical approach used: the direct lateral approach (DLA group, 120 hips) and the posterior lateral approach (PLA group, 201 hips).

In the DLA group, the operation was performed using a modified Hardinge approach. The gluteus medius and minimus muscles were bluntly separated in the anterior middle 1/3 of the gluteus medius. ¹⁹ The hip was flexed and externally rotated to create anterior dislocation of the hip joint.

In the PLA group, the operation was performed using a Gibson approach. The gluteus maximum was split, and the short external rotators and capsule were taken as a unit. The short external rotators and capsule were repaired to the greater trochanter after implantation. ²⁰ The head of the femur was dislocated posteriorly by flexion and internal rotation of the hip joint.

Measurement of EOS

EOS radiographs were acquired on the day before surgery and on the third day after surgery. The patient stood with staggered lower limbs in a standard posture. Preoperative and postoperative images were analyzed using the EOS 3D software by two independent senior radiologists (J.H. and D.J.) trained to use the procedure (Fig. 2). EOS scans were performed on patients from L4 to the planta postoperatively. The 2D image we refer to is a two‐dimensional X‐ray film taken by EOS, and 3D is a three‐dimensional reconstruction model of the target joint obtained by computer software. All the data in this study are obtained by 3D reconstruction of two‐dimensional images. The EOS measurements were performed by two blinded, experienced independent senior radiologists using CorelDRAW (version 12.0; Corel; Canada), and the results were averaged (Figure 3). When the pubic symphysis was posterior to the anterosuperior iliac spine (ASIS), the anterior pelvic plane inclination (APPI) measurement was negative. When the pubic symphysis was anterior to the ASIS, APPI was positive. The measurement procedure using EOS biplane images is shown in Figure 4.

Fig. 2.

The representative image shows 3D model after reconstruction using the EOS biplanar images.

Fig. 3.

The representative image shows the measurement workflow based on EOS biplanar images, which include pelvic incidence (°); sacral slope (°); sagittal pelvic tilt (°); lateral pelvic tilt (mm); acetabular cup inclination (°); acetabular cup anteversion (°); femoral offset (mm); neck shaft angle (°); femoral torsion (°); stem anteversion (°); femur length (cm).

Fig. 4.

(A) Draw a line from the posterior corner of the sacral plate to the anterior corner of the sacral plate; (B) Confirmed the position of the pubic symphysis and the right; (C) Left anterior iliac spines to determine the pelvic positioning; (D) Identified the key landmarks on the femur, including the center of the femoral head, the center of the trochlea notch; (E) The diameter of the prosthetic head; (G) the distal third of the femoral diaphysis, the trochlear notch, the medial and lateral condyles; (F) The position of the prosthetic neck; (H) On the frontal view, used the following eight marker points to identify the proximal diaphysis to identify the contour of the stem.

Measurements

The sacral slope was defined as the angle in the sagittal plane between the sacral plate and horizontal axis. The femoral offset was defined as between the center of the prosthetic head and orthogonal projection of the center of the prosthetic head on the femoral anatomical axis. The pelvis axial rotation was defined as between the acetabular axis and frontal plane. The pelvic incidence was defined as the angle between the line perpendicular to the sacral plate at its midpoint and line connecting the midpoint of the sacral plate and the midpoint of the acetabular axis in the patient's sagittal plane. The sagittal pelvic tilt was defined as the angle between the line connecting the midpoint of the sacral plate and midpoint of the acetabular axis and vertical axis in the patient's sagittal plane. The anterior pelvic plane (APP) was defined by the three following anatomical landmarks: left anterosuperior iliac spine, right anterosuperior iliac spine, and pubic symphysis. The APPI was defined as between the APP and the patient's frontal plane. Patient's plane was defined as vertical plane containing the line connecting the centers of the acetabula (acetabular axis). The patient's plane (PP) is called the “functional plane,” and the anteversion in this plane is defined as “functional anteversion”; the anterior pelvis plane is called “anatomical plane,” and the anteversion in this plane is defined as “anatomical anteversion.” The acetabular cup inclination was defined as the angle between the acetabular axis (line connecting the center of the cup with the center of the contralateral acetabulum) and the intersection line of the acetabular cup plane and APP. The acetabular cup anteversion was defined as between the line orthogonal to the acetabular axis and the intersection line of the acetabular cup plane and the patient's axial plane or axial plane of the APP. The stem anteversion was defined as between the femoral stem's neck axis and the posterior condylar axis (axis passing through the most posterior points of the medial and lateral condyles), which projected in the plane orthogonal to the femoral mechanical axis. The combined anteversion was defined as the algebraic sum of stem anteversion and acetabular cup anteversion in both the PP and APP. Combined anteversion safe zone: the position of the acetabular cup is relatively safe; the inclination of the cup is 40° ± 10° and the anteversion of the cup is 15° ± 10°.

Statistical Analysis

Data and statistical analyses were performed using SPSS 26.0 (IBM Corp., Armonk, NY, USA). Age, BMI, pelvic incidence, sacral slope, sagittal pelvic tilt, lateral pelvic tilt, pelvis axial rotation, anterior pelvic plane inclination, acetabular cup inclination, acetabular cup anteversion, femoral head diameter, combined anteversion, neck shaft angle, femoral offset, femoral torsion, and stem anteversion were continuous variables and are presented as means and standard deviations; between‐group comparisons of these values were performed using Student's t‐tests. Gender, laterality were categorical variables and are presented as numbers (n) and frequencies (%); between‐group comparisons were performed using the chi‐square test. One‐way analysis of variance was used to compare the differences in indicators, and the indicators that might be meaningful were used for multiple linear regression analysis. In analyzing the influence factors for anteversion, inclination, and combined anteversion, we performed a multiple linear regression analysis. The dependent variables are acetabular cup inclination, stem antetorsion, acetabular cup anteversion (A and P), and combined anteversion (A and P). The independent variables are age, approach, gender, laterality, BMI, anterior pelvic plane inclination, femoral head diameter, femoral offset, lateral pelvic tilt, pelvic incidence, pelvis axial rotation, sacral slope, sagittal pelvic tilt, and surgery time. Correlation coefficients were determined by the Spearman rank correlation test using two‐tailed P values. This process was performed using R software (version 3.5.3; The R Project for Statistical Computing). P < 0.05 was considered statistically significant.

Results

We retrospectively analyzed the imaging data of 321 patients who had EOS imaging data of both lower extremities after total hip arthroplasty in our hospital. There were 120 (37.4%) and 201 (62.6%) patients in the DLA and PLA groups, respectively. There was no statistical difference in average age, sex, BMI, and laterality of all patients in this study (Table 1). No patients had a history of dislocation during hospitalization or at any follow‐up time point. All measurements were performed in both the APP and PP.

TABLE 1.

Clinical description of the patients

Clinical data	PLA (n = 201)	DLA (n = 120)	p‐Value
Age (Years)	60.0 ± 15.0	62.6 ± 13.5	0.160
Gender (Male/Total)	91/201	48/120	0.337
BMI (kg/m2)	24.9 ± 15.0	24.7 ± 13.4	0.366
Laterality (Left/Total)	90/201	57/120	0.663

Abbreviations: BMI, body mass index; DLA, the direct lateral approach; PLA, the posterior lateral approach.

Primary Outcomes

We compared the number of prostheses distributed in Lewinnek's safe zone between the PLA and DLA. The distribution of DLA and PLA was 48.3% (58/120) versus 40.3% (81/201) (APP) and 51.7% (62/120) versus 39.3% (79/201) (PP) within the safe zone based on Lewinnek's criteria (Fig. 5). The distribution of DLA and PLA was 40.0% (48/120) versus 47.3% (95/201) (APP) and 30.8% (37/120) versus 42.3% (85/201) (PP) within the safe zone of combined anteversion (Fig. 5).

Fig. 5.

The results of this study, which included all patients enrolled following primary THA with two different surgical approaches and two different planes scatter distributions and the safe zone based on an inclination of 40° ± 10° and a combined anteversion angle from 25° to 50°. PLA, posterior lateral approach; DLA, direct lateral approach.

Secondary Outcomes

Acetabular Inclination Angle

There was no significant difference between the mean inclination angle of PLA and that of DLA, (41.60° ± 6.66° [range, 12.2° to 59.5°] versus 40.75° ± 6.41° [25.8° to 63.4°]) (p = 0.991), respectively (Table 2). There was no statistical difference in the THA inclination angle between the two surgical approaches (Fig. 6).

TABLE 2.

Counts and percentages of variables comparing the direct lateral approach with the posterior lateral approach

Variable	PLA (n = 201)	DLA (n = 120)	p‐Value
Pelvic incidence (°)	47.67 ± 11.26	48.63 ± 13.13	0.581
Sacral slope (°)	39.50 ± 10.60	40.78 ± 8.59	0.027
Sagittal pelvic tilt (°)	8.17 ± 9.62	7.95 ± 9.77	0.501
Lateral pelvic tilt (mm)	8.98 ± 7.96	8.33 ± 6.20	0.036
Pelvis axial rotation (°)	0.14 ± 6.88	−1.56 ± 6.83	0.833
Anterior pelvic plane inclination (°)	0.89 ± 10.02	0.96 ± 10.30	0.524
Acetabular cup inclination (°)	40.75 ± 6.41	41.60 ± 6.66	0.991
Acetabular cup anteversion (°)	24.83 ± 12.10	21.03 ± 13.78	0.037
Acetabular cup anteversion (°) A	25.34 ± 12.76	21.33 ± 17.31	0.038
Femoral head diameter (mm) H	43.52 ± 3.37	43.53 ± 4.21	0.048
Combined anteversion A	42.37 ± 18.85	33.71 ± 20.85	<0.001
Combined anteversion P	41.86 ± 18.54	33.70 ± 18.79	<0.001
Neck shaft angle (°) A	127.8 ± 14.45	130.59 ± 5.13	0.015
Femoral offset (mm) A	47.73 ± 16.37	47.82 ± 15.87	0.79
Femoral torsion (°)	17.45 ± 10.58	16.06 ± 9.92	0.494
Stem anteversion (°)	17.71 ± 15.15	12.38 ± 12.48	<0.001
Femur length (cm)	39.61 ± 5.14	39.40 ± 2.66	0.474
HKS (°) H	5.38 ± 3.10	5.38 ± 1.50	0.202

Note: The bolded values indicate a p‐value <0.05; (A), Anterior pelvic plane; (P), Patient's plane.

Abbreviations: BMI, body mass index; DLA, the direct lateral approach; PLA, the posterior lateral approach.

Fig. 6.

The violin plot indicates the difference between the two surgical approaches in the above angles. (A), Anterior Pelvic Plane; (P), Patient's Plane; NS, No significance; *p < 0.05; ***p < 0.001.

Acetabular Cup Anteversion

With APP as the reference plane, there was a significant difference in the mean acetabular cup anteversion between (PLA and DLA [25.34° ± 12.76° vs. 21.33° ± 17.31°, p = 0.038]) (Table 2). With PP as the reference plane, the mean acetabular cup anteversion of PLA was significantly different from that of DLA (21.03° ± 13.78° vs. 24.83° ± 12.10°, p = 0.037). There was a difference in anteversion between the two surgical approaches (Fig. 6).

Stem Anteversion

There was a significant difference between the mean stem anteversion with PLA and that with DLA at (17.71° ± 15.15° [−19° to 50.1°] vs. 12.38° ± 12.48° [−20.4° to 46.2°]) (p < 0.001), respectively (Table 2).

Combined Anteversion

There was a statistically significant difference in the combined anteversion between the two approaches when either APP or PP was used as the reference plane. In APP as the reference plane, the mean combined anteversion with PLA and that with DLA was (42.37° ± 18.85° [−8.7° to 84.7°] vs. 33.71° ± 20.85° [−38.8° to 77.6°]) (p = 0.001), respectively. In PP as the reference plane, the angle of PLA and that of DLA were (33.70° ± 18.79° [−6.2° to 83.6°] vs. 41.86° ± 18.54° [−9.0° to 73.0°]), respectively (Table 2). All measurements of the combined anteversion angle were based on the safe zone (25°–50°) suggested by Dorr et al. ⁹ The differences in combined anteversion between the two surgical approaches were compared (Figure 6) and there was a significant difference (p < 0.001) regardless of the reference plane (Table 2).

Correlations between Prosthesis Position and 14 Other Factors

To identify prosthesis position‐related factors, 14 variables were analyzed for correlations with prosthesis position using multiple linear regression analysis. In this study, the mean APPI in the PLA and DLA groups were 0.89° ± 10.02° and 0.96° ± 10.30°, respectively (Table 2). The correlation between prosthesis position and factors is presented in Figure 7. In Table 3, we found that both approaches and APPI significantly affected the acetabular cup anteversion, acetabular cup inclination, and combined anteversion (p < 0.05). In Tables 4 and 5, we found that APPI is the only influencing factor of anatomic anteversion and combined anteversion but had no significant effect on functional anteversion and combined anteversion. In other words, for a person's pelvic position, the anteversion of the acetabular cup based on APP is a fixed value. The anatomic anteversion does not change with the change of posture. In the PLA group, our study found a 0.622° and 0.694° increase in cup and combined anteversion, respectively, with each degree of change in APPI. In the DLA group, there was a 0.950° and 0.834° increase in cup and combined anteversion, respectively, with each degree of change in APPI. Multiple linear regression analysis was conducted separately for the DLA and PLA groups, and we can still conclude that APPI remained an important factor influencing acetabular cup inclination, acetabular cup anteversion and combined anteversion (p < 0.05, Tables 4 and 5). Age was illustrated in Table 5 as an influencing factor for the combined anteversion and acetabular cup anteversion (p < 0.05). We found that APPI was positively correlated with acetabular cup anteversion (Figure 8).

Fig. 7.

The correlation coefficients were calculated by the Spearman rank correlation test using two‐tailed P values and shown in the plot. Red dots represent positive correlations, while blue dots represent negative correlations. BMI, body mass index; *p < 0.05; **p < 0.01; ***p < 0.001.

TABLE 3.

Multiple linear regression analysis with implant orientation and 14 factors (P value)

Factors	Acetabular cup inclination	Stem antetorsion	Acetabular cup anteversion (A)	Combined anteversion(A)	Acetabular cup anteversion (P)	Combined anteversion (P)
Approach	0.766	0.015	0.003	0.032	<0.001	<0.001
Gender	<0.001	<0.001	0.642	<0.001	0.003	< 0.001
Laterality	0.076	0.802	0.116	0.277	0.201	0.070
Age	0.009	0.212	0.729	0.048	0.276	0.122
Surgery time	0.202	0.656	0.775	0.491	0.921	0.785
BMI	0.420	0.792	0.227	0.669	0.500	0.575
Femoral head diameter	<0.001	<0.001	0.125	<0.001	<0.001	<0.001
Femoral offset	0.786	0.555	0.936	0.383	0.732	0.52
Lateral pelvic tilt	0.1	0.189	0.02	0.075	0.012	0.004
Pelvic incidence	0.947	0.704	0.006	0.77	0.092	0.072
Pelvis axial rotation	0.021	0.257	0.643	0.172	0.275	0.218
Sacral slope	0.901	0.568	0.006	0.644	0.115	0.088
Sagittal pelvic tilt	0.817	0.869	0.004	0.959	0.051	0.037
Anterior pelvic plane inclination	<0.001	<0.001	0.199	0.614	<0.001	0.209

Note: The bolded values indicate a p‐value <0.05; (A), Anterior pelvic plane; (P), Patient's plane.

Abbreviations: BMI, body mass index; DLA, the direct lateral approach; PLA, the posterior lateral approach.

TABLE 4.

Multiple linear regression analysis with implant orientation and 13 factors in the direct lateral approach (P value)

Factors	Acetabular cup inclination	Stem antetorsion	Acetabular cup anteversion (A)	Combined anteversion (A)	Acetabular cup anteversion (P)	Combined anteversion (P)
Age	0.454	0.626	0.239	0.573	0.144	0.441
Surgery time	0.018	0.004	0.752	0.09	0.249	0.36
BMI	0.803	0.101	0.863	0.224	0.866	0.208
Pelvic incidence	0.917	0.059	0.888	0.175	0.718	0.125
Sacral slope	0.779	0.074	0.753	0.342	0.989	0.223
Sagittal pelvic tilt	0.95	0.075	0.685	0.138	0.505	0.094
Lateral pelvic tilt	0.91	0.924	0.93	0.998	0.866	0.904
Pelvis axial rotation	0.233	0.262	0.21	0.094	0.215	0.114
Anterior pelvic plane inclination	0.007	0.522	<0.001	<0.001	0.257	0.702
Femoral head diameter	0.127	0.035	0.653	0.083	0.934	0.460
Femoral offset	0.617	0.467	0.692	0.437	0.673	0.344
Gender	0.123	0.448	0.755	0.846	0.879	0.769
Laterality	0.241	0.170	0.819	0.53	0.636	0.89

Note: The bolded values indicate a p‐value <0.05; (A), Anterior pelvic plane; (P), Patient's plane.

Abbreviations: BMI, body mass index; DLA, the direct lateral approach; PLA, the posterior lateral approach.

TABLE 5.

Multiple linear regression analysis with implant orientation and 13 factors in the posterior lateral approach (P value)

Factors	Acetabular cup inclination	Stem antetorsion	Acetabular cup anteversion (A)	Combined anteversion (A)	Acetabular cup anteversion (P)	Combined anteversion (P)
Age	0.737	0.105	0.041	0.01	0.063	0.016
Surgery time	0.655	0.369	0.948	0.535	0.828	0.683
BMI	0.803	0.495	0.434	0.978	0.43	0.97
Pelvic incidence	0.808	0.46	0.825	0.691	0.774	0.752
Sacral slope	0.832	0.482	0.801	0.727	0.751	0.787
Sagittal pelvic tilt	0.802	0.399	0.863	0.612	0.815	0.669
Lateral pelvic tilt	0.047	0.542	0.893	0.719	0.265	0.24
Pelvis axial rotation	0.083	0.382	0.815	0.626	0.584	0.746
Anterior pelvic plane inclination	0.018	0.009	<0.001	<0.001	0.34	0.221
Femoral head diameter	0.35	0.01	0.176	0.315	0.11	0.493
Femoral offset	0.745	0.507	0.333	0.866	0.412	0.464
Gender	0.562	0.033	0.063	0.752	0.032	0.952
Laterality	0.871	0.875	0.737	0.91	0.366	0.61

Note: The bolded values indicate a p‐value <0.05; (A), Anterior pelvic plane; (P), Patient's plane.

Abbreviations: BMI, body mass index; PLA, the posterior lateral approach; DLA, the direct lateral approach.

Fig. 8.

The scatter plot shows the correlation of anterior pelvic plane inclination with the acetabular cup anteversion and combined anteversion in anterior pelvic plane. A, Anterior pelvic plane; P, Patient's plane; PLA, posterior lateral approach; DLA, direct lateral approach. A: the relationship between APPI and Acetabular Cup Anteversion; B: the relationship between APPI and Combined Anteversion.

Intra‐Rater and Inter‐Rater Reliability

Table 6 has presented intra‐observer and inter‐observer intraclass correlation coefficient (ICC) values. The ICC values showed excellent agreement (>0.9) for all measurements. None of the differences were statistically significant. Therefore, it is accurate to use EOS imaging system to measure the imaging data of patients with DDH after THA. In Table 7, we verified that the measurements of the two radiologists were consistent (p > 0.9).

TABLE 6.

Inter‐observer and intra‐observer intraclass correlation coefficients for all DDH parameters

Imaging data	Inter‐observer intraclass correlation coefficients (ICC)		Intra‐observer intraclass correlation coefficients (ICC)
	3 day (95% CI)	3 mouth (95% CI)	3 day (95% CI)	3 mouth (95% CI)
Acetabular cup inclination (°)	0.98 (0.98,0.99)	0.97 (0.99,0.96)	0.97 (0.99,0.98)	1.00 (0.99,1.00)
Acetabular cup anteversion (°) (A)	0.99 (0.98,0.99)	0.98 (0.98,0.99)	1.00 (0.98,0.97)	0.98 (0.98,0.99)
Combined anteversion (°) (A)	1.00 (0.99,1.00)	1.00 (1.00,1.00)	0.99 (0.96,0.98)	0.98 (0.97,0.99)
Acetabular cup anteversion (°) (P)	0.99 (0.98,1.00)	0.98 (0.97,0.97)	0.99 (0.98,0.99)	0.97 (0.96,0.98)
Combined anteversion (°) (P)	0.97 (0.99,0.99)	1.00 (1.00,1.00)	1.00 (1.00,1.00)	0.95 (0.94,0.97)
Stem anteversion (°)	1.00 (1.00,1.00)	0.99 (0.98,0.99)	1.00 (1.00,1.00)	1.00 (1.00,1.00)
Anterior pelvic plane inclination (°)	1.00 (0.99,0.99)	0.99 (0.98,0.99)	1.00 (1.00,1.00)	1.00 (1.00,1.00)
Femoral head diameter (mm)	1.00 (1.00,1.00)	0.99 (0.98,0.99)	1.00 (0.98,0.99)	1.00 (1.00,1.00)
Femoral offset (mm)	0.97 (0.99,0.99)	1.00 (1.00,1.00)	1.00 (1.00,1.00)	0.97 (0.96,0.98)
Lateral pelvic tilt (mm)	1.00 (1.00,1.00)	0.99 (0.98,0.99)	0.96 (0.99,0.98)	0.99 (0.99,0.99)
Pelvic incidence (°)	0.98 (0.99,0.96)	1.00 (1.00,1.00)	0.99 (0.99,0.98)	0.99 (0.99,1.00)
Pelvis axial rotation (°)	1.00 (1.00,1.00)	0.97 (1.00,0.99)	0.99 (0.98,0.99)	0.96 (0.93,0.97)
Sacral slope (°)	1.00 (1.00,1.00)	0.98 (0.99,1.00)	1.00 (1.00,1.00)	0.96 (0.95,0.97)
Sagittal pelvic tilt (°)	0.99 (0.98,0.99)	0.99 (0.96,1.00)	0.99 (0.98,0.99)	0.93 (0.95,0.97)

Abbreviation: (A), Anterior pelvic plane; (P), Patient's plane.

TABLE 7.

Inter‐observer intraclass correlation coefficients in all cases

Imaging data	ICC	95% CI
Acetabular cup inclination (°)	0.978	0.973–0.982
Acetabular cup anteversion (°) (A)	0.993	0.992–0.995
Acetabular cup anteversion (°) (P)	0.995	0.993–0.996
Stem anteversion (°)	0.992	0.990–0.994
Anterior pelvic plane inclination (°)	0.992	0.987–0.995
Femoral head diameter (mm)	0.911	0.886–0.930
Femoral offset (mm)	0.995	0.994–0.996
Lateral pelvic tilt (mm)	0.971	0.964–0.976
Pelvic incidence (°)	0.989	0.987–0.991
Pelvis axial rotation (°)	0.994	0.990–0.998
Sacral slope (°)	0.985	0.982–0.988
Sagittal pelvic tilt (°)	0.993	0.991–0.997

Abbreviations: (A), Anterior pelvic plane; (P), Patient's plane; ICC, Intraclass correlation coefficients.

Discussion

There is different acetabular cup inclination, acetabular cup anteversion, stem anteversion, safe zone, and combined anteversion between DLA and PLA. We found that combined anteversion, acetabular cup anteversion, and safe zone with PLA were greater than those with DLA. The distribution of our research results in the safe zone is different from Lewinnek. This is probably because we chose APP as the reference plane. In addition, the surgical approach, anterior pelvic plane inclination, gender, and femoral head diameter were factors for cup orientation. Based on the previous literature and the results of our study, we conclude that APP has higher stability as the reference plane of THA. We recommend APP as a reference plane.

The Differences in Postoperative Imaging between the Two Surgical Approaches

Different approaches may require different prosthesis orientations to avoid dislocation after THA because different tissues are broken in the use of different surgical approaches. Combined anteversion and safe zones have been proposed and used as important methods for prosthesis orientation during THA. However, previous studies on the combined anteversion had focused on THA using PLA. Therefore, it is worthwhile investigating whether these concepts are still applicable to the DLA approach and whether the ranges of the different surgical approaches are the same.

Proper prosthesis orientation is a crucial factor in determining stability after THA, including acetabular cup inclination and acetabular cup, stem, and combined anteversions. Because femoral prostheses tend to dislocate from the direction of soft tissue destruction after total hip arthroplasty, different approaches may require different implant orientations. Using PP as a reference plane, 51.7% of THA patients on DLA and 39.3% of THA patients on PLA were inside the Lewinnek safe zone. ⁷ This DLA result is similar to the 52.9% distribution rate obtained by Li et al. ¹⁹ but lower than that (57%) measured by Timothy et al. ²¹ using CT. This difference may be due to the biological reasons for the different positions in CT and EOS examinations. Using APP as the reference plane, there was a difference in the distribution rate of the two surgical approaches in the safe zone (DLA, 48.3% vs. PLA, 40.3%). Many studies have demonstrated that THA using DLA has a lower rate of dislocation than THA using PLA, including a study by Masonis et al. (DLA, 0.55% vs. PLA, 3.23%) ²² that found that DLA (0.55%) had the lowest rate of dislocation compared to PLA (3.23%). Hernandez et al. ²³ found that repairing the anatomical posterior capsule during THA is important; therefore, the DLA dislocation rate may be due to limited posterior capsular destruction. Therefore, proper prosthesis orientation for different approaches is worth investigating.Changes in body positions caused different measurements of anteversion in X‐rays or CT scans. We selected the APP as a reference plane to reduce the effect of body position and measured cup inclination, cup anteversion, and combined anteversion with the EOS imaging system in the standing position. Fujishiro et al. ²⁴ measured the THA with PLA by CT and found that the cup anteversion, cup inclination, and combined anteversion were 24.7° ± 11.3°, 41.0° ± 6.2°, and 65.0° ± 15.7°, respectively. Li et al. ¹⁹ measured the THA with DLA using CT and found that the cup anteversion, femoral stem anteversion, cup inclination, and combined anteversion were 9.26° ± 11.19°, 13.38° ± 10.7°, 38.83° ± 5.04°, and 23.1° ± 13.4°, respectively. Using EOS, we found a significantly smaller femoral stem anteversion (17.71° ± 15.15°) and combined anteversion (42.37° ± 18.85°) with PLA, and greater cup anteversion (21.33° ± 17.31°) and cup inclination (41.60° ± 6.66°) with DLA than their results. This difference may be due to the working principle of the equipment. The EOS imaging system allows the simultaneous acquisition of two orthogonal X‐ray images ²⁵ without magnification factor. ²⁶ It also allows better handling of lower limb deformities when the patient is in a functional standing position at the time of the exam. ²⁷ In case of torsional malalignment, ²⁸ EOS system is uncorrelated with CT, and shows advantages because it is not dependent on the position of the leg. ²⁹ In the 1970s, Lewinnek proposed the safe zone for THA, but its accuracy was limited in PLA owing to the small number of cases and the short follow‐up period ⁷ and was used to compare PLA with other surgical approaches. A different safe zone may be required for the DLA, especially in terms of anteversion. Abdel et al. ³⁰ demonstrated differences in the distribution of dislocation cases within the safe zone for different surgical approaches. Lewinnek's study was based on anteroposterior radiographs without lateral radiographs, and femoral torsion and position were not evaluated. The errors caused by anatomical deformities can be reduced by performing 3D reconstruction of the acetabulum and femur. Esposito et al. ³¹ reported a 2% dislocation rate at 6 months which varied with the duration of follow‐up, and this limitation persisted in our study. Moreover, the distribution of the two surgical approaches in the safe zone was different (Figure 5). This phenomenon may be caused by differences in the surgeon's perspective and the direction of intraoperative artificial dislocation because it is difficult to properly assess acetabular anteversion using surgical instruments and anatomical landmarks.

Additionally, we discovered that no dislocation occurred in 30.8% of the THAs utilizing DLA (33.70° ± 18.79°) and 42.3% of the THAs using PLA (41.86° ± 18.54°), which both had combined anteversions in the safe zone (25°–50°) using PP as the reference plane. When APP was used as the reference plane, the results for the distribution (PLA47.3%>DLA40.0%) of the two angles were similar. APPI (p < 0.01) was an important influencing factor of stem anteversion, which influences combined anteversion. However, compared to earlier research, ²¹ we used EOS biplane images, suggesting that THA using different approaches requires various safety zones. Several studies have indicated that the Lewinnek safe zone may no longer be applicable. Abdel et al. reported that 58% of dislocated THAs had prostheses within the Lewinnek safe zone for both inclination and anteversion. ³⁰

Our data showed that the mean anteversion (p = 0.038) and combined anteversion (p < 0.001) were greater in PLA than in DLA, suggesting that a wider safe zone can be achieved using PLA. When using the PLA approach, the anteversion of the acetabular cup placed by the joint surgeon was larger than that of the DLA approach to ensure surgical safety. Therefore, hip prostheses should be placed in different orientations when using different surgical approaches for THA.

Analysis of Factors Affecting the Position of Prosthesis Installation

This study involved two planes: the anterior pelvic plane (APP) and patient's plane (PP). Different reference planes resulted in different measurement results. Many previous studies have used PP as a reference plane, while we used APP, which is widely applied in THA, as the reference plane, since all CT examinations were performed with the patient in the horizontal supine position. When the patient's position changes during daily life, the measurement of anteversion does not correspond to the actual state. Because the pelvic structure is the most stable, the pelvis is always in a neutral position regardless of the true pelvic position, ³² and measurement of the anteversion differs between body positions. ³³ Therefore, using the APP as a reference plane reduces the influence of postural changes and may represent the true state of the human body.

Anterior pelvic plane inclination (APPI) was the most significant factor affecting acetabular cup anteversion (Figure 7). Many previous studies have revealed that the pelvic tilt angle affects the degree of acetabular cup anteversion. ³⁴ , ³⁵ , ³⁶ We found that APPI was an important factor affecting the acetabular cup anteversion without considering what kind of surgical approach to use (Table 3). When surgical approach was considered, we found that APPI remained an important factor influencing acetabular cup anteversion, and that its size was positively correlated with acetabular cup anteversion (Tables 4 and 5). Lembeck et al. used CT measurements to find that a 1° change in pelvic inclination can lead to a change in the anteversion of the acetabular cup by about 0.7°. ³⁵ Moreover, our study found a 0.622° and 0.694° increase in cup and combined anteversion, respectively, with each degree of change in APPI in the PLA group, and a 0.950° and 0.834° increase, respectively, with each degree of change in APP in the DLA group. Different examination equipment, patient positions, and reference planes were used, dependent on individual pelvic mobility and initial 3D cup positioning. ³³ APP, a more stable reference plane, was the focus of this study. The results of several studies ³⁷ indicate that APPI is a potential factor influencing acetabular anteversion. We found that APPI (p < 0.05) was an influencing factor of acetabular cup anteversion and combined anteversion in both DLA and PLA groups.

We also found that gender and femoral head diameter were factors influencing acetabular cup anteversion (Table 3). Hartel's study also showed that gender is a small but significant factor in acetabular cup anteversion in THA, ³⁸ differing from previous studies. ¹⁹ This may be caused by chance, as there was no difference in patient gender or femoral head diameter. To place the acetabular cup in the proper position, a relatively fixed plane for preoperative planning is crucial. In total hip arthroplasty, which plane is used as the reference of naked eye observation during the installation of acetabular prosthesis will directly affect the measurement of acetabular cup anteversion after operation. During the operation, the visual anteversion is usually the patient's functional anteversion, which may be the reason why the standard deviation of acetabular anteversion is too large. We suggest that in addition to fixing the pubic symphysis and sacrum, the position of the anterior pelvic plane should be clearly marked during the preoperative stage. In this way, when locating the prosthesis during the operation, the ideal anteversion can be obtained according to the angle formed between the surgical instrument and the anterior pelvis plane. Based on these results, we propose paying more attention to the APPI during the operation to avoid dislocation.

Limitations and Strengths

Our study has some limitations. First, most patients undergoing hip replacement are over 65 years old ³⁹ and have varying degrees of osteoporosis. When the EOS imaging system was used for postoperative examinations, poor visualization of bone structures and obscuration of surrounding tissue caused errors in localization of modeling marker sites. Second, all patients underwent EOS imaging only after the drainage tube was removed on the third postoperative day, and vascular ultrasonography of both lower limbs revealed no substantial vascular abnormalities. From the first postoperative day, all patients had been receiving oral celecoxib and intravenous flurbiprofen for analgesia. Patient apprehension following total hip arthroplasty and individual variability in pain tolerance ⁴⁰ may make it difficult for patients to cooperate with standard body shots during EOS tests, causing unavoidable error owing to body position. Third, the follow‐up in this study was within 1–1.5 years after surgery, and since we did not find any dislocated patients during the follow‐up period, we lacked positive control cases. Fourth, EOS is rarely utilized as a standard postoperative review test because of its high cost. This study lacks longitudinal comparisons of hip prosthesis orientation and pelvic position over time, as well as follow‐up imaging data of EOS after 1 year. Fifth, the hip joint was the focus for measurements in this study, extending up to the superior sacral endplate plane and down to the distal femur. Although the radiographic range extended from the sole of the foot to the upper endplate plane of L4, data below the knee was lacking.

We believe that our study makes a significant contribution to the literature because it suggests the anterior pelvic plane inclination may be more useful to assess the position of the prosthesis using EOS, and the anterior pelvic plane should be considered as a reference plane for total hip arthroplasty. Also, we used, EOS system, a new technology for research.

Conclusions

This is the first report of radiographic outcomes between the direct lateral approach and the posterior lateral approach with the EOS 2D/3D X‐ray imaging system after THA. In total hip arthroplasty, different prosthesis installation directions should be made according to different surgical approaches. Compared with the direct lateral approach, the acetabular anteversion can be intentionally enlarged when using the posterolateral approach. In addition, the surgical approach, anterior pelvic plane inclination, gender, and femoral head diameter were factors for cup orientation. In the future, studies on the correlation between the anterior pelvic plan inclination and orientation of the prosthesis are important. APP reference planes should be considered for THA‐related parameter measurements.

Author Contributions

Conception and design of study: L.X. Huang, L.S. Li, D.H. Jiang, J. Huang; acquisition of data: R. Xie, Y.F Qian, J. Huang, Q. Wu, D.H. Jiang; analysis of data: R. Xie, Y.F. Qian, L.S Li; drafting the manuscript: R. Xie, J. Huang; revising the manuscript critically for important intellectual content: L.S Li, L.X. Huang.

Conflict of Interest

The authors declare that we have no competing interests.

Ethics Statement

The study was approved by the ethical committees, and all participants provided written informed consent.

Data Availability Statement

All data related to this case report are contained within the manuscript.