Introduction

Sagittal spinal deformity has been proven to affect the function of acetabular components [1], increasing the risk of dislocation [2] and burden on the quality of life after THA [3]. Recently, orthopedic surgeons are becoming increasingly aware of the association between spinal factors and hip joint dislocation. The safe zone was defined by Lewinnek et al.’s seminal work [4] in 1978, and Delsole et al. [5] found that there was still potential instability even in the traditional safe zone. Other research showed that acetabulum cups implanted in what is considered a safe coronal position but not a safe sagittal zone may cause impingement [6]. Therefore, the exact mechanism of lumbar-pelvic-femoral dislocation after THA has not been fully elucidated. The accurate analysis of sagittal balance parameters is necessary for evaluating hip stability.

A large sample screening [7] of older Chinese patients (>65 years) shows that according to the Meyerding grading system, the incidence of lumbar spondylolisthesis is 24.84%, of which grade I or above is 22.03%, and 2.80% with grade II or higher level of slippage. The progression of osteoarthritis of the hip increases degenerative lumbar spondylolisthesis (DSPL) leading to lumbar kyphosis and pelvic retroversion [8]. The prevalence of degenerative spondylolisthesis associated with osteoarthritis of the hip joint in Japanese patients is as high as 31–36% [9]. In another study [10] comparing four degenerative lumbar spine conditions, the average complication risk ratios for spondylolisthesis were highest at 90 days after primary THA. At present, compared with the focus on lumbar spinal fusion [11, 12] or lumbar degenerative disc disease [13, 14], there are few studies of postoperative complications of THA in patients with DSPL. Therefore, we researched the effect, in the sagittal plane, of patients with lumbar spondylolisthesis, when changing from standing to sitting, on acetabular anteversion before and after THA, analyzed the risk of lumbar spondylolisthesis on postoperative dislocation.

Material and methods

Study population

Approved by IRB of the authors’ affiliated institutions, the cross-sectional study recruited 91 patients between January 2018 and November 2019, all of whom met these inclusion criteria: (1) older than 18 years, no previous history of lower limb surgery; (2) diagnosed with primary end-stage hip osteoarthritis. A flow diagram is presented as Fig. 1. We excluded 26 patients who had a high risk of dislocation and factors affecting lumbar mobility as well as the spinopelvic kinematics [15]. These included three patients with a history of spinal surgery [16], four patients with scoliosis, which was defined as an abnormal lateral curvature of the spine greater than 10° in the coronal plane, two patients with vertebral compression fractures (stable fracture of the fourth and third lumbar, respectively), four patients with spinal ankylosis (with complete bridging osteophytes, ankylosing spondylitis, or diffuse idiopathic skeletal hyperostosis), five patients with hip contracture, four patients with neurologic or musculoskeletal disorders [17] and four patients with a BMI > 35 kg/m2 [18]. The main characteristics of the population are shown in Table 1.

Fig. 1
figure 1

Flow diagram showing the selection process for the study. THA total hip arthroplasty, BMI body mass index

Full size image
Table 1 Demographic characteristics of the patients included in the study
Full size table

Defining lumbar spine diseases

All surgeries were performed under the standardized protocol of a THA procedure at our institution so that the only variable during the study period was the change of DSPL. Patients whose vertebral bodies moved forward less than 50% relative to the lower vertebral bodies were included in the DSPL group by the Meyerding classification [19], which was used for grading as follows: grade 0, no slip; grade 1, translation up to 25%; grade 2, translation of 26–50%. The mean (±SD) Meyerding grades were 1.3 (±0.6) for patients with DSPL. The interobserver kappa coefficient between the two observers, who were spine surgeons with more than 10 years of experience with classifying the grade of patients with DSPL was 0.80 (95% confidence interval, CI 0.61–0.94); the intraobserver kappa coefficient was 0.91 (95% CI, 0.82–0.99), which were considered substantial for interobserver and intraobserver variability.

Biplanar imaging and spine parameters

A total of 65 patients underwent standing and sitting biplanar frontal and lateral plane two-dimensional radiographs from thoracolumbar junction (T12–L1) to the distal femur before and 6 months after THA, using the Star-PACS imaging system (YiLianZhong, Xiamen, China). The positions of all patients in the imaging system were adjusted by the same musculoskeletal radiologist (F. Lin) to ensure image quality and quantify the hip–spine relationship (Fig. 2). Patients were asked to keep their arms held parallel to the floor in front of their chest with horizontal gaze to minimize the effect of torso posture changes of the mobility of the spinopelvic complex [20]. Patients were then positioned consistently in a relaxed seated position on the wooden stool with the torso not exerting pressure on the back of the chair. We adjusted the height of chairs so that the femora were aligned parallel to the floor to achieve 90° of apparent hip flexion with the aim of achieving a natural and significant motion trajectory of the spine-pelvis-femur.

Fig. 2
figure 2

Lateral radiographs of the spine and pelvis were made with the subject in a controlled standing (a) and sitting position (b)

Full size image

For each patient, six alignment parameters (Fig. 3) of standing and sitting postures were measured independently by two observers (K. Shen and L. Lin), using digital imaging analysis software (Materialise interactive medical image control system, ©2014 Materialise NV, Leuven, Belgium): (1) pelvic incidence (PI) is the angle formed between a line perpendicular to the upper end plate at its midpoint of S1 and a line connecting this point to the femoral head axis; (2) sacral slope (SS) is the angle between the upper plate of S1 and a horizontal line; (3) pelvic tilt (PT) is the angle between the vertical and the line joining the midpoint of the S1 endplate as well as the axis of the femoral heads [21]; (4) it has been shown that overall lumbar lordosis (LL), as a Cobb angle from the superior end plate of L1 to the upper endplate of S1 [22]; (5) pelvic–femoral angle (PFA) is the angle of femoral extension when standing (range from 170° to 190°) and of flexion when sitting (range from 120° to 144°) in relationship to the sacrum and can be used to define femoral motion [23]; (6) the method of combined anteversion, which we name in our study acetabular anteversion (AA), is a combination of both the anteversion and inclination of the acetabular component. The AA is a dynamic angle between a line tangent to the anterior and posterior edges of the acetabular cup and a horizontal line parallel to the margin of the lateral radiograph and is considered an appropriate method for assessing sagittal imbalance [24]. The change of AA in standing and sitting positions in the patient reflects the vertical and the anteroposterior synergistic relationship between the opening of the inclination and anteversion of acetabular components and the two-dimensional femoral head coverage in concert as the pelvis tilts [25].

Fig. 3
figure 3

Diagram showing the method of measurement of spinopelvic relationship in a sitting and b standing positions. SS sacral slope, PT pelvic tilt, PI pelvic incidence, AA acetabular anteversion, PFA pelvic–femoral angle, LL lumbar lordosis

Full size image

To assess the sagittal functional hip motion to predict impingement and dislocation, the functional safe zone was first proposed in 2019 by Tezuka et al. using the combined sagittal index (CSI), which is a new value for standing and sitting X‑rays [6]. The CSI outliers are confirmed to be at risk of impingement and defined as follows: standing outliers have a standing CSI > 243° (upper range of standing ante-inclination (AI) 45° + upper range of standing PFA 197°); sitting outliers have a sitting CSI < 151° (lower range of sitting AI 41° + lower range of sitting PFA 110°).

Statistical methods

The reliability of the spinopelvic complex alignment parameters was estimated using the intraclass correlation coefficient (ICC) [26]. Reliability of both intrarater and interrater values were interpreted as excellent (≥0.75) in all parameters evaluated, with ICC values of 0.809–0.965. Although few studies have concerned the analysis of the sagittal balance in patients with DSPL in definite populations, we chose 5° in the approximate difference of preoperative pelvic movement. This is similar to a previous report [13] comparing control and degenerative disc disease patients because variations of sagittal balance are performed in the same way by patients with degenerative disc disease and DSPL [27]. At the least, 60 patients in our study (control and patients with DSPL) were required to achieve 80% power to detect a difference of 5° in sacral slope angle from standing to sitting between preoperative and postoperative measurements with an estimated group mean of 30°.

The two-sample and paired-sample t-test, Pearson correlation coefficients as well as multiple linear regression analysis were used to examine whether demographics, radiographic parameters, and correlations differed significantly among the two groups. P values less than 0.05 were considered significant with a 95% confidence interval. All statistical analyses were performed using SPSS Statistics, version 20.0 (IBM Corp, Armonk, NY, USA).

Result

There were differences between standing to sitting position in preoperative lumbar-pelvic-femoral alignment measurements in patients undergoing THA (Fig. 4). Lumbar-pelvic-femoral alignment parameters’ change before and 6 months after THA appear in Table 2. After controlling for age, sex, and BMI, we found that patients with DSPL before THA had 16° more posterior PT (24° DSPL patients versus 8°control; p < 0.01), 8° less LL (35° DSPL patients versus 43°control; p < 0.05), and 20° more PFA (194° DSPL patients versus 174°control; p < 0.05) when standing than patients with normal spine. For example, as shown in Fig. 5, increased PT and decreased LL of patient 2 may be explained by the differences in lumbar-pelvic-femoral alignment in patients with DSPL; however, there was no difference in sitting PT before (p = 0.181) or after (p = 0.201) THA. Postoperatively, DSPL patients stood with a mean 15° more posterior PT (22° DSPL patients versus 7° control; p < 0.01), 8° more AA (31° DSPL patients versus 23° control; p < 0.05), and 8° more PFA (187° DSPL patients versus 179° control; p < 0.05).

Fig. 4
figure 4

Graph showing the differences from standing to sitting position preoperative lumbar-pelvic-femoral alignment measurements in patients undergoing THA. SS sacral slope, PT pelvic tilt, PI pelvic incidence, AA acetabular anteversion, PFA pelvic-femoral angle, LL lumbar lordosis

Full size image
Table 2 Differences in lumbar-pelvic-femoral alignment parameters between patients with normal spines (control) and patients with degenerative lumbar spondylolisthesis (DSPL) before and 6 months after total hip arthroplasty
Full size table
Fig. 5
figure 5

Lateral radiographs of the pelvis of patient 1 without spine arthrosis before total hip arthroplasty: a standing, b seated; preoperative c standing and d seated lateral radiographs of patient 2 with degenerative spondylolisthesis of the fifth lumbar vertebra

Full size image

Preoperative standing PT revealed a significant correlation with postoperative standing PFA in patients with DSPL (R2 = 0.9180; p < 0.01, Fig. 6) and in control patients (R2 = 0.6176; p < 0.01, Fig. 6). When considering differences between patients with DSPL and those with normal spines, the relative contributions of posterior PT and hip extension necessary to achieve a standing position were also different postoperatively.

Fig. 6
figure 6

The scatter plot shows a positive correlation between preoperative standing pelvic tilt and postoperative standing pelvic–femoral angle (THA) in control patients (R2 = 0.6176) with normal spines and in patients with DSPL (R2 = 0.9180). DSPL degenerative lumbar spondylolisthesis

Full size image

The changes of standing AA of patients with DSPL and those with normal spines before (p = 0.354) and after surgery (p < 0.05) were associated with a change in sagittal pelvic alignment following THA. There was a positive correlation between preoperative standing PT and postoperative standing AA in patients with DSPL (R2 = 0.8416; p < 0.01, Fig. 7) and in patients with normal spines (R2 = 0.6872; p < 0.01, Fig. 7).

Fig. 7
figure 7

The scatter plot shows a positive linear correlation between preoperative standing pelvic tilt and postoperative standing acetabular anteversion in (THA) control patients with normal spines (R2 = 0.6872) and in patients with DSPL (R2 = 0.8416). DSPL degenerative lumbar spondylolisthesis

Full size image

A significant positive correlation was found between preoperative standing PT and postoperative standing CSI in patients with DSPL (R2 = 0.9459; p < 0.01, Fig. 8) and that in patients with normal spines (R2 = 0.7129; p < 0.01, Fig. 8). As shown in Fig. 9, preoperative and postoperative CSI (PFA + AA) of a 74-year-old female with DSPL were significantly higher than that of the 48-year-old male without spinal arthrosis. Patients with DSPL exhibited more hip extension and required more posterior pelvic tilt in a standing position to prevent spinal flexion due to reduction of lumbar lordosis compared to control patients.

Fig. 8
figure 8

The scatter plot shows a positive linear correlation between preoperative standing pelvic tilt and postoperative standing CSI (THA) in control patients (R2 = 0.7129) with normal spines and in patients with DSPL (R2 = 0.9459). DSPL degenerative lumbar spondylolisthesis, CSI combined sagittal index

Full size image

Discussion

To achieve sagittal balance, a normal posture keeps the pelvis tilt back at 20°, hip extension about 55–70°, and LL decreases when changing the stance from upright standing to sitting [21], which may explain the variation of position in spine-pelvic motion in our study (Fig. 4). We discovered that compared to control patients, patients with DSPL had significantly increased preoperative PT (p < 0.01) and decreased LL (p < 0.05) in the standing position. Although the pelvic incidence was not statistically significant, the patients with DSPL still showed an increasing trend. Barrey et al. [27] analyzed the spinopelvic alignment of the pelvic–spine complex in three degenerative lumbar diseases and concluded that patients with DSPL have variations of sagittal alignment such as greater PI, less global LL, and increased PT. Lumbar slippage in the DSPL populations increases tendency to lose lordosis, resulting in a significant anterior displacement of the center of gravity and pelvis back tilt, which accommodates an increased flexion moment applied to the spine (Fig. 5, patient 2).

Compared with lumbar fusion and other fixed spinopelvic alignment, DSPL populations demonstrated flexible mobility of the pelvis with a much wider range through which adaptation can occur. In the sitting position, the sagittal imbalance, which is clearly compensated for by the pelvic back tilt, is due not only to increased lumbar kyphosis [28], but also to the stiff arthritic hip joint [21], so less change will occur in the seated movement arc, although there may still be significant impingement based on the original imbalance. The phenomena of more hip flexion and insufficient backward tilt of the pelvis in patients with fixed spinopelvic alignment [29] would not occur in patients with DSPL. Therefore, in this study, there was no significant difference in the measurement parameters between the control patients with pelvic retroversion due to hip arthritis and the patients with DSPL in the sitting position.

Spine deformity has consequence on sagittal balance of the coxofemoral joint in standing position and its extension capacity, especially for pelvic retroversion, which increases cup inclination and anteversion, leading to edge loading or impingement [30]. Despite the improvement in hip stiffness after THA, patients with DSPL while standing still kept a higher PT (p < 0.01) than control patients (Fig. 9). THA helped restore the hip anatomy, resulting in a clearer trajectory of the acetabular component and lumbar-pelvic-femoral movement. With back tilt of the pelvis, our study found that postoperative standing AA and PFA of the patients with DSPL increased (p < 0.05), and there was increased likelihood of posterior hip impingement with maximum hip extension.

Fig. 9
figure 9

Standing lateral radiographs of the pelvis of a 48-year-old male without spinal arthrosis: a preoperative, c postoperative; standing lateral radiographs of the pelvis of a 74-year-old female with degenerative spondylolisthesis of the first sacral vertebra: b preoperative, d postoperative

Full size image

The validation of the correlation between the combined sagittal index (CSI), which defined the functional safe zone [6], and dislocation is given in the study by Heckmann et al. [1] in which 90% of patients had the outliers of CSI associated with late dislocation. Our study found that the variation in the trend of postoperative standing CSI was more significant in patients with DSPL, among which 5 patients with CSI outliers (including only abnormal standing femoral position or abnormal acetabular position for standing CSI outliers) accounted for 16.13%, and 1 patient in the control patients accounted for 2.94%. In addition, we also found that there were positive correlations in both of the two study groups between preoperative standing PT and postoperative standing AA and PFA, of which the patients with DSPL, whose preoperative standing PT and PFA were higher, were more significant than those of the control patients. For patients with DSPL, preoperative posterior PT due to lumbar spondylolisthesis was associated with decreased coverage of the femoral head by the acetabular component and increased hip extension after surgery. Therefore, the patients with DSPL were more likely to develop the potential risk of anterior dislocation due to posterior impingement during hip extension.

The risk of impingement of the patients in standing position is relevant to the daily activities. A previous study [31] in a THA patient population has shown that anatomic AA increased 7.5° from the supine to standing positions, and the PFA had higher values in the standing position. When moving from sitting to standing, an acetabulum with normal mobility is able to cover approximately 15° [6] of anteversion, while the DSPL patients is about 29° of anteversion in our study; however, the risk of posterior impingement in standing position does not decrease due to the pelvis back tilt and femoral hyperextension of DSPL patients. In addition, potential posterior impingement often occurs while the patients are walking or getting in or out of bed [15]. It is necessary to include standing and sitting lateral spine-pelvic-hip radiographs (from thoracolumbar junction to the distal femur) that may predict the risk and direction of impingement into a standardized preoperative diagnostic pathway. Based on these radiographs, the sagittal change influenced by positioning in the DSPL patient with excessive AA and PFA can be visualized. Therefore, orthopedic doctors need to carry out education of daily activities for patients to avoid greater physical activity and poor posture, strengthen the training of dorsal lumbar muscles and decrease loads on the lumbar spine.

Phan et al. [24] summarized the current research on the influence of sagittal spinal deformity on AA during THA, and divided patients into four patterns. In this study, the patients with DSPL, with a potential increase in extension of hip and an increased AA during standing, belong to a flexible and unbalanced type whose abnormal pelvic parameters are PT > 25° and PI–LL > 10°. Phan et al. concluded that there are two preoperative treatment options for the management of the flexible and unbalanced type. Patients who have abnormal lumbopelvic parameters should preferentially consider adjustment for spinal realignment. Patients with DSPL with lumbar instability should be converted to the rigid and balanced orientation by fusion surgery in a balanced position that puts the acetabulum in a more predictable position before THA. The other possible option is to proceed with THA, in which placement of the acetabular component of patients with DSPL should be more retroverted to help correct the relative AA, especially when standing. This decreased anteversion compensates for a greater extension arc of the hips to decrease the likelihood of impingement and dislocation. It is worth noting that there is a risk of revision THA to accommodate the acetabular cup reorientation if patients with DSPL undergo spinal surgery following THA.

Although a retrospective cohort study was conducted to regulate the accuracy of data collection and measurement as much as possible, limitations still exist in this study. First, the follow-up time was short, and none of the patients in this study experienced hip dislocation at the 6‑mouth follow-up. Although we did not correlate measurements with clinical outcomes, we are following these patients for complications, including dislocation, to identify those at high risk of instability in combination with postoperative imaging parameters. Second, most patients in our study also had contralateral hip arthritis to different degrees, which affected the study results because osteoarthritis of the hip restricted pelvic mobility as it was defined earlier. Moreover, the sample size was small, and we need more patients to demonstrate the reliability of our conclusions. Finally, although static imaging in standing and sitting may not fully replicate the patient’s pelvic orientation during activities of daily living, it was of referential significance as the most common position for hip dislocation [32], and previous studies used the same method [13, 14, 20]. Despite these limitations in the present study, we tried to conduct a retrospective observational cohort study and demonstrated that DSPL can affect hip stability, and more attention should be paid to the risk of dislocation caused by the increase of AA in the postoperative standing position. We are looking forward to maximum samples and multicenter prospective controlled trials, which may better characterize this relationship in outliers, as well as the possible effects on stability after THA.

Conclusion

The mechanism by which patients with DSPL achieve a standing posture is different from those with a normal spine, with more hip extension and posterior tilt of the pelvis. The imbalance of the seated sagittal plane in patients with DSPL is usually insignificant and compensable, but more attention should be paid to the risk of impingement caused by the increase of AA in the postoperative standing position.