Association Between Facet Joint Orientation and Degenerative Spondylolisthesis: A Radiological Study of Double-Level Versus Single-Level Degenerative Spondylolisthesis
Article information
Abstract
Objective
To evaluate the correlation between lumbar degenerative spondylolisthesis (LDS) and facet joint orientation, and to examine the factors influencing facet joint orientation in patients with double-level LDS (dLDS).
Methods
A total of 40 patients with L3–5 dLDS (mean age, 64.1 years) and 106 patients with L4–5 single-level LDS (sLDS; mean age, 63.5 years) were included. Besides, 100 age-matched healthy participants were recruited as the control group. Facet joint angles at each level from L2–3 to L5–S1 were measured on axial computed tomogrpahy images. Slippage and spinopelvic sagittal parameters were measured using lateral full-spine x-rays.
Results
Both dLDS and sLDS groups had significantly larger facet joint angles from L2–3 to L5–S1 than those in the control group, except for left L5–S1. In patients with spondylolisthesis, the facet joint angles at the L2–3 and L3–4 levels in the dLDS group were significantly greater than those in the sLDS group, while the angles at the L4–5 and L5–S1 levels showed no significant differences. In contrast to the sLDS group, the dLDS group had significantly greater pelvic tilt, sagittal vertical axis, L3 slope, and L4 slope, as well as smaller sacral slope, lumbar lordosis, L3–4 disc height, L4–5 disc height, L4–5 slippage angle, and L3–S1 height. Age and dLDS were identified as independent factors influencing the changes in the L3–4 facet joint angles between the 2 LDS groups.
Conclusion
Spondylolisthesis and aging are associated with facet joint sagittalization. The present study provides evidence that the combined effects of preexisting degeneration and spondylolisthesis alter the morphology of the facet joints.
INTRODUCTION
Lumbar degenerative spondylolisthesis (LDS) is characterized by the degenerative slippage of a lumbar vertebra relative to the one below it, with an intact neural arch [1]. Single-level LDS (sLDS) is commonly seen among women above 50 years and it occurs more frequently at L4–5 segment [2,3]. According to a large-sample multicenter study, multilevel LDS accounts for 12% while sLDS accounts for the rest 88% [4]. The morphisms and segments of multilevel LDS can vary widely, and the double-level LDS (dLDS) with “step-shaped” anterolisthesis of 2 contiguous vertebrae being one of them [5].
Age, sex, and facet joints are commonly discussed as predisposing factors for LDS [3,6]. As parts of the lumbar 3-joint complex, facet joints possess a critical role in resisting anterior shear forces. Thus defects of the facet joints’ function were considered to be one of the factors contributing to the occurrence of spondylolisthesis [7]. Other than degenerative discs, some scholars considered facet joint changes as initiators in the cascade of vertebral slippage [3,8]. A more sagittal facet joint was proposed to be associated with LDS in many studies [2,9-14]. Samartzis et al. [15] identified a cutoff value of transverse facet joint angle 58° at the L4–5 level, the likelihood of LDS increased 3 times when one facet met the threshold.
However, whether the sagittal positioning of facet joints works as a preexisting morphological factor or results from remodeling after the occurrence of spondylolisthesis remains a subject of great debate. Berlemann et al. [9] held the secondary remodeling theory for they found that sagittal-oriented facet joints existed only in the spondylolisthesis segments of L4–5 LDS patients when comparing with the nonspondylolisthesis younger patients. They believed that “congenital” changes of the facet joints could not be detected in younger patients who presented LDS later in their lives. Boden et al. [10] found that sagittal facet joint orientation existed not only in the spondylolisthesis segment but also in the neighboring segments, strongly suggesting a predisposing morphological joint configuration. Few studies have focused on the facet joint orientation changes in dLDS. Hasegawa et al. [16] identified facet joint angles as one of the risk factors for dLDS compared with sLDS, but they only described the facet joint angles of greater slippage segment and neglected those of the other segments.
UP until now, there has been a lack of solid data with regard to the patterns of changes in facet joint angles. Therefore, the present study was performed, measuring facet joint angles level by level among the dLDS, sLDS, and control groups. The purposes of this study were (1) to evaluate the morphologic changes of facet joint angles in greater detail and further discuss the role of spondylolisthesis in the process of facet sagittalization, (2) to explore the possible factors contributing to changes in facet joint orientation in LDS. We formulated the hypotheses that LDS is associated with facet joint sagittalization and that facet joint sagittalization precedes the progression of spondylolisthesis.
MATERIALS AND METHODS
1. Study Population
With the approval of Institutional Review Board of Nanjing Drum Tower Hospital, Medical School of Nanjing University (ethical number: 2025-0272-01), we conducted a single-center, retrospective, cross-sectional study. The medical records and radiological images of patients who had undergone surgical intervention due to LDS from September 2016 to September 2021 were reviewed. The inclusion criteria were as follows: LDS patients with low back pain and/or radicular symptoms lasting over 6 weeks that were unresponsive to conservative treatment; age above 55 years at the time of surgery; had preoperative standing lateral full-spine x-ray examination and lumbar computed tomography (CT) scan. For the dLDS group, patients with a nonisthmic L3 anterior slip over L4 and an L4 anterior slip onto L5 by at least 5% were included. A particular “step-shaped” morphology is formed through the alignment of 2 contiguous spondylolisthesis segments. The sLDS group comprised patients with L4–5 anterior slip of at least 5%. Exclusion criteria were as follows: history of spine fractures, previous spine surgery, primary or metastatic spinal tumors, spinal tuberculosis, and scoliosis with a coronal Cobb angle greater than 20°. A total of 40 L3–5 dLDS (39 females and 1 male, average 64.1 years old) and 106 L4–5 sLDS (81 females and 25 males; average, 63.5 years old) patients were included. Besides, one hundred outpatients (68 females and 32 males; average, 63.2 years old) over 55 years old who had undergone lumbar CT scans for routine health examinations or back pain were recruited to constitute the control group. In addition to meeting the above exclusion criteria, the control group did not exhibit any detectable lumbar disc herniation, lumbar spinal stenosis, or spondylolisthesis.
2. Radiographic Measurements
CT images of intervertebral disc slices were obtained for segments from L2 to S1. Axial images were favored if they palpated at the posterior-superior corner of the caudal vertebral body. According to the literature, this slice most closely bisected the facet joints [15]. A line was drawn along the posterior border of the vertebral body, the second line connected the anterosuperior and posteroinferior tips of the left or right superior articular facet [17]. Measurements relating to the angulation of the 2 lines were made and defined as facet joint angle (Fig. 1). Facet joint tropism was defined as a difference >8° between the right and left facet joint angles [18].
(A) showing a full lateral x-ray image of a 66 years old female dLDS patient with an 80.4 mm sagittal vertical axis. (B) showing the regional lumbosacral x-ray image of this patient. Disc height=(a+b)/2; L3–S1 height (c): perpendicular distance between the 2 horizontal lines passing through the midpoint of the L3 upper endplate and S1. (C) Computed tomography image of this patient, α and β: the right and left facet joint angle at the L2–3 level. Facet joint tropism was defined as [(α+β)/2]>8°. dLDS, double-level lumbar degenerative spondylolisthesis.
Sagittal parameters were measured on standing full-spine films. Slip Parameters incorporated slip distance (SD, the anterolisthesis interval distance of an upper vertebra relates to its lower neighbor), slip angle (SA, angulation of the inferior endplate of the cephalad vertebra and the superior endplate of the caudal vertebra), slip percentage (SP, SD ratio to superior endplate of the caudal vertebra); pelvic parameters consisted of pelvic incidence (PI), pelvic tilt (PT) and sacral slope (SS). Spinal parameters included lumbar lordosis (LL, measured from the upper endplate of L1 to the upper endplate of S1), thoracic kyphosis (TK, measured from the upper endplate of T5 to the lower endplate of T12), C7 sagittal vertical axis (SVA), regional lumbosacral parameters included disc height (DH, average perpendicular interval of the anterior and posterior intervertebral distance), L3–S1 height (perpendicular distance between the 2 horizontal lines passing through the midpoint of the L3 upper endplate and S1 upper endplate) (Fig. 1), L3–L4–L5 slope (angulation of the L3–L4–L5 upper endplate and the horizontal line).
3. Statistical Analysis
IBM SPSS Statistics ver. 21.0 (IBM Corp., USA) was applied to perform the statistical analyses. All continuous variables were presented as the mean±standard deviation. Categorical variables were expressed as numbers with percentages. Independent t-test was used to compare the spondylolisthesis and sagittal parameters between sLDS and dLDS. The age and facet joint angles were analyzed using one-way analysis of variance (1-way analysis of variance [ANOVA]) followed by least significant difference post hoc test. Prior to conducting a 1-way ANOVA, the issue of unequal variances was addressed by applying a natural logarithmic transformation (ln[x]) to the relevant facet joint angles. To compare the incidence of facet joint tropism, we categorized all spondylolisthesis levels into one group and all nonspondylolisthesis levels into another, then performed a Pearson chi-square test. We also compared the facet joint tropism ratio for nonspondylolisthesis levels among the 3 groups using the same test. Univariate linear analysis was employed, and all variables with a p-value <0.10 were included in the multivariate linear regression analysis to determine the correlation between L3–4 facet joint angles and age, the occurrence of dLDS, and sagittal malalignment. A p-value lower than 0.05 was regarded as significant.
RESULTS
1. Demographic Characteristics
There were 39 women and 1 man in the dLDS group, with the mean age of 64.1 years. The sLDS included 81 women and 25 men, with an average age of 63.5 years. And the control group consisted of 32 males and 68 females, with an average age of 63.2 years. There was no difference (p=0.721) of age distribution between groups. Female preponderance was obviously evident among the LDS patients, especially in the dLDS group (Table 1).
Patient demographics and slip parameters of the groups
2. Sagittal Parameters
In terms of slip parameters, the average SD in the dLDS and the sLDS group was 7.17 mm and 6.22 mm at the L4–5 level, respectively. At this level, no significant difference in SD or SP was observed between the dLDS and the sLDS group, but the SA was significantly smaller in the dLDS group (p<0.001). The dLDS group exhibited average 5.01 mm SD at the level of L3–4, while the sLDS group exhibited no spondylolisthesis at this level. Both LDS groups exhibited no spondylolisthesis at the L2–3 and L5–S1 levels (Table 1).
As for spinopelvic parameters, the dLDS group exhibited a significantly larger SVA by an average of 22.2 mm compared to the sLDS group (p<0.001). The dLDS group also had smaller SS, LL, and TK, but a larger PT (all p<0.05), with no significant difference in PI (Table 2).
Comparison of spinal, pelvic and regional lumbosacral parameters between LDS groups
Moreover, the dLDS group had significantly smaller L3–4 DH, L4–5 DH, and L3–S1 height compared to the sLDS group, while the L3 slope and L4 slope were significantly larger (p<0.001) (Table 2).
3. Facet Joint Orientation and Facet Joint Tropism
In the control group, facet joint angles were gradually coronally changed from L2–3 to L5–S1. Compared to the control group, the facet joint angles of the 2 LDS groups were larger at all levels (all p<0.05), except for left L5–S1. When averaging the right and left facet joint angles, the 2 LDS groups exhibited more sagittal alteration in facet joint angles at L5–S1 as well (p< 0.05). From L2–3 to L5–S1, the dLDS group demonstrated facet joint angles that exceeded those of the control group by 7.67°, 16.22°, 14.92°, and 4.04°, respectively. Similarly, the sLDS group exhibited facet joint angles that were greater than those of the control group by 3.60°, 6.76°, 15.91°, and 3.18°, respectively (Table 3, Fig. 2).
Comparison of segmental facet joint angles between groups
Trend of changes in average facet joint angles from L2–3 to L5–S1. *p<0.05, ***p<0.001, comparing the facet joint angles of the dLDS/sLDS group with the control group. ††p<0.01, †††p<0.001, comparing the facet joint angles of the dLDS group with the sLDS group. ns, no significant difference when comparing the facet joint angles of the dLDS group with the sLDS group. One-way analysis of variance followed by least significant difference post hoc test. The L4–5 and L5–S1 facet joint angles were log-transformed (ln[x]) to address unequal variances. LDS, lumbar degenerative spondylolisthesis; dLDS, double-level LDS; sLDS, single-level LDS.
At the L2–3 level, both sLDS and dLDS exhibited no signs of spondylolisthesis. The mean facet joint angles were 66.6° on the right and 66.2° on the left for dLDS. For sLDS, they were 63.3° on the right and 61.3° on the left, respectively. Significantly larger facet joint angles were observed on both right and left sides in dLDS at this level (p<0.05). At the L3–4 level, spondylolisthesis was seen in dLDS but not in sLDS. The facet joint orientation in dLDS was more sagittally inclined than in sLDS (p<0.001). At the L4–5 level, both dLDS and sLDS presented spondylolisthesis. No difference in facet joint angles was identified between dLDS and sLDS (p=0.911 for right and p=0.249 for left). At the L5–S1 level, both LDS groups exhibited no signs of spondylolisthesis, and the facet joint angles did not significantly differ between the 2 groups (p>0.05) (Table 3, Fig. 2).
Since females dominated the dLDS group, we included only female participants from all 3 groups to compare facet joint angles. The results matched those in Table 3 (Supplementary Table 1).
When separating spondylolisthesis segments and nonspondylolisthesis segments, we found that the facet joint tropism ratio was 34.4% for spondylolisthesis segments and 19.1% for nonspondylolisthesis segments. The difference was significant (p< 0.001). Besides, facet joint tropism ratio in nonspondylolisthesis segments of 2 LDS groups (25.0% in dLDS and 21.1% in sLDS) were also significantly higher than the control group (14.0%) (p=0.014 and p=0.013, respectively) (Fig. 3).
Degenerative spondylolisthesis (DS) segments, degenerative spondylolisthesis segments. Non-DS segments, segments without spondylolisthesis. (A) The percentage of the facet joint tropism between all spondylolisthesis segments and all nonspondylolisthesis segments. (B) The percentage of facet joint tropism between nonspondylolisthesis segments of the dLDS, sLDS, and control group. LDS, lumbar degenerative spondylolisthesis; dLDS, double-level LDS; sLDS, singlelevel LDS; ns, no significant difference. *p<0.05, ***p<0.001. Pearson chi-square test.
4. Uni-and Multivariate Analysis
Between the dLDS and sLDS groups, facet joint angles changed most remarkably at the level of L3–4 (Fig. 3). We further explored the correlation between patient characteristics, sagittal parameters and the L3–4 facet joint angles. Univariate analysis demonstrated that age (p<0.001), double-level or single-level slip (p< 0.001), PT (p=0.032), SS (p=0.003), LL (p=0.003), L3–4 DH (p=0.002), L3-slope (p=0.031), and SS (p=0.047) were significantly related to the L3–4 facet joint angles. Multivariate linear regression analysis revealed that age (p<0.001) and double-level slip (p<0.001) served as independent factors predicting the change in the L3–4 facet joint angles in the 2 LDS groups. Every additional year of age corresponded to a 0.47° increase in the average facet joint angles at L3–4, which rose by 7.98° when transitioning from L4–5 sLDS to L3-5 dLDS (Table 4).
Analysis of factors associated with L3–4 facet joint angle changes in dLDS and sLDS groups
Similarly, in the female-only subgroup analysis, multivariate linear regression also demonstrated that age (p<0.001) and double-level slips (p<0.001) independently predicted variations in L3–4 facet joint angles across the 2 LDS cohorts (Supplementary Table 2).
DISCUSSION
In the present study, facet joint angles were measured level by level to explore the relationship between spondylolisthesis and facet joint orientation. The 2 LDS groups showed significantly sagittally altered facet joints, not limited to the spondylolisthesis levels. The facet joint angles exhibited an abnormal increase at the L3–5 spondylolisthesis levels of dLDS and the L4–5 spondylolisthesis level of sLDS. Multivariate linear regression analysis revealed that age and L3–5 double-level spondylolisthesis were significantly associated with facet joint angle changes at the L3–4 level.
In this study, the dLDS group had significantly greater SVA and PT along with smaller SS and LL compared to the sLDS group, which was similar to the previous research [4,19]. It indicated that dLDS patients showed increased forward trunk shift, with reduced LL compensated by pelvic retroversion. DH loss was reported to be an independent predictor of sLDS [20]. Our results further revealed that the intervertebral space in the dLDS group was significantly more collapsed than that in the sLDS group. This may be due to dLDS having more severe disc degeneration and entering the restabilization phase, as DH loss is considered a sign of restabilization [21].
Biomechanical studies confirmed that sagittal-oriented facet joints reduced facet joint contact forces in extension and anterior shear [22,23]. Nonetheless, the statement regarding changes in facet joint orientation occurring before or after spondylolisthesis is ambiguous. Morphological changes of the facet joints may occur with increased levels of spondylolisthesis, potentially elucidating its underlying mechanism. The L3–5 dLDS were introduced here for the convenience of level-by-level comparison with L4–5 sLDS. Spondylolisthesis occurred at the L4–5 level in both LDS groups, while at the L3–4 level, spondylolisthesis was observed in the dLDS group but not in the sLDS group.
Previous studies of L4–5 sLDS have demonstrated that sagittal change of facet joint angles happened not merely at the spondylolisthesis levels [10,13,24]. In our investigation, both the dLDS and the sLDS group showed larger facet joint angles than the control group, not only at the spondylolisthesis segments (namely L3–4 and L4–5 in the dLDS, L4–5 in the sLDS), but also at the nonspondylolisthesis segment of L2–3. These results suggested that spondylolisthesis might not be the only contributing factor to the facet joint angles changes. Interestingly, in line with Rai’s [13] measurement, for the left angle at L5–S1, neither LDS group was significantly larger than the control group. This might be explained by the additional stabilizing structures at L5–S1, such as iliac crest [6] and iliolumbar ligament [25].
At the level of L4–5, where both the dLDS group and the sLDS group presented spondylolisthesis, the facet joint angles did not differ between the 2 groups. At the level of L2–3, neither the dLDS group nor the sLDS group presented spondylolisthesis. However, dLDS had significantly larger facet joint angles at this level. This further indicated that facet joint angle changes were not affected by spondylolisthesis alone. Some other factors may work preceding spondylolisthesis. At the level of L3–4, spondylolisthesis was only seen in dLDS but not in sLDS. Significantly larger facet joint angles of dLDS were observed at this level. Also, in the sLDS group, the facet joint angles at the L4–5 spondylolisthesis level demonstrated a significant increase compared to the control group at the corresponding segment. Nevertheless, we could not simply conclude here that remodeling process of spondylolisthesis changed the shape of facet joint angles for the reasons that preexisting factors were not eliminated. To solve this problem, we calculated the average facet joint angles of the right and left sides, drew a line plot to demonstrate the changes in facet joint angles at each level. The following points highlighted the impact of spondylolisthesis on the sagittal alteration of facet joint angles. Firstly, at the L3–4 level, the facet joint angles of the dLDS group were significantly larger than those of both the control group (16.2° difference) and the sLDS group (9.5° difference). In contrast, sLDS demonstrated only a mild increase compared to control at this level (6.8° difference). Secondly, while the dLDS group maintained notably larger facet joint angles than the sLDS group at L3–4 (9.5° difference), this intergroup disparity substantially diminished at L2–3 (4.1° difference). Thirdly, facet joint angles increased progressively from L5–S1 to L2–3 in the control group, and showed a similar trend from L5–S1 to L3–4 in the dLDS group. However, they decreased from L3–4 to L2–3 in the dLDS group. Similarly, in the study of L4–5 sLDS, Cinotti et al. [26] found that the facet joint angles at the spondylolisthesis segment (i.e., L4–5) more closely resembled those of its upper adjacent level rather than being coronally oriented. Collectively, these observations lend tentative support to the concept that sagittally oriented facet joints might constitute a preexisting morphological feature, while spondylolisthesis development appears associated with progressive sagittalization.
To our knowledge, no study has yet analyzed the correlation between sagittal parameters and facet joint angles regarding dLDS. Although univariate linear regression analysis revealed significant correlations between PT, SS, LL, L3–4 DH, L3-slope, and changes in the L3–4 facet joint angles, these correlations were not observed in multivariate linear regression analysis. Age and double-level spondylolisthesis were identified as independent factors influencing L3–4 facet joint angle changes in the multivariate linear regression analysis model. Several studies have reported that more sagittal-oriented facet joints were correlated with age [27,28]. Aging leads to global spinal degeneration, including facet joints and lumbar discs, which may trigger the development of spondylolisthesis [29-31]. Thus, we considered that aging might work as a preexisting factor. Each additional year of age resulted in a 0.47° increase in the facet joint angles at L3–4, suggesting that aging acts as a facilitator in a gentle manner. The presence of double-level spondylolisthesis being an independent factor affirmed the influence of spondylolisthesis progression. Weinberg et al. [32] described that sagittal facet joints were more prevalent in patients with facet arthritis. This condition is indicative of secondary remodeling of the facet joints. More sagittally oriented facet joints failed to provide anterior-posterior translation restraint, and the unstable motion led to an inflammatory cascade and eventual osteophyte formation. Spondylolisthesis may further remodel the already age-altered facet joints.
This study emphasized the role of preexisting factor (age) and spondylolisthesis in the sagittal alteration of the facet joints. Additionally, facet joint orientation has been reported to be associated with superior facet joint violation in pedicle screw implantation [33,34]. Comprehending the changes in facet joints can help in developing customized operative plans and avoiding postoperative complications during surgery.
This study has several limitations. First, the recruitment of patients could potentially introduce selection bias, as females made up the majority the dLDS group, although a study indicated that facet orientation is not influenced by sex [35]. While this study has partially addressed the focus on female subgroup, future research will be necessary to conduct a more comprehensive investigation into male subgroup. Second, to ensure a balanced age distribution in our comparison of facet joint angles, we included subjects aged above 55, which may limit the generalizability of our findings. Third, this study failed to follow the progression of spondylolisthesis, and further longitudinal study is needed to be taken to make the conclusion solid.
CONCLUSION
In conclusion, this study supports the correlation between spondylolisthesis and aging with the sagittal alteration of the facet joints in LDS. The segmental comparison of facet joint angles between dLDS and sLDS provides evidence that sagittal facet joint orientation may already exist before spondylolisthesis occurs, and the development of spondylolisthesis exacerbates the progression of facet joint sagittalization. The combined effects of preexisting degeneration and spondylolisthesis change the morphology of the facet joints.
Supplementary Materials
Supplementary Tables 1-2 are available at https://doi.org/10.14245/ns.2550654.327.
Comparison of segmental facet joint angles between groups in female cohorts
Analysis of factors associated with L3–4 facet joint angle change in female cohorts of dLDS and sLDS group
Notes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This work was supported by the Development Project of Nanjing Science and Technology Commission and Foundation (ZKX20020), and the Jiangsu Provincial Medical Innovation Center of Orthopedic Surgery (CXZX202214).
Acknowledgments
The authors thank all the patients who participated in this study.
Author Contribution
Conceptualization: ZTZ, XS; Data curation: BW, ZZZ, XS; Formal Analysis: ZTZ; Investigation: HZ; Methodology: ZTZ; Software: ZTZ; Validation: QZ, XS; Project Administration: YQ; Writing – original draft: ZTZ, HZ, QZ; Writing – review & editing: BW, YQ, ZZZ, XS.